The Lipid Handbook with CD-ROM Third Edition
The Lipid Handbook with CD-ROM Third Edition Edited by
Frank D. Gunstone John L. Harwood Albert J. Dijkstra
Boca Raton London New York
CRC Press is an imprint of the Taylor & Francis Group, an informa business
CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2007 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number-10: 0-8493-9688-3 (Hardcover) International Standard Book Number-13: 978-0-8493-9688-5 (Hardcover) This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use. No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC) 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data The lipid handbook with CD-ROM / [edited by] Frank D. Gunstone, John L. Harwood, Albert J. Dijkstra. -- 3rd ed. p. ; cm. Includes bibliographical references and index. ISBN-13: 978-0-8493-9688-5 (alk. paper) ISBN-10: 0-8493-9688-3 (alk. paper) 1. Lipids--Handbooks, manuals, etc. I. Gunstone, F. D. II. Harwood, John L. III. Dijkstra, Albert J. [DNLM: 1. Lipids. QU 85 L7633 2007] OP751.L547 2007 572’57--dc22 Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com
2006052237
CONTENTS
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Editors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
1
Fatty Acid and Lipid Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 C.M. Scrimgeour and J.L. Harwood 1.1 Fatty acid structure (CMS). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Lipid structure (JLH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2
Occurrence and Characterisation of Oils and Fats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 F.D. Gunstone and J.L. Harwood 2.1 Introduction (FDG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 2.2 Major oils from plant sources (FDG). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 2.3 Minor oils from plant sources (FDG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 2.4 Milk fats, animal depot fats and fish oils (FDG). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 2.5 Waxes (JLH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .108 2.6 Egg lipids (JLH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .115 2.7 Milk lipids (JLH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .117 2.8 Liver and other tissue lipids (JLH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 2.9 Cereal lipids (JLH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .121 2.10 Leaf lipids (JLH). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .123 2.11 Algal lipids (JLH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125 2.12 Fungal lipids (JLH). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .128 2.13 Bacterial lipids (JLH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .134 2.14 Lipids of viruses (JLH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .141
3
Production and Refining of Oils and Fats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .143 A.J. Dijkstra and J.C. Segers 3.1 Introduction (AJD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .143 3.2 Production of animal oils and fats (AJD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .147 3.3 Production of vegetable oils and fats (AJD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .155 3.4 Degumming of oils and fats (AJD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .177 3.5 Alkali refining of oils and fats (AJD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .191 3.6 Soapstock and by-product treatments (AJD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .204 3.7 Bleaching of oils and fats (AJD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .212 3.8 Dewaxing of oils (AJD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .231 3.9 Vacuum stripping of oils and fats (AJD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .235 3.10 HACCP for oils and fats supply chains (JCS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .251
4
Modification Processes and Food Uses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .263 A.J. Dijkstra 4.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .263 4.2 Hydrogenation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .264
v
Contents
4.3 4.4 4.5 4.6
Interesterification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .285 Fractionation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .300 Food grade emulsifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .315 Food uses of oils and fats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .333
5
Synthesis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .355 M.S.F. Lie Ken Jie, J.L. Harwood and F.D. Gunstone (with W.H. Cheung and C.N.W. Lam) 5.1 Unsaturated fatty acid synthesis via acetylene (MSFLKJ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .355 5.2 Fatty acid synthesis by the Wittig reaction (MSFLKJ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .359 5.3 Isotopically labelled fatty acids (MSFLKJ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .363 5.4 Synthesis of acylglycerols (MSFLKJ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .368 5.5 Fullerene lipids (MSFLKJ). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .375 5.6 Glycerophospholipids (JLH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .386 5.7 Sphingolipids (JLH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403 5.8 Glycosylglycerides (JLH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .406 5.9 Bulk separation procedures (FDG). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .410
6
Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .415 A. J. Dijkstra, W.W. Christie and G.Knothe 6.1 Introduction (AJD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .415 6.2 Requirements stemming from quality control and process investigation (AJD) . . . . . . . . . . . . . . . . . . . .420 6.3 Some selected analytical methods (AJD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .423 6.4 Chromatographic analysis of lipids (WWC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .426 6.5 Nuclear Magnetic Resonance Spectroscopy (GK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .455
7
Physical Properties: Structural and Physical Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .471 I. Foubert, K. Dewettinck, D. Van de Walle, A.J. Dijkstra and P.J. Quinn 7.1 Introduction (IF) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .471 7.2 Crystallisation and melting (IF) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .472 7.3 Phase behaviour (KD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .491 7.4 Lipid/water interactions (DVdW) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .495 7.5 Interaction between lipids and proteins (PJQ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .503 7.6 Biological membranes (PJQ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .509
8
Chemical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .535 G. Knothe, J.A. Kenar and F.D. Gunstone 8.1 Autoxidation and photo-oxidation (GK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .535 8.2 Enzymatic oxidation (GK). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .542 8.3 Epoxidation, hydroxylation and oxidative fission (GK). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .546 8.4 Halogenation and halohydrins (GK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .551 8.5 Oxymercuration (JAK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .552 8.6 Metathesis (JAK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .554 8.7 Stereomutation (JAK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .555 8.8 Double-bond migration and cyclisation (JAK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .557 8.9 Cyclisation (GK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .559 8.10 Dimerisation (JAK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .564 8.11 Chain branching and extension (GK). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .566 8.12 Hydrolysis, alcoholysis, esterification and interesterification (GK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .570 8.13 Acid Chlorides, Anhydrides and Ketene Dimers (JAK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .576 8.14 Peroxy acids and related compounds (JAK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .577 8.15 Nitrogen-containing compounds (JAK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .579 8.16 Other reactions of the carboxyl group (JAK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .583 8.17 Oleochemical carbonates (JAK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .585 8.18 Guerbet compounds (GK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .587
9
Nonfood Uses of Oils and Fats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .591 F.D. Gunstone, J. Alander, S.Z. Erhan, B.K. Sharma, T.A. McKeon and J.-T. Lin 9.1 Introduction (FDG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .591 9.2 Basic oleochemicals (FDG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .592 9.3 Surfactants (FDG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .597 9.4 Cosmetics and personal care products (JA). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .604 9.5 Lubricants (SZE and BKS). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .610 9.6 Biofuels (FDG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .625 9.7 Surface coatings and inks (FDG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .629 9.8 Castor oil products (TAMcK and J-TL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .632 vi
Contents
10
Lipid Metabolism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .637 J.L. Harwood 10.1 Fatty acids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .637 10.2 Glycerophospholipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .668 10.3 Glyceride metabolism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .680 10.4 Glycosylglycerides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .686 10.5 Sphingolipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .689 10.6 Lipids as signalling molecules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .694 10.7 Sterol esters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .698 10.8 Control mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .699
11
Medical and Agricultural Aspects of Lipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .703 J.L. Harwood, M. Evans, D.P. Ramji, D.J. Murphy and P.F. Dodds 11.1 Human dietary requirements (JLH). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .703 11.2 Lipids and cardiovascular disease (ME) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .710 11.3 Clinical aspects of lipids with emphasis on cardiovascular disease and dyslipaemia (DPR) . . . . . . . . . .721 11.4 Skin lipids and medical implications (JLH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .742 11.5 Sphingolipidoses (JLH). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .746 11.6 Other disorders of lipid metabolism (JLH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .749 11.7 Pulmonary surfactant (lung surfactant) (JLH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .751 11.8 Agricultural aspects (DJM, JLH and PFD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .756
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i Dictionary Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Name Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 617
vii
PREFACE
of which can be detailed in the text. We are grateful to Taylor & Francis for allowing us to include this information and we thank Fiona Macdonald for assistance in selecting and organising it. In order to make our task manageable in the time scale agreed between the publishers and the editors and to present authoritative coverage of our topics, we have secured the assistance of several contributors from Europe, Hong Kong and the United States. Only one contributor (P. J. Quinn) and two of the editors (F. D. Gunstone and J. L. Harwood) were involved with the previous edition and almost the entire text now has different authors. This brings fresh minds to the volume. By bringing a wide range of information into a single volume, we hope that the book will be useful to all who work in the lipid field as scientists or technologists, in industrial or academic laboratories, as newcomers, or as those who already know their way around the field. Lipid science is of increasing interest for metabolic, nutritional, and environmental reasons and we offer this revised and updated volume as a contribution to that growth. For 20 years the book has provided assistance to a generation of those working with lipids and we offer LH-3 (our acronym for this work) to the next generation. The third edition is also available on a CD-ROM (included with the book). This will provide a compact form of the so-called “Handbook” and will be easily searchable, thereby providing easy access to material hidden in tables and figures and in the extensive list of references, which now come with full titles.
The Lipid Handbook was first published in 1984, with a second edition in 1994. We now present the third edition of this successful book, with Albert Dijkstra replacing Fred Padley as a member of the editorial team. The decision to revise this book was made late in 2004 and most of the writing was completed during 2005. We planned the book to take account of the many changes in lipid science and technology that have occurred in the past 10 years, but we sought to maintain the approach and organisation of material used in the earlier editions. Compared to the second edition, some chapters have been combined — “Fatty acid structure” with “Lipid structure” (Chapter 1), “Separation and isolation” with “Analytical methods” (Chapter 6), along with the two chapters on “Physical properties” (Chapter 7). Other chapters have been divided — The former chapter on “Processing” now appears as separate chapters devoted to “Production and refining of oils and fats” (Chapter 3) and to “Modification processes and food uses” (Chapter 4). One new chapter —“Nonfood uses” (Chapter 9) has been introduced. All chapters have been rewritten (often by a new author) and we have sought to present information on the basis of thinking and practice in the present day. One interesting change is that the processing sections refer to patents now easily accessible through espacenet.com or uspto.gov. In addition, the Dictionary section has been extended on the basis of the latest Taylor & Francis Group databases. This contains a wealth of information covering chemical structures, physical properties, and references to hundreds of lipid and lipid-related molecules, only some
F. D. Gunstone J. L. Harwood A. J. Dijkstra
ix
EDITORS
currently editor of four journals, including executive editor of Progress in Lipid Research. Dr. Harwood has published nearly 500 scientific papers and communications, plus authoring three books (including Lipid Biochemistry) and editing 14 others. He has given many plenary and named lectures, received his D.Sc. in 1979 and is in receipt of personal prizes. He also has awards for his publications and those of his students. He is an honorary visiting scientist at the Malaysian Palm Oil Board (Kuala Lumpur), Centre d’Etudes Nucléaires (Grenoble), and the Hungarian Academy of Sciences (Szeged).
Frank D. Gunstone, Ph.D., is professor emeritus of the University of St. Andrews (Scotland) and holds an honorary appointment at the Scottish Crop Research Institute (Invergowrie, Dundee, Scotland). He received his Ph.D. from the University of Liverpool (England) in 1946 for studies with the late Professor T. P. Hilditch, and subsequently, there followed an academic career in two Scottish Universities: Glasgow (1946 to 1954) and St. Andrews (1954 to 1989). He continues to be professionally active and has spent over 60 years studying fatty acids and lipids with many publications to his credit. Since his retirement in 1989, Dr. Gunstone has written or edited several books. He has given many invited lectures and has received distinguished awards in the United States (1973, 1999, 2005, and 2006), Britain (1962 and 1963), France (1990), Germany (1998), and Malaysia (2004). For many years he has been the editor of Lipid Technology, an activity that gives him continued contact with lipid scientists of many differing interests.
Albert J. Dijkstra, Ph.D., specialised in gas kinetics with Professor A. F. Trotman-Dickenson at University College of Wales, Aberystwyth, before defending his Ph.D. thesis at Leyden University in 1965. He joined ICI, first at the Petrochemical & Polymer Laboratory in Runcorn, Cheshire, then at the ICI Holland Rozenburg Works, The Netherlands, and finally at the ICI Europa headquarters in Everberg, Belgium. He became involved in edible oils and fats in 1978 when he joined the Vandemoortele Group in Izegem, Belgium, as its R&D director. Dr. Dijkstra is the inventor in a dozen patents and has published numerous articles on edible oil processing. He was the first nonAmerican to receive the American Oil Chemists’ Society (AOCS) Chang Award (1997) and the first to receive the EuroFedLipid Technology Award (2002). Although officially retired, he continues to be active in the field of edible oils and fats as author and scientific consultant.
John L. Harwood, Ph.D., is head of the School of Biosciences at Cardiff University (Wales, United Kingdom). He received his Ph.D. from the University of Birmingham in 1969, with studies on the metabolism of inositol lipids with Professor J. N. Hawthorne and, subsequently, learned about plant fatty acid synthesis at the University of California with Professor P. K. Stumpf. Following a tenure at the University of Leeds, he moved to Cardiff where he was promoted via reader to professor in 1984. He is
xi
CONTRIBUTORS
J. Alander
F. D. Gunstone
D. J. Murphy
AarhusKarlshamn, Sweden AB
The Scottish Crop Research Institute Invergowrie Dundee, Scotland
School of Applied Sciences University of Glamorgan Pontypridd, Wales
W. W. Christie The Scottish Crop Research Institute Invergowrie Dundee, Scotland
K. Dewettinck Laboratory of Food Technology and Engineering Ghent University Ghent, Belgium
A. J. Dijkstra Scientific Consultant St. Eutrope-de-Born, France
P. F. Dodds Department of Biology Imperial College at Wye Wye, Ashford, U.K.
S. Z. Erhan USDA, ARS, NCAUR Peoria, Illinois USA
M. Evans Llandough Hospital Llandough, Cardiff, Wales
P. J. Quinn J. L. Harwood Cardiff School of Biosciences Cardiff University Cardiff, Wales
D. P. Ramji J. A. Kenar USDA, ARS, NCAUR Peoria, Illinois USA
G. Knothe USDA, ARS, NCAUR Peoria, Illinois USA
School of Biosciences Cardiff University Cardiff, Wales
C. M. Scrimgeour The Scottish Crop Research Institute Invergowrie Dundee, Scotland
J. C. Segers M. S. F. Lie Ken Jie Hong Kong University Hong Kong, China
Jacques Segers Consultancy Nieuwerkerk aan den IJssel, The Netherlands
B. K. Sharma Jiann-Tsyh Lin USDA, ARS, WRRC Albany, California USA
I. Foubert Laboratory of Food Technology and Engineering Ghent University Ghent, Belgium
Department of Life Sciences King’s College London, UK
Department of Chemical Engineering Pennsylvania State University University Park, Pennsylvania USA
D. Van De Walle
T. A. McKeon USDA, ARS, WRRC Albany, California USA
xiii
Laboratory of Food Technology and Engineering Ghent University Ghent, Belgium
1 FATTY ACID AND LIPID STRUCTURE
C. M. Scrimgeour and J. L. Harwood
1.1
Fatty acid structure
1.1.1
Introduction and nomenclature of fatty acids
substituents. However, the terms cis and trans (abbreviated c and t) are widely used to describe double bond geometry, as with only two types of substituents there is no ambiguity that requires the systematic Z/E convention (Figure 1.1). However, a recent proposal for systematic naming for use in lipidomic and bioinformatic databases requires the use of Z or E (Fahy et al., 2005a, 2005b). Systematic names for fatty acids are cumbersome in general use and both shorthand alternatives and trivial names are widely used. Trivial names seldom convey any structural information, often reflecting a common or early source of the acid. The shorthand names use two numbers separated by a colon for the chain length and number of double bonds, respectively. Octadecenoic acid with 18 carbons and 1 double bond is, therefore, 18:1. The position of double bonds is indicated in a number of ways — explicitly, defining the position and configuration or locating double bonds relative to the methyl or carboxyl ends of the chain. In the biomedical literature, it is common to number the chain from the methyl end rather than the systematic numbering from the carboxyl end, to emphasise the biosynthetic relationship of different double bond patterns. Numbering from the methyl end is written n-x or ωx, where x is the double bond carbon nearest the methyl end. If there is more than one double bond, a cis configuration, methylene-interrupted pattern is implied. Although the n-x notation is recommended, both n-x and ωx are widely used in the current biomedical literature and wider nutritional contexts. The ∆ notation is used to make it explicit that the numbering is from the carboxyl end. Other substituents may also be included in the shorthand notation; for example 12-OH 18:1 9c for ricinoleic acid (12-hydroxy-9-cis-octadecenoic acid). The order and style used for shorthand names varies widely in the literature.
Fatty acids are aliphatic, usually straight chain, monocarboxylic acids. The broadest definition includes all chain lengths, but most natural fatty acids have even chain lengths between C4 and C22, with C18 the most common. Natural fatty acid structures reflect their common biosynthesis — the chain is built in two-carbon units and cis double bonds are inserted at specific positions relative to the carboxyl carbon. Over 1000 fatty acids are known with different chain lengths, positions, configurations and types of unsaturation, and a range of additional substituents along the aliphatic chain. However, only around 20 fatty acids occur widely in nature; of these, palmitic, oleic, and linoleic acids make up ~80% of commodity oils and fats. Figure 1.1 shows the basic structure of fatty acids and a number of the functional groups found in fatty acids. A list of many of the known structures, sources, and trivial names is available online (Adlof and Gunstone, 2003). Table 1.1 illustrates the naming of some commonly encountered fatty acids (additional examples are found in the following sections). Fatty acids are named systematically as carboxylic acid derivatives, numbering the chain from the carboxyl carbon (IUPAC-IUB, 1976). Systematic names for the series of saturated acids from C1 to C32 are given in Table 1.2. The -anoic ending of the saturated acid is changed to -enoic, -adienoic, -atrienoic, -atetraenoic, -apentaenoic, and -ahexaenoic to indicate the presence of one to six double bonds, respectively. Carbon–carbon double bond configuration is shown systematically by Z or E, which is assigned following priority rules for the
1
1.1
Fatty acid structure
or
COOH
CH3(CH2)16 COOH
Fatty acid with saturated alkyl chain R
R′
R
R
Methylene interrupted double bonds
R′
H
R
R
H
H
H
R′
trans (E)
cis (Z)
Conjugated double bonds
Double bond configuration OH
. Acetylene
Allene
Methyl branch
Hydroxyl
O O Cyclopropane
FIGURE 1.1 TABLE 1.1
Furan
Fatty acid structure and some functional groups found in fatty acids. Structure, systematic, trivial, and shorthand names of some common fatty acids
Structure CH3(CH2)l0COOH CH3(CH2)12COOH CH3(CH2)14COOH CH3(CH2)5CH=CH(CH2)7COOH CH3(CH2)16COOH CH3(CH2)7CH=CH(CH2)7COOH CH3(CH2)5CH=CH(CH2)9COOH CH3(CH2)5CH=CH(CH2)9COOH CH3(CH2)3(CH2CH=CH)2(CH2)7COOH CH3(CH2CH=CH)3(CH2)7COOH CH3(CH2)3(CH2CH=CH)3(CH2)4COOH CH3(CH2)18COOH CH3(CH2)3(CH2CH=CH)4(CH2)3COOH CH3(CH2CH=CH)5(CH2)3COOH CH3(CH2)20COOH CH3(CH2)7CH=CH(CH2)11COOH CH3(CH2CH=CH)6(CH2)2COOH CH3(CH2)22COOH CH3(CH2)7CH=CH(CH2)13COOH a
Epoxide
Cyclopropene
Trivial Name/ Abbreviation
Systematic Name Dodecanoic Tetradecanoic Hexadecanoic Z-9-hexadecenoic Octadecanoic Z-9-octadecenoic Z-11-octadecenoic E-11-octadecenoic Z,Z- 9,12-octadecadienoic Z,Z,Z- 9,12,15-octadecatrienoic Z,Z,Z- 6, 9,12-octadecatrienoic eicosanoica Z,Z,Z,Z- 5,8,11,14-eicosatetraenoica Z,Z,Z,Z,Z- 5,8,11,14,17eicosapentaenoica docosanoic Z-13-docosenoic Z,Z,Z,Z,Z,Z- 4,7,10,13,16,19docosahexaenoic tetracosanoic Z-15-tetracosenoic
Shorthand Name
lauric myristic palmitic palmitoleic stearic oleic cis-vaccenic vaccenic linoleic (LA) α-linolenic (ALA) γ-linolenic (GLA) arachidic arachidonic (ARA) EPA
12:0 14:0 16:0 16:1 9c 18:0 18:1 9c 18:1 11c 18:1 11t 18:2 9c,12c 18:3 9c,12c,15c 18:3 6c,9c,12c 20:0 20:4 5c,8c,11c,14c 20:5 5c,8c,11c,14c,17c
behenic erucic DHA
22:0 22:1 13c 22:6 4c,7c, 10c,13c,16c,19c 24:0 24:1 15c
lignoceric nervonic
n- or ω
7 9 7 7 6 3 6 6 3
9 3
9
Icosa- replaced eicosa- in systematic nomenclature in 1975, but the latter is still widely used in the current literature.
rationalised by considering their biosynthesis; a few basic processes build and extend the chain and insert double bonds, producing the common families of fatty acids. We do not consider the details of these biochemical processes here (see Section 10.1), but the reader should be aware of the result of the various enzyme processes that build and modify fatty acids. Saturated fatty acids are built from two carbon units, initially derived from acetate, added to the carboxyl end of the molecule, usually until there are 18 carbons in the chain. Double bonds are introduced by desaturase enzymes at specific positions relative to the carboxyl group. Elongases further
The following sections describe classes of naturally occurring fatty acids, emphasising acids that are nutritionally and biologically important, are components of commodity oils and fats, or are oleochemical precursors. The structures of many fatty acids are contained in the dictionary section of this book. Up to date information on fatty acid occurrence in seed oils can be found online (Aitzetmuller et al., 2003) and this is the source of much of the data in Section 1.1.2. Further information on fatty acid structure is available online at http://www.lipidlibrary.co.uk/ and http://www.cyberlipid.org/. The structures of naturally occurring fatty acids are most easily 2
Fatty Acid and Lipid Structure
TABLE 1.2
Systematic, trivial, and shorthand names and melting points of saturated fatty acids
Systematic Name
Trivial Name
Shorthand Name
methanoic ethanoic propanoic butanoic pentanoic hexanoic heptanoic octanoic nonanoic decanoic undecanoic dodecanoic tridecanoic tetradecanoic pentadecanoic hexadecanoic heptadecanoic octadecanoic nonadecanoic eicosanoic heneicosanoic docosanoic tricosanoic tetracosanoic pentacosanoic hexacosanoic heptacosanoic octacosanoic nonacosanoic triacontanoic hentriacontanoic dotriacontanoic
formic acetic propionic butyric valeric caproic enanthic caprylic pelargonic capric
1:0 2:0 3:0 4:0 5:0 6:0 7:0 8:0 9:0 10:0 11:0 12:0 13:0 14:0 15:0 16:0 17:0 18:0 19:0 20:0 21:0 22:0 23:0 24:0 25:0 26:0 27:0 28:0 29:0 30:0 31:0 32:0
a
lauric myristic palmitic margaric stearic arachidic behenic lignoceric cerotic carboceric montanic melissic lacceric
Melting Pointa (°C) 8.4 16.6 –20.8 –5.3 –34.5 –3.2 –7.5 16.5 12.5 31.6 29.3 44.8 41.8 54.4 52.5 62.9 61.3 70.1 69.4 76.1 75.2 80.0 79.6 84.2 83.5 87.8 87.6 90.9 90.4 93.6 93.2 96.0
Data from The Lipid Handbook, 2nd Edition (1994), Chapman & Hall, London. With permission.
extend the chain in two carbon units from the carboxyl end. These processes produce most of the fatty acids of commercial importance in commodity oils and fats, and which are considered to be of most value in food and nutrition. A great diversity of fatty acid structures is produced by variations on the basic process. The start, particularly, of the chain elongation process may be derived from acids other than acetate, resulting in odd or branched chains. Enzymes closely related to the desaturases may introduce functional groups other than double bonds, but usually with similar positional patterns. The result is a great variety of fatty acid structures, often restricted to a few related plant genera in which the altered enzymes have evolved. Additional structural variety is introduced by subsequent modification of fatty acids, e.g., oxidation at or near the carboxyl or methyl end. The Euphorbiacae and Compositae (Asteracae) are particularly adept at producing many and varied fatty acid structures. Fatty acids may be modified further, producing other groups of natural products, such as polyacetylenes, ecosanoids, and oxylipins. The following sections illustrate these various structures, but are not exhaustive.
References Adlof, R.O. and Gunstone, F.D. (2003) Common (non-systematic) names for fatty acids. http://www.aocs.org/member/ division/analytic/fanames.asp Aitzetmuller, K. et al. (2003) A new database for seed oil fatty acids — the database SOFA, Eur. J. Lipid Sci. Technol., 105, 92–103. http://www.bagkf.de/sofa/ Fahy, E. et al. (2005a) A comprehensive classification system for lipids. J. Lipid Res., 46, 839–861. Fahy, E. et al. (2005b) A comprehensive classification system for lipids. Eur. J. Lipid Sci. Technol., 107, 337–364. IUPAC-IUB (1976) Nomenclature of Lipids, World Wide Web version, prepared by G.P. Moss. http://www.chem. qmul.ac.uk/iupac/lipid/
1.1.2 1.1.2.1
Fatty acids Saturated acids
Saturated fatty acids form a homologous series of monocarboxylic acids (CnH2n+1COOH). Table 1.2 lists the saturated acids from C1 to C32 with their systematic and trivial names and melting points. Naturally occurring saturated acids are mainly of even chain length between C4 3
1.1
Fatty acid structure
The most common monoene is oleic acid (18:1 9c). Oleic acid (1) is found in most plant and animal lipids and is the major fatty acid in olive oil (70 to 75%) and several nut oils, e.g., macadamia, pistachio, pecan, almond, and hazelnut (filbert) contain 50 to over 70%. High oleic varieties of sunflower and safflower contain 75 to 80% oleic acid.
and C24. Fats rich in saturated acids are high melting and are characteristic of many tropical species. Odd chain acids are usually minor or trace components of plant and animal lipids, but some are more abundant in bacterial lipids. Short chain acids, particularly butyric (4:0), are found mainly in ruminant milk fats. Medium chain fatty acids (8:0, 10:0, 12:0, and 14:0) occur together in coconut and palm kernel oils, both tropical commodity oils. In both of these oils, lauric acid (12:0) predominates (45 to 55%), with 14:0 next most abundant. A number of Lauracae and Myristacae species contain in excess of 80% of 12:0 or 14:0, respectively. Cuphea, a temperate genus, has species rich in individual medium chain acids, e.g., C. pulcherrima >90% 8:0, C. koehneana >90% 10:0, and C. calophylla ~85% 12:0. These include some of the highest levels of single fatty acids in seed oils. Palmitic acid (16:0) is the most abundant and widespread natural saturated acid, present in plants, animals, and microorganisms. Levels of 20 to 30% are common in animal lipids, 10 to 40% in seed oils. Palm oil is a rich commodity oil source and contains over 40% of palmitic acid. Stearic acid (18:0) is also ubiquitous, usually at low levels, but is abundant in cocoa butter (~34%) and some animal fats, e.g., lard (5 to 24%) and beef tallow (6 to 40%). A few tropical plant species contain 50 to 60+% of 18:0, e.g., Shorea, Garcinia, Allanblackia, and Palaquium. Arachidic acid (20:0) is 20 to 30% of the seed oils of some tropical Sapindaceae species, but is usually a minor component of plant and animal lipids. Groundnut oil is the only commodity oil with significant amounts (~1%). Saturated acids are often most easily obtained by hydrogenation of more readily available unsaturated acids, e.g., docosanoic acid (22:0) could be obtained by hydrogenation of erucic acid (22:1). Chain shortening and chain extension reactions give access to odd or even chain lengths not readily found in natural sources. Saturated acids with 10 or more carbons are solids, and melting points increase with chain length (see Table 1.2). Melting points alternate between odd and even chain length, with odd chain lengths having a lower melting point than the preceding even chain acid. Polymorphism occurs, where one or more lower melting, metastable forms exist. 1.1.2.2
COOH Oleic acid (1)
Cis-vaccenic acid (18:1 11c, n-7) is common in bacterial lipids and a minor component of plant and animal lipids, co-occurring with the more abundant oleic acid. Cis-vaccenic is relatively abundant in sea buckthorn pulp, which is also rich in its n-7 biosynthetic precursor 16:1 9c. Petroselinic acid (18:1 6c) makes up over 50% of seed oil fatty acids of Umbelliferae species, such as carrot, parsley, and coriander, and is also found in the Araliaceae, Garryaceae, and Geraniaceae species. The biosynthesis of petroselinic acid involves a ∆4 desaturase acting on palmitic acid (16:0) followed by two carbon chain elongation (Cahoon et al., 1994). Palmitoleic acid (16:1 9c, n-7) is a ubiquitous minor component in animal lipids; somewhat more abundant in fish oils. A few plant oils are richer sources, e.g., nuts such as macadamia (20 to 30%) and the pulp of sea buckthorn (25 to 40%). C20 monoenes (11c and 13c) are present in brassica seed oils and the 9c and 11c isomers are found in fish oils. 20:1 5c is >60% of meadowfoam (Limnanthes alba) seed oil fatty acids. Erucic acid (22:1 13c, n-9) is up to 50% of Cruciferae oils, e.g. rape, mustard, crambe and over 70% in some Tropaeolum species. Nervonic acid (24:1 15c, n-9) occurs at 15 to 20% in Lunaria annua seed oil, along with higher levels of erucic acid. Some monoenes are used as or have potential use as oleochemicals. Erucic acid, as the amide, is used as an antislip agent for polythene film. 20:1 5c from meadowfoam oil can be used to prepare estolide lubricants and other novel materials. ω-Olefins, such as 10-undecenoic acid available from pyrolysis of castor oil, are useful oleochemical intermediates. Cis-monoenes with 18 or less carbons are liquids at room temperature or low-melting solids; higher homologues are low-melting solids. Trans-monoenes are higher melting, closer to the corresponding saturated acids. Double bond position also influences the melting point; both cis- and trans-C18 monoenes are higher melting when the double bond is at even positions than at odd positions; a pattern most distinct for double bonds between C4 and C14. The solid acids may exist as a number of polymorphs, with different melting points, resulting from subtly different packing in the crystal (Table 1.3).
Monoenoic acids
Straight-chain, cis-monoenoic acids with an even number of carbons are common constituents of many lipids and commodity oils. Trans- monoenes are rare components of natural oils and fats (see Section 1.2.6). The cis (Z) double bond is usually inserted by a ∆9-desaturase enzyme into preformed saturated acids; this may be followed by twocarbon chain extension at the carboxyl end. Starting with 16:0, this results in n-7 monoenes, while desaturation of 18:0 leads to the n-9 family. Monoenes may also result from desaturation at the ∆4 or ∆5 positions since oils with unsaturation at these positions occur in a few plant genera.
1.1.2.3
Methylene-interrupted polyunsaturated acids
Most unsaturated fatty acids with two or more double bonds show a characteristic methylene-interrupted pattern
4
Fatty Acid and Lipid Structure
fatty acids found in most plants, animals, and commodity oils and fats. Linoleic acid (18:2 n-6, 2) is present in most plant oils and is abundant (>50%) in corn, sunflower, and soybean oils, and exceeds 70% in safflower oil. γ-linolenic acid (18:3 n-6, 3) is usually a minor component of animal lipids, but is relatively abundant in some plant oils, e.g., evening primrose (~10%), borage (~20%), blackcurrant (~15%), and echium (~25%). Other n-6 acids, dihomo-γ-linolenic acid (20:3 n-6) and arachidonic acid (20:4 n-6) are present in animal tissues, but do not usually accumulate at significant levels in storage fats. These two C20 acids are the precursors of the PG1 and PG2 prostaglandin families, respectively. Some fungi, e.g., Mortierella species produce up to 50% arachidonic acid in storage lipids and are a commercial source of this acid (Ratledge, 2004).
TABLE 1.3 Trivial names and melting points of some monoene fatty acids Fatty Acid 16:1 9c (n-7) 16:1 9t (n-7) 18:1 9c (n-9) 18:1 9t (n-9) 18:1 6c 18:1 11c (n-7) 18:1 11t (n-7) 20:1 5c (n-16) 20:1 11c (n-9) 22:1 13c (n-9) 24:1 15c (n-9) a
Trivial Name palmitoleic palmitelaidic oleic elaidic petroselinic cis-vaccenic vaccenic gondoic erucic nervonic
Melting Pointa (°C) 0.5 32 16.2, (13.3) 45.5 31, (29) 15.5 44.1 27 25 33.5 (–52, –7, 2, 14) 45, 41
Data from The Lipid Handbook, 2nd Edition (1994), Chapman & Hall, London. With permission. Also references in Section 1.1.3. Polymorph melting points in parentheses.
COOH
of unsaturation, with one CH2 between cis double bonds. This pattern results from the operation of a few specific desaturases and chain-elongation enzymes. Plants generally insert double bonds at the ∆9, ∆12, and ∆15 positions in C18 fatty acids, giving n-9, n-6, and n-3 compounds, respectively. Animals can also insert double bonds at the ∆9 position, but not at ∆12 or ∆15; instead, further double bonds are introduced between the carboxyl group and the ∆9 position by ∆5 and ∆6 desaturase enzymes and the chain can then be extended in two carbon units at the carboxyl end of the molecule. The resulting n-6 and n-3 polyenes are shown in Figure 1.2. The step leading to DHA appears to be the result of a ∆4 desaturase, but is usually the net result of two elongations, a ∆6 desaturase and subsequent two-carbon chain shortening. Leonard et al. (2004) have reviewed the biosynthesis of long chain polyenes. Along with a few saturates (mainly 16:0 and 18:0, but also 10:0 to 14:0) and oleic acid, the n-6 and n-3 polyenes make up the n–9 18:1 9c oleic
Linoleic acid (2) COOH γ-linolenic acid (3)
α-linolenic acid (18:3 n-3, 4) is ubiquitous in plant leaf lipids and is present in several commodity seed oils: 8 to 10% in soybean and canola, >50% in linseed oil, and 65 to 75% of perilla oil. The seed oils of many Labiatae species are >50% α-linolenic acid. In plant leaves, chloroplast lipids contain up to 50% α-linolenic acid accompanied, in some species, by its C16 homologue, 16:3 7c, 10c,13c (Mongrand et al., 1998). Stearidonic acid (18:4 n-3, 5) is a minor component of animal lipids and fish oils and is found in some seed oils, e.g., blackcurrant (up to 5%) and echium (~7%). The n-3 long-chain, polyunsaturated fatty
n–6 ∆12 D
n–3
18:2 9c, 12c linoleic
18:3 9c, 12c, 15c ∆15 D
∆6 D
α–linolenic ∆6 D
18:3 6c, 9c, 12c
18:4 6c, 9c, 12c, 15c stearidonic
γ-linolenic E
E
20:3 8c, 11c, 14c
20:4 8c, 11c, 14c, 17c
∆5 D
∆5 D
20:4 5c, 8c, 11c, 14c arachidonic
20:5 5c, 8c, 11c, 14c, 17c EPA E, E, ∆6 D, –2C 22:6 4c, 7c, 10c, 13c, 16c, 19c DHA
FIGURE 1.2
Biosynthesis of n-6 and n-3 polyenes (D = desaturase, E = elongase, -2C = two-carbon chain shortening).
5
1.1
Fatty acid structure
∆5t double bond occur in Thalictrum species (see Section 1.2.6).
acids (LC-PUFA) 20:5 (EPA, 6) and 22:6 (DHA, 7) are important nutritionally and are mainly obtained from oily fish and fish oils where they are present at levels from 5 to 20%. EPA is the precursor of the PG3 prostaglandin series. Attempts are being made to produce EPA and DHA in plant lipids by the incorporation of appropriate enzymes because of the desire to have new sources of these important acids. Two types of microorganisms, a dinoflagellate Crypthecodinium cohnii and marine protist Schizochytrium species, are commercial single-cell oil sources of DHA (Ratledge, 2004).
COOH Pinolenic acid (8)
Sponges and some other marine invertebrates contain a wide range of fatty acids with 5c,9c double bonds, with chain lengths (both odd and even) ranging from C16 to C34, known as demospongic acids. Additional double bonds are usually n-7 or n-9 and methyl branching may also be present (Dembitsky et al., 2003).
COOH α-linolenic acid
1.1.2.5
(4)
Fatty acids with two or more conjugated double bonds are found in some plants and animals. Ruminant fats contain small amounts (~1%) of “conjugated linoleic acid” (CLA), resulting from bio-hydrogenation of linoleic and α-linolenic acids in the rumen, which gives mainly the 18:2 9c,11t isomer (rumenic acid, 9). The only reported long chain, conjugated diene from a plant is 18:2 10t,12t (~10%), which occurs in Chilopsis linearis along with the more abundant conjugated triene 18:3 9t,11t,13c. Estolides in stillingia oil (Sapium sebiferum) and Sebastiana species contain 10:2 2t,4c linked to a short chain allenic hydroxy acid (Spitzer et al., 1997; Figure 1.3). Conjugated dienes (and higher polyenes) are prepared chemically from methylene-interrupted fatty acids by alkaline isomerisation. Under controlled conditions, linoleic acid produces a mixture containing only the 9c11t and 10t12c CLA isomers (Sæbø, 2001). These isomers have potential uses in modifying body composition and as anticancer agents.
COOH Stearidonic acid (5)
COOH EPA (6) COOH DHA (7)
While the n-3 and n-6 polyenes are the most widely occurring and of prime biological and nutritional interest, a large number of other methylene-interrupted polyenes are known, produced by the same desaturation and elongation steps, but starting with fatty acids of different chain length and initial unsaturation. For example, animals deprived of linoleic or linolenic acids can use oleic acid as substrate for the ∆6 desaturase and subsequent steps, resulting in an n9 polyene series. The accumulation of 20:3 n-9 (Mead’s acid) in animals is considered to be a symptom of essential fatty acid (i.e., linoleic acid) deficiency. The presence of two or more cis double bonds results in a large lowering of the melting point compared to saturates of the same chain length and these polyenes are all liquid at room temperature. Linoleic acid melts at −5°C. 1.1.2.4
Conjugated acids
COOH Rumenic acid (9)
Conjugated trienes and tetraenes are found in several plant species. They are produced biologically from methylene-interrupted polyenes by a conjugase enzyme similar to ∆12 desaturase, which shifts an existing double bond into conjugation with a new double bond (Dyer et al., 2002). Table 1.4 gives the structure, common name, source, and melting point of the known conjugated trienes and tetraenes from plants. Conjugated trienes and tetraenes containing cis double bonds readily isomerise to the all trans form on heating or on exposure to light. Tung oil, containing >60% α-eleostearic acid (10), oxidises and
Bis- and polymethylene-interrupted acids
Fatty acids with bis- or polymethylene-interrupted double bonds, or a mixture of methylene and polymethylene separated unsaturation, occur in some plant species and marine organisms. Often these have a double bond inserted at the ∆5 position in addition to one or more double bonds in more usual positions. Bis-methylene-interrupted acids with a ∆5c double bond are common in gymnosperms (conifers), a typical example being pinolenic acid (18:3 5c,9c,12c) (8), occurring at levels of 25 to 30% in a number of pine and larch species (Wolff et al. 2001). Among angiosperms, Limnanthes alba (meadowfoam) seed oil contains the polymethylene-interrupted 22:2 5c,13c (~20%) and other ∆5 acids. Bis-methylene-interrupted acids with a
O O R
O
R′
O
. O
O O
FIGURE 1.3 18:2, 18:3.
6
O
Estolide from stilingia oil. R, R’ 16:0, 18:0, 18:1,
Fatty Acid and Lipid Structure
tion of commodity oils may result in low levels of trans isomers, particularly of polyenes. The undesirable nutritional properties of trans acids have led to alternative ways of producing hardened fats, such as interesterification or blending with fully saturated fats, and to the use of milder deodorisation procedures.
polymerises readily and is used as a drying agent in paints and varnishes. Along with CLA, there has been recent interest in the biological and nutritional properties of conjugated polyenes. COOH
α-eleostearic acid (10)
1.1.2.7 Trans acids
1.1.2.6
Fatty acids with acetylenic and allenic unsaturation are rare. The two types of unsaturation are isomeric and can be interconverted. In the allenic function, the double bonds are rigidly held at right angles and introduce a twist in the molecule, resulting in optical activity when they are asymmetrically substituted. The estolide oil in stillingia oil contains the allenic hydroxy acid 8-hydroxy-5,6-octadienoic acid (Spitzer et al., 1997; Figure 1.3). The (R,E) form of 2,4,5-tetradecatrienoic acid is an insect sex pheromone. Fatty acids with a 5,6 allene are found in the seed oils of a few Labiatae species: laballenic acid (18:2 5,6; 11) is up to 25% of Phlomis tuberosa and some Leucas species; lamenallenic acid (18:3 5,6,16t) is up to 10% in Lamium purpureum.
Monoenes and methylene-interrupted polyenes are predominantly cis. A few trans monoenes and dienes with typical double bond positions are known, e.g., 18:1 9t in Butyrospermum parkii (12.5%) and Dolichos lablab (15%), co-occurring with 18:1 9c, and 18:2 9t12t, (~15%) in Chilopsis linearis, associated with conjugated acids. Thalictrum (and some other Ranunculaceae species) contain several acids with a ∆5t bond, 16:1 5t (~2%), 18:1 5t (~20%), 18:2 5t,9c (~6%), and 18:3 5t,9c,12c (~45%). A similar pattern with ∆3t unsaturation is seen in some Aster species. 16:1 3t occurs widely in leaves associated with chloroplast lipids. Vaccenic acid, 18:1 11t, is the most abundant trans monoene in ruminant lipids, which contain a complex mixture of both cis and trans positional isomers resulting from biohydrogenation of linoleic and linolenic acids. Conjugated acids usually contain one or more trans double bonds (see Section 1.2.5). Trans isomers, mainly monoenes, are produced during catalytic partial hydrogenation, and can be present in substantial amounts in hardened fats, generally as a mixture of positional isomers. Heat treatment during deodorisaTABLE 1.4
Acetylenic and allenic acids
.
COOH
Laballenic acid (11)
Fatty acids containing an acetylenic group are tariric acid (18:1 6a, 12), up to 85% of some Picramnia species and crepenynic acid (18:2 9c,12a, 13) 50 to 75% of some
Common name, source, and melting point of some conjugated fatty acids
Fatty Acid
Common Name
10:2 2t,4c
Source
Melting Pointa (°C)
Sapium sebiferum (stillingia oil) (~5 to 10%)
18:2 18:2 18:2 18:2 18:2 18:2 18:2
8t,10t 9t,11t 9c,11c 9c,11t 10t,12t 10t,12c 10c,12c
CLA CLA CLA CLA CLA CLA CLA
18:3 18:3 18:3 18:3 18:3 18:3 18:3 18:3
8t,10t,12t 8t,10t,12c 8c,10t,12c 9t,11t,13t 9c,11t,13tb 9t,11t,13c 9c,11c,13t 9c,11t,13c
β-calendic calendic jacaric β-eleostearic α-eleostearic catalpic – punicic
Calendula officinalis (tr) Calendula officinalis (60%) Jacaranda mimosifolia (36%) Aleurites fordii (11%) Aleurites fordii (Tung oil), Parinarium spp., Momordica sp. (>60%) Catalpa spp. (~40%) – Punica granatum (~70%), Momordica balsamina (~60%)
78 40 44 72 49 32 62 45
α-parinaric β-parinaric
Parinarium laurinum (>50%), Impatiens spp. (>20%) –
86 96
18:4 9c,11t,13t,15cc 18:4 9t,11t,13t,15t a b
c
ruminant fats Chilopsis linearis (~10%)
56 54 43 20 56 23 39
Data from The Lipid Handbook, 2nd Edition (1994), Chapman & Hall, London. With permission. Occurs also as the 18-hydroxy (kamlolenic acid, Mallotus philippinensis (70%)) and 4-oxo (licanic acid, Licania rigida (80%)) derivatives. Occurs also as the 4-oxo derivative (Chrysobalanus icaco (18%)).
7
1.1
Fatty acid structure
are, however, abundant in the surface waxes of plant leaves. Fatty acids with a mid-chain methyl branch are characteristic of some bacteria. For example, 10-R-methyloctadecanoic acid (tuberculostearic acid) (14) is the major normal chain length fatty acid in Mycobacterium tuberculosis, the causative agent of tuberculosis, and is found in a number of other actinomycetes. The biosynthesis involves methylation of oleic acid, the methyl carbon being derived from the C-1 pool. C16 to C24 mid-chain methyl branched acids are also found in Mycobacterium species.
Crepis species. In C. alpina, the acetylenic bond is introduced by a ∆12-desaturase-like enzyme (Lee et al., 1998). Crepenynic acid is the starting point for the biosynthesis of a large number of fatty acid-derived acetylenic and polyacetylenic secondary natural products (e.g., matricaria ester). Stearolic acid (18:1 9a), the acetylenic analogue of oleic acid (from which it is easily prepared), is not often found in nature, other than as a minor component. However, it is more abundant in some Pyrularia species, P. edulis containing over 50%. Tariric acid (12)
H
COOH
COOH Tuberculostearic acid (14)
COOH Crepenynic acid (13)
1.1.2.8
Polymethyl fatty acids include those of isoprenoid origin, derived from partial metabolism of the phytyl chain from dietary chlorophyll. Phytanic (15) and pristanic acids (16) are the most common examples and are minor components of fish oils. A different pattern is seen in fatty acids from bird uropygial glands where the methyl groups are found on alternate, usually even, carbons, with two to four methyl groups present, e.g., (17) found in the preen gland wax of the graylag goose. Dimycocerosate esters, found in mycobacteria, contain a range of polyketidederived polymethyl fatty acids. These also have the methyls on alternate even carbons (Onwueme et al., 2005).
Branched chain acids
Straight chain fatty acids are the norm, but a wide variety of branched chain structures are known, mainly from bacterial and some animal sources. These acids are usually saturated or monoenes and the alkyl branch is a methyl group. Acids with a methyl group on the n-2 or n-3 carbon (iso and anteiso, respectively; Figure 1.4) are common in bacteria; their occurrence and distribution being strong taxonomic indicators. The biosynthesis of these acids involves the normal two-carbon chain extension, but instead of starting with a two-carbon acetate-derived unit, they start with 2-methyl propionic acid (from valine) or 2-methyl butanoic acid (from leucine), respectively. The resulting iso and anteiso acids, thus, have an even and odd total number of carbons, but α-oxidation may subsequently shorten the chain resulting in both odd and even carbon iso and anteiso acids. The shorthand nomenclature for these acids can be confusing, as the total number of carbons is shown, while the systematic name uses the number of carbons in the longest alkyl chain. For example, 15-methyl hexadecanoic acid is iso-17:0. Iso and anteiso acids found in animal fats, particularly ruminant fats, are mostly derived from bacteria in the diet or digestive system. However, some specific acids are of animal origin: 18-methyleicosanoic acid is the major thioester-bound fatty acid on the surface of wool and mammalian hair fibres, producing a continuous hydrophobic layer (Jones and Rivett, 1997). Iso and anteiso acids are rarely found in plant oils, apart from 14-methylhexadecanoic acid, which is found as a taxonomically useful minor component (~1%) in the Pinacae family. These acids
iso
FIGURE 1.4
CH3
COOH Phytanic acid (15)
COOH Pristanic acid (16)
COOH (17)
1.1.2.9
Cyclic fatty acids
Cyclic fatty acids, with a carbon ring along or at the end of the alkyl chain, occur naturally in some bacteria and plants. In addition, a variety of carbocyclic structures are formed from methylene-interrupted polyenes during heating, for example, during deep frying. The sources, synthesis, and biological properties of cyclic fatty acids have been reviewed by Sebedio and Grandgirard (1989). Fatty acids with a mid-chain cyclopropane group are found mainly in bacteria, with cis-9,10-methylenehexadecanoic (9,10 cpa 17:0); cis-9,10-methyleneoctadecanoic acid (9,10 cpa 19:0; dihydrosterculic acid); and cis-10,11-methyleneoctadecanoic acid (10,11 cpa 19:0; lactobacillic acid, 18) most common. They are found in diverse bacterial species, both aerobic and anaerobic, and in both Gram-negative
anteiso
Iso and anteiso branched-chain structures.
8
Fatty Acid and Lipid Structure
various chain lengths. The most abundant is usually the C16 hydnocarpic acid (22), but in Oncoba and Caloncoba species the C18 chaulmoogric acid (23) predominates (~70%). Gorlic acid (24), C18 with a ∆6 double bond, is usually 10 to 20% of these oils. Related homologues from C6 to C20 are often found at low levels. Arum maculatum seed oil contains ~20% of 13-phenyltridecanoic acid (25)
and Gram-positive species. Depending on culture conditions, they may be up to 35% of the membrane lipids. COOH Lactobacillic acid (18)
Biosynthesis of the cyclopropane ring involves addition of a methylene group, derived from S-adenosylmethione (the “C1 pool”), to an existing double bond, for example, lactobacillic acid is derived from cis-vaccenic acid, the most abundant monoene in many bacteria. The cyclopropane acids that have been found in protozoa, slime moulds, and invertebrates are most likely derived from bacteria in their diet. The distribution and biosynthesis of cyclopropane acids in bacteria has been reviewed by Grogan and Cronan (1997). Cyclopropane acids are often found at low levels (~1%) in plant oils containing cyclopropene acids (see below). Litchi chinensis, however, contains ~40% dihydrosterculic acid (9,10 cpa 19:0) along with small amounts of shorter chain homologues. Cyclopropene acids are found in plant oils of the Malvalaceae, Sterculiaceae, Bombacaea, Tiliaceae, and Sapicidacaea families. These are mainly sterculic acid (9,10methyleneoctadec-9-enoic acid; 9,10 cpe 19:1; 19) and malvalic acid (8,9-methyleneheptadec-8-enoic acid; 8,9 cpe 18:0; 20). Sterculic acid is usually the more abundant (>50% in Sterculia foetida oil) accompanied by smaller amounts of malvalic acid. 2-hydroxysterculic acid may also occur in these oils, probably an intermediate in the biosynthesis of malvalic acid by α-oxidation of sterculic acid. 9,10-methyleneoctadec9-en-17-ynoate (sterculynic acid) occurs in Sterculia alata (~8%). The biosynthesis of the cyclopropene ring is not fully understood, but is thought to proceed from oleic acid to the cyclopropane, produced by the same mechanism as in bacteria, followed by further desaturation. Long chain cyclopropane and cyclopropylidene fatty acids have been found in sponges, for example, (21) from the Amphimedon species (Nemoto et al., 1997). Their biosynthesis is unknown.
COOH Hydnocarpic acid (22)
COOH Chaulmoorgic acid (23) COOH Gorlic acid (24)
COOH
13-phenyltridecanoic acid (25)
Bacteria isolated from the extreme environment of hot springs produce fatty acids with a terminal cyclohexyl group. In strains of the acidophilic and thermophilic Bacillus acidocardarius, 11-cyclohexylundecanoic acid and 13-cyclohexyltridecanoic acid (26) account for 70 to over 90% of the fatty acids in the bacteria (Oshima and Ariga, 1975). One of the most unusual fatty acid structures reported to date is a terminal concatenated cyclobutane or ladderane, containing up to five cis-fused four membered rings (e.g. 27). These occur as glycerol and methyl esters in the unusually dense membranes of anammox bacteria (Damste et al., 2002).
COOH
COOH
Sterculic acid (19)
13-cyclohexyltridecanoic acid
COOH
(26)
Malvalic acid (20)
COOH
COOH (21)
(27)
Fatty acids with terminal rings are thought to be produced by incorporating a cyclic acid rather than acetate at the start of the chain, although the biosynthetic origin of the cyclic acid has not always been unequivocally established. Up to 80% of the seed oils of Hydnocarpus species and other genera of the Flacourtiaceae are terminal cyclopentenyl acids of
1.1.2.10
Fatty acids with oxygen-containing functional groups
Most fatty acids contain only double bonds, but a number of fatty acids and their metabolites have oxygen-containing functional groups, most commonly a hydroxyl or epoxide. 9
1.1
Fatty acid structure
Cutin, a cross-linked polyester constituent of plant cuticle, contains a number of C16 and C18 mono, di, and trihydroxy fatty acids. The C16 acids, derived from palmitic acid contain a terminal hydroxyl group and a mid-chain hydroxyl between C7 and C10. The predominant C18 acids, derived from oleic acid, are 18-hydroxyoleic, 9,10,18-trihydroxystearic and 9,10-epoxy-18-hydroxystearic acids. The primary hydroxyls are mainly ester linked, while the midchain hydroxyls are only partially esterified. Polyhydroxy acids are not usually found in seed oils; however, 9,10,18trihydroxy-12-cis-octadecenoic acid occurs as ~14% of Chamaepeuce afra oil. 2-hydroxy or α-hydroxy acids occur in sphingolipids, skin lipids, wool wax, bacterial cell wall lipids, and in a few seed oils. In some Thymus species 2-hydroxylinolenic occurs up to ~13%, along with linolenic acid and its C17 homologue (17:3 8c,11c,14c). The hydroxy acid is probably an intermediate in the biosynthesis of the C17 acid (see also hydroxysterculic acid, Section 1.2.9). Salvia nilotica oil contains α-hydroxy oleic, linoleic, and linolenic acids along with traces of C17 acids. 3-hydroxy or β-hydroxy fatty acids are found in bacterial lipids, both medium to normal chain-length saturates and in mycolic acids. Mycolic acids are very long chain compounds, typically C60 to C90, branched at C2, with unsaturation or cyclopropane groups along the long chain in addition to the 3-hydroxy group. Vernolic acid (12-epoxy-9-cis-octadecenoic acid, 30) is the most widespread epoxy acid in plant oils occurring in a number of Compositae, Malvaceae and Euphorbiaceae species. It makes up 60 to 80% of Vernonia oils and is over 90% of Bernardia pulchella oil. (+)-vernolic acid with the 12S,13R configuration is the most usual form, but the other optical isomer (–)-vernolic acid, has been isolated from some seed oils of the Malvaceae. In Crepis palaestina and Vernonia galamensis, the epoxide group is introduced by a ∆12-desaturase-like enzyme (Lee et al., 1998). However, in Euphorbia lagascae, the epoxygenase is a cytochrome P450 acting on linoleic acid (Cahoon et al., 2002).
Some of these are introduced by enzyme-mediated oxidation of methylene-interrupted fatty acids, e.g., by lipoxygenase or the initial stages of fatty acid catabolism, the latter giving hydroxyl groups near the carboxyl or methyl end of the chain. Autoxidation, occurring in the absence of enzymes also gives oxygen-containing products (hydroxy, keto, epoxy, etc.) with less positional specificity. In a few plant oils, hydroxy and epoxy groups are introduced in mid-chain positions by enzymes with the same positional specificity as desaturases. Castor oil, rich in ricinoleic acid (12-OH 18:1 9c), is the only commodity oil containing a fatty acid with a functional group other than double bonds. Oils containing vernolic acid (an epoxy acid) have been investigated as oleochemical precursors. Ricinoleic acid (R-12-hydroxy-9-cis-octadecenoic acid; 12-OH 18:1 9c; 28) is 80 to 90% of castor oil (from Ricinus communis). It occurs at similar levels in Hiptage species and is found in a number of other species. In Azima tetracantha, Argyreia cuneata, and Anogeissus latifolia, it occurs at levels of 10 to 25% along with lower amounts of the cyclopropene, which contain malvalic and sterculic acids (see Section 1.2.9). The sclerotia of the ergot fungus (Claviceps purpurea) contain up to 50% ricinoleic acid (see below). Isoricinoleic acid (R-9-hydroxy-12-cis-octadecenoic acid; 9-OH 18:1 12c) is over 70% of the Wrightia species. Lesquerolic acid (R-14-hydroxy-11-cis-eicosenoic acid; 29), the C20 homologue of ricinoleic acid, occurs in Lesquerella species (50 to 70%). It is produced from ricinoleic acid by an elongase specific for hydroxy acids (Moon et al., 2001). Related acids found in Lesquerella species include densipolic acid (12-OH 18:2 9c, 15c) and auricolic acid (14-OH 20:2 11c, 17c). Hydroxy (and keto) acids are also found with conjugated double bonds (see Table 1.4). These include kamlolenic acid (18-OH 18:3 9c, 11t, 13t) in Mallotus philippinensis (70%) and coriolic acid (13-OH 18:2 9t,11c) in Coriaria species (~70%). OH COOH Ricinoleic acid (28)
O COOH OH
Vernolic acid (30)
COOH Lesquerolic acid (29)
Other epoxy acids include coronoric acid (9,10-epoxy12-cis-octadecenoic acid), which occurs in a number of mainly Compositae species and is ~15% of Chrysanthemum coronarium oil. It is also found in sunflower and other oils after prolonged storage of the seeds. 9,10-epoxyoctadecanoic acid is found at low levels in Tragopogon porrifolius oil, and alchornoic acid (14,15-epoxy-11-cis-eicosanoic acid), the C20 homologue of vernolic acid, occurs in Alchornea cordifolia (~50%). A number of oxygen-containing fatty acid derivatives are produced from methylene-interrupted fatty acids following
A hydroxyl group along the acyl chain can be esterified to other fatty acids, forming an estolide. In castor oil, ricinoleic acid is present only in simple triacylglycerols, but in the ergot fungus Claviceps purpurea, ricinoleic acid is extensively esterified with both nonhydroxy acids and other molecules of ricinoleic acid in polyestolide groups (Batrakov and Tolkachev, 1997). Seed oils of Lesquerella and related species rich in lesquerolic acid contain estolides (Hayes et al., 1995).
10
Fatty Acid and Lipid Structure
R2
O
R1
COOH PGE1
lipoxygenases R2
R2
R1
R1
HO 20:3 n–6
OOH
HOO
OH
HO COOH PGF1α
jasmonates, leukotrienes (a)
HO
OH
O COOH
C OOH PGE2 HO
arachidonic acid 20:4 n–6 cyclooxygenase
OH
HO COOH
O
PGF2α
COOH HO
O OOH
OH
O
cyclic endoperoxide
C OOH PGE3 HO
prostaglandins, prostacyclins, thromboxane (b)
OH
20:5 n–3 HO
FIGURE 1.5 Formation of hydroperoxides and cyclic endoperoxides catalysed by (a) lipoxygenase and (b) cyclooxygenase enzymes.
COOH
HO
the formation of a hydroperoxide or cyclic endoperoxide catalysed by lipoxygenase and cyclooxygenase enzymes, respectively (Figure 1.5). Subsequent cyclisation and modification leads to physiologically active products, such as eicosanoids (in mammals) and jasmonates and divinyl ether fatty acids (in plants), and also to furanoid fatty acids. Although the enzyme-catalysed, oxygen addition is stereo and regiospecific, the range of starting acids and subsequent modifications results in many different products; only a few representative structures are shown here. Eicosanoids are biologically active C20 fatty acid metabolites acting as short-lived hormones or mediators, and include prostaglandins, thromboxanes, and leukotrienes. The PG1, PG2, and PG3 families of prostaglandins are derived from dihomo-g-linolenic acid (20:3 n-6), arachidonic acid (20:4 n-6), and eicosapentaenoic acid (EPA, 20:5 n-3), respectively, via their cyclic endoperoxides (Figure 1.6). Christie (2005) has recently reviewed ecosanoid structure and function. Among other functions, prostaglandins are involved in the inflammatory response, platelet aggregation, vasodilation, and smooth muscle function. Jasmonates are produced in plants following lipoxygenase catalysed conversion of 16:3 n-3, 18:2 n-6, and
FIGURE 1.6
PGF3α
OH
The PG1, PG2, and PG3 families of prostaglandins.
18:3 n-3 to conjugated hydroperoxides, which are then converted to a range of metabolites. The most widely studied is (–)- jasmonic acid (31), which is derived from 13-hydroperoxylinolenic acid. The cyclised product is chain shortened to C12 by β-oxidation. Jasmonates have hormone properties, regulating plant growth and development and are involved in leaf senescence and in defence against pathogens and in wound signalling (Farmer et al., 2003). O
COOH Jasmonic acid (31)
Divinyl ether synthase in plant leaves and roots converts hydroperoxides generated by lipoxygenase to divinyl ethers. In the potato, the 9-hydroperoxides of linoleic and linolenic acids lead to colneleic (32) and colnelenic acids
11
1.1
Fatty acid structure
e.g., (36) from the Parmelia species. A number of toxic ωfluoro fatty acids have been isolated from the South African plant Dichapetalum toxicarium. The origin of the fluorine is fluoroacetic acid, which can accumulate in the leaves of Dichapetalum species. The most abundant is 18-fluoro-oleic acid.
(33), respectively. Structurally similar compounds are derived from the 13-hydroperoxides in Ranunculus leaves, e.g., etherolenic acid (34) (Hamberg, 1998). These compounds are thought to be plant defence compounds protecting against pathogen attacks. O
COOH
Colneleic acid (32)
COOMe Br
O
COOH
(36)
Colnelenic acid (33) O
Sulfur-containing fatty acids have been reported at trace levels (,90% α-linolenic acid in chloroplastic 26
Fatty Acid and Lipid Structure
TABLE 1.11
Some bacterial sn-O-glycosyldiacylglycerols
Glyceride
Structure of Glycoside Moiety
Monoglucosyldiacylglycerol Diglucosyldiacylglycerol Diglucosyldiacylglycerol Dimannosyldiacylglycerol Galactofuranosyldiacylglycerol Galactosylglucosyldiacylglycerol Glucosylgalactosylglucosyldiacylglycerol
Occurrence
α-D-Glucopyranoside β-D-Glucopyranosyl(1→6)-O-β-D-glucopyranoside α-D-Glucopyranosyl(1→2)-O-β-D-glucopyranoside α-D-Mannopyranosyl(1→3)-O-D-mannopyranoside α-D-Galactofuranoside α-D-Galactopyranosyl(1→2)-O-α-D-glucopyranoside α-D-Glucopyranosyl(1→6)-O-α-Dgalactopyranosyl(1→2)-O-α-D-glucopyranoside
Pneumococcus, Mycoplasma Staphylococcus Mycoplasma, Streptococcus Microccus lysodeikticus Mycoplasma, Bacteroides Lactobacillus Lactobacillus
Adapted from Kates (1972).
A number of gluco- and galactoglycerolipids have been isolated in small quantities from animal tissue. Their structures are given in Murray and Narasimhan (1990). The majority of galactoglycerolipids contain a single galactose residue, which is linked in a β-glycosidic link between the C-1 of galactose and the C-3 of glycerol. The glucoglycerolipids constitute a large number of compounds with up to eight glucose residues linked α(1→6). Alkylacyl and diacyl lipids as well as sulfated forms have been reported (Slomiany et al., 1987).
TABLE 1.12 danica
Derivative MonochloroDichloroPentachloroHexachloro-
References
Position of substitution in 1,14dicosane-disulfate 13 11,15 3,3,11,13,16 3,3,11,13,15,16
Position of substitution in 1,15-tetracosanedisulfate 14 2,12,14,16,17 2,2,12,14,16,17
Chlorosulpholipids are found in some fungi (e.g., caldariomycin) and certain algae. Ochromonas danica contains particularly high amounts of chlorosulpholipids, where they represent almost half of the total membrane lipids. An entire family of compounds can be found with two sulfate ester functions and from one to six chlorines (Haines, 1973), as shown in Table 1.12. Some other structures (e.g., trichloro derivatives) have been reported but not characterized (Haines, 1973). An unusual glycolipid sulfate ester (36) has been reported in extremely halophilic bacteria by Kates and coworkers (cf. Kates, 1972), and a glycolipid sulfate (2,3,6,6′-tetraacetyl-α-α-tetrahalose-2′-sulfate) in Mycobacterium tuberculosis (Goren, 1970). For a review of sulfated glycolipids in Archaebacteria, see Kates (1990), and for mycobacterial sulphoglycolipids, see Goren (1990). Sulfated glycolipids and sterols are minor components of animal tissues (see Murray and Narasimhan, 1990). Lactosyl sulfatide and seminolipid are examples of the former (Ishizuka, 1997). A number of novel taurine-containing lipids have been isolated from the ciliated protozoan, Tetrahymena. These have the structures shown in (37).
Harwood, J.L. (1980) Plant acyl lipids: structure, distribution and analysis. In Biochemistry of Plants, vol. 4, P.K. Stumpf and E. E. Conn, Eds., New York: Academic Press, pp. 1–55. Harwood, J.L. and Okanenko, A.A. (2003) Sulphoquinovosyl diacylglycerol — the sulpholipid of higher plants. In Sulfur in Plants, Y.P. Abrol and A. Ahmad, Eds., Dordrecht: Kluwer, pp. 189–219. Kates, M. (1972) Techniques in Lipidology, 2nd ed., Amsterdam: Elsevier. Kates, M. (1990) Glyco-, phosphoglyco- and sulfoglycoglycerolipids from bacteria. In Handbook of Lipid Research, vol.6, M. Kates, Ed., New York: Plenum, pp. 1– 122. Murray, R.K and Narasimhan, R. (1990) Glycerolipids in animal tissues. In Handbook of Lipid Research, vol. 6, M. Kates, Ed., New York: Plenum, pp. 321–361. Slomiany, B.L., Murty, V.L.N., Liau, Y.N., and Slomiany, A. (1987) Animal glycoglycerolipids. Prog. Lipid Res., 26, 29– 51.
1.2.6
Structures of the sulfolipids of Ochromonas
Sulfur–containing lipids
In addition to cerebroside sulfates, other sulfur-containing sphingolipids (Section 1.2.4) and diacylsulpho-quinovosylglycerol and sulfated gluco- or galactogly-cerolipids (Section 1.2.5), various other sulfur-containing lipids have been reported. These include alkyl sulfates (Mayers et al., 1969) in microorganisms (35).
–O SO 3
– 3-Gal(1
6)Man(1
2)Glc(1
1') –OCH2 ROCH
OSO3–
ROCH2
H3C[CH2]7CH[CH2]12CH2OSO3–
(36) Glycolipid sulfate ester
(35) (1,14S)-Docosanediol-1,14-disulfate
27
1.2 Lipid structure
development. Small quantities of such diols are found in mammalian and fish liver, mammalian adipose tissue, egg yolk, corn seed, and yeast. Diesters of butane-1,3-diol and butane-1,4-diol are produced by various yeasts (Ratledge and Wilkinson, 1988) and mixed acyl and alk-1-enyl derivatives of these and other simple diols (e.g., ethylene glycol) have been reported (Batrakov et al., 1974). Extracellular acyl esters of arabinitol, xylitol, or mannitol have been reported (Stodola et al., 1967) and an acylated diol phospholipid has been isolated from the yeast Lipomyces starkeyi, after growth on propane-1,2-diol (Suzuki and Hasegawa, 1974).
CH3(CH2)24COO R1 CONHCH2CH2SO3H R2
R3
(37) Taurine-containing lipids Lipid Taurolipid A Taurolipid B Taurolipid C 7-Acyltaurolipid A
R1 OH OH OH CH3(CH2)14COO
R2 H OH OH H
R3 H H OH H
H2COCOR1
An alkaline-stable, taurine-containing lipid, lipo-taurine (2-(7,13-dihydroxy-2-trans-octadecenoylamino)ethanesulphonic acid) was also detected and probably plays a role as a metabolic intermediate. Other related compounds, such as 2-(octadecanoylamino)ethanesulphonic acid, were also identified. The isolation, characterization, and biochemistry of the taurolipids have been reviewed (Kaya, 1992).
CH2 H2COCOR2
(38) Diacylpropane-1,3-diol
References Batrakov, S.G. et al. (1974) Identification of threo-butane-2,3diol phospholipid from Actinomyces olivaceus., Biochim. Biophys. Acta, 337, 29– 40. Ratledge, C. and Wilkinson, S.G. (1988) An overview of microbial lipids. In Microbial Lipids, vol. 1, C. Ratledge and S.G. Wilkinson, Eds., Academic Press, London, pp. 3–22. Stodola, F.H. et al. (1967) Extracellular lipids of yeasts. Bacterial Rev , 31, 194–213. Suzuki, T. and Hasegawa, K. (1974) Diol lipids in the phospholipid fraction of Lipomyces starkeyi grown in the medium containing 1,2-propanediol. Agric. Biol. Chem., 38, 613–620.
References Goren, M.B. (1970) Sulfolipid I of Mycobacterium tuberculosis, strain H37Rv. II Structural studies. Biochim. Biophys. Acta, 210, 127– 138. Goren, M.B. (1990) Mycobacterial fatty acid esters of sugars and sulfosugars. In Handbook of Lipid Research, vol. 6, M. Kates, Ed., Plenum, New York, pp. 363– 461. Haines, T.H. (1973). Halogenated sulphatides. In The Biochemistry of Lipids, T.W. Goodwin, Ed., MTP International Reviews of Science, Biochemistry Series 1, vol. 4, Butterworths, London, pp. 271–286. Ishizuka, I (1997) Chemistry and functional distribution of sulfoglycolipids, Prog. Lipid Res, 36, 245–319. Kates, M. (1986) Techniques of Lipidology,2 nd ed. Elsevier, Amsterdam. Kates, M. (1990) Glyco-, phosphoglyco-, and sulfoglycoglycerolipids of bacteria. In Handbook of Lipid Research, vol. 6 (M. Kates, Ed.), Plenum, New York, pp. 1–122. Kaya, K. (1992) Chemistry and biochemistry of taurolipids. Prog. Lipid Res., 31, 87–108. Lederer, E. (1967). Glycolipids of mycobacteria and related microorganisms. Chem. Phys. Lipids, 1, 294–315. Mayers, G.M. et al. (1969). Microbial sulfolipids. III. Disulfate of (+) -1, 14-docosanediol in Ochromonas danica. Biochemistry, 8, 2981–2986. Murray, R.K. and Narasimhan, R. (1990) Glycerolipids of animal tissues. In Handbook of Lipid Research, vol. 6, M. Kates, Ed., Plenum, New York, pp. 321–361.
1.2.8
Other esters
A wide variety of other lipid esters have been reported. Wax esters are a typical example. Although the term “wax” should, strictly speaking, only be used for esters of long-chain fatty acids with long-chain primary alcohols, common usage, unfortunately, often equates “wax” with an entire mixture of lipids of which the true waxes are but a part. Ester waxes are found in animals and plants where they form part of the water-repellent surface coating (i.e., skin surface of animals and the leaf cuticle (cf. Section 1.2.11)). The general formula for a simple wax is shown in (39). The preen glands of birds, in addition, contain esters of normal alcohols with mono- or multibranched fatty acids (Odham, 1967). H3C[CH2]xCOO[CH2]yCH3
1.2.7
Diol lipids
(39) A simple wax
Complex waxes are compounds where either the fatty acid or the alcohol component or both has a complex structure. For example, the waxes of Mycobacterium spp. are diesters of phthiocerols (C33-C35 branched-chain diols)
Only recently has the presence of diol lipids, such as diacylpropane-1,3-diol (38), been confirmed for a wide variety of tissues. This is probably because techniques for the elucidation of their structures are a recent
28
Fatty Acid and Lipid Structure
with mycocerosic acids (C29-C32 branched-chain acids) (Asselineau, 1966; see Barry et al., 1998).
C22H45 CH2OCOCHCH[CH2]17CH CH[CH2]17CH3 OH O
H HO
OH
H
H
OH
RCOO
H
H O
H
OH
OH H H O OH O H OH O CCHCH[CH2]17CH CH[CH2]17CH3 C22H45
(40) Cholesterol ester FIGURE 1.15
Cord factor from Mycobacterium smegmatis.
CH2OCOR
Esters are found in most of the commonly occurring sterols (40), including those from plant tissues (Mudd, 1980). The fatty acid constituents usually reflect those of the acylglycerols from the same source. Likewise, esters are found in vitamin alcohols, such as vitamin A, the D vitamins, and vitamin E. Examples are shown in (41) to (43). Carotenoid esters have been reported from a few plant sources (Hitchcock and Nichols, 1971). Flower pigments, for example, are known to contain saturated fatty acid esters of carotenoid alcohols. These compounds are also found in green algae. Where dihydroxy alcohols are involved, then both substitutents are usually esterified, e.g., (44). In addition, acyl esters of terpenoid alcohols have been reported. For example, Dunphy and Allcock (1971) showed that 30 to 60% of the total monoterpenoid alcohol content of rose petals occurred as acyl esters with geranyl stearate (45) predominating. An important example of a carbohydrate ester is the so-called cord factor from Mycobacteria spp. This contains an ester of the disaccharide, trehalose, with two molecules of a complex acid, mycolic acid. The latter is a general term embracing a whole series of fatty acids containing 60 to 90 carbons. They are hydroxy fatty acids that differ in their degree of unsaturation and chain branching (see Section 1.1.2.10). In the example given in Figure 1.15, the mycolic acid is the 60-carbon compound found in Mycobacterium smegmatis. Various other esters have been reported in different bacteria. For example, propionibacteria contain diacyl myo-inositol mannosides in which the mannose is glycosidically linked to position 2 of myo-inositol. The 1 and 6 positions of inositol are esterified with fatty acids. Other bacteria and yeasts contain esters of glucose and certain other sugars (Lederer, 1967; Weete, 1980). Not only are simple carbohydrate esters found in Nature, but fatty acyl derivatives of amino acids have also been reported. These include serratamic acid (N-(D-3-hydroxydecanoyl)-L-serine), siolipin A (46), and siolipin B. The latter is the ornithine analogue of siolipin A.
(41) Acyl retinol
CH3 RCOO
(42) Acyl cholecaldiferol CH3 RCOO H3C
O CH3
H 3
(43) Acyl-α-tocopherol
OCO[CH2]14CH3
H3C[CH2]14COO
(44) Luteol dipalmitate
CH2OCO[CH2]16CH3
(45) Geranyl stearate
Animal skin-surface lipids have two types of diester waxes. In the first, a hydroxy acid has its hydroxyl group esterified to a normal fatty acid and its carboxyl group to a fatty alcohol. The second wax type consists of an alkane α, βdiol in which both hydroxyls are esterified with fatty acids (Nicolaides et al. 1970).
29
1.2 Lipid structure
Mycosides from mycobacteria are glycosides of methylated sugars with a long-chain, highly branched hydroxylated hydrocarbon terminated by a phenol group. The hydroxy groups of the long-chain glycol are esterified with mycocerosic and palmitic acids (Kates, 1972). Acylated steryl glycosides of plants usually contain Dglucose attached via a β-glycosidic linkage to the 3-position of sterols, such as sitosterol, stigmasterol, campesterol, and cholesterol (see Figure 1.17). The 6-position of the glucose is esterified with fatty acids, such as palmitic, stearic, oleic, linoleic, and linolenic (Mudd and Garcia, 1975). In some tissues, such as potato tubers, acylated sterol glycosides are major components (around 20% total) (Mudd, 1980). Although glucose is the major esterified sugar, galactose, mannose, xylose, and gentiobiose have been found in isolated cases and an α-glycosidic link also reported (Harwood, 1980). Nitrogen-fixing cyanobacteria produce heterocyst cells containing characteristic glycolipids (see Murata and Nishida, 1987). The chemical structures of major components of Anabaena cylindrica are shown in (47) and (48). Glycerol ester glycolipids (49) and (50) are also present (Murata and Nishida, 1987). The heterocyst lipids of Nodularia harveyana (a marine cyanobacterium) have been purified
[CH2]4NH2 R1CONHCHCOOR2 R1COOH = normal, branched and β-hydroxy-branched acids R2OH = long-chain polyalcohols
(46) Siolipin A
The peptidolipids that occur in mycobacteria and Nocardia spp. are N-acyl oligopeptides. They often occur as glycoside derivatives (Kates, 1972).
References Asselineau, J. (1966) The Bacterial Lipids, Hermann, Paris. Barry, C.E. et al. (1998) Mycolic acids: structure, biosynthesis and physiological functions. Prog. Lipid. Res. 37, 143–179. Dunphy, P.J. and Allcock, C. (1971) quoted by Hitchcock and Nichols (1971). Hitchcock, C. and Nichols, B.W. (1971) Plant Lipid Biochemistry, Academic Press, London. Kates, M. (1986) Techniques in Lipidology, 2 nd ed., Elsevier, Amsterdam. Lederer, E. (1967) Glycolipids of mycobacteria and related microorganisms. Chem. Phys. Lipids, 1, 294–315. Mudd, J.B. (1980) Sterol interconversions. In The Biochemistry of Plants, P.K. Stumpf and E.E. Conn, Eds., vol. 4. Academic Press, New York, pp. 509–534. Nicolaides, F., Fu, H.C., and Ansari, M.N.A. (1970) Diester waxes in surface lipids of animal skin. Lipids, 5, 299–307. Odham, G. (1967) Fatty acids in the feather waxes of some water birds. Fette Seifen Anstrichm. 69, 164–172. Weete, J.D. (1980) Lipid Biochemistry of Fungi and Other Organisms, Plenum, New York.
HO
O
CH3 OCHCH2COOCHCH2COOH
O HO
OH
OH
C7H15
C7H15
(a) CH2OH O HO
1.2.9
O
HO
CH3
Glycosides
CH2OH O
Several types of glycosides can be identified — those of hydroxy fatty acids, aromatic glycols and sterols. Glycolipids of some microorganisms, particularly yeasts, are extracellular products. There has been increasing interest in several of these compounds as biosurfactants (Solaiman, 2005). Emulsan, a polyanionic heteropolysaccharide having acyl chains esterified to the sugar moieties, is produced by Acinetobacter calcoaceticus. Often the lipids contain a mono- or disaccharide glycosidically linked to a hydroxy acid. Examples would be a rhamnolipid from Pseudomonas aeruginosa, a sophorolipid from Candida bombicola (see Figure 1.16) and cellobiolipids from Ustilago maydis (see Ratledge and Wilkinson, 1988). Emulsan is used as a degreasing agent and detergent. Rhamnolipid is used for oil recovery, in the printing industry and in a multitude of applications as a detergent in the agrochemical, food and cosmetic industry, and as a component of germicidal solutions for food and medical uses (Solaiman, 2005).
HO
OCH[CH2]15COOH
OH
CH3 O
OH OH (b)
FIGURE 1.16 (a) Rhamnolipid of Pseudomonas aeruginosa; (b) sophorolipid (yeasts).
RCOOCH2 O HO
O
OH OH
FIGURE 1.17
30
Acylated steryl glucoside (ASG).
Fatty Acid and Lipid Structure
recently and fully characterized as 1-(O-α-D-glucopyranosyl)-3R, 25R-hexacosanediol, 1-(O-α-D-glucopyranosyl)-3S, 25R-hexa-cosanediol, and 1-(O-α-D-glucopyranosyl)-3keto-25R-hexacosanol (Soriente et al., 1992).
Ratledge, C. and Wilkinson, S..G. Eds., (1988) Microbial Lipids, vol. 1, Academic Press, London. Solaiman, D.K.Y. (2005) Applications of microbial biosurfactants. INFORM 16, 408–410. Soriente, A., Sodano, G., Gambacorta, A., and Trincone, A. (1992) Structure of “heterocyst glycolipids” from the marine cyanobacterium, Nodularia harveyana. Tetrahedron, 48, 5375–5384.
Glycosidic glycolipids HOCH2
HOCH2 O HO
OH
O
HO OH
and
O
OH
(90%)
1.2.10
O CH2
OH
OH
OH
The occurrence of lipid-soluble arsenic compounds in marine organisms was first reported over 30 years ago (Lunde 1973). At the present time, more than 100 naturally occurring arsenolipids have been reported, as reviewed recently (Dembitsky and Levitsky, 2004). They are found in a wide variety of organisms ranging from lichens, fungi, and plants, to freshwater and marine algae, and invertebrates, fishes, and animals. The primary analytical technique used for separation, identification, and quantification of arsenolipids is LC coupled to various types of mass spectrometry and a full discussion is given in Dembitsky and Levitsky (2004). Many species of bacteria seem to be active in metabolising arsenic compounds and, in particular, have been shown to be capable of producing methylated derivatives, such as trimethylarsine. These include soil organisms, such as Flavobacterium or Pseudomonas spp. (Shariatpanahi et al., 1981) as well as microorganisms from the deep sea (1000 to 3500 m) where they seem important for the metabolism of arsenic compounds eventually to simpler metabolites and inorganic arsenic (Hanaoka et al., 1997). Other notable microorganisms with high contents of arsenic, including arsenobetaine and arsenocholine compounds, are various halophytes (Oremland and Stolz, 2003). Some examples of arsenolipids found in freshwater and marine algae are given in Figure 1.18. Marine brown algae have been well studied (Dembitsky and Levitsky, 2004), but green and red algae have also been noted to actively metabolise arsenic compounds. From freshwater environments, Chlorella spp. and Chlamydomonas reinhardtii have been studied. Most algal species seem well capable of methylating arsenic as part of the conversion to compounds isolated in the polar lipid fraction. A variety of arsenolipids were identified and these could accumulate in the range of 1.5 to 33.8 µg/g dry weight. Arsenolipids have been identified in marine invertebrates (Benson, 1989), including jellyfish, crustacea, worms and molluscs. Freshwater molluscs, crustacea, and earthworms have also been studied. The major arsenic compound in marine fish and animals is arsenobetaine, first detected in lobsters (Edmonds et al., 1977). Other animals studied include the sperm whale. In plants, arsenic-containing lipids have been identified in species ranging from higher plants through ferns and lichens.
(10%)
(47) 3,25-Dihydroxyhexacosanyl α-D-glycopyranoside HOCH2
HOCH2
O HO
OH
O
HO O
OH
and
OH
O CH2 OH
(90%)
HO
HO OH
(10%)
(48) 3,25,27-Trihydroxyoctacosanyl α-D-glycopyranoside
Glycosyl ester glycolipids HOCH2 O HO
O
OH
O C OH
OH
(49) α-D-Glucopyranosyl 25-hydroxyhexacosanate HOCH2 O HO
OH
O O C
OH
Arsenolipids
HO OH
(50) α-D-Glucopyranosyl 25,27-dihydroxyoctacosanate
References Harwood, J.L. (1980) Plant acyl lipids: structure, distribution and analysis. In Biochemistry of Plants, Vol. 4, P.K. Stumpf and E.E. Conn, Eds., Academic Press, New York, pp. 1–55. Kates, M. (1986) Techniques in Lipidology, 2nd ed. Elsevier, Amsterdam. Mudd, J.B. (1980) Sterol interconversions. In The Biochemistry of Plants, P.K. Stumpf and E.E. Conn, Ed., vol. 4, Academic Press, New York, pp. 509–534. Mudd, J.B. and Garcia, R.E. (1975) Biosynthesis of glycolipids. In Recent Advances in the Chemistry and Biochemistry of Plant Lipids, T. Galliard and E. I. Mercer, Eds., Academic Press, New York, pp. 161–201. Murata, N. and Nishida, I. (1987) Lipids of blue-green algae (cyanobacteria). In The Biochemistry of Plants, vol. 9, P.K. Stumpf and E. E. Conn, Eds., Academic Press, New York, pp. 315–347.
31
1.2 Lipid structure
Shariatpanahi, M. et al. (1981) Biotransformation of the pesticide sodium arsenate. J. Environ. Sci. Health, 16, 35–47.
O H3C–As–CH2 O CH3
OR
18.R = Me OH
19.R = OH
OH OH 20.R =
OH
P O O O
21.R =
OH
22.R =
OH O
26. R = OH
OSO3H OH
OH
SO3H OH
P O O
OCOR1 OCOR
23.R =
SO3– NH3+
24.R =
27. R = OH
OH P O O O
+
O O O
OH OH
OH OH
O 29. R =
OH
28. R =
SO3–
NH3
O
OH OH
OH O
O
O
30. R =
N H
31. R =
Waxes
Plant wax is the general term used to describe the lipid components of the cuticle that covers the outer surface of aerial plant tissues or is associated with the suberin matrix of underground or wound tissues. The components of plant cuticular waxes have been reviewed by Kolattukudy (1980, 1987) and Walton (1990). Major components include hydrocarbons, very long-chain fatty acids, alcohols and monoesters (Table 1.13 and Table 1.14). Surface waxes are exposed to the environment and, therefore, are chemically rather stable. Thus, there is an absence of functional groups, which might be susceptible to attack by atmospheric agents. Furthermore, the very long carbon chains of most wax components reduces their volatility. In addition, many of compounds present in surface waxes are rather stable metabolically and are not readily susceptible to microbial degradation (Kolattukudy, 1976). Certain general structural features of natural waxes have been described by Kolattukudy (1976), and these are summarized in Table 1.13. However, it must also be stressed that the structure and composition of surface waxes vary considerably from organism to organism. Thus, with regard to Table 1.13, the longer aliphatic chains are more abundant in plant waxes than in animal surface waxes, whereas bird waxes may contain appreciable amounts of chains of less than 16 carbons. With regard to branching, methyl branches are the most common, but, in birds, ethyl and propyl branches are found. Although polyunsaturated carbon chains are nearly always absent from surface waxes, in insects substantial proportions of di-unsaturated hydrocarbons have been found. In this case, autoxidation may be reduced by the simultaneous presence of cuticular phenolics. So far as the general composition of surface waxes is concerned, very long chain hydrocarbons are common in insects and plants, but rare in animals. Higher plant waxes contain the most complex
OH
OH 25. R =
1.2.11
OH
OH OH
FIGURE 1.18 Arsenolipids isolated from freshwater and marine algae. (From Dembitsky, V.M. and Levitsky D.O. (2004) Arsenolipids. Prog. Lipid Res., 43, 403–448. With permission.)
Some plants are hyper-accumulators and may take up and accumulate more than 1 mmol As g-1 dry weight (Brooks et al., 1977). Details of all these reports of different arsenolipids in various species and their possible metabolism are given in Dembtisky and Levitsky (2004).
References Benson, A.A. (1989) In Marine Biogenic Lipids, Fats and Oils, R.G. Actaman, Ed., vol. 1, Boca Raton, FL: CRC Press, pp. 243– 250. Brooks R.R. et al. (1977) Detection of nickeliferous rocks by analysis of herbarium specimens of indicator plants. J. Geochem. Explor. 7, 49–57. Dembitsky, V.M. and Levitsky D.O. (2004) Arsenolipids. Prog. Lipid Res., 43, 403–448. Edmonds, J.S. et al. (1977) Isolation, crystal-structure and synthesis of arsenobetaine, arsenical constituent of western rock lobster Panulirus longipes cygnus George. Tetrahedron Lett. 18, 1543–1546. Hanoaka, K. et al. (1997) Arsenobetaine-decomposing ability of marine microorganisms occurring in particles collected at depths of 1100 and 3500 metres. Appl. Organomet. Chem. 11, 265–271. Lunde, G. (1973) Synthesis of fat and water-soluble arseno organic compounds in marine and limnetic algae. Acta Chem. Scand. 27, 1586–1594. Oremland, R.S. and Stolz, J.F. (2003) The ecology of arsenic. Science, 300, 939–944.
TABLE 1.13 Chain length Branching Unsaturation
Functional types
32
General structural features of natural waxes Very long chains (up to C62) are common Branched carbon chains common, with methyl branches frequent Polyunsaturated chains nearly always absent; double bonds, when present, at different position from those of internal lipids Saturated hydrocarbons, olefins, wax esters, aldehydes, ketones, primary and secondary alcohols, and terpenoids can be present; the bulk of the surface lipid is distinctly different from the major internal lipids of the same organism
Fatty Acid and Lipid Structure
TABLE 1.14 Major classes of plant aliphatic wax components Wax Class n-Alkanes Secondary alcohols
Chain-Length Range in Plants C21-C35 C21-C35
Ketones Fatty alcohols
C21-C35 C22-C34
Fatty acids
C16-C34
Aldehydes
C21-C35
Wax esters
C32-C64
Major Arabidopsis Components C29,C31,C27 C29,C31,C27
Notes for Plants in General Common; usually C29, C31 About as common as ketones Not as common as alkanes Common, even chains predominate Very common; even-chain saturated usually Usually minor; not as common as alcohols Common
C29 C28,C30, C26
% in Arabidopsis Stems 38 10 30 12
C30, C28
3
C30, C28
6
-
1
Source: See Kolattukudy, P.E. (1980) In Biochemistry of Plants, vol. 4 (P.K. Stumpf and E. E. Conn, Eds.), Academic Press, New York, pp. 571–645; Kunst, L. and Samuels, A.L. (2003) Prog. Lipid. Res. 42, 51–80.
mixture of components, while insects and birds have the simplest. The wax associated with suberin has also been examined, and very long-chain fatty acids, alcohols and terpenes have all been found. These are all typical components of cuticular wax, but certain differences have been noted. The hydrocarbons in suberin have a broader chain-length distribution with a predominance of shorter carbon chains and more even-numbered carbon chains than cuticular wax. Suberin-associated wax also contains a high proportion of free fatty acids. Free and esterified alkan-2-ols are also present (Kolattukudy, 1980). Further details of the wax components of other organisms will be found in Section 2.5.
Kunst, L. and Samuels, A.L. (2003) Biosynthesis and secretion of plant cuticular wax. Prog. Lipid. Res. 42, 51–80. Walton, T.J. (1990) Waxes, cutin and suberin. In Methods in Plant Biochemistry, vol. 4, J.L. Harwood and J.R. Boyer, Eds., Academic Press, London, pp. 105–158.
1.2.12
In many organisms the outer envelope or covering consists of polymers of carbohydrate or amino acids. In plants, however, the covering (cuticle) consists of a hydroxy fatty acid polymer called cutin. The underground parts and healed wound surfaces of plants are covered with an analogous material, suberin. Both cutin and suberin are embedded in or associated with a complex mixture of lipids, which is termed wax (see Section 1.2.11). The structure and composition of cutin and suberin are reviewed by Kolattukudy (1980, 1987) and by Walton (1990). Cutin contains C16 and C18 families of acids. The former is predominate in rapidly growing plants, while both are present in the thicker cuticle of slower- growing plants. The C16 family is based on palmitic acid, while the C18 family is based on oleic acid (Table 1.15) (see also Kolattukudy, 1980). In the cutin structure, a polyester intramolecular structure exists where crosslinking is mainly influenced by the availability of secondary hydroxyl groups. Thus, cutins that contain large amounts of epoxy, oxo, and ω-hydroxy
References Hamilton, R.J. Ed., (1996) Waxes: chemistry, molecular biology and functions. The Oily Press, Dundee, Scotland. Kolattukudy, P.E. Ed., (1976) Chemistry and Biochemistry of Natural Waxes, Elsevier, New York. Kolattukudy, P.E. (1980) Cutin, suberin and waxes. In Biochemistry of Plants, vol. 4, P.K. Stumpf and E.E. Conn, Eds., Academic Press, New York, pp. 571– 645. Kolattukudy, P.E. (1987) Lipid-derived defensive polymers and waxes and their role in plant-microbial interaction. In Biochemistry of Plants, vol. 9, P.K. Stumpf and E.E. Conn, Eds., Academic Press, New York, pp. 291–314. TABLE 1.15
Cutin and suberin
The major components of cutin, the cutin acids
C16 Family H3C[CH2]14COOH HOCH2[CH2]14COOH HOCH2[CH2]XCHOH[CH2]yCOOH (x + y = 13; y = 5-8)
C18 Familya H3C[CH2]7CH = CH[CH2]7COOH HOCH2[CH2]7CH = CH[CH2]7COOH
HOCH2[CH2]7CH–CH[CH2]7COOH O HOCH2[CH2]7CHOHCHOH[CH2]7COOH
∆12 unsaturated analogues also occur. Note: For further details, see Harwood (1980) and Kolattukudy (1977). a
33
1.2 Lipid structure
Typical cutin monomers and their ability to form polyesters
TABLE 1.16
Cutin Monomers (Acids) Capable of Cross-Linking a Polyester Polymer a
HO[CH2]6CHOH[CH2]8COOH HOOC[CH2]5CHOH[CH2]8COOH OHC[CH2]6CHOH[CH2]7COOHa HO[CH2]8CHOHCHOH[CH2]7COOHa HO[CH2]5CHOHCHOHCH2CHOHCHOH[CH2]7COOH
10,16-dihydroxyhexadecanoic acid (and other positional isomers) 7-hydroxyhexadecanedioic acid (and other positional isomers) 9-hydroxy-16-oxohexadecanoic acid (and other positional isomers) 9,10,18-trihydroxyoctadecanoic acid (and its ∆12 analogue) 9,10,12,13,18-pentahydroxyoctadecanoic acid
Cutin Monomers (Acids) Capable of Forming Only a Linear Polyester Monobasic α, ω-Dibasic ω-Hydroxymonobasic namely HO[CH2]6CO[CH2]8COOHa
16:0, 18:0, 18:1(9), 18:2 (9,12) 16-hydroxy-10-oxohexadecanoic acid (and other positional isomers)
O HO[CH2]8CH – CH[CH2]7COOHa
9,10-epoxy-18-hydroxyoctadecanoic acid (and its ∆ 12 analogue)
a Major components of cutin. Source: Adapted from Deas and Holloway (1977).
TABLE 1.17
Polymeric form of dihydroxyhexadecanoic acid and related C18 acids in four plant cutins Total Monomers (%)
Polymeric Form Tomato Cutin-O[CH2]15COO-Cutin Cutin-OOC[CH2]5CHOH[CH2]8COO-Cutin
Cutin
O
Cutin-OOC[CH2]5CH[CH2]8COO-Cutin HO[CH2]6CHOH[CH2]8COO-Cutin Cutin-O[CH2]6CHOH[CH2]8COO-Cutin
Cutin-O[CH2]6CH[CH2]8COO-Cutin
O
Lemon
7 6 5
18 5
14 2 1
2 48
5 50
3 38
2 25
36
24
30
5
3 2
3 1
4 2
1 51
Cutin
HO[CH2]6CH[CH2]8COO-Cutin
O
Rosehip
Cutin
O
Cutin
Blackcurrant leaf
5 1 4
[CH2]6CO[CH2]8COO
Cutin
Source: Adapted from Deas and Holloway (1977).
acid is the usual dicarboxylic acid, and 18-hydroxyoleic, the major hydroxy fatty acid. The proportion of very longchain fatty acids (>C20) is usually much greater in the ωhydroxy acid fraction than in the dicarboxylic acid fractions. Among the α,ω-diols, fatty alcohols and fatty acids, which are often found as significant components of suberin, long chains are common. Kolattukudy has suggested some basic rules for the classification of hydroxy acid phytopolymers as cutin or suberin (Table 1.18). However, these rules must be regarded only as a guide, since the examination of individual plant species has provided exceptions. Indeed, it should be noted that, apart from species or varietal differences, environmental conditions may cause large changes in surface lipids. Thus, light, temperature, and age have all been found to affect leaf cuticular components (Harwood. 1980).
monomers must be predominantly linear (Table 1.16) (Deas and Holloway, 1977). Esterification appears to occur chiefly through the primary hydroxy groups of the monomers. A significant portion (up to 40%) of the monomers is also cross-linked through secondary hydroxyl groups (Table 1.16 and Table 1.17). Considerable diversity is evident when cutins from different sources are compared in detail. It is of interest, though, that the cutin composition of delicate tissues, such as spinach leaves, is essentially similar to that of much more substantial membranes. The wax part of the epidermal layer (see Section 1.2.11) is also usually similar between species, but with differences in detail (Harwood. 1980; Kolattukudy, 1980). The major aliphatic components of suberins are ωhydroxy acids and dicarboxylic acids. Octadec-9-enedioic
34
Fatty Acid and Lipid Structure
TABLE 1.18 suberin
Biochemistry of Plants, vol. 9, P.K. Stumpf and E.E. Conn, Eds., Academic Press, New York, pp. 291–314. Walton, T.J. (1990) Waxes, cutin and suberin. In Methods in Plant Biochemistry, vol. 4, J.L. Harwood and J.R. Boyer, Eds., Academic Press, London, pp. 105–158.
Compositional differences between cutin and
Monomer Dicarboxylic acids In-chain substituted acids Phenolics Very long-chain (C20-C26) acids Very long-chain alcohols
Cutin Minor Major Low Rare and minor Rare and minor
Suberin Major Minor High Common and substantial Common and substantial
1.2.13
A full discussion of the varied lipid structures found in bacterial cell walls is beyond the scope of this book, but the reader will find detailed accounts in Rogers et al. (1980) and Goldfine (1982). More recent detailed updates on the biochemistry and distribution of the cell-wall lipids of mycobacteria and other actinomycetes and of Gram-negative bacteria will be found in Brennan (1988) and Wilkinson (I988, 1996), respectively. The structure, biosynthesis, and physiological functions of mycolic acids are reviewed in Barry et al. (1998). The unique lipid-rich cell walls of mycobacteria contribute to their resilience and contain many compounds known to increase pathogenicity. The dimycocerosate esters (also called phthiocerol diesters) are particularly important and have been reviewed recently (Onwueme et al., 2005). Gram-negative bacteria have a cell envelope containing two membranes, with the outer membrane having lipopolysaccharide in its outer leaflet. Lipopolysaccharide is complex and consists of four parts. On the outside is the O-antigen, which is a polysaccharide of variable structure. This is attached to a core polysaccharide, which is in two parts, an outer core and a backbone. The backbone is connected to a glycolipid, called lipid A, through a short “link” usually composed of 3-deoxy-D-manno-octulosonic acid (KDO). These structures are shown in Figure 1.19. The role of lipid A as bacterial endotoxin and further details of different structures are given in Raetz and Whitfield (2002).
Source: Kolattukudy, P.E. (1975) In Recent Advances in the Chemistry and Biochemistry of Plant Lipids (T. Galliard and E.I. Mercer, Eds.), Academic Press, New York, pp. 203–246. With permission.
References Deas, A.H.B. and Holloway. P.J. (1977) The intracellular structure of some plant cutins. In Lipids and Lipid Polymers in Higher Plants, M. Tevini and H.K. Lichtenthaler, Eds., Springer-Verlag, Berlin. pp. 293–299. Harwood, J.L. (1980) Plant acyl lipids: structure, distribution and analysis. In Biochemistry of Plants. vol. 4, P.K. Stumpf and E.E. Conn, Eds., Academic Press, New York. pp. 1–55. Kolattukudy, P.E. (1975) Biochemistry of cutin, suberin and waxes on the lipid barriers of plants. In Recent Advances in the Chemistry and Biochemistry of Plant Lipids, T. Galliard and E.I. Mercer, Eds., Academic Press, New York, pp. 203–246. Kolattukudy. P.E. (1977) Biosynthesis and degradation of lipid polymers. In Lipids and Lipid Polymers in Higher Plants, M. Tevini and H.K. Lichtenthaler, Eds., Springer-Verlag, Berlin. pp. 271–292. Kolattukudy, P.E. (1980) Cutin, suberin and waxes. In Biochemistry of Plants, vol. 4, P.K. Stumpf and E.E. Conn, Eds., Academic Press, New York, pp. 571–645. Kolattukudy, P.E. (1987) Lipid-derived defensive polymers and waxes and their role in plant-microbe interaction. In GlcN
Gal
Glc
Gal
Glc
Abe Man
Hep
Hep EtN
Gal n
Rha
Bacterial-wall lipids
P
P
Outer core
O-antigen
(KDO)3 P
lipid A
EtN
Backbone R-core
(a) Lipopolysaccharide M P
CH2O M0 O
P
O
O
CH2 O
EtN
P
KDO O KDO KDO
NH
M O
O
M0
M
O P
NH M0
(b) Lipid A–KDO link region
FIGURE 1.19 Generalized structures of lipopolysaccharide and lipid A. Abbreviations: Abe, abequose; Man, mannose, Rha, rhamnose; Gal, galactose; Glc, glucose; Hep, heptose; KDO, 3-deoxy-D-manno-octulsonic acid; P , phosphate; EtN, ethanolamine; M, myristate; M0, β-hydroxymyristate. (From Harwood and Russell (1984). With permission.).
35
1.2 Lipid structure
O–
OCH2
H
CHOCOR2 D-alanine O CH O or CH2OPO K OCH2 D-glucose O– n n = 28–35
CH2CHCH2OPO O
CH2OCOR1
O
D-glucose or D-alanine
(a) Teichoic acid (general formula)
(b) Lipoteichoic acid (from Streptococcus lactis) CH2OCOR1 CHOCOR2
D-mannose
OCH2
O
n = 52–75 Succinic acid n (c) Lipomannan (from Micrococcus lysodeikticus)
FIGURE 1.20 Structures of some anionic polymers in bacteria. Abbreviations: K, kojibiose (6,O,β-D-glucosyl-D-glucose); R1 and R2, fatty acids. (From Harwood and Russell (1984) With permission.)
The cell walls and membranes of most Gram-positive bacteria contain a series of highly anionic polymers. Quantitatively, one of the most important of these is teichoic acid, which can be covalently linked to a glycolipid to give a lipoteichoic acid (Figure 1.20). An alternative type of anionic polymer. which is found in Gram-positive bacteria such as Micrococcus lysodeikticus, is succinylated lipomannan (Figure 1.20). Like teichoic acid, the lipomannan is embedded in the membrane by linkage to a diacylglycerol moiety.
S.G. Wilkinson, Eds., Academic Press, London. pp. 203–285. Goldfine, H. (1982) Lipids of prokaryotes — structure and distribution. Curr. Top. Membr. Transp, 17, 1–43. Harwood, J.L. and Russell, N.J. (1984) Lipids in Plants and Microbes, Allen and Unwin, Hemel Hempstead, U.K. Onwueme, K.C. et al. (2005) The dimycocerosate ester polyketide virulence factors of mycobacteria. Prog. Lipid. Res. 44, 259–302. Raetz, C.R.H. and Whitfield, C. (2002) Lipopolysaccharide endotoxins. Annu. Rev. Biochem. 71, 635–700. Rogers, H.J. et al. (1980) Microbial Cell Walls and Membranes, Chapman & Hall, London. Wilkinson, S.G. (1988) Gram-negative bacteria. In Microbial Lipids. vol. 1, C. Ratledge and S.G. Wilkinson, Eds., Academic Press, London, pp. 299–488. Wilkinson, S.G. (1996) Bacterial lipopolysaccharides — themes and variations. Prog. Lipid Res. 35, 283–343.
References Barry, C.E. et al. (1998) Mycolic acids: structure, biosynthesis and physiological function. Prog. Lipid. Res. 37, 181–207. Brennan, P.J. (1988) Mycobacterium and other actinomycetes. In Microbial Lipids. vol. 1, C. Ratledge and
36
2 OCCURRENCE AND CHARACTERISATION OF OILS AND FATS
F. D. Gunstone and J. L. Harwood
2.1
dramatic example relates to the attempts — which must surely be ultimately successful — to grow plants that generate long chain PUFA (polyunsaturated fatty acids), such as eicosapentaenoic (EPA) and docosahexaenoic (DHA) in their seed lipids. • For a long time, fats were considered to be useful only as a source of calories, but now, in addition to the recognition of essential fatty acids and to the important minor components present in oils and fats, it is recognised that many fatty acid derivatives and lipids act as signalling molecules in the complex interactions that make up life in both animals and plants. • Gunstone (2005) has calculated the annual fatty acid production during 2004 and 2005 on the basis of the production in that year of 17 commodity oils and fats. Of the 136.4 million tonnes produced, he calculated levels for the following acids: lauric (3.4 million tonnes, 2.5% of total production), myristic (2.6, 1.9%), palmitic (27.4, 20.0%), stearic (7.2, 5.3%), oleic (47.8, 35.1%), linoleic (37.5, 27.5%), linolenic (4.5, 3.3%), and other (6.0, 4.4%). Other figures relate to oils and fats used for food purposes and show the changes resulting from industrial hydrogenation. Attention is drawn to the serious consequence of hydrogenation for the level of linolenic acid, which is the major source of omega-3 PUFA.
Introduction
The first chapter in this book was concerned with the wide range of fatty acids that occur naturally and with the various natural lipids of which the acids are major constituents. This chapter is devoted to information about the natural occurrence of the lipids covering the important materials (mainly triacylglycerols) that furnish our food lipids and are the basis of the (growing) oleochemical industry and also the less common lipids, such as those occurring in leaves, in algae, etc. These may seem to be mature topics, but they are being developed in many new and important ways. The following are typical: • Oils and fats are being produced in ever-increasing quantities. In Chapter 3 of the second edition of this book (published in 1994), it was forecast that the average annual supply of oils and fats (from 17 commodity sources) in the 5-year period of 2003 to 2007 would be 104 metric tonnes. In the harvest year 2004–2005, the supply was 136 million tonnes, with soybean oil and palm oil predominating, each at levels of 33 million tonnes. • We now recognise a close relation between health and disease on the one hand and dietary intake of lipids on the other. Our increasing ability to modify lipid composition through seed breeding with or without genetic modification is leading to extensive changes in fatty acid composition, driven in large part by nutritional influences. These new products are considered to be healthier fats. This will be illustrated in several ways in the following section. The most
References Gunstone, F.D., Fatty acid production for human consumption, Inform, 16, 736–737, 2005.
37
2.2 Major oils from plant sources
2.2
Major oils from plant sources
2.2.1
Introduction
Akoh, C.C. and Min, D.B. Eds. Food Lipids — Chemistry, Nutrition and Biotechnology, 2nd ed., Marcel Dekker, New York, 2002. Ching Kuang Chow Ed. Fatty Acids in Foods and Their Health Applications, Marcel Dekker, New York, 2000. Firestone, D. Ed. Official Methods and Recommended Practices of the American Oil Chemists’ Society: Physical and Chemical Characteristics of Oils, Fats and Waxes, AOCS Press, Champaign, IL, 1997, reprinted as a book and disc, 1999. Gunstone, F.D. Ed. Vegetable Oils in Food Technology — Composition, Properties and Uses, Blackwell Publishing, Oxford, U.K., 2002. Gunstone, F.D. Ed. Modified Foods For Use in Lipids, Woodhead Publishing Ltd., Cambridge, U.K., 2006. Murphy, D.J., Ed. Plant Lipids — Biology, Utilisation and Manipulation, Blackwell Publishing, Oxford, U.K., 2005. O’Brien, R.D., Fats and Oils — Formulating and Processing for Applications, 2nd ed., CRC Press, Boca Raton, FL, 2004. Rossell, J.B. and Pritchard, J.L.R., Ed. Analysis of Oilseeds, Fats, and Fatty Foods, Elsevier Applied Science, London, 1991. Seed oil fatty acids database (SOFA) www.bagkf.de/SOFA Shahidi, F. Ed. Bailey’s Industrial Oil and Fat Products, 6th ed., Wiley-Interscience, Hoboken, New Jersey, USA, 2005. Ucciani, E., Nouveau Dictionnaire des Huiles Végétales, Compositions en acides gras, Lavoisier Tec & Doc, Paris, 1995. Websites (general) www.codexalimentarius.net www.cyberlipid.org www.fosfa.org www.ncaur.usda.gov/currentres www.usda.gov/nas www.margarine.org www.nal.usda.gov/fnic www.oilworld.de www.gafta.com
This discussion on vegetable fats is divided arbitrarily into major and minor oils, so it is necessary to consult Section 2.3 for a long list of minor oils. The oils in each section are presented in alphabetical order. The major oils are discussed where possible in the following terms: production; harvest yields; trade (exports and imports); major components (fatty acids and triacylglycerols); minor components (phospholipids, sterols, tocols, other); major uses; and sources of information. Information is tabulated where appropriate and data for several oils may be collated in a single table. Tables at the end of this section contain information on many different oils. They include: Table 2.43a: Past, present, and future production of oils and fats. Table 2.43b: Past and present production of oils and fats. Table 2.44: Predicted total (million tonnes) and per capita consumption (kg per annum) on a global basis and for selected countries/regions throughout the century. Table 2.45: Production, consumption, imports, and exports of 17 oils and fats (million tonnes) by country/region for the calendar years 2000 to 2004 by country/region. Table 2.46a and Table 2.46b: Range of fatty acid composition for some major oils taken from Codex Alimentarius. Table 2.47: Sterols (mg/100 g oil) in a range of crude vegetable oils. Table 2.48: Content of tocols in selected vegetable oils, animal fats, and nuts and berries. Table 2.49: Some physical and chemical properties of major vegetable oils adapted from Firestone (1997).
2.2.2
Castor oil (Ricinus communis)
Castor oil is unique among commodity oils in that it is rich in a hydroxy acid (ricinoleic, 12-hydroxyoleic) and is used only for industrial and cosmetic purposes. The distinct physical and chemical properties of the oil depend on the unusual chemical nature of this acid. The hydroxyl group provides additional functionality and polarity in a mid-chain position. Compared with common vegetable oils, castor oil is more viscous, less soluble in hexane, more soluble in ethanol, and is optically active. It can be converted to a range of interesting and useful materials (see Section 9.8). The castor plant is grown mainly in India, China, and Brazil (Table 2.1). Extraction by pressing and with solvent furnishes castor oil and residual meal. The latter contains a mildly toxic alkaloid (ricinine), an extremely poisonous protein (ricin), and a heat-stable allergen. Castor oil contains about 90% ricinoleic acid and small amounts of palmitic, stearic, oleic, linoleic, and 9,10dihydroxystearic acids. Most of the triacylglycerols are triricinolein or glycerol esters with two ricinoleic and one other acyl chain. In contrast to some other (less common) hydroxy acid-containing oils, the hydroxyl groups in castor oil remain free and are not themselves acylated. Ricinoleic
Useful general information is available in the following sources: Rossell and Pritchard, 1991; Ucciani, 1995; Ching Kuang Chow, 2000; Gunstone, 2002 and 2006; Akoh and Min, 2002; O’Brien, 2004; Shahidi, 2005; and Murphy, 2005. Aitzetmüller and his colleagues have prepared a valuable database on seed oil fatty acids (www.bagkf.de/ SOFA) and have described this in Aitzetmüller et al (2003a and 2003b). An older database is also available (www.ncaur.usda.gov/nc/ncdb).
References Aitzetmüller, K., and Matthäus, B., Potential uses of the seed oil fatty acids database ‘SOFA’, Lipid Technol. Newsl., 9, 123–127, 2003a. Aitzetmüller, K. et al. A new database for seed oil fatty acids — the database SOFA, Eur. J. Lipid Sci. Technol.105, 92–103, 2003b.
38
Occurrence and Characterisation of Oils and Fats
dibasic acid, when reacted appropriately, produces a nylon (polyamide) and efficient lubricants (esters). Splitting ricinoleic acid with steam yields C7 and C11 products. This splitting process has been much improved by the development of a continuous steam-cracking process. Heptanal is used in perfumes and 10-undecenoic acid shows antifungal properties and can be converted, via 11-amino-undecanoic acid, to a polyamide (Rilsan). A new plasticiser made from fully hydrogenated castor oil and acetic acid is particularly effective with PVC and, unlike the presently used phthalates, shows no hormone disrupting effects. It is metabolised like other vegetable oils and is fully biodegradable (Anon., 2005). For further information, see Table 2.43 to Table 2.49.
TABLE 2.1 Major countries/regions involved in the production (from indigenous or imported seed), consumption, export, and import of castor oil in 2004/05 (1000 tonnes) Total (kt) Production Consumption
522 519
Exports Imports Seed yield (t/ ha)
258 260 0.99
Countries/Regions India 316, China 106, Brazil 70 China 146, EU-25 110, India 87, Brazil 65, US 39, Japan 25, Thailand 20 India 224 EU-25 110, China 45, US 39, Japan 24 India 1.07, China 0.96, Brazil 0.91
Source: Adapted from Oil World Annual 2005, ISTA Mielke GmbH, Hamburg, 2005.
acid is produced in nature by hydroxylation of oleic acid, probably present in a phosphatidylcholine molecule. The oil contains some sterols and some tocols, but since it is not used for food purposes these are not considered to be very important. Castor oil differs from other commodity oils and fats in that it contains high levels of ricinoleic acid (12hydroxyoleic acid) and the oil or castor acids is a starting point for several useful chemicals (Caupin, 1997). Sulfation converts the secondary hydroxyl group (>CHOH) to a sulfate (>CHOSO2OH) with improved surfactant properties. Apart from soap, this is the earliest anionic surfactant (1874) and is still used in textile processing, leather treatment, and as an additive for cutting oils and hydraulic fluids. The sulfated hydrogenated oil has the consistency of an ointment and gives adjustable viscosity to water-based formulations with excellent skin compatibility. Castor oil has been converted to estolides by acylation of the free hydroxyl groups with oleic acid at 175 to 250°C in the absence of any catalyst (Isbell and Cermak, 2002). Reaction with other acids has been achieved using tetrabutyl titanate as catalyst (Kulkarni and Sawant, 2003). Dehydration of castor oil and of castor acids gives products rich in diene acids (mainly 9,11- and 9,12-18:2), some of which have conjugated unsaturation. These products are valuable alternatives to drying oils, such as tung oil, which contain conjugated trienoic acids (see Section 2.3.109). Hydrogenated castor oil and hydrogenated castor acids, with higher melting points than the nonhydrogenated material, are used in cosmetics, coatings, and greases. Greases prepared from tallow are much improved when salts of 12-hydroxystearic acid are added. Castor oil reacts with isocyanates to give polyurethanes, which are frequently used for wood preservation and have been developed as encapsulating materials. Splitting ricinoleic acid with caustic soda gives C8 and C10 products. At 180 to 200°C with a 1:1 caustic/castor ratio, the major products are 2-octanone and 10-hydroxydecanoic acid. At 250 to 275°C and a 2:1 ratio, the products are 2-octanol and sebacic (decanedioic) acid. The
References Anon., Danisco’s plasticizer base on vegetable oil gets EU approval, Lipid Technol., 17, 51–52, 2005. Caupin, H-J., Products from castor oil in Gunstone, F.D. and Padley, F.B. Eds. Lipid Technologies and Applications, Marcel Dekker, New York, 1997, pp. 787–795. Isbell, T.A. and Cermak, S.C., Synthesis of triglyceride estolides from lesquerella and castor oils, J. Amer. Oil Chem. Soc., 79, 1227–1233, 2002. Kulkarni, M.G. and Sawant, A.B., Some physical properties of castor oil esters and hydrogenated castor oil esters, Eur. J. Lipid Sci. Technol., 105, 214–218, 2003.
2.2.3
Cocoa butter (Theobroma cacao)
The commercial exploitation of cacao or cocoa beans was probably first practised by the Aztecs. The Spanish transferred the bean from Mexico to Europe in the 16th century, where it was consumed as a drink. Chocolate was developed only in the 19th century. The plant is an evergreen tree growing to 5 to 10 metres. The fruit is a large pod approximately 15 to 20 cm long and 7 cm in diameter containing 25 to 50 seeds embedded in a soft sweet edible pulp (Nickless, website). Production figures for cocoa butter are not included in the statistics generally cited for oil and fat production, but according to information cited in www.gobi.co.uk, world consumption of cocoa butter was over 700 kt in 2003 and is growing at a rate around 2% a year. Europe is the largest consuming region accounting for 60% of world consumption and Germany, the U.S., and France are the main importing countries. Cocoa is grown mainly in West Africa (Ghana, Ivory Coast, Nigeria), Malaysia, Brazil, Central America, India, and Sri Lanka. The composition of cocoa butter from these different sources varies somewhat as shown in Table 2.2 for cocoa butter from Ghana, Ivory Coast, Brazil, and Malaysia. Small differences in fatty acid composition are reflected in the iodine value, but more significantly in the triacylglycerol composition and, consequently, in the melting profile. The average content 39
2.2 Major oils from plant sources
behaviour gives it properties that are significant in chocolate. At ambient temperature, it is hard and brittle giving chocolate its characteristic snap, but also it has a steep melting curve with complete melting at mouth temperature. This gives a cooling sensation and a smooth creamy texture. For example, the content of solids falls from 45 to 1% between 30 and 35°C. The hardness of cocoa butter is related to its solid fat content at 20 and 25°C. This melting behaviour is related in turn to the chemical composition of cocoa butter. The fat is rich in palmitic (24 to 30%), stearic (30 to 36%), and oleic acids (32 to 39%) and its major triacylglycerols are of the kind SOS, where S represents saturated acyl chains in the 1 and 3 positions and O represents an oleyl chain in the 2 position. There are three major components: POP, POSt, and StOSt (P = palmitic acid and St = stearic acid). These triacylglycerols have 50, 52, or 54 carbon atoms in their 3 acyl chains and the levels of these can be determined by high temperature gas chromatography (GC) with the ratio of these being used to detect adulteration of cocoa butter. (Triacylglycerol molecular species are detailed in Table 2.2.) Cocoa butter has a high content of saturated acids that raises health concerns, but it has been argued that much of this is the noncholesterolemic stearic acid. Chocolate is also a rich source of flavonoids, which are considered to be powerful antioxidants (Beckett, 1999 and 2000). Triacylglycerol analysis of cocoa butter is generally carried out by capillary GC and the results of an interlaboratory study have been reported (Buchgruber et al., 2003). Seventeen triacylglycerol species were recognised, including POSt (39.8%), StOSt (28.0%), POP (15.6%), PLSt (3.2%), StLSt/StOO (2.9%), POO (1.9%), PLP (1.8%), OOO (1.5 %), StOA (1.0%), and seven others (total 3.3%). The crystal structure of cocoa butter has been studied extensively because of its importance in understanding the nature of chocolate (Section 4.6.5). The solid fat has been identified in six crystalline forms designated I to VI. Some crystals show double chain length (D) and some triple chain length (T) (Sato et al 1989). The six forms have the following melting points (°C) and D/T structure: I (17.3, D), II (23.3, D), III (25.5, D), IV (27.3, D), V (33.8, T), and VI (36.3, T). Form V is the one preferred for chocolate. This crystalline form gives good molding characteristics and has a stable gloss and favourable snap at room temperature. It is desirable to promote the formation of form V and to inhibit its conversion to form VI. Form V is usually obtained as a result of extensive tempering (putting molten chocolate through a series of cooling and heating processes), which have been found to optimise production of the appropriate polymorph. Alternatively, molten chocolate can be seeded with cocoa butter already crystallised in form V. Transition from form V to the more stable form VI leads to the appearance of white crystals of fat on the surface of the chocolate. This phenomenon (“bloom”) is promoted
TABLE 2.2 Composition and properties of cocoa butter from different countries
Iodine value Melting point °C Diacylglycerols (%) Free acid (%)
Ghana
Ivory Coast
Brazil
Malaysia
35.8 32.2 1.9 1.53
36.3 32.0 2.1 2.28
40.7 32.0 2.0 1.24
34.2 34.3 1.8 1.21
23.7 32.9 37.4 4.0 1.0
24.8 37.1 33.2 2.6 1.1
Component acids (%) Palmitic Stearic Oleic Linoleic Arachidic
24.8 37.1 33.1 2.6 1.1
25.4 35.0 34.1 3.3 1.0
Component triacylglycerols (%) Trisaturated Monounsaturated POP POSt StOSt Diunsaturated Polyunsaturated
0.7 84.0 15.3 40.1 27.5 14.0 1.3
0.6 82.6 15.2 39.0 27.1 15.5 1.3
trace 71.9 13.6 33.7 23.8 24.1 4.0
1.3 87.5 15.1 40.4 31.0 10.9 0.3
Solid content (pulsed NMR) — after tempering for 40 hours at 26°C 20°C 25°C 30°C 35°C
(%) (%) (%) (%)
76.0 69.6 45.0 1.1
75.1 66.7 42.8 0.0
62.6 53.3 23.3 1.0
82.6 77.1 57.7 2.6
Source: Adapted from Shukla, V.J.S., Inform, 8, 152–162, 1997. The original paper contains more details along with information on cocoa butter from India, Nigeria, and Sri Lanka.
of the important SOS triacylglycerols (S = saturated, O = oleic) varies between 87% in Malaysian and 72% in Brazilian cocoa butter, with the African samples midway between these extremes (Shukla, 1995 and 1997, see also Kurvinen et al, 2002). There is, however, some evidence that the cocoa butters of different geographical origins are becoming more alike. Harvested pods are broken open and left in heaps on the ground for about a week during which time the sugars ferment. The beans are then sun dried and are ready for transportation and storage. To recover the important components, the beans are roasted at ~150°C, shells are separated from the cocoa nib, and the latter is ground to produce cocoa mass. When this is pressed, it yields cocoa butter and cocoa powder still containing some fat (10 to 24%). Typically, 100 g of beans produce 40 g of cocoa butter by pressing, expelling, or solvent extraction; 40 g of cocoa powder; and 20 g of waste material (shell, moisture, dirt, etc.). Increasingly the beans are processed in the country where they grow and cocoa liquor, cocoa powder, and cocoa butter (usually in 25 kg parcels) are exported to the chocolate-producing countries. Cocoa butter carries a premium price and is sometimes adulterated (Crews 2002). Cocoa butter is a solid fat melting at 32 to 35°C (Table 2.2). It is in high demand because its characteristic melting 40
Occurrence and Characterisation of Oils and Fats
because of their high content of lauric acid. Some other lauric oils occur among the minor oils (Section 2.3). This oil has been reviewed by Pantzaris and Yusof Basiron, 2002; O’Brien, 2004; Canapi et al., 2005; and Gervajio, 2005. Coconuts grow in coastal regions between 20º N and 20º S of the equator. The trees bear fruit after 5 or 6 years and for up to 60 years thereafter. The shell is split open and allowed to dry. The “meat” on the inside of the shell is called copra and is the source of coconut oil in a yield of ~65%. The oil is extracted by pressing, usually followed by solvent extraction. As indicated in Table 2.3, coconut oil is produced mainly in the Philippines, Indonesia, and India and exported from the first two countries to EU-25 and U.S., in particular. Both lauric oils are used for a similar range of food and nonfood purposes. They are used to make soaps and other surface-active products and in the production of spreads and other food products. They are also the source of the C8 and C10 acids required to make MCT (medium chain triglycerides). These liquid products are used as lubricants in food-making equipment and, because they are easily metabolised, they appear in food preparations for invalids and athletes. Both oils can be fractionated into oleins and stearins, and hydrogenated to modify their properties and extend their range of uses. Palm kernel stearin is used as a chocolate substitute fat. Coconut stearin is somewhat softer and is used as a confectionery filling fat. The two lauric oils differ slightly from one another mainly in that coconut oil is the richer in the 6:0 to 10:0 acids and palm kernel oil is the richer in unsaturated C18 acids (Table 2.4). This is reflected in the triacylglycerol composition usually expressed in terms of carbon number (the sum of the carbon atoms in the three acyl groups and ignoring the three glycerol carbon atoms). The C36 triacylglycerols, dominant in both oils, will be mainly, but not
by fluctuations in temperature during storage and by migration of liquid oils from nut centres. This change is undesirable because it detracts from the appearance of the chocolate and may be mistaken for microbiological contamination. Bloom can be inhibited by the addition of a small amount of 2-oleo 1,3-dibehenin (BOB), milk fat, or other form V stabilisers to the cocoa butter. This phenomenon is discussed in more detail by Padley, 1997; Smith, 2001; Timms, 2003; and Longchampt and Hartel, 2004. Minor components include phospholipids (0.050.13%), tocopherols (~200 ppm, — mainly γ-tocopherol), sterols, 4-methylsterols, and triterpene alcohols. Cocoa butter is also used in cosmetics (Section 9.4). For further information, see Table 2.43 to Table 2.49. For information on cocoa butter replacers and cocoa butter substitutes, see the appropriate minor vegetable oils and the section on chocolate.
References Beckett, S.T., Ed. Industrial Chocolate Manufacture and Use, Blackwell Science, Oxford, 1999. Beckett, S.T., The Science of Chocolate, The Royal Society of Chemistry, Cambridge, 2000. Buchgruber, M. et al. Capillary GC: a robust method to characterise the triglyceride profile of cocoa butter — results of an intercomparison study, Eur. J. Lipid Sci. Technol., 105, 754–760, 2003. Crews, C., Authenticity of cocoa butter, in Oils and Fats Authentication, Jee, M. Ed. Blackwell Publishing, Oxford, U.K., 2002. Kurvinen, J.-P. et al. Rapid MS method for analysis of cocoa butter TAG, J. Amer. Oil Chem. Soc.,79, 621–626, 2002. Longchampt, P.L. and Hartel, R.W., Fat bloom in chocolate and compound coatings, Eur. J. Lipid Sci. Technol., 106, 241–274, 2004. Nickless, H., Cocoa butter quality, Guest contribution: www. britanniafood.com Padley, F.B., Chocolate and confectionery fats, in Lipid Technologies and Applications, Gunstone, F.D. and Padley, F. B. Eds. Marcel Dekker, New York, 1997, pp. 391–432. Sato, K. et al., Polymorphism of POP and SOS. I Occurrence and polymorphic transformation, J. Am. Oil Chem. Soc., 66, 664–674, 1989. Shukla, V.J.S., Chocolate — the chemistry of pleasure, Inform, 8, 152–162, 1997. Shukla, V.J.S., Cocoa butter — properties and quality, Lipid Technol., 7, 54–57, 1995. Smith, K.W., Cocoa butter and cocoa butter equivalents, in Structured and Modified Lipids, Gunstone, F.D. Ed. Marcel Dekker, New York, 2001. Timms, R.E., Confectionery Fats Handbook — Properties, Production, and Application, The Oily Press, Bridgwater, England, 2003.
2.2.4
TABLE 2.3 Major countries/regions involved in the production (from indigenous or imported seed), consumption (food and nonfood uses), export, and import of coconut oil in 2004/05 (million tonnes)
Coconut oil (Cocos nuciferus)
Coconut oil and palm kernel oil differ from other commodity oils and are known collectively as lauric oils
Total (mt)
Countries/Regions
Production
3.01
Consumption
2.99
Exports
1.86
Imports
1.87
Seed yield (t/ha)
0.52
Philippines 1.27, Indonesia 0.74, India 0.40, Mexico 0.11 EU-25 0.69, India 0.43, US 0.36, Philippines 0.28, Indonesia 0.20, China 0.12, Mexico 0.12 Philippines 0.98, Indonesia 0.56, Malaysia 0.13 EU-25 0.72, US 0.38, Malaysia 0.17, China 0.12 Philippines 0.90, Indonesia 0.46, India 0.35, Mexico 1.36
Source: Adapted from Oil World Annual 2005, ISTA Mielke GmbH, Hamburg, 2005.
41
2.2 Major oils from plant sources
TABLE 2.4
Fatty acid composition (% weight) of coconut oil and palm kernel oil Coconut oil
Palm kernel oil
Mean (b)
Range (b)
Range (c)
Mean (d)
Range (d)
0.4 7.3 6.6 47.8 18.1 8.9 2.7 6.4 1.6
0–0.6 4.6–9.4 5.5–7.8 45.1–50.3 16.8–20.6 7.7–10.2 2.5–3.5 5.4–8.1 1.0–2.1
0–0.7 4.6–10.0 5.0–8.0 45.1–53.2 16.8–21.0 7.5–10.2 2.0–4.0 5.0–10.0 1.2–2.5 0–0.2 0–0.2 0–9.2 6.3–10.6
0.2 3.3 3.5 47.8 16.3 8.5 2.4 15.4 2.4
0–0.8 2.1–4.7 2.6–4.5 43.6–53.2 15.3–17.2 7.1–10.0 1.3–3.0 11.9–19.3 1.4–3.3
0–0.8 2.4–6.2 2.6–5.0 45.0–55.0 14.0–18.0 6.5–10.0 1.0–3.0 12.0–19.0 1.0–3.5
17.5 26.4
14.1–21.0 24.0–28.3
14.1–21.0
6:0 8:0 10:0 12:0 14:0 16:0 18:0 18:1 18:2 18.3 20:0 20.1 IV (a) SMP (ºC)
0.1
0–0.2
8.5 24.1
6.3–10.6 23.0–25.0
Range (c)
Source: Adapted from Pantzaris, T.P. and Yusof Basiron, in Gunstone, F.D. (Ed.) Vegetable Oils in Food Technology — Composition, Properties and Uses, Blackwell Publishing, Oxford, pp 157–202, 2002. (a) Iodine value calculated from fatty acid composition (b) Leatherhead Food Research Association (LFRA) survey, 35 samples (c) Codex Alimentarius values (d) LFRA survey, 71 samples Values in the original references cited as trace and as not detected have been replaced by 0 in this table.
entirely, trilaurin because of the very high level of this acid. Careful study of Table 2.5 shows small differences in triacylglycerol composition between the two oils reflecting the differences in fatty acid composition referred above. Interesting results reported by Caro et al. (2004) show that the sn-2 position is enriched in lauric acid and the two unsaturated C18 acids, and that the remaining saturated acids are enriched at the sn-1/3 positions. Other work cited by Caro et al. (Table 2.6) shows that the 1 and 3 positions differ in their fatty acids with 6:0, 8:0, and 10:0 occurring particularly at sn-3. Coconut oil is highly saturated with an iodine value between 7 and 10 and this is probably associated with the low levels of tocols in the oil. These have been cited at a mean level of 10 ppm (ranging between 0 and 44 ppm) with α-tocopherol and α-tocotrienol the major members. Among the sterols (mean 836 ppm, range 470 to 1140 ppm) β-sitosterol, ∆ 5 -avenasterol, stigmasterol, and campesterol predominate and account typically for 46, 27, 13, and 9% of total sterols, respectively. Polycyclic aromatic hydrocarbons (PAH) are usually present at levels of 150 ppb in crude vegetable oils and 98% of the cis-isomer. Another effective hydrogenation catalyst is the borohydride-reduced nickel (P-2 Ni) developed by Brown and Ahuja (1973), which reduces alkynes to cisalkenes selectively. This catalyst was used by Nunn et al. (1992) in the synthesis of leukotrienes (a class of conjugated polyunsaturated fatty acids). Conversion of an acetylenic fatty acid to the trans olefinic fatty acid can be readily achieved with sodium or lithium dissolved in pure liquid ammonia (by condensing gaseous ammonia with solid CO2 in a cold finger). Cis-olefinic fatty esters can also be transformed (stereomutation) to the corresponding trans-olefinic fatty ester by refluxing the cisolefinic ester in dioxan with a catalytic amount of freshly prepared sodium p-toluenesulphinic acid (Snyder and Scholfield, 1982). Another method is to expose the cis-olefinic fatty acid or ester to 2-mercaptoethanol in iso-propanol under ultraviolet light as was used to obtain the all-trans-arachidonic acid from arachidonic acid or its methyl ester (Anagnostopoulos et al., 2005). Impurities present after reduction may include saturated acids, traces of the undesired geometric isomers (cis or trans), which are best removed by silverion silica chromatography. Trans fatty acids can be purified by crystallization when available in sufficiently large quantities. Gunstone and Ismail (1967) and Barve and Gunstone (1971) prepared all the octadecynoic acids and from them all the cis- and trans-octadecenoic acids. Lie Ken Jie and Lam (1974) synthesised all positional isomers of methyl undecynoate and cis-undecenoate. Gilman and Holland (1974) synthesised all the C10, C11, C12, C13, and C14 monoynoic acids, while Pomonis and Hakk (1990) prepared long-chain monoene acids (C24 and C28). Monoynoic acids can also be produced from olefinic acids by partial synthesis via bromination followed by dehydrobromination (KOH/EtOH) (Gunstone and Hornby, 1969; Lie Ken Jie and Kalluri, 1998). However, this procedure is not suitable for the preparation of poly-ynoic fatty acids from the corresponding poly-enoic acids.
Br MgBr
(1)
(2)
Cl
To improve the yield of these multistep reaction sequences to arachidonic acid, Van der Steen et al. (1963) proposed the use of 2,5-hexadiyne-1-ol as a C6 fragment, which allowed the introduction of two acetylenic bonds in a single step. This approach was exploited by Beerthuis et al. (1971) and by Belosludtsev et al. (1986). Kunau (1971a, 1971b) developed a useful two-step route to 1-bromo2,5,8-tetradecatriyne (3) for the synthesis of arachidonic acid. A further development led to the production of arachidonic acid from the coupling of a C9 fragment (4) and a C11 fragment (5). This approach was also applied to the synthesis of fatty acids containing five and six olefinic bonds. 357
5.1 Unsaturated fatty acid synthesis via acetylene
Br COOH I
(3)
i
(4)
(5)
ii
THPO
THPO
Cl
HO
Cl
Cl Br
Cl
iii
ii
iv THPO
Cl
Cl
v HO
Br (A)
Br
vi
viii
vii
ix
x Cl
Cl
COOH
SCHEME 5.1 (iv) Synthesis of arachidonic acid. (Adapted from Rachlin, A.I., Wasyliw, N., and Goldberg, M.W. (1961) J. Org. Chem., 26, 2688–2693.) Reagents: i, NaNH2, Br(CH2)3Cl; ii, H3O+; iii, PBr3; iv, THP-OCH2C≡CH, NaNH2 (THP = tetrahydropyranyl); v, PBr3; vi, NaC≡CH; vii, NaNH2, BrCH2C≡CH; viii, NaNH2, A, ix, H2, Lindlar catalyst/quinoline; x, Mg, CO2.
i, ii
i, ii
Br
Br
Br
i, ii Br
iv
iii
COOH COOH
SCHEME 5.1 (v) 2779–2787.)
Synthesis of arachidonic acid. (Adapted from Osbond, J.M., Philpott, P.G., and Wickens, J.C (1961) J. Chem. Soc.,
Reagents: i, BrMgC≡CCH2OMgBr, CuCl; ii, PBr3; iii, NaC≡CH(CH2)3COONa, NH3; iv, H2, Lindlar catalyst/quinoline
358
Synthesis
i
THPO
ii, iii
THPO
THPO TBDPS
OH
OH iv
v
Br
vi
OH TBDPS
TBDPS TBDPS
I TBDPS
+ PPh3I TBDPS _
vii, viii
ix
x
O
O OMe
xi
OMe
TBDPS
OH
SCHEME 5.1 (vi) Synthesis of 14-hydroxy-(all-cis-)5,8,11-tetradecatrienoate. (Adapted from Han, L. and Razdan, R.K. (1998) Tetrahedron Lett., 39, 771–774.) Reagents: i, n-BuLi, THF, ethylene oxide; ii, TBDPSCl, imidazole (TBDPSCl = tert-butyldiphenylsilylchloride); iii, Dowex-MeOH; iv, CBr4, Ph3P, CH2Cl2; v, CuI, n-Bu4NI, Na2CO3, DMF, 3-butyn-1-ol; vi, 2Ni(OAc)2-NaBH4, H2; vii, MsCl, Et3N, CH2Cl2; viii, NaI, acetone; ix, Ph3P, MeCN; x, n-BuLi, HMPA, THF, CH3OOC(CH2)3CHO; xi, n-Bu4NF-THF, AcOH.
An elegant procedure for preparing 14,15-dihydroarachidonic acid was based on the coupling of propargylic halides or tosylates with 1-alkynes catalyzed by Cu(I) iodide (Jeffery et al., 1992). Adopting a similar approach 14-hydroxy-(all-cis)-5,8,11-tetradecatrienoate, a useful intermediate for the synthesis of arachidonic acid analogues, was prepared by Cu(I) catalyzed propargylic substitution (Han and Razdan, 1998) (Scheme 5.1(vi)). Several sets of C18 cis,cis- and trans,trans-diene esters were prepared by the acetylenic route where the number of methylene groups between the two unsaturated centres in the isomer varied from 0 to 4 (Gunstone and Lie Ken Jie, 1970; Gunstone, et al., 1971). The complete series of C18 dimethylene-interrupted diacetylenic fatty ester isomers was prepared by Lam and Lie Ken Jie (1975) from
which the corresponding cis,cis- and trans,trans-olefinic C18 isomers were derived by reduction (Lam and Lie Ken Jie, 1976). Examples in Schemes 5.1(vii-ix) illustrate the methods used in obtaining the diunsaturated fatty esters with varying number of methylene groups between the unsaturated centres. (See References for Section 5.1 at the end of Section 5.5.)
5.2
Fatty acid synthesis by the Wittig reaction
The Wittig reaction is comparable in importance to the acetylenic route as a means of synthesizing olefinic fatty acids and has also been used extensively to prepare the long-chain alcohols and their esters that are important as COOH
ii, iii
i
Br
COOH
+
iv, v
COOMe
SCHEME 5.1 (vii) Synthesis of methyl octadeca-10,12-diynoate. (Adapted from Gunstone, F.D. and Lie Ken Jie, M. (1970) Chem. Phys. Lipids, 4, 1–14.) Reagents: i, NaOBr; ii, Br2; iii, KOH/EtOH; iv, NH2NH2, CuCl, Et2NH; v, MeOH/H3O+
359
5.2 Fatty acid synthesis by the Wittig reaction
COOH OH v
i
COOH
BrMg OMgBr ii–iv
i
Br
BrMg
+
COOMgBr iii, vi
COOMe
vii
viii, ix COOMe COOH
SCHEME 5.1 (viii)
Synthesis of methyl octadeca-9,12-diynoate, octadeca-9-cis,12-cis-dienoate and octadeca-9-trans,12-trans-dienoic acid. (Adapted from Gunstone, F.D. and Lie Ken Jie, M. (1970) Chem. Phys. Lipids, 4, 1–14.) Reagents: i, EtMgBr; ii, CH3(CH2)4Br; iii, CuCl; iv, PBr3/Et2O; v, Br2, KOH/EtOH; vi, BF3/MeOH; vii, H2, Lindlar catalysis/quinoline; viii, KOH, EtOH; ix, Li, distilled NH3
i
ii
iii
Cl COOMe
iv, v
SCHEME 5.1 (ix) Synthesis of methyl octadeca-9,13-diynoate. (Adapted from Lam, C.H. and Lie Ken Jie, M.S.F. (1975) J. Chromatogr., 115, 559–570.) Reagents: i, Br2, NaNH2; ii, NaNH2, CH3(CH2)3Br; iii, LiNH2, I(CH2)7Cl; iv, KCN, DMSO; v, HCl(g)/MeOH
of the ylid. Stabilized ylids with an α-double bond or an α-electron-withdrawing group (such as an ester) give the trans alkene in >95% purity, while nonstabilized ylids with electron-donating groups (such as alkyl) give cis-alkenes in >98% purity. An example of the use of the Wittig reaction in organic synthesis is described in Scheme 5.2(i), which describes the synthesis of 2cis- and 2-trans-octene (Schlosser and Christmann, 1966). When the reaction is conducted in a polar aprotic solvent (such as dimethylformamide, DMF) the production of cis-alkene is favoured. The stereochemistry and mechanism of the Wittig reaction have been reviewed (Vedejs and Peterson, 1994; Murphy and Lee, 1999). Vatèle (1999) has described the various synthetic methods for the preparation of β,γ-unsaturated fatty acids for use in the Wittig reaction.
insect pheromones. The Witting reaction leads to the alkene (6) by reaction of an aldehyde (or ketone) and an alkyl triphenylphosphonium salt or its ylid (derived from an alkyl bromide). The group X is selected to give the necessary functionality at the end of the long chain.
R
X n
O + Ph3P or
R
PPh3
R +
O
X n
(6)
X n
For effective use of this reaction, it is important to appreciate the factors that influence the configuration of the double bond, of which the most important is the structure 360
Synthesis
Ph + – CH 3 Ph P C H Ph i
Ph Ph P+ – O
Ph Ph + Ph P – O
Ph H Li
iii
H H
CH3 (CH2)4CH3
ii
H
CH3 (CH2)4CH3
erythro betaine
Ph Ph H H3C
Ph P+ – O
Ph Ph + Ph P _ O
iv
H H3C
Li (CH2)4CH3
H3 C
(CH2)4CH3 H
cis-alkene
v
H (CH2)4CH3
threo betaine
H3C H
H (CH2)4CH3
trans-alkene
SCHEME 5.2 (i) Wittig reaction — synthesis of 2-octene. (Adapted from Schlosser, M. and Christmann, K.F. (1966) Angew. Chem. Int. Ed. Engl. 5, 126.) Reagents: i, CH3(CH2)4CHO, THF; ii, warm; iii, PhLi, THF; iv, HCl, Et2O; v, tert. BuOH.
and Vostrowsky 1979; and Vatèle 1999). Gravier-Pelletier et al. (1990) have reviewed synthetic routes to longchain hydroxy acids, many of which involve the Wittig reaction. Viala and Santelli (1988) described the preparation of the phosphonium salt (Ph3P+[CH2]2CH[OiPr]2Br–) from acrolein (propenal) and found it to be very effective as a three-carbon homologating agent under Wittig conditions. They converted hexanal to methyl arachidonate in four steps with an overall yield of 58% (Scheme 5.2(viii)). Heitz et al. (1989) followed a similar approach in the synthesis of all-cis-1-bromo-4,7,10,13-nonadecatetraene (a precursor to C-1 labelled arachidonic acid), using instead the more readily available diethyl acetal phosphonium salt (Ph3P+[CH2]2CH[OEt]2Br–) as the C3 building block. The β,γunsaturated diethyl acetals intermediate from the Wittig coupling were hydrolyzed under very mild conditions (trifluoroacetic acid in chloroform at 0oC) to
Formation of the ylid requires a suitable base and, while several have been used (see examples in Scheme 5.2(ii) to Scheme 5.2(x)), particularly good results are claimed for sodium bis-trimethylsilylamide at –78oC. To obtain highly steroselective cis-olefins in Wittig reactions, salt-free ylid solutions are prepared with sodium amide (Bestmann, 1965) or sodium or potassium hexamethyldisilazide (Bestmann et al., 1976; Viala and Santelli, 1988; Labelle et al., 1990) in THF at low temperature, or by generating the ylid with t-butyllithium (Bestmann and Stransky, 1974), LiN(SiMe3)2 or LDA (lithium dimethylamide) (Dussault and Lee, 1995) in HMPT (hexamethylphosphoric triamide) or HMPA (hexamethylphosphoric amide) (Corey et al., 1980). Aldehydes required for the Wittig reaction are prepared from unsaturated fatty acids or cycloalkenes by ozonolysis, from α,ω-diols by partial oxidation and from α,ω-alkadienes by partial hydroboration followed by oxidation (Bergelson and Shemyakin 1964; Bestmann
i–iv O
+
Ph3P
COOEt
PPh3
i
v, vi
COOH COOH
SCHEME 5.2 (iii) Octadec-9-cis-en-12-ynoic acid (crepenynic acid). (Adapted from Bradshaw, R.W. et al. (1971) J. Chem. Soc., C, 1156–1158.)
SCHEME 5.2 (ii) Dodec-4-cis-enoic acid. (Adapted from Bestmann, H.J. and Vostrowsky, O. (1979) Chem. Phys. Lipids, 24, 335–389.)
Reagents: i, LiNH2, ethylene oxide; ii, PBr3; iii, PPh3, C6H6; iv, BuLi, Et2O; v, MeOOC(CH2)7CHO; vi, KOH, H3O+.
Reagent: i, Wittig reaction, ester hydrolysis by KOH, H3O+.
361
5.2 Fatty acid synthesis by the Wittig reaction
O
i, ii Ph3P
+
COOEt
COOH
SCHEME 5.2 (iv) Octadeca-9-cis,11-trans,13-trans-trienoic acid (catalpic acid). (Adapted from Bergelson, L.D. and Shemyakin, M.M. (1964) Angew. Chem. Int. Ed., 3, 250–260.) Reagents: i, DMF, I; ii, KOH, H3O+.
i
ii
HO
HO
OH
Br
OTHP
iii, iv
OTHP
Ph3P
OTHP
v
vi, vii OTHP
OAc
SCHEME 5.2 (v) Tetradeca-9-cis,11-trans-dienyl acetate. (Adapted from Hall, D.R., Beevor, P.S., Lester, R., Poppi, R.G. and Nesbitt, B.F. (1975) Chem. Ind., 216–217.) Reagents: i, 2,3-dihydro-2H pyran; ii, PBr3, iii, PPh3, CH3CN, K2CO3; iv, dimsyl sodium, DMSO; v, CH3CH2CH=CHCHO; vi, MeOH, H3O+; vii, Ac2O, pyridine.
O
+
O
Ph3P
i
O
i
AcO
SCHEME5.2 (vi) Pentadeca-11-trans,13-cis-dienyl acetate. (Adapted from Bestmann, H.J. and Vostrowsky, O. (1979) Chem. Phys. Lipids, 24, 335–389.) Reagents: i, CH3CH=PPh3; ii, 9-BBN (9-borabicyclo[3.3.1]nonane); Ac2O, pyridine.
O
+
i, ii
Ph3P
OAc O
SCHEME5.2 (vii) 335–389.)
Hexadeca-6-cis,11-cis-dienal. (Adapted from Bestmann, H.J. and Vostrowsky, O. (1979) Chem. Phys. Lipids, 24,
Reagents: i, KOH, H3O+; ii, C5H5NHCrO3Cl.
O P+Ph3Br– +
COOMe
i
COOMe
ii, iii
OH
SCHEME 5.2 (viii)
Triacontanol. (Adapted from Subramanian, G.B.V. and Rastogi, A. (1991) Chem. Ind., 436–440.)
Reagents: i, K2CO3, 1,4-dioxan, formamide (trace); ii, H2, Pd/C; iii, LiAlH4.
O
i, ii
O
i, ii
O
iii, iv O
SCHEME 5.2 (ix)
i, ii
COOMe
Methyl arachidonate. (Adapted from Viala, J. and Santelli, M. (1988) J. Org. Chem., 53, 6121–6123.)
Reagents: i, Ph3P+(CH2)2CH(OPri)2Br –, NaN(SiMe3)2, THF, HMPA; ii, H2O, TsOH; iii, Ph3P+(CH2)4COOH Br –, NaN(SiMe3)2, THF, HMPA; iv, CH2N2.
362
Synthesis
PhCO2
PhCO2
OH OH
i
CHO
HO
OAc
OH PhCO2 PhCO2
OH
ii, iii
CHO
COOEt PhCO2 PhCO2
D-ribose
iv
PhCO2
PhCO2 OAc
PhCO2
viii, ix
OTs
v–vii
CH=CHCH2COOEt
O
O
O
x, xi
COOCH3
O
PhCO2
O
(CH2)3COOCH3
COOCH3
COOCH3
xii
SCHEME 5.2 (x) Synthesis of methyl ester of Leukotriene A. (Adapted from Corey, E.J., Clark, D.A., Goto, G., Marfat, A., Mioskowski, C, Samuelsson, B. and Hammarström, S., (1980) J. Am. Chem. Soc., 102, 1436–1439.) Reagents: i, pyridine, benzoyl chloride, HBr, acetic acid, Na2CO3, Ag2CO3; ii, Ph3P=CHCOOEt, benzoic acid, dimethoxyethane; iii, Ac2O, trace H2SO4; iv, Zn amalgam, Et2O, HCl; v, Pd/C, H2, MeOH; vi, HCl, MeOH; vii, tosyl chloride, pyridine; viii, K2CO3, MeOH; ix, Collin's reagent, CH2Cl2; x, LiCH=CHCH=CHOEt, Na2CO3; xi, CH3SO2Cl, Et3N, CH2Cl2; xii, I–Ph3P+CHCH2CH=CH(CH2)4CH3, n-BuLi, HMPA (hexamethylphosphoric acid), THF.
synthesis of DHA (all-cis-4,7,10,13,16,19-docosahexaenoic acid) (Scheme 5.2(xiii)). (See References for Section 5.2 at the end of Section 5.5.)
yield the requisite unsaturated aldehyde intermediate for use in the next Wittig coupling reaction. Sandri and Viala (1995) developed two new C6 cis,cis-1,4diene homologating units, which can react in turn either as ylid or aldehyde in a subsequent Wittig reaction to yield methylene-skipped polyunsaturated fatty acids, such as α-linolenic acid (Scheme 5.2(xi)) and EPA (all-cis5,8,11,14,17-eicosapentaenoic acid) (Viala and Sandri, 1992). Eynard et al. (1998) developed two new C6 homologating units (Bu t Ph 2 SiO(CH 2 ) 2 CH=CHCH 2 CHO and Ph3P+(CH2)2CH=CHCH2CH(OEt)2I– from a common intermediate, cis-HO(CH2)2CH=CHCH2CH(OEt)2, for the synthesis of EPA (Scheme 5.2(xii)). A C9 building block was employed by Taber and You (1995) in the Br– + Ph3PO
i
OiPr
O
OiPr
O
iii
ii
OiPr
iPrO
iPrO
Isotopically labelled fatty acids
Fatty acids labelled with stable isotopes, such as deuterium (2H) or carbon-13 (13C) or with radioactive isotopes, such as tritium (3H) and carbon (11C, 14C), are used in studies of reaction mechanism (e.g., hydrogenation, oxidation), nuclear magnetic spectroscopy, mass spectrometry, lipid biosynthesis, and metabolism. Compounds containing radioactive isotopes (3H, 14C) are examined by liquid
iPrO
OiPr iPrO
5.3
iv
iPrO iPrO
v–vii
Br– + Ph3P
viii, ix
COOH
SCHEME 5.2 (xi)
Synthesis of α-linolenic acid. (Adapted from Sandri, J. and Viala, J. (1995 Synthesis, 271–275.)
Reagents: i, NaN(SiMe3)2, O2; ii, HCOOH, CH3COCH3/H2O; iii, NaN(SiMe3)2, Ph3P+ –CHCH2CH3; iv, HCOOH, THF/H2O; v, LiALH4; vi, Ph3PBr, pyridine; vii, Ph3P; viii, NaN(SiMe3)2, CHO(CH2)7COOMe; ix, LiOH, THF/H2O.
363
5.3 Isotopically labelled fatty acids
i–iii
OEt
HO
CHO
v, vi HO
iv
OEt OEt
OEt
OTBDPS ix
OEt OEt OEt OEt
vii, viii I–P+Ph3
OTBDPS OEt OEt
x
P+Ph3I–
xiii, xiv
xi, xii
OH OEt OEt
COOMe
OEt OEt COOMe
xv
OEt OEt COOMe
xvi
O
SCHEME 5.2 (xii) Synthesis of all-cis-5,8,11,14,17-eicosapentaenoic acid. (Adapted from Eynard, T. et al. (1998) J. Labelled Cpd. Radiopharm, XLI, 411–421.) Reagents: i, n-BuLi; ii, BF3⋅Et2O; iii, ethylene oxide; iv, H2, Lindlar catalyst/ quinoline; v, TBDPSCl, imidazole (TBDPSCl = tert-butyldiphenylsilylchloride); vi, 50% CF3COOH/CH2Cl2; vii, PPh3, I2, imidazole; viii, PPh3; ix, n-BuLi, HMPA/THF; x, n-Bu4NF; xi, PPh3, I2, imidazole; xii, PPh3, CaCO3/CH3CN; xiii, n-BuLi, HMPA; xiv, CHO(CH2)3COOMe; xv, 50% CF3COOH/CH2Cl2; xvi, n-Pr+PPh3I–, HMPA/THF.
OTHP OH
OH
i, ii
iii, iv
OH v–vii
OH
viii
O
O
O
O EtOOC
ix, x O
O
P+PH3 –OTs
EtOOC xiii, xiv
xii HO
xi O
O
EtOOC
OH
SCHEME 5.2 (xiii) Synthesis of ethyl all-cis-4,7,10,13,16,19-docosahexaenoic acid. (Adapted from Taber, D.F. and You, K (1995) J. Org. Chem., 60, 139–142.) Reagents: i, EtMgBr, CuBr; ii, CH2=CHCH2Br; iii, TsCl (4-CH3C6H4SO2Cl)/KOH; iv, BrMgC≡C(CH2)2OTHP; v, OsO4, K3Fe(CN)6; vi, Dowex/ MeOH; vii, H2SO4/ CH3COCH3; viii, P-2 Ni, H2; ix, TsCl, pyridine; x, PPh3; xi, NaN(SiMe3)2/ THF, CHO(CH2)2COOEt; xii, HCl/THF; xiii, NaIO4/H2O-CH2Cl2/SiO2; xiv, PPh3+OTs–-(CH)2CH=CHCH2CH=CHCH2CH3.
labelled acids are commercially available, it is frequently necessary to synthesise these compounds as required, sometimes by suitable adaptation of the synthetic procedures
scintillation counting or radio gas chromatography. Nonradioactive labels are studied by mass spectrometry or by nuclear magnetic resonance spectroscopy. Though some 364
Synthesis
a deuterium atom. This method is also used to label the terminal methyl group with other isotopes (3H, 13C or 14C) (Tamvakopoulos and Anderson, 1990) (Scheme 5.3.1(ii)). Wilkinson’s catalyst [(Ph3)3RhCl(I)] and deuterium gas have been used to reduce olefinic or acetylenic fatty acid methyl esters to saturated, deuterium-labelled fatty esters. Reduction of methyl linolenate by this method furnished methyl octadecanoate-[9,10,12,13,15,16-2 H6 ] (Rakoff, 1982). Lie Ken Jie and Choi used deuterated picolinyl esters of polyunsaturated fatty esters obtained by a similar procedure to determine the positions of double bonds by mass spectrometric analysis (Lie Ken Jie and Choi, 1992). When other metal catalysts are used instead of Wilkinson’s catalyst extensive isotopic scrambling occurs. Acidic hydrogens alpha to carbonyl or keto groups are readily exchanged with deuterium by using Na and CH3O2H (Aasen et al., 1970) or 2H2O and pyridine (Tucker et al., 1971). To incorporate deuterium atoms at a specific position in the alkyl chain, reduction of the methyl ester function of a fatty ester with lithium aluminum deuteride gives the corresponding 1,1-2H2-alkanol, which is converted to the bromide and chain extended to the required deuterated fatty acid (DasGupta et al., 1982) (Scheme 5.3.1(iii)).
described in Section 5.1 and Section 5.2. This account is confined to fatty acids labelled with one or more of the hydrogen (2H, 3H) or carbon (11C, 13C, 14C) isotopes. The subject has been reviewed by Crombie (1996), Kunau (1973), Sprecher (1977), Emken (1978), Tulloch (1979), Rakoff (1982), Westerman and Ghrayeb (1982), and Lie Ken Jie et al. (1997). Adlof (1999) has written a very comprehensive review with lists of over 170 stable isotopically labelled fatty acids and a number of suppliers of isotopically labelled fatty acids. Descomps (1995), Sauerwald et al. (1996), and Demmelmair et al. (1997) have reviewed the use of stable isotopes in the study of human lipid metabolism.
5.3.1 5.3.1.1
Deuterium-labelled fatty acid Synthesis of deuterium-labelled saturated fatty acids
Depending on the position on the alkyl chain and the number of isotopic labels to be incorporated into the saturated fatty acid, several methods are available. Starting with commercially available deuterated acetic acid, total replacement of the hydrogen atoms of the terminal carbon atom (ω-carbon) of the alkyl chain by deuterium is achieved by the Kolbe reaction with diacids (Klok et al., 1974) (Scheme 5.3.1(i)). A versatile procedure has been developed by treating the tosyl derivative of ωhydroxy-fatty acids with sodium borodeuteride or labelled lithium dimethyl cuprate to substitute one of the three hydrogens of the resulting methyl group by
5.3.1.2
Synthesis of deuterium-labelled unsaturated fatty acids
The most successful way of preparing deuterium-labelled unsaturated fatty acids of the cis-configuration with 2H atoms on the olefinic carbons is to reduce the corresponding acetylenic ester with deuterium gas in the
COOH (i), (ii)
MeOOC
2H 2
H
COOH
2
H
Synthesis of hexadecanoic acid-16-2H3. (Adapted from Klok, R. et al. (1974) Recl. Trav. Chim. Pays-Bas, 93,
SCHEME 5.3.1 (i) 222–224.)
Reagents: i, C2H3COO2H; ii, 2Na (-2e).
(i) COOCH3
TsO
CH22H
COOMe
SCHEME 5.3.1 (ii) Synthesis of hexadecanoic acid-16-2H1. (Adapted from Tamvakopoulos, C.S. and Anderson, V.E. (1990) J. Labelled Cpd. Radiopharm., 28, 187–191.) Reagent: i, NaB2H4.
COOMe
2H
(i)
O2H
(ii), (iii)
2H 2H
MgBr
2H
(iv), (v)
COOH
2H
SCHEME 5.3.1 (iii)
2H
Synthesis of palmitate-7-2H2. (Adapted from DasGupta, S.K. et al. (1982) J. Lipid Res., 23, 197–200.)
Reagents: i, LiAl2H4; ii, HBr; iii, Mg; iv, Br(CH2)5COOMgCl; v, H3O+.
365
5.3 Isotopically labelled fatty acids
2
H
2H 2H 2H
2
H
2
(i) COOH
COOH H
2
H
2
H
SCHEME 5.3.1 (iv) Synthesis of deuterated arachidonic acid. (Adapted from Taber, D.F. et al. (1982) in Methods in Enzymology, Academic Press, New York, pp. 366–369.) Reagent: i, 2H2, Lindlar catalyst.
reductive ozonization of methyl oleate). Wittig coupling of the two intermediates furnished a mixture of deuterated CLA isomers, which were separated by reversed-phase and silver resin chromatography. Once a 2H-labelled compound or an intermediate has been made, it can be incorporated into a long-chain acid by chain extension (via the nitrile, by malonation, or by enamine synthesis) or by involvement in a Wittig coupling synthesis as illustrated in the synthesis of CLA isomers (Scheme 5.3.1(vii)) (Adlof, 1997). Reduction of acetylenic acids with lithium aluminium deuteride in 2H2O gives the corresponding trans-olefinic-2H2 fatty acid. This method was successfully applied to the synthesis of β-parinaric acid, (all-trans)-octadeca-9,11,13,15tetraenoic acid-9,10,11,12,13,14,15,16-2H8 (Goerger and Hudson, 1988). Replacement of all hydrogen atoms of a long chain fatty acid by deuterium yields the corresponding perdeuterated fatty acid. This reaction can be accomplished by reaction of the fatty acid with either PtO2 in 2H2O (Hsiao et al., 1974), 2H2O, PtO2 and Na2O2 (Dinh-Nguyen et al., 1972) or the potassium salt of the fatty acid with 2H2O, PtO2 and Na2O2 (Reeves et al., 1979).
presence of Lindlar catalyst. Taber et al. (1982) prepared labelled arachidonic acid (all-cis-5,8,11,14-eicosatetraenoic acid-5,6,8,9,11,12,14,15-2H8) from the corresponding acetylenic precursor (Scheme 5.3.1(iv)). To probe the reaction mechanism of prostaglandin H synthase, Peng et al. (2002) synthesised both 13(R)- and 13(S)-deuterium-labelled arachidonic acids in high enantiomeric purity (Scheme 5.3.1(v)). Other useful catalysts reported in the reduction of acetylenic intermediates to the corresponding cis-olefinic fatty acid are bis-(2-deuteriocyclohexyl)borane-B-2H (Svatos, et al.,1994; Tamvokopoulos and Anderson, 1990), deuterated disiamylborane/acetic acid-d4 for the synthesis of a conjugated linoleic acid (CLA) isomer, viz. 7-trans,9-cis-octadecadienoic acid–9,10- 2H 2 (Broustal and Loreau, 2004) (Scheme 5.3.1(vi)) and P-2 Ni (Taber et al., 1982). In the synthesis of methyl 9-cis,11-trans- and 9-trans,11trans-octadecadienoate-17,17,18,18-2H2 (CLA isomers), two key intermediates were prepared: viz. trans-2-nonenyl8,8,9,9-2H2 bromide and MeOOC(CH2)7CHO (from the
O
(i)
O 2
H
H
2
H
OH
O
COOMe H
O
O
H
2
H
H
(ii) H
(v), (vi)
O
H
COOMe
2H
(iii), (iv) O
COOH
2H
O
SCHEME 5.3.1 (v) Synthesis of 13(R)- and 13(S)-deuterium-labelled arachidonic acids. (Adapted from Peng, S. et al. (2002) J. Am. Chem. Soc., 124, 10786–10796.) Reagents: i, PCC; ii, NaHMDS; iii, BF3-HOAc; iv, Pb(OAc)4; v, BrPh3P(CH2)5CH3; vi, LiOH, CH3OH, THF.
OTHP (i)-(v)
2H
COOH
2
H
SCHEME 5.3.1 (vi) Synthesis of a deuterium labelled conjugated linoleic acid (7-trans,9-cis-octadecadienoic-9,10-2H2 acid). (Adapted from Broustal, G. and Loreau, O. (2004) J. Labelled Cpd. Radiopharm., 47, 875–880.) Reagents: i, NaB2H4, 2-methyl-but-2-ene, BF3.Et2O; ii, acetic acid-d4; iii, PPh3.Br2; iv, Mg; v, CO2.
366
Synthesis
(iii), (iv)
(i), (ii) OH
(v), (vi)
2
O
2H 2
2
O
O
H
OH
(viii)
2
H
2
H 2H
H
H H
2
H
H 2H
2
H2 H
H
2
I
H 2H
2
O
(vii)
H
H
2 2H
OH
(ix)
H
H
2
H
H 2H
2
(x)
Br
2
H 2
H2 H
2
H
H
2
H 2H
2H
H
Br–Ph3P+
COOMe
2H
(xi) 2 2 H H H H 2H
2
H
2
COOMe
SCHEME 5.3.1 (vii) Synthesis of deuterated conjugated linoleic acid (CLA) isomers. (Adapted from Adlof, R. (1997) Chem. Phys. Lipids, 88, 107–112.) Reagents: i, dihydropyran; ii, p-toluenesulfonic acid; iii, 2H2, benzene; iv, (Ph3P)RhCl; v, H3PO4/P2O5; vi, KI; vii, HC≡CCH2OH; viii, Li/NH3; ix, Li/NH3; x, Ph3PBr2; xi, Ph3P; xii, n-BuLi, MeOOC(CH2)7CHO.
5.3.2
13
flow, membrane transport, and as an imaging agent for cardiac fatty acid metabolism (Buckman et al., 1994). Wüst et al. (2000) have developed a new approach to the synthesis of 11C-labelled fatty acids (Scheme 5.3.3(i)).
C-labelled fatty acids 13
Reagents used for the incorporation of C atoms are 13CO , K13CN, or 13CH I, all of which are very expensive 2 3 and therefore limited to small scale synthesis. Syntheses of 13C labelled fatty acids follow similar procedures to those described in Sections 5.1 (acetylenic intermediates) and Section 5.2 (Wittig reaction). A large-scale preparation of octadeca-9-cis,12-trans-dienoic acid-1-13C, octadeca-9-cis,12-cis,15-trans-trienoic acid-1- 13C and the (all-cis)-octadeca-9,12,15-trienoic acid-1-13C was reported (Loreau et al., 2000).
5.3.3
5.3.4
C-labelled fatty acids
The application of 11C-labelled fatty acids is limited by the short half-life (22 min) of this isotope. However, the short half-life allows 11C-labelled fatty acids to be used in following the kinetics of biological processes, such as blood
(ii)
C-labelled fatty acids
Reagents used to incorporate a 14C atom are K14CN, 14CO , 14CH I. Fatty esters with the label at C-1 position 2 3 are produced by the standard 1-carbon chain extension procedure from the alkyl chloride or bromide (RCl or RBr) with K14CN followed by treatment of the resulting nitrile with HCl (g) in anhydrous methanol to yield the requisite R14COOCH3. Carbonation reaction of the Grignard reagent (from alkyl bromide and magnesium) with 14CO2 (generated from Ba14CO3) furnishes R14COOH. A comprehensive list of synthetic 14C-labelled fatty acids is available (Adlof, 1999).
11
(i)
ButOOC
14
ButOOC
I
Zn 2
CuI(MgCl)2.Zn(CH3)2
tBuO C 2
2
(iii), (iv)
11C
HOOC
11
C
HOOC
SCHEME 5.3.3 (i) 1289–1300.)
Synthesis of [11C]-labelled fatty acids. (Adapted from Wüst, F. et al. (2000) J. Labelled Cpd. Radiopharm., 43,
Reagents: i, ZnEt2; ii, Me2CuI(MgCl)2; iii, [1-11C] alkyl iodide; iv, TFA.
367
5.4 Synthesis of acylglycerols
OH
(i)
(ii)
(iii)
OMs 14
CN
SCHEME 5.3.4 (i)
OH
(iv), (v)
14COOH
Synthesis of octadec-9-trans-enoic acid-1-14C. (Adapted from Valicenti, A.J. et al. (1985) Lipids, 20, 234–242.
Reagents: i, Li/NH3; ii, MsCl/pyridine; iii, K14CN/DMSO; iv, hydrolysis; v, H3O+.
2- or 3-14C-Labelled fatty acids have been prepared from 1- C-labelled fatty acids by a series of standard one-carbon chain extension reactions as was the case for the preparation of tetracosa-9,12,15,18,21-pentaenoic acid-[3-14C] (Voss et al., 1991). Unsaturated fatty acids with a 14C-label at the C-1 position can be obtained by decarboxylation of the corresponding unlabelled analogs to shorten the chain by one carbon (to the nor-alken-1-ol) and “reconstitution” of the carboxyl group by a one-carbon-14C chain extension procedure via the Grignard reagent of the corresponding bromide derivative of the alken-1-ol with 14CO (Valicenti et al., 1985) (Scheme 5.3.4(i)). 2 Unsaturated fatty acids labelled at an internal carbon atom are prepared by the Wittig coupling of an alkyl or alkenyltriphenylphosphonium halide and the appropriate 14C-labelled aldehydic ester, as exemplified in the synthesis of methyl oleate-9-14C (Barley et al., 1973) (Scheme 5.3.4(ii)). Incorporating a 14C label at the terminal methyl end of the fatty acid chain is achieved using 14CH3I. Methyl oleate-[18-14C] was prepared by Pichat et al. (1969) as illustrated in Scheme 5.3.4(iii). (See References for Section 5.3 at the end of Section 5.5.)
5.4
14
Efforts have focused on preparing specifically structured and configurationally pure triacylglycerols. Mono- and diacylglycerols are normally obtained as intermediates in these processes. General methods for synthesizing mono, di-, and triacylglycerols are reviewed by Quinn et al. (1967), Smith (1972), Jensen and Pitas (1976), Bhati et al. (1980), Gunstone and Norris (1983), Jensen (1995), and Sonnet (1999).
5.4.1
Acylglycerols by chemical synthesis
Acyl chlorides (RCOCl) or acid anhydrides (RCO)2O) provide rapid and reliable acylation. Saturated acid chlorides are prepared by reaction of fatty acid with thionyl chloride (Jensen and Pitas, 1976), while unsaturated fatty acid chlorides are best produced with oxalyl chloride (ClOCCOCl) (Mattson and Volpenhein, 1962). Notes on precautions to be taken during the preparations are detailed by Buchnea (1978). A very successful esterification method employs 1,1′-dicyclohexylcarbodiimide (DCC) and 4-dimethylaminopyridine (DMAP), which permits the carboxylic acid
(i)
I
Ph3P+I–
14
C
SCHEME 5.3.4 (ii) 151–154.)
Synthesis of acylglycerols
(ii)
COOMe
Synthesis of methyl oleate [9-14C]. (Adapted from Barley, G.C. et al. (1973) J. Chem. Soc. (Perkin I),
Reagents: i, PPh3; ii, MeOOC(CH2)7 14CHO.
O
(i), (ii)
14CH
O
14
(iii)
O
3
O O
O
(iv)
CH3
(v)
14
CH3
O
H O
14
CH3
SCHEME 5.3.4 (iii) 1198–1200.)
OMe 14
Synthesis of methyl oleate-[18- C]. (Adapted from Pichat, L. et al. (1969) Bull. Soc. Chim Fr., 4,
Reagents: i, C6H5Li; ii, 14CH3I; iii, H2, Pt; iv, H2SO4; v, (Ph)3P=CH(CH2)7COOCH3.
368
Synthesis
OH OH
6:0
i
OH OH OH
6:0 6:0
OH
SCHEME 5.4.1 (i) Synthesis of glycerol trihexanoate. (Adapted from Lie Ken Jie, M.S.F. and Lam, C.C. (1995a) Chem. Phys. Lipids, 77, 155–171.)
synthesised numerous mixed rac- triacylglycerols containing short and long chain saturated acyl chains at specific positions of the glycerol “backbone” (Lie Ken Jie and Lam, 1995a) (Schemes 5.4.2(i) and (ii)), and mixed triacylglycerols containing saturated, acetylenic and olefinic acyl moieties (Lie Ken Jie and Lam, 1995b, 1995c) (Scheme 5.4.2(iii)). The question of the stereospecific description of the absolute configuration of glycerol derivatives has received considerable attention. Various methods are discussed by Buchnea (1978), who also emphasizes the advantages of the stereospecific numbering (sn) system, now recommended by the IUPAC-IUB Commission. This system recognizes that the two primary methylene groups of glycerol are prochiral and are not interchangeable in their reactions with chiral structures, such as enzymes. If the secondary hydroxyl group is shown to the left of C-2 in a Fischer projection, then the carbon atom above is called C-1 and that below is termed C-3. The problem of 1,2-acyl migration (Mattson and Volpenhein, 1962) led to the need for protection of the selective hydroxyl groups of the glycerol substrate during the course of preparation of mono-, di- and triacylglycerols. Isomerization occurs readily under basic or acidic condition. For example, (S)-glycerol 1,2-dipalmitate equilibrates with the 1,3-isomer at 40 to 65oC in about 3 h (Kodali and Duclos, 1992). Migration is also observed in the solid state (Dorset and Pangborn, 1979; Dorset, 1987) and the rate of isomerization depends on chain length (Boswinkel et al. 1996). Thus, during the course of
Interest in relating the structure of dietary fat molecules (triacylglycerols) to human health has led to efforts to synthesise triacylglycerols of specific structure (Bell et al., 1997). Publications on the chemistry, nutrition and biotechnology of food lipids (Akoh and Min, 2002), structurally modified lipids (Gunstone, 2001), modified food fats (Christophe, 1998), functional foods and nutraceuticals (Gunstone, 2003), functional lipids (Akoh, 2005), lipids in infant nutrition (Huang and Sinclair, 1998), and healthful lipids (Akoh and Lai, 2005) have recently received considerable attention. Commercial applications have focused on the production of triacylglycerols containing readily metabolized medium-chain acids and other potentially desirable fatty acids (such as γ-linolenic acid, arachidonic acid) at appropriate position(s) of glycerol (Akoh, 1995; Bell et al., 1997; Haumann, 1997). Low-calorie fats, such as Salatrim (Smith et al., 1994; Softly et al., 1994; Klemann, 1994) and Caprenin (Haumann, 1997), contain short (C2-C4) or medium-chain C8-C10 fatty acids that have been incorporated into the triacylglycerol molecule by chemical interesterification. Lie Ken Jie and Lam have
iii
i
O OH O OB(OEt3)3 6:0 OB(OEt3)3
18:0 18:0 6:0
iv
Re a g e n t s : i , C H 3 C O C H 3 , p - t o l u e n e s u l fo n i c a c i d ; i i , CH3(CH2)4COOH, DCC (1,1'-dicyclohexylcarbodiimide), DMAP (4-dimethylaminopyridine), CH2Cl2; iii, boric acid, triethyl borate, H2O; iv, CH3(CH2)16COOH, DCC, DMAP, CH2Cl2.
Structured triacylglycerols by chemical synthesis
OH OH OH
OH 6:0
O O 6:0
ii
Scheme 5.4.2 (i) Synthesis of rac-glycerol 1,2-distearate-3hexanoate via isopropylidene glycerol. (Adapted from Lie Ken Jie, M.S.F. and Lam, C.C. (1995a) Chem. Phys. Lipids, 77, 155-171. )
(fatty acid) to be acylated to the glycerol molecule (Neises and Steglich, 1978; Kodali, 1987; Lie Ken Jie and Lam, 1995a, 1995b, 1995c; Mazur et al., 1991) (Scheme 5.4.1(i)). Fatty acid anhydrides can be prepared via mixed-acid anhydride intermediates using either trifluoroacetic anhydride or acetyl chloride (Mattson et al., 1964). After refluxing, the excess reagent is removed by distillation and the mixed-acid anhydrides converted to fatty acid anhydrides by heating under pressure.
5.4.2
OH
iii
Reagents: (i) hexanoic acid, 1,1'-dicyclohexylcarbodiimide (DCC) and 4-dimethylaminopyridine (DMAP).
O O OH
i
O 6:0 O
ii CHC6H5 OH 6:0
iv
OH
v
CHC6H5 18:0 6:0 18:0
SCHEME 5.4.2 (ii) Synthesis of glycerol 1,3-distearate-2-hexanoate via benzylidene glycerol. (Adapted from Lie Ken Jie, M.S.F. and Lam, C.C. (1995a) Chem. Phys. Lipids, 77, 155–171.) Reagents, i, benzaldehyde, sulfuric acid; ii, CH3(CH2)4COOH, DCC (1,1′-dicyclohexylcarbodiimide), DMAP (4-dimethylaminopyridine), CH2Cl2; iii, boric acid, triethyl borate; iv, H2O; v, CH3(CH2)16COOH, DCC, DMAP.
369
5.4 Synthesis of acylglycerols
OH OH OH
i
iv
OH OH 18:0
O O 18:0
O O OH
ii
v
18:1 (9-cis) OH 18:0
vi
iii
OB(OEt3)3 OB(OEt3)3 18:0 18:1(9-cis) 18:2(9-cis, 12-cis) 18:0
SCHEME5.4.2 (iii) Synthesis of rac-glycerol 1-oleate-2-linoleate-3-stearate. (Adapted from Lie Ken Jie, M.S.F. and Lam, C.C. (1995a) Chem. Phys. Lipids, 77, 155–171; and Lie Ken Jie, M.S.F. and Lam, C.C. (1995c) Chem. Phys. Lipids, 78, 1–13.) Reagents, i, CH3COCH3, p-toluenesulfonic acid; ii, CH3(CH2)16COOH, DCC (1,1'-dicyclohexylcarbodiimide), DMAP (4-dimethylaminopyridine), CH 2 Cl 2 ; iii, boric acid, triethyl borate; iv, H 2 O; v, CH 3 (CH 2 ) 7 CH=CH(CH 2 ) 7 COOH, DCC, DMAP; vi, CH3(CH2)4CH=CHCH2CH=CH(CH2)7COOH, DCC, DMAP, CH2Cl2.
been extensively used for acyglycerol preparations (Verkade, 1953). This blocking group is removed by catalytic hydrogenolysis, with hydrogen chloride in diethyl ether or petroleum ether, or with a molar equivalent of boron trifluoride-methanol in dichloromethane at 22oC for 15 min (Iwama and Foglia, 1988). Detritylation leads to the formation of by-products, which must be removed by crystallization or by silicic acid column chromatography. Therefore, the benzylidene protective group (PhCH=) is more useful for many purposes (Buchnea, 1978).
preparation of specific acylglycerols, it is necessary to incorporate protective groups for specific hydroxyl groups of the glycerol, which can be removed with minimal acyl migration. In addition, since racemic derivatives of snglycerol cannot be resolved, at least one protective group must be used during the initial preparation of the enantiomeric glycerol. Eibl and Woolley (1986) compared the benzyl, allyl and trityl groups. 5.4.2.1
Benzyl group (PhCH2-)
This group was first used as a blocking group for the synthesis of monoacid diacyl-sn-glycerols when the group was added by reacting sodium or potassium isopropylidene-snglyceroxide with benzyl chloride (PhCH2Cl). 1-Benzyl- and 3-benzyl-sn-glycerols can be further protected by trityl groups (see below) so that mixed acid diacyl- and triacyl-snglycerols can be produced. Removal of the benzyl group requires catalytic hydrogenolysis, so its use is limited to preparations with saturated fatty acids, although, of course, unsaturated fatty acids can be introduced after the benzyl group has been removed (Buchnea, 1978). 4-Methylbenzyl (Golding, 1988) is reported to be a better protecting group than benzyl. It is removed by reaction with 2,3-dichloro-5,6dicyano-1,4-benzo-quinone (DDQ). 5.4.2.2
5.4.2.4
This protective group can be used during the preparation of unsaturated 1,2- and 2,3-diacyl-sn-glycerols. The 2,2,2trichloroethoxycarbonyl group is introduced onto the free primary hydroxyl group of isopropylideneglycereol using 2,2,2-trichloroethyl chloroformate. This protecting group is retained, without migration, when the isopropylidene group is removed and is itself removed with zinc in acetic acid at room temperature (Pfeiffer et al., 1968). 5.4.2.5
Other blocking groups
Various other blocking groups are discussed by Buchnea (1978) and Gunstone and Norris (1983) with respect to their use in acylglycerol synthesis. Moss et al. (1987) recommend the use of 2-iodobenzoyl as a protecting group. This can be removed by chlorination followed by mild basic hydrolysis with methanolic sodium carbonate or bicarbonate. Greene and Wuts (1999) have described numerous protecting groups for alcohols and 1,2-diols and procedures for deprotecting these groups. For example, in the preparation of 1,2-diacylglycerols, dimethylboron dibromide can be used to cleave the 1,2-acetonide of glycerol (Kodali, 1987). This reagent can also be used for detritylation and debenzylation (Kodali and Duclos, 1992), thus circumventing the catalytic hydrogenolysis procedure. Detritylation can be achieved by the use of methanolic borontrifluoride (Hermetter and Paltauf, 1981), while Nbromosuccin-imide in dimethyl sulfoxide removes readily the t-butyldimethylsilyl protecting group (Batten et al., 1980; Burgos et al., 1987) in the preparation of 1,2diacylglycerols.
Isopropylidene group (Me2C =)
This is a widely used protective group for chemical reactions involving multiple hydroxyl functions as in the acidcatalysed reaction between glycerol and acetone (Me2C = O). Its first use for the preparation of intermediates in acylglycerol synthesis was by Fischer and Baer (1937), who synthesise 1,2- and 2,3-isopropylidene-sn-glycerols. Such substances can then be acylated or blocked (e.g., by benzyl groups) in the remaining hydroxyl moiety. Care must be taken that no racemization or loss of the protective group takes place on storage (Buchnea, 1967). The isopropylidene group is best removed by trimethyl borate in the presence of boric acid (Mattson and Volpenhein, 1962). Kodali (1987) claims improved results with dimethylboron bromide (Me2BBr) at –50oC. 5.4.2.3
2,2,2-Trichloroethoxycarbonyl group (Cl3CCH2OCO-)
Triphenylmethyl (trityl) group (Ph3C-)
Triphenylmethyl chloride (Ph3CCl) reacts mainly with the primary hydroxyls of glycerol, and the trityl group has 370
Synthesis
OH OH OH
i
O Ph
OH
O
ii
Ph
O
16:0 O
iii
HO 16:0 HO
SCHEME 5.4.3 (i) Synthesis of glycerol 2-palmitate. (Adapted from Jensen, R.G. and Pitas, R.E. (1976) in Advances in Lipid Research, Vol. 14 (Eds., R. Paoletti and D. Kritchevsky), Academic Press, New York, pp. 213–247.) Reagents: i, benzaldehyde, sulphuric acid; ii, palmitoyl chloride, pyridine; iii, boric acid, triethyl borate, H2O.
5.4.3 5.4.3.1
problems associated with the inherent instability of the isomer and its isomerization to 1,3-diacylglcyerol. In addition, it is difficult to crystallize 1,2-diacylglycrols because of their high solubility in appropriate solvents. Again, if only small amounts of 1,2-diacylglycerols are needed, then it is more convenient to use pancreatic lipase and to isolate 1,2-diacylglycerol by TLC (Jensen and Pitas, 1976). Also, passage of a 1,3-diacylglycerol isomer through neutral alumina will result in conversion of about 50% to the 1,2-diacylglycerol isomer, which can then be purified by TLC (Jensen et al., 1966). A convenient method to prepare 1,2-diacylglycerol makes use of the tetrahydropyranyl ether of glycerol, which is obtained by reaction of dihydropyran with allyl alcohol to form allyl tetrahydropyranyl ether. The latter is oxidized with potassium permanganate to give tetrahydropyranylglycerol (Barry and Craig, 1955). Alternatively, glycerol carbonate can be converted to the tetrahydropyranyl ether and the carbonate removed with potassium hydroxide (Cunningham and Gigg, 1965). Examples are the preparation of glycerol 1,2-dioleate (Krabisch and Borgstrom, 1965; Turner et al., 1968) and glycerol 1,2dipalmitate (Jensen and Pitas, 1976). Acylation of tetrahydropyranyl-glycerol is performed with an acid chloride. Purification of the tetrahydropyranyl protected 1,2-diacylglycerol is achieved by crystallization. The blocking ether can be removed by boric acid (Gigg and Gigg, 1967). To prepare specifically structured triacylglycerols, methods to obtain the R- and S-enantiomers of the corresponding aldehyde of 1,2-glycerol acetonide have been developed (Jurczak et al., 1986). D-mannitol is converted to a 1,2,5,6-di-O-isopropylidene derivative and then oxidatively cleaved to yield an intermediate aldehyde, which is reduced with NaBH4 to give the requisite (S)-1,2O-isopropylidene glycerol (Scheme 5.4.3(ii)) (Eibl, 1981). Mikkilineni et al. (1988) prepared the (S)-(protected)alcohol from D-isoascorbic acid. To prepare the (R)-(protected)-alcohol, diazotization of L-serine gave chiral glyceric acid, which was then methylated. Acetonization of the latter compound was followed by lithium aluminum hydride reduction of the ester group to furnish the (R)-1,2-O-isopropylideneglycerol (Scheme 5.4.3(iii)) (Lok et al., 1976). Methods for the preparation of the (R)-1,2-O-isopropylidene glycerol from ascorbic acid (Jung and Shaw, 1980) or from L-arabinose (Kanda and Wells, 1980; Kodali, 1987) were also reported. 1-Benzyl- and 3-benzylsn-glycerols can be formed from the appropriate 2,3- or 1,2-isopropylidene-sn-glycerol by reaction with benzyl
Racemic acylglycerols 1-Acylglycerols
Glycerol is reacted with acetone in the presence of ptoluenesulphonic acid to yield 1,2-isopropylidene glycerol, which can be acylated with a fatty acid. Excess isopropylidene glycerol is removed with sodium acetate solution, and the extent of the acetonization and acylation reactions can be determined by quantification of the amount of water released during the reaction (Jensen and Pitas, 1976). The isopropylidene ester can then be cleaved with boric acid and the 1-acylglycerol purified by crystallization. Purity can be checked by thin layer chromatography using a solvent system where 1- and 2-monoacylglycerols are separated (Thomas et al., 1965) or by NMR spectroscopy. An alternative method of cleavage is the use of hydrogen chloride (Buchnea, 1971). 5.4.3.2
2-Acylglycerols
The two primary alcohol groups of glycerol are protected by condensing benzaldehyde with glycerol in catalytic amount of p-toluenesulphonic acid to yield 1,3-benzylideneglycerol. In this compound, only the free 2-hydroxyl is available for acylation (Mattson and Volpenhein, 1962). Benzylideneglycerol is crystallized and purified (Jensen and Pitas, 1976). The free 2-hydroxyl group is acylated with an acid chloride and the resulting acyl-1,3-benzylideneglycerol is crystallized. Deprotection of the benzylidene group is achieved by treatment with boric acid and water (Scheme 5.4.3 (i)). The presence of boric acid inhibits isomerization to the more stable 1-acylglycerol (the equilibrium ratio is 9:1 in favour of the 1-isomer). Small amounts of 2-acylglycerol can be prepared by hydrolysis of an appropriate triacylglycerol with pancreatic lipase. This enzyme is specific for the 1and 3-postions and, thus, yields a 2-acylglycerol product. By keeping the incubation times short (e.g., 5 min), acyl migration is minimized. The 2-acylglycerol can then be purified by a boric acid TLC system (Jensen and Pitas, 1976). 5.4.3.3
1,3-Diacylglycerols
1,3-Diacylglycerols can be synthesised by acylation of a 1acylglycerol with an acid chloride in the presence of pyridine and chloroform at room temperature for 24 h. The requisite product is crystallized from hexane or hexane-ethanol mixtures in 50 to 70% yield (Jensen and Pitas, 1976). 5.4.3.4
1,2-Diacylglycerols
Several methods are available for synthesizing 1,2-diacylglycerols (Jensen, 1972). All procedures have to overcome the
371
5.4 Synthesis of acylglycerols
HO HO
OH
O
O i
OH
ii
HO
O
OH
HO
iii
O
O
O
CH2OH
CHO OH
O
O (S)
SCHEME 5.4.3 (ii) Synthesis of (S)-1,2-O-isopropylidene glycerol from D-mannitol. (Adapted from Eibl, H. (1981) Chem. Phys. Lipids, 28, 1–5. With permission.) Reagents: i, ZnCl2/acetone; ii, NaIO4, pH = 8; iii, NaBH4.
OH
CH2OH H2N
i
ii
O
OH
COOH
iii
O
COOMe
O
O
COOMe
OH (R)
SCHEME 5.4.3 (iii) Synthesis of (S)-1,2-O-isopropylidene glycerol from L-serine. (Adapted from Lok, C.M. et al. (1976) Chem. Phys. Lipids, 16, 115–122.) Reagents: i, HNO2, (CH3O)2C(CH3)2/p-TsOH; ii, acetone/p-TsOH; iii, LiAlH4.
be evaluated as acyl migration is likely to occur during the heating process in the above reaction. To overcome acyl migration, Burgos et al. (1987) treated (S)-glycidol with stearic acid in the presence of Ti(Oi-Pr)4 at 0oC to give (S)-1-stearoylglycerol. Sonnet’s group (1991, 1994) reacted (S)-glycidol with an acyl chloride in pyridine at 0oC to form the chiral glycidol ester, which on treatment with an acid anhydride and LiBr gave a bromodiester intermediate. Villeneuve et al. (1994) used Bu4NBr in addition to LiBr to improve ring opening of the epoxy ring. The bromodiester intermediate was then reacted with a carboxylate to displace the bromide to yield a chiral triacylglycerol (Scheme 5.4.3(v)).
chloride under alkaline conditions (Buchnea, 1978). The 3-benzyl-1,2-isopropylidene- or 1-benzyl-2,3-isopropylidene-sn-glycerol products are separated by reduced pressure distillation and the isopropylidene protective group removed by hydrolysis with 10% acetic acid. Glycidol is also used to prepare a variety of structurally specific diacylglycerols (Lok, 1978, Lok et al., 1985, Zlatanos et al., 1985). Glycidol ester was prepared from acid chlorides in pyridine, which was heated with carboxylic acids in the presence of quaternary ammonium salts to furnish 1,3-diacylglycerols (Scheme 5.4.3(iv)) (Lok et al., 1985). When enantiomers of glycidol are used, the configurational purity of the 1,3-diacylglycerols needs to i
O
O
OH
OH
ii
O
O
O
R′
O
O
SCHEME 5.4.3 (iv) 329–334.)
R
CR O
Synthesis of 1,3-diacylglycerol from glycidol. (Adapted from Lok, C.M. et al. (1985) Chem. Phys. Lipids, 36,
Reagents: i, RCOCl, pyridine; ii, R1COOH, Et4NBr.
O i
O
O
O O
OH
CR
O
ii Br
R′ O
R O
O iii
O O
R′′ O
R′ O
R O
SCHEME 5.4.3 (v) Synthesis of chiral triacylglycerol from glycidol. (Adapted from Sonnet, P.E. and Dudley, R.L. (1994) Chem. Phys. Lipids, 72, 185–191.) Reagents: i, RCOCl, pyridine; ii, (R'CO)2O, LiBr, PhH; iii, CsO2CR".
372
Synthesis
5.4.3.5
rare lipid molecules (from unsaturated fatty acids to complex lipid metabolites, such as prostaglandins, leukotrienes, thromboxanes, and prostacyclins, very long chain fatty acids) have been produced in the laboratory. Some have found applications in the oleochemical industry (viz. surfactants and speciality triacylglycerols). Nevertheless, the application of traditional organic synthesis and production of some lipids suffers distinct drawbacks. Organic reactions accelerated by organic or inorganic catalysts do not always have a high degree of selectivity and require the use of complex multistep reactions to achieve stereospecificity and regioselectively in the final product. The past 2 decades have witnessed a dramatic increase in the use of enzymes (lipases) as catalysts in lipid synthesis. Enzymes are highly selective in their catalytic action compared to organic or inorganic catalysts. Many of these enzymatic reactions are conducted at moderate temperatures and frequently in the absence of organic solvents. Immobilized lipases are readily available from commercial sources, which make the application of enzymic processes in the synthesis of acylglycerol molecules attractive, reliable and repeatable. Successful enzymatic methods have been developed to produce structured lipids. “Structured lipids” are triacylglycerols that have been modified by incorporation of specific fatty acids to alter the fatty acid profile of natural triacylglycerols, by restructuring triacylglycerols through changing the positions of the acyl moieties within the same lipid molecule, or by enzymatic synthesis to yield a novel triacylglycerol. Structured lipids are modified from their natural state to achieve a desired nutritional, physical and chemical outcome. All these reactions involve a biocatalyst (lipase). Gandhi (1997), Lee and Akoh. (1998), Akoh (2001, 2002a, 2002b), Osborn and Akoh (2002), Iwasaki and Yamana (2000), Kim and Yoon (2003), Willis et al. (1998), McNeill (1999), and Xu (2000) have reviewed the synthesis and applications of structured lipids with medical, nutraceutical, and food applications. Most common vegetable seed oils are composed mainly of saturated (palmitic and stearic acid), monounsaturated (oleic), and diunsaturated (linoleic acid) fatty acids. In view of the highly beneficial effects of polyunsaturated fatty acids (PUFA) to health, especially eicosapentaenoic (EPA), docosahexaenoic acid (DHA), arachidonic acid (AA), conjugated linoleic acid (CLA), α-linolenic acid (ALA), and γ-linolenic acid (GLA), great efforts have been directed to design enzymatic procedures to incorporate such polyunsaturated fatty acids into traditional triacylglycerols to yield nutraceuticals for supplementation in infant formula (Udell et al., 2005) or as food supplement for adults to enhance overall health (Gandhi, 1997). These designer lipids may replace conventional fats and oils in certain specialty applications because of their role as structure-health (nutraceuticals or medical lipids) and structure-function (functional lipids) attributes. The position of the highly unsaturated fatty acid in the glycerol moiety is key to their
Triacylglycerols
5.4.3.5.1 Monoacid These triacylglycerols are readily prepared by reaction of glycerol with an appropriate acid chloride. Purification can be effected by crystallization or by TLC. They are often commercially available. 5.4.3.5.2 Diacid The general principle of the synthesis of racemic diacid triacylglycerols is to react an appropriate monoacylglycerol or diacylglycerol with the desired acid chloride. Jensen and Pitas (1976) have detailed the formation of glycerol 1-palmitate2,3-distearate; glycerol 1-palmitate-2,3-dioleate; glycerol 1palmitate-2,3-dilinoleate; glycerol 1-oleate-2,3-dipalmitate; and glycerol 2-palmitate-1,3-dioleate. For the acylation, additional amounts (50% molar excess) of pyridine and acid chloride are beneficial. The reaction takes from hours to days at room temperature. The triacylglycerol product can be crystallized with good purity in some cases, but in others it is necessary to use an alumina column in addition. Jensen and Pitas (1976) emphasize that, for the latter to be used efficiently, the triacylglycerol must be soluble in hexane-anhydrous diethyl ether. Many saturated triacylglycerols are not soluble in this mixed solvent and so have to be purified by crystallization alone. When the latter is used, then ethanol is included in the crystallization solvent so that excess acid chloride is converted to the ester, which remains in solution. When alumina columns are employed, care must be taken to eliminate all traces of alcohol, chloroform, or water from the eluting solvents, since these can ruin the separation. Alumina columns can be run rapidly and recoveries are 80 to 90% (Jensen et al., 1966). Diacid triacylglycerols containing one acid in the 1-position and another in the 2- and 3-positions are prepared by acylation of the appropriate 1-mono-acylglycerol. When the single acid is at the 2-position, then the 1,3-diacylglycerol is acylated. The purity of the triacylglycerols is determined by TLC and the proper ratio of the acyl components can be established by GLC or determined by 13C NMR spectroscopy (Lie Ken Jie et al., 1996a, 1996b). 5.4.3.5.3 Triacid Jensen and Pitas (1976) have described the preparation of 1-palmitoyl-2-oleoyl-3-stearoylglycerol. The general principle is the same as for the synthesis of diacid triacylglycerols. In this case glycerol 1-palmitate-3-stearate is reacted with oleoyl chloride. After crystallization and purification on an alumina column, the triacylglcerol was recovered at better than 90% purity. By an appropriate selection of 1,3diacylglycerols it is possible to prepare triacid triacylglycerols with any desired combination of acids.
5.4.4
Acylglycerols by enzymic processes
There are few lipid molecules, if any, which cannot be made by modern organic synthesis. Many common and 373
5.4 Synthesis of acylglycerols
diacyglycerol (6%) to give monoacylglycerol (90%) (Watanabe et al., 2005b). Enzymatic synthesis of symmetrical 1,3-diacylglycerols by direct esterification of glycerol is exemplified by the reaction of oleic acid and glycerol in a solvent-free system with dehydration (3 mm Hg pressure), and Lipozyme (Rhizopus miehei lipase) at 25 to 45oC to give 1,3-diolein (61%) (Rosu et al., 1999). Using Novozym 435 and Lipase PS-D, Kristensen et al. (2005) prepared diacylglycerols in 60% yield from triacylglycerols by enzymatic glycerolysis. Lipase from Rhizopus delemar hydrolyzes common saturated and unsaturated acyl moieties, which are located exclusively at the sn-1 and sn-3 positions of triacylglycerols (Kosugi et al., 2004). To prepare structured triacylglycerols containing eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) at the 1-/3- or position of the glycerol “backbone” of a triacylglycerol molecule, a chemo-enzymatic synthesis method for glycerol 2-EPA/DHA-1,3-distearate and by a two-step enzymatic approach for glycerol 1-EPA/DHA-2EPA/DHA-3-stearate was designed (Haraldsson et al., 2000) (Scheme 5.4.4.1(i)).
functionality in foods and absorption when consumed. In most cases insertion of the desired highly unsaturated fatty acid at the sn-2 position of the triacylglycerol molecule will provide maximum nutritional benefits. The best studied enzymes for lipid synthesis is the lipase family (triacylglycerol acylhydrolase, EC 3.1.1.3), which catalyze the hydrolysis of triacylglycerol to give mixtures of free fatty acids, diacylglycerols, monoacylglycerol, and glycerol when excess water is present. At low water activity, fatty acid and alcohols are esterified in the presence of lipases as exemplified by the esterification of glycerol and polyunsaturated fatty acids (PUFA) from fish oil. The main factors influencing the degree of esterification and triacylglycerol yield are the amount of enzyme, water content, temperature, and glycerol/fatty acid ratio. An example of reaction conditions established follows: 100 mg of Novozym 435 (from Candida antartica), 9 ml hexane, 50oC, glycerol/PUFA concentrate molar ratio 1.2:3 in the absence of water, 1 g molecular sieves added at the start of the reaction, and agitation at a rate of 200 rpm. Under these conditions, a triacylglycerol yield of 93.5% was obtained from cod liver oil polyunsaturated fatty acid concentrate. The product contained 25% eicospentaenoic acid (EPA) and 45% docosahexaenoic acid (DHA) (Robles Medina et al, 1999). 5.4.4.1
5.4.4.2
Esterification of fatty acids by lipases
Enzymatic enrichment of special polyunsaturated fatty acids
Fish oil contains about 10 to 15% of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA). Enzymatic enrichment procedures have been developed to enable these health benefiting fatty acids to be concentrated or isolated. The fish oil triacylglycerols are normally hydrolyzed to the free fatty acids. Lipases are then used to selectively esterify the common fatty acids (saturated, monoene, and diene fatty acids) with alcohols (methanol, ethanol, butanol, or lauryl alcohol) to give the corresponding esters, which can be readily isolated by solvent extraction (using n-hexane or other nonpolar organic solvents) leaving the free fatty acid fraction high in the desired polyunsaturated fatty acids (such as EPA and DHA).
Acid oil (a mixture of mainly free fatty acids and some triacylglycerols, FFA/TAG) is a by-product in the neutralization step of vegetable oil refining and an alternative source of biodiesel fuel. Watanabe et al. (2005a) describes a two-step enzymatic reaction of acid oil (FFA/ TAG, 1:1, wt/wt) with methanol: (1) FFA/TAG, 10% wt methanol, and 0.5% wt immobilized lipase (Candida antarctica) at 30oC for 24 h resulted in the esterification of FFA; (2) the dehydrated reaction mixture, and 5% wt methanol using 6% wt immobilized lipase furnished methyl esters of fatty acids. The activity of the enzyme reached a maximum after six cycles and furnished >98.5% of methyl esters (biodiesel). A continuous production process involving a three-step flow reaction for the conversion of vegetable oil to 93% yield of methyl esters (biodiesel) using immobilized C. antarctica lipase was developed (Watanabe et al., 2000). Immobilized lipase from Rhizomucor miehei is nonselective, but a lipase from Geotrichum candidum esterified the octadeca-9-cis,11-trans-dienoic acid (9c,11t-CLA) selectively (91%) leaving an unreacted free fatty acid fraction consisting of 82% of octadeca-10-trans,12-cis-dienoic acid (10t,12c-CLA) (McNeill et al., 1999). Esterification of a mixture of (9c,11t-CLA) and (10t,12c-CLA) with glycerol using lipase from Penicillium camembertii without dehydration gave equal amounts of the corresponding monoacyl- and diacylglycerols. Esterification with dehydration at 5mm Hg reduced pressure not only achieved a high degree of esterification (95%) but also suppressed the formation of
OH OH OH
(i), (ii)
18:0 OH 18:0
(iii), (iv)
18:0 20:5/22:6 18:0
(a)
OH OH OH
(i), (v)
18:0 18:0 18:0
(iii), (vi)
20:5/22:6 18:0 18:0
(b)
SCHEME 5.4.4.1 (i) Synthesis of structured triacylglycerols containing EPA/DHA at sn-specific positions of the triacylglycerol. (Adapted from Haraldsson, G.G. et al. (2000) J. Am. Oil Chem. Soc., 77, 1139–1145.) Reagents: i, stearic acid; ii, Rhizomucor miehei lipase; iii, eicosapentaenoic (EPA) or docosahexaenoic acid (DHA) acid; iv, 1-(3-dimethylamino-propyl)-3-ethylcarbodiimide; 4-dimethylaminopyridine (DMAP); v, Candida antarctica lipase; vi, Rhizomucor miehei lipase.
374
Synthesis
Ethyl 20:5
(i)
20:5 (FFA)
(ii)
20:5 20:5 20:5
(iii)
8:0 20:5 8:0
SCHEME 5.4.4.3 (i) Enzymatic synthesis of glycerol 1,3-dioctanoate-2-eicosapentaenoate. (Adapted from Iwasaki, Y. et al. (2000) J. Am. Oil Chem. Soc., 77, 501–506.) Reagent: (i) Novozyme, H2O; (ii) Novozyme, glycerol; (iii) ethyl octanoate, Lipozyme.
which gave 65% of GLA in their acylglycerols (Huang et al., 1997). GLA can also be purified from borage oil (GLA content 22%) in two steps by hydrolysis with Pseudomonas sp. lipase (Liposam) to yield 91% of free fatty acids (GLA content 22.5%), which on selective esterification (74% yield) with lauryl alcohol by using Rhizopus delemar lipase gave the corresponding lauryl ester (GLA content 70%). To further elevate the GLA content, unesterified fatty acids were extracted and esterified in the same manner. GLA was purified to 93.7% with a recovery of 67% of its initial content (Shimada et al., 1997b). By a similar approach, DHA was purified from tuna oil using Pseudomonas sp. Lipase (Lipase-AK) was used in the hydrolysis step, while Pseudomonas delemar lipase was used for the esterification step. The DHA content was raised from 24 to 72% (yield of 83%) (Shimada et al., 1997c). Glycerol 1,3-dioctanoate -2-eicosapentaenoate was synthesisedsynthesised by interesterification of glycerol trieicosapentaenoate with ethyl octanoate catalyzed by Novozyme and Lipozyme (Iwasaki et al., 2000, Kawashima et al., 2001; Irimescu et al., 2001) (Scheme 5.4.4.3(i)). Similar approaches were adopted for the synthesis of glycerol 1,3-dioctanoate-2-γ-linolenate from borage oil using Candida rugosa lipase and Rhizopus oryzae lipase (Kawashima et al., 2002). Also prepared were glycerol 1,3dilaurate-2-oleate using Lipozyme IM (Miura et al., 1999) and glycerol 1,3-dioleate-2-palmitate using Novozym 435 under reduced pressure (Chen et al., 2004). Lipase B (GCB) produced by the fungus Geotrichum candidum CMICC 335426 is known for its high specificity towards 9-cis unsaturated fatty acids. When sunflower oil was hydrolyzed in the presence of G. candidum lipase, the level of unsaturates (all of which have a 9-cis double bond) in the free fatty acid fraction was >99% w/w (Diks and Lee, 1999). (See References for Section 5.4 at the end of Section 5.5.)
Esterification of the fatty acids in sardine cannery effluent with 1-butanol with n-hexane in the presence of lipozyme allowed up to 80% DHA enrichment (Schmitt-Rozieres et al., 2000). Pseudomonas lipases have high activity toward the saturated and monounsaturated fatty acids in fish oil and low activity toward EPA and DHA (with double bonds close to the acid/ester function). Thus, treatment of fish oil in anhydrous ethanol in the presence of Pseudomonas lipase resulted in the production of the ethyl esters of the saturated and monounsaturated fatty acids leaving EPA and DHA (total of 50%) unreacted in the residual mixture as mono-, di- and triacylglycerols (Haraldsson, et al., 1997). Using the lipase Rhizopus delemar to esterify the fatty acids in hydrolysed tuna fish oil (containing 23% of DHA) with lauryl alcohol, the process allowed 84% DHA to be recovered from the unesterified fatty acid fraction (Shimada et al., 1997a). A two-step selective esterification of CLA isomers with lauryl alcohol and using Candida rugosa lipase that acted on 9c,11t-CLA more strongly than on 10t,12c-CLA, was reported by Nagao et al. (2002). The 9c,11t-CLA content as the lauryl ester increased to 73% with 79% recovery. 5.4.4.3
Incorporation of special fatty acids to triacylglycerols
Enzymatic transesterification (interesterification) has been widely used to incorporate health benefiting fatty acids into conventional (common) triacylglycerol molecules (Hunter, 2001). Lipase regioselectivity provides the ability to distinguish between the sn-1,3 and sn-2 position of the triacylglycerol molecule, which is important in the manufacture of structured lipids. Hence, taking advantage of the selectivity and specificity of lipases in their hydrolysis or esterification behaviour, numerous structured lipids have been produced. Two immobilized lipases, IM60 (from Rhizopus miehei) and QLM (from Alcaligenes sp.), were used for the modification of the fatty acid composition of palm oil by incorporating EPA and DHA. Acidolysis and interesterification reactions were conducted in hexane. After a 24 h incubation in hexane, there was an average incorporation of 20.8% EPA and 15.6% DHA into palm oil. QLM discriminates against EPA (Fajardo et al., 2003). γ-Linolenic acid (GLA) in borage oil was enriched by a three-step process: (1) selective hydrolysis in iso-octane by Candida rugosa lipase immobilized on microporous polypropylene, (2) selective esterification of borage fatty acids with 1-butanol by Lipozyme IM-20, and (3) acidolysis of the products of the previous reactions, that is, unhydrolyzed acylglycerols and unesterified free fatty acid,
5.5
Fullerene lipids
During experiments aimed at understanding the mechanisms by which long-chain carbon molecules are formed in interstellar space, Kroto and coworkers (Kroto et al., 1985) vapourized graphite by laser irradiation under a helium atmosphere. A remarkable stable carbon cluster (in trace amount in the gas phase) was produced, which was shown by time-of-flight mass spectrometry to consist
375
5.5
Fullerene lipids
The unique properties of fullerene and their derivatives (especially the hydrophobic nature, the spheroid structure, and the electronic properties) have attracted the attention of many life scientists, material technologists, and that of the pharmaceutical industry. In order to perform systematic testing in vivo or in vitro, the hydrophobic nature of fullerenes was overcome by encapsulation in special carriers (Jin et al., 1996; Cataldo, 2002), through suspension of fullerene derivatives in water or co-solvents (Mchedlov-Petrossyan et al., 1997; Scrivens et al., 1994) or by chemical functionalization through introduction of hydrophilic appendages to the fullerene moiety (Meier and Kiegiel, 2001; Isobe et al., 2003; Zhang et al., 2004; Cerar and Skerjanc, 2003; Goswami et al., 2004; Liu et al., 2005; Bar-Shir et al., 2005). Synthetic lipid bilayer membranes exhibit similar physicochemical properties to those of bio-membranes and can be immobilized as molecular lipid films (Fendler, 1994; Kuo and O’Brien, 1988). A triple-C 18 fatty acid chain lipid containing a [60]fullerene moiety, forming multibilayer membrane films that possess main and subphase transitions with the subtransition regulating the electronic properties of the fullerene, was studied (Murakami et al., 1996; Nakanishi et al., 2002). Analogues of the same fullerene lipid containing different saturated and unsaturated fatty acid chains were also prepared (Lie Ken Jie and Cheung, 1999) to study the NMR properties (Scheme 5.5(i)). Fullerene and its derivatives possess properties that hint at their use in biomedicine, such as the inhibition of HIV protease (HIV-P) by the C60 carbon cage that fits into the catalytic site of the proteases (Sijbesma et al., 1993; Friedman et al., 1993). The fullerene derivative (8) used was linked via a cyclopropane system to malonic acid to improve its water solubility.
of 60 carbon atoms. The structure for this extremely stable allotrope of carbon was predicted and subsequently proven as consisting of a truncated icosahedron, a polygon with 60 vertices and 32 faces, 12 of which are pentagonal and 20 hexagonal. The shape of this compound resembled a football or soccer ball (7).
(7)
As the naming of such a 60-carbon cluster would fill four lines according to the IUPAC nomenclature system, the discoverers coined the name “Buckminsterfullerene” for this type of carbon molecules. The name for this molecule stemmed from the resemblance of its shape to the dome structure built for the 1967 Montreal Exposition by the famous architect Buckminster Fuller. In current chemical literature, these carbon cage compounds are simply referred to as “fullerenes.” To the laymen the word “bucky ball” has also been used to describe such molecules. To specify the number of carbon atoms in the cage structure of a fullerene molecule, a number inside the square bracket, for example, “[60]fullerene,” denotes the number of carbon atoms (in this case 60) involved. The [60]fullerene is also referred to as “C60 fullerene.” Fullerenes could have different sizes and consist of more than 60 carbon atoms. By mid 1990, a remarkable simple technique was developed for the production of a black soot by vaporizing inexpensive graphite electrodes in an atmosphere of helium (Krätschmer et al., 1990). The black soot contained about 5 to 10% of a mixture of fullerenes, consisting of C60 (about 80%), C70 (about 15%), and higher fullerenes (C76, C78, C84, etc., about 5%). Recent experiments show the presence of giant fullerenes (C540, C960 and C1500) that could form by coalescence during condensation and soot agglomeration (Rietmeijer et al., 2004) and a reported giant C 6 5 0 allotrope (Homann, 19 9 8) . [ 60 ]F u l le re ne and [70]fullerene are commercially available at reasonable prices. In recognition of the discovery and the pioneering work on the physics and chemistry of this unusual class of carbon allotropes, the 1996 Nobel Prize in Chemistry was awarded to Kroto, Smalley, and Curl. Many books have been published during the past 20 years on the chemistry and physics of fullerenes (Kroto et al., 1993; Kadish and Ruoff, 2000; Guldi and Martin, 2002; Prassides, 2004; Hirsch and Brettreich, 2005). The chemistry of higher fullerenes (C70 and beyond) was reviewed by Thilgen et al. (1997).
O HO
O OH
(8)
Utilizing the two carboxylic groups on the cyclopropane ring structure in compound 8, Lie Ken Jie and Cheung (1998) esterified the carboxylic acid groups with long chain alcohols (from saturated, olefinic, and acetylenic fatty acids) to give a comprehensive series of fullerene lipids (viz. dialkyl 1,2-[6,6]-methano[60]-fullerene dicarboxylate derivatives) (Scheme 5.5(ii)). 376
Synthesis
CH2OH
CH2OH H2N
i
N
CH2OH
OH CH2OH
CH2OH
CH2OCOR ii
N
CH2OCOR
iii, iv
CH2OCOR
CH2OCOR
H2N
OH CH2OCOR
CH2OCOR
CH2OCOR
O Cl
v
N H
CH2OH
CH2OCOR
O N3
vi
CH2OCOR
CH2OCOR
N H
CH2OCOR
CH2OCOR
CH2OCOR
O N
N H
vii
CH2OCOR CH2OCOR
R = (CH2)10CH3, (CH2)12CH3, (CH2)14CH3, (CH2)16CH3 (CH2)7CH=CH(CH2)7CH3, (CH2)7CH=CHCH2CH=CH (CH2)4CH3 or (CH2)7C≡C(CH2)7CH3
SCHEME 5.5 (i) Synthesis of a triple-acyl chain [60]fullerene lipid. (From Lie Ken Jie, M.S.F. and Cheung, S.W.H. (1999) Lipids, 34, 1223–1230. With permission.) Reagents: i, 2-hydroxynaphthaldeyde; ii, RCOOH, DCC, DMAP (DCC = 1,1'-dicyclohexyl carbodiimide; DMAP = 4-dimethylamino pyridine); iii, bromoacetic acid, THF; iv, n-butylamine; v, chloroacetyl chloride, Et3N; vi, NaN3; vii, C60.
O RO
O
O
O
RO
OR
O ii
i OR
O
RO
OR Br
H
R = (CH2)nCH3 where n = 1,3,5,7,8,9,17 (CH2)9CH=CH2, (CH2)5CH=CH(CH2)10CH3, (CH2)8CH=CH(CH2)7CH3, (CH2)8CH=CHCH2CH=CH (CH2)4CH3, (CH2)5C≡C(CH2)10CH3 (CH2)8C≡C(CH2)7CH3 or (CH2)12C≡C(CH2)7CH3
SCHEME 5.5 (ii) Synthesis of dialkyl 1,2-[6,6]-methano-[60]-fullerene dicarboxylate derivatives. (From Lie Ken Jie, M.S.F. and Cheung, S.W.H. (1998) Lipids, 33, 729–732. With permission.) Reagents: i, Br2; ii, NaH, C60, toluene.
377
5.5
Fullerene lipids
O
AcO O
O
O
O O
OH
O O
O
N H
OH O
O Ph
NH Ph
O
OAc Ph
O
(9)
N
N H H N
O
N
O
O
N
O
O
O
N H
O
O
N
H N O
O
(10) HO O
O
OH O
O
HO O O
OH O
HO
OH HN
HO H N
O
O OH
O O
H N
O
HO O
O
OH
H N
O
O
O OH
O
O
NH
OH
HN
O HO
O
H N
O
O
NH
O O
O OH
OH
HO
O HO
O
OH
(11)
are also produced when the carbene anion of various malonate esters are reacted with [60]fullerene to give hexakisadducts (Hirsch and Vostrowsky, 2001; Djojo et al., 2000, Felder-Flesch et al., 2005).
An interesting and potential lipophilic chemotherapeutic has been reported, which makes use of the carboxylic acid function of compound 8 to incorporate an anticancer drug, paclitaxel, in an effort to produce a slow-release drug-delivery system to combat lung cancer (compound 9) (Zakharian et al., 2005). Two pyropheophorbide a moieties have also been linked to the two carboxylic acid functions of compound 8 to yield a sensitizer (compound 10) for photodynamic therapy (Rancan et al., 2005). Dendritic methano[60]fullerene octadeca acid (compound 11) shows evidence for aggregation of the acid form in water. It has a high potential of scavenging deleterious radicals in biological systems (Quaranta et al., 2003; Krusic et al., 1991). This molecule exhibits antiviral activity against HIV-infected human lymphocytes in vitro in the micromolar range with no cell toxicity. Multiadducts
HOOC
COOH
HOOC HOOC HOOC
COOH
(12)
[60]Fullerene containing three methano-dicarboxylic acids (compound 12) was found to block apoptotic
378
Synthesis
signaling of transforming growth factor-β in human hepatoma cells (Huang et al., 1998). Reactions of various saturated ω-azido fatty esters with [60]fullerene gave fullerene lipids with a [5,6]-open substructure in the fullerene cage (Lie Ken Jie et al., 2001a). However, when methyl 11-azido-7-undecynoate was reacted with [60]fullerene, a mixture of [6,6]-closed substructure (13) and [5,6]-open type (14) in the fullerene cage was obtained (Scheme 5.5(iii)). Several structured rac-triacylglycerols containing an acyl group with a [60]fullerene tethered to the distal end of the acyl chain at either the 1/3-acyl (Scheme 5.5(iv)) or 2-acyl position were synthesised (Lie Ken Jie et al., 2001b) (Scheme 5.5(v)).
N
N
(14)
(13)
N R
i
O N
OMe
SCHEME 5.5 (iii) Synthesis of nitrogen-bridged [60]fullerene fatty ester derivatives. (From Lie Ken Jie, M.S.F. et al. (2001a) Lipids, 36, 421–426. With permission.) Reagents: i, N3R where R = (CH2)5COOMe; (CH2)7COOMe; (CH2)11COOMe; N3(CH2)3C≡C(CH2)5COOMe.
ii
i OH
iv, v
OH
OTHP
OCOR ROCO OTHP
OTHP
OCOR
vi
ROCO O
OCOR ROCO O
Br
N3 O
O
vii
iii
HO
OCOR ROCO O
N O
R = (CH2)16CH3, (CH2)7CH=CH(CH2)7CH3, (CH2)7CH=CHCH2CH=CH (CH2)4CH3 or (CH2)7C C(CH2)7CH3
SCHEME 5.5 (iv) Synthesis of structured triacylglycerols containing an aza-[60]fullerene unit at the 1/3-acyl position. (From Lie Ken Jie, M.S.F. et al. (2001b) Lipids, 36, 649–654. With permission.) Reagents: i, dihydropyran, pyridinium, p-toluenesulfonate (PPTS), CH2Cl2; ii, KMnO4, acetic anhydride, H2O; iii, RCOOH, DCC (1,1'-dicyclohexylcarbodiimide), DMAP (4-dimethylaminopyridine), CH2Cl2; iv, PPTS, EtOH; v, Br(CH2)5COOH, DCC, DMAP; vi, NaN3/DMF; vii, C60, toluene.
379
5.5
Fullerene lipids
OH O
i
OCOR
OH
iii
HO
OCOR
O Br
OCOR
ii
O
OCOR
OCOR
O
iv
N3
O OCOR
v
OCOR
O N
OCOR O
OCOR O OCOR
R = (CH2)16CH3, (CH2)7CH=CH(CH2)7CH3, (CH2)7CH=CHCH2CH=CH (CH2)4CH3 or (CH2)7C≡C(CH2)7CH3
SCHEME 5.5 (v) Synthesis of structured triacylglycerols containing an aza-[60]fullerene unit at the 2-acyl position. (From Lie Ken Jie, M.S.F. et al. (2001b) Lipids, 36, 649–654. With permission.) Reagents: i, RCOOH, DCC, DMAP; ii, NaBH4/THF, H2O; iii, Br(CH2)5COOH, DCC, DMAP (DCC = 1,1'-dicyclohexyl carbodiimide; DMAP = 4-dimethylamino pyridine); iv, NaN3/DMF; v, C60, toluene.
Barry, P.J. and Craig, B.M. (1955) Synthesis of symmetrical diglycerides from dihydroxy acetone and allyl alcohol. Can. J. Chem., 33, 716–721. Bar-Shir, A. et al. (2005) Synthesis and water solubility of adamantyl-OEG-fullerene hybrids. J. Org. Chem., 70, 2660–2666. Barve, J.A. and Gunstone, F.D. (1971) The synthesis of all the octadecynoic acids and all the trans-octadecenoic acids. Chem. Phys. Lipids, 7, 311–323. Batten, R.J. et al. (1980) A new method for removing the t-butyldimethylsilyl protecting group. Synthesis, 3, 234–236. Bhati, A. et al. (1980) Prospects and retrospects of glyceride synthesis, in Fats and Oils: Chemistry and Technology, Eds, R.J. Hamilton and A. Bhati), Applied Science, London, pp. 59–107. Beerthuis, R.K. et al. (1971) Synthesis of a series of polyunsaturated fatty acids, their potencies as essential fatty acids and as precursors of prostaglandins. Recl. Trav. Chim. Pays-Bas, 90, 943–960. Bell, S.J. et al. (1997) The new dietary fats in health and disease. J. Am. Diet Assoc., 97, 280–286. Belosludtsev, Y.Y. et al. (1986) A new synthesis of 5,8,11,14eicosatetraynoic acid. Bioorg. Khim., 12, 1425–1426. Bergelson, L.D. and Shemyakin, M.M. (1964) Synthesis of naturally occurring unsaturated fatty acids by sterically controlled carbonyl olefination. Angew. Chem. Int. Ed., 3, 250–260. Bestmann, H.J. (1965) New reactions of alkylidenephosphoranes and their preparative uses. 1. Acid-base character of phosphonium salts and alkylidenephosphoranes. Angew. Chem. Int. Ed., 4, 583–587. Bestmann, H.J. and Stransky, W. (1974) Reaktionen mit Phosphinalkylenen: XXXII. Herstellung und Reaktionen von Alkylidenphosphoranen (Phosphinalkylenen) in Hexamethylphosphorsauretriamid unter Verwendung von OP[N(CH3)2]2- and N(CH3)2- als Base. Synthesis, 798–800.
References Aasen, A.J. et al. (1970) Mass spectrometry of triglycerides: II. Specifically deuterated triglycerides and elucidation of fragmentation mechanisms. Lipids, 5, 869–877. Adlof, R. (1997) Preparation of methyl cis-9, trans-11- and trans9, trans-11-octadecadienoate-17,17,18,18-d4, two of the isomers of conjugated linoleic acid. Chem. Phys. Lipids, 88, 107–112. Adlof, R.O. (1999) Isotopically labelled fatty acids, in Lipid Synthesis and Manufacture, Ed. F.D. Gunstone, Sheffield Academic press, Sheffield, U.K., 46–93. Akoh, C. (1995) Lipid-based fat substitutes. Crit. Rev. Food Sci. Nutr., 35, 405–430. Akoh, C.C. (2001) Structured lipids containing omega-3 highly unsaturated fatty acids. ACS Symposium Series, 788 (Omega-3 fatty acids), 151–161. Akoh, C.C. (2002a) Structured lipids. Food Sc. Technol. 117, 877–908. Akoh, C.C. (2002b) Structured lipids, in Food Lipids, Chemistry, Nutrition, and Biotechnology, Eds. C.C. Akoh and D.B. Min, Marcel Dekker, New York, pp. 877–909. Akoh, C.C. Ed. (2005) Handbook of Functional Lipids, Taylor & Francis Group, Boca Raton, FL. Akoh, C.C. and Lai, O.M. Eds. (2005) Healthful Lipids, Am. Oil Chem. Soc., Champaign, IL. Akoh, C.C. and Min, D.B. Eds. Food Lipids, Chemistry, Nutrition, and Biotechnology (2002) Marcel Dekker, New York. Anagnostopoulos, D. et al. (2005) Synthesis of all-trans arachidonic acid and its effect on rabbit platelet aggregation. Bioorg. Med. Chem. Lett., 15, 2766–2770. Barley, G.C. et al. (1973) Natural acetylenes: Part XXXIX. Synthesis of methyl [1,9-14 C]-, [9-14 C]- and [10-3 H]crepenynate, methyl [1,9-14C]-, and [10-3H]-linoleate and methyl [9-14C]- and [10-3H]-oleate. J. Chem. Soc. (Perkin I), 151–154.
380
Synthesis
Bestmann, H.J. et al. (1976) Darstellung lithiumsalzfreier Ylidlosungen mit Natrium-bis (trimethylsilyl) amid als Base. Chem. Ber., 109, 1694–1700. Bestmann, H.J. and Vostrowsky, O. (1979) Synthesis of pheromones by stereoselective carbonyl olefination: a unitised construction principle. Chem. Phys. Lipids, 24, 335–389. Boswinkel, G. et al. (1996) Kinetics of acyl migration in monoglycerides and dependence on acyl chain length. J. Am. Oil Chem. Soc., 73, 707–711. Bradshaw, R.W. et al. (1971) Natural acetylenes. A synthesis of crepenynic acid (octadec-9-en-12-ynoic acid). J. Chem. Soc., C, 1156–1158. Broustal, G. and Loreau,. O. (2004) Stereocontrolled synthesis of (7E,9Z)-[9,10-2H]-conjugated linoleic acid. J. Labelled Cpd. Radiopharm., 47, 875–880.
Lipoprotein Analyses Eds. J.L. Sébédio and E.G. Perkins, AOCS Press, Champaign, IL, 299–308. Diks, R.M.M. and Lee, M.J. (1999) Production of a very low saturated oil based on the specificity of Geotrichum candidum lipase. J. Am. Oil Chem. Soc., 76, 455–462. Dinh-Nguyen, N. et al. (1972) Perdeuteriated organic compounds. I. Normal long-chain saturated deuteriocarbons, monocarboxylic acids and methyl esters. Chem. Scr., 2, 171–178. Djojo, F. et al. (2000) Fullerene dendrimers and lipofullerenes with an inherently chiral hexaaddition pattern. Eur. J. Org. Chem., 1051–1059. Dorset, D.L. (1987) Is the initial solid state acyl shift of 1,2diglyceride intermolecular? Chem. Phys. Lipids, 43, 179–191. Dorset, D.L. and Pangborn, W.A. (1979) Solid state acyl shift of diglycerides: electron-diffraction study. Chem. Phys. Lipids, 25, 179–189. Dussault, P. and Lee, I.Q. (1995) A chemoenzymatic approach to hydroperoxy-eicosatetraenoic acids. Total synthesis of 5(S)-HPETE. J. Org. Chem., 60, 218–226. Ege, S.N. et al. (1961) Synthesis of methyl 5,8,11,14-eicosatetraenoate (methyl arachidonate). J. Am. Chem. Soc., 83, 3080–3085. Eibl, H. (1981) An improved method for the preparation of 1,2isopropylidene-sn-glycerol. Chem. Phys. Lipids, 28, 1–5. Eibl, H. and Woolley, P. (1986) Synthesis of enantiomerically pure glyceryl esters and ethers. I. Methods employing the precursor 1,2-isopropylidene-sn-glycerol. Chem. Phys. Lipids, 41, 53–63. Emken, E.A. (1978) Synthesis and analysis of stable isotope- and radioisotope-labelled fatty acids, in Handbook of Lipid Research — Fatty Acids and Glycerides, Ed. A. Kuksis, Plenum, New York, pp. 77–121. Eynard, T. et al. (1998) Synthesis of methyl (5Z,8Z,11Z,14Z,17Z)and (5Z,8Z,11Z,14Z,17E)-[18-14C] eicosapentaenoate. J. Labelled Cpd. Radiopharm, XLI, 411–421. Fajardo, A.R. et al. (2003) Lipase-catalyzed incorporation of n-3-PUFA into palm oil. J. Am. Oil Chem. Soc., 80, 1197–1200. Felder-Flesch, D. et al. (2005) Interfacial behavior and filmforming properties of an amphiphilic hexasubstituted [60]fullerene. Tetrahedron Lett., 46, 6507–6510. Fendler, J.H. (1994) Membrane — Mimetic Approach to Advanced Materials. Advances in Polymer Science, Series 113, Springer-Verlag, Berlin. Fischer, H.O.L. and Baer, E. (1937) Synthese optisch-activer Glyceride. Naturwissenschaften, 25, 588–589. Friedman, S.H. et al. (1993) Inhibition of the HIV-1 protease by fullerene derivatives: model building studies and experimental verification. J. Am. Chem. Soc., 115, 6506–6509. Gandhi, N.N. (1997) Applications of lipase. J. Am. Oil Chem. Soc., 74, 621–634. Gigg, J. and Gigg, R. (1967) The preparation of unsymmetrical diglycerides. J. Chem. Soc. C, 431–434. Gilman, N.W. and Holland, B.C. (1974) Synthesis of some acetylenic acids. Chem. Phys. Lipids, 13, 239–248. Goerger, M.M. and Hudson, B.S. (1988) Synthesis of all-transparinaric acid-d8 specifically deuterated at all vinyl positions. J. Org. Chem., 53, 3148–3153.
Brown, C.A. and Ahuja, V.K. (1973) “P2-Nickel” Catalyst with ethylenediamine, a novel system for highly stereospecific reduction of alkynes to cis-olefins. J. Chem. Soc. Chem. Commun.,
553–554. Buchnea, D. (1967) Acyl migration in glycerides. A bimolecular resonant ion complex as intermediate in acyl migration of monoglycerides. Chem. Phys. Lipids, 1, 113–127. Buchnea, D. (1971) Synthesis of C-18 mixed acid diacyl-sn-glycerol enantiomers. Lipids, 6, 734–739. Buchnea, D. (1978) Stereospecific synthesis of enantiomeric acylglycerols, in Handbook of Lipid Research, Vol.1, Ed. A. Kuksis, Plenum, New York, pp. 233–287. Buckman, B.O. et al. (1994) Synthesis and tissue biodistribution of [ω-11C]palmitic acid. A novel PET imaging agent for cardiac fatty acid metabolism. J. Med. Chem., 37, 2481–2485. Burgos, C.E. et al. (1987) A new, asymmetric synthesis of lipids and phospholipids. J. Org. Chem., 52, 4973–4977. Cataldo, F. (2002) Encapsulation of C60 fullerene in γ-cyclodextrin: a new concept in the protection of organic substrates and polymers from ozone attack. Kinetic aspects on the reactivity between C60 and O3. Poly. Degrad. Stabil., 77, 111–120. Cerar, J. and Skerjanc, J. (2003) Water-soluble fullerenes. 3. Alkali salts of fullerenehexamalonic acid Th-C66(COOH)12. J. Phys. Chem. B., 107, 8255–8259. Chen, M.L. et al. (2004) Synthesis of the structured lipid 1,3dioleoyl-2-palmitoylglycerol from palm oil. J. Am. Oil Chem. Soc., 81, 525–532. Christophe, A.B. Ed. (1998) Structural modified food fats: synthesis, biochemistry, and use. Am. Oil Chem. Soc., Champaign, IL. Corey, E.J. et al. (1980) J. Am. Chem. Soc., 102, 1436–1439. Crombie, L. (1996) Synthesis in the isotopic labelling of plant fatty acids: their use in biosynthesis, in Synthesis in Lipid Chemistry, Ed. J.H.P. Tyman, Royal Society of Chemistry, Cambridge, U.K., pp. 34–56. Cunningham, J. and Gigg, R. (1965) Glycerol 1,2-carbonate. J. Chem. Soc., 1553–1554. DasGupta, S.K. et al. (1982) Synthesis of isotopically labelled saturated fatty acids. J. Lipid Res., 23, 197–200. Demmelmair, H. et al. (1997) New insights into lipid and fatty acid metabolism via stable isotopes. Europ. J. Ped., 156, S70–S74. Descomps, B. (1995) Utilization of stable isotopes to study lipid metabolism in humans, in New Trends in Lipid and
381
5.5
Fullerene lipids
Homann, K.H. (1998) Fullerenes and soot formation — new pathways to larger particles in flames. Angew. Chem. Int. Ed. Engl., 37, 2434–2451. Hsiao, C.Y. et al. (1974) Preparation of fully deuterated fatty acids by simple method. Lipids, 9, 913–915. Huang, F.C et al. (1997) Enrichment of γlinolenic acid from borage oil via lipase-catalyzed reactions. J. Am. Oil Chem. Soc., 74, 977–981. Huang, Y.L. et al. (1998) Blockage of apoptotic signaling of transforming growth factor-β in human hepatoma cells by carboxyfullerene. Eur. J. Biochem., 254, 38–43. Huang, Y.S. and Sinclair, A.J. Eds. (1998) Lipids in infant nutrition, Am. Oil Chem. Soc., Champaign, IL. Hunter, J.E. (2001) Studies on effects of dietary fatty acids as related to their position on triglycerides. Lipids, 36, 655–668. Irimescu, R. et al. (2001) Comparison of acyl donors for lipase-catalyzed production of 1,3-dicapryloyl-2eicosapentenoylglycerol. J. Am. Oil Chem. Soc., 78, 65–70. Isobe, H. et al. (2003) Synthesis of fullerene glycoconjugates through sulfide connection in aqueous media. Org. Lett., 5, 4461–4463. Iwama, T. and Foglia, T.A. (1988) An improved procedure for the detritylation of 1-alkyl 2-acyl 3-trityl-sn-glycerols. J. Am. Oil Chem. Soc., 65, 954–958. Iwasaki, Y. et al. (2000) Enzymatic synthesis of 1,3-dicapryloyl2-eicosapentaenoylglycerol. J. Am. Oil Chem. Soc., 77, 501–506. Iwasaki, Y. and Yamana, T (2000) Enzymatic synthesis of structured lipids. J. Mol. Cat. B: Enzym., 10, 129–140. Jeffery, T. et al. (1992) An efficient route to skipped diynes and triynes, (Z,Z) dienes and (Z,Z,Z) trienes, Tetrahedron Lett., 33, 5757–5760. Jensen, R.G. (1972) Synthetic glycerides, in Topics in Lipid Chemistry, 3, 1–35, ELEK Science, London. Jensen, R.G. (1995) Discoveries in synthesizing glycerides. INFORM, 6, 440–445. Jensen, R.G. et al. (1966) Purification of triglycerides with an alumina column. Lipids, 1, 451-452. Jensen, R.G. and Pitas, R.E. (1976) Synthesis of some acylglycerols and phosphoglycerides, in Advances in Lipid Research, Vol. 14, Eds. R. Paoletti and D. Kritchevsky, Academic Press, New York, pp. 213–247. Jin, C.Y. et al. (1996) Studies on novel cyclodextrans: Inclusion of C60 and C70. J. Inclusion Phenom. Mol. Recog. Chem., 24, 301–310. Jung, M.E. and Shaw, J.T. (1980) Total synthesis of (R)-glycerol acetonide and the antiepileptic and hypotensive drug (-)gamma-amino-β-hydroxybutyric acid (GABOB): use of vitamin C as a chiral starting material. J. Am. Chem. Soc., 102, 6304–6311. Jurczak, J. et al. (1986) (R)- and (S)-2,3-O-isopropylidene-glyceraldehyde in stereoselective organic synthesis. Tetrahedron, 42, 447–488. Kadish, K.M. and Ruoff, R.S. (2000) Fullerenes — Chemistry, Physics and Technology, John Wiley & Sons, New York. Kanda, P. and Wells, M.A. (1980) Simplified procedure for the preparation of 2,3-O-isopropylidene-sn-glycerol from Larabinose. J. Lipid Res., 21, 257–258.
Golding, B.T. (1988) Synthesis and reactions of chiral C3-units. Chem. Ind., 19, 617–621. Goswami, T. H. et al. (2004) One-pot synthesis of a novel water-soluble fullerene-core starlike macromolecule via successive Michael and nucleophilic addition reaction. Chem. Mater., 16, 2442–2448. Gravier-Pelletier, C. et al. (1990) Methods for the total synthesis of acyclic hydroxylated fatty-acids. Progr. Lipid Res., 29, 229–276. Greene, T.W. and Wuts, P.G.M (1999) Protective Groups in Organic Synthesis, 3rd ed., John Wiley & Sons, New York. Guldi, D.M. and Martin, N. (2002) Fullerenes: from Synthesis to Optoelectronic Properties, Kluwer Academic Pub., Dordrecht. Gunstone, F.D., Ed. (2001) Structured and modified lipids, Marcel Dekker, New York. Gunstone, F.D., Ed. (2003) Lipids for functional foods and nutraceuticals, The Oily Press, Bridgwater, U.K. Gunstone, F.D. and Hornby, G.M. (1969) The conversion of alkenoic acids to alkynoic acids by bromination-dehydrobromination. Chem. Phys. Lipids, 3, 91–97. Gunstone, F.D. and Ismail, I.A. (1967) The synthesis of all the cis n-octadecenoic acids. Chem. Phys. Lipids, 1, 209–224. Gunstone, F.D. and Lie Ken Jie, M. (1970) The synthesis of ten octadecadiynoic acids and of the related cis,cis- and trans,trans-octadecadienoic acids. Chem. Phys. Lipids, 4, 1–14. Gunstone, F.D. et al. (1971) The synthesis and chromatographic and spectroscopic properties of some methyl cis,cis- dimethylene octadecadienoates. Chem. Phys. Lipids, 6, 147–151. Gunstone, G.D. and Norris, F.A. (1983) Synthesis of acylglycerols, in Lipids in Foods, Pergamon, Oxford, U.K., pp. 85–91. Hall, D.R. et al. (1975) Synthesis of the major sex pheromone of the Egyptian cotton leafworm Spodoptera littoralis (Boisd). Chem. Ind., 216–217. Han, L. and Razdan, R.K. (1998) Total synthesis of methyl 14hydroxy-(all-cis)-5,8,11-tetradecatrienoate: a useful intermediate for the synthesis of arachidonic acid analogues. Tetrahedron Lett., 39, 771–774. Haraldsson, G.G. et al. (2000) Chemoenzymatic synthesis of structure triacylglcyerols containing eicosapentaenoic and docosahexaenoic acid. J. Am. Oil Chem. Soc., 77, 1139–1145. Haraldsson, G.G. et al. (1997) The preparation of concentrates of eicosapentaenoic acid. J. Am. Oil Chem. Soc., 74, 1419–1424. Haumann, B.F. (1997) Structured lipids allow fat tailoring. INFORM, 8, 1004–1011. Heitz, M.P. et al. (1989) Synthesis of all-cis-1-bromo-4,7,10,13nonadecatetraene: A precursor to C-1 labelled arachidonic acid. J. Org. Chem., 54, 500–503. Hermetter, A. and Paltauf, F. (1981) A new method for the detritylation of 1,2-diradyl-3-O-tritylglycerols. Chem. Phys. Lipids, 29, 191–195. Hirsch, A. and Brettreich, M. (2005) Fullerenes — Chemistry and Reactions, Wiley-VCH Verlag, Weinheim. Hirsch, A. and Vostrowsky, O. (2001) C60 Hexakisadducts with an octahedral addition pattern – a new structure motif in organic chemistry. Eur. J. Org. Chem., 829–848.
382
Synthesis
Kawashima, A. et al. (2001) Enzymatic synthesis of highpurity structured lipids with caprylic acid at 1,3-positions and polyunsaturated. J. Am. Oil Chem. Soc., 78, 611–616. Kawashima, A. et al. (2002) Production of structured TAG rich in 1,3-dicapryloyl-2-γ-linolenoyl glycerol from borage oil. J. Am. Oil Chem. Soc., 79, 871–877. Kim, E.J. and Yoon, S.H. (2003) Recent progress in enzymatic production of structured lipids. Food Sci. Biotechnol., 12, 721–726. Klemann, L.P. et al. (1994) Production of fat mixtures enriched with triacylglycerols bearing short, medium and long chain fatty acids. U.S. Patent 5,286,512, Feb. 15, 1994 (Chem. Abstr., 122, 54727y, 1994). Klok, R. et al. (1974) Synthesis of 19-cis-docosenoic, 17-ciseicosenoic and 15-cis-octadecenoic acid. Recl. Trav. Chim. Pays-Bas, 93, 222–224. Kodali, D.R. (1987) Improved method for the synthesis of 1- or 3-acyl-sn-glycerols. J. Lipid Res., 28, 464–469. Kodali, D.R. and Duclos, R.I. Jr. (1992) Debenzylation and detritylation by bromodimethylborane: synthesis of mono-acid or mixed-acid 1,2- or 2,3-diacyl-sn-glycerols. Chem. Phys. Lipids, 61, 169–173. Kosugi, Y. et al. (2004) Use of Rhizopus delemar lipase as compared with other lipases for determination of sn-2 fatty acids in triacylglycerol. J. Am. Oil Chem. Soc., 81, 235–239. Krabisch, L. and Borgstrom, B. (1965) Synthesis of racemic 1,2diolein. J. Lipid Res., 6, 156–157. Krätschmer, W. et al. (1990) Solid C-60 — A new form of carbon. Nature, 347, 354–358. Kristensen, J.B. et al. (2005) Diacylglycerol synthesis by enzymatic glycerolysis: screening of commercially available lipases. J. Am. Oil Chem. Soc., 82, 329–334. Kroto, H.W. et al. (1993) The Fullerenes, Pergamon Press, Oxford, U.K. Kroto, H.W. et al. (1985) C60: Buckminsterfullerene. Nature, 318, 162–163. Krusic, P.J. et al (1991) Radical reactions of C60. Science, 254, 1183–1185. Kunau, W.H. (1971a) Chemical synthesis of highly unsaturated fatty acids. I. Preparation of (n-4), (n-1)-alkadiynoic acids. Chem. Phys. Lipids, 7, 101–107. Kunau, W.H. (1971b) Chemical synthesis of highly unsaturated fatty acids, II. Preparation of substituted propargyl halides. Chem. Phys. Lipids, 7, 108–120. Kunau, W.H. (1973) Synthesis of unsaturated fatty acids. Chem. Phys. Lipids, 11, 254–269. Kuo, T. and O’Brien, D.F. (1988) Free standing polydiacetylene films cast from bilayer membranes. J. Am. Chem. Soc., 110, 7571–7572. Labelle, M. et al. (1990) Synthesis of two analogues of arachidonic acid and their reactions with 12-lipoxygenase. Tetrahedron, 46, 6301–6310. Lam, C.H. and Lie Ken Jie, M.S.F. (1975) Synthesis of all the dimethylene-interrupted methyl octadecadiynoates and a study of their gas-liquid chromatographic properties. J. Chromatogr., 115, 559–570. Lam, C.H. and Lie Ken Jie, M.S.F. (1976) The gas-liquid chromatographic properties of all dimethylene interrupted
methyl cis,cis-octadecadienoates. J. Chromatogr., 117, 365–374. Lee, K.T. and Akoh, C.C. (1998) Structured lipids: synthesis and applications, Food Rev. Inter., 14, 17–34. Lie Ken Jie, M.S.F. and Cheung, S.W.H. (1998) Synthesis of dialkyl 1,2-[6,6]-methano-[60]-fullerene dicarboxylate derivatives. Lipids, 33, 729–732. Lie Ken Jie, M.S.F. and Cheung, S.W.H. (1999) Synthesis of C-60 fullerene derivatives bearing a long-chain saturated or unsaturated triester system. Lipids, 34, 1223–1230. Lie Ken Jie, M.S.F. et al. (2001a) Fullerene lipids: synthesis of novel nitrogen-bridged [60]fullerene fatty ester derivatives. Lipids, 36, 421–426. Lie Ken Jie, M.S.F. et al. (2001b) Fullerenoid lipids: first synthesis of structured triacylglycerols containing an aza[60]fullerene unit. Lipids, 36, 649–654. Lie Ken Jie, M.S.F. and Choi, Y.C. (1992) Mass spectral studies of deuterium-labelled picolinyl fatty esters in the determination of double bond positions. J. Am. Oil Chem. Soc., 69, 1245–1247. Lie Ken Jie, M.S.F. and Kalluri, P. (1998) Ultrasound in fatty acid chemistry: facile dehydrobromination of dibromo fatty esters to acetylenic ester derivatives. Chem. Phys. Lipids, 91, 79–83. Lie Ken Jie, M.S.F. and Lam, C.H. (1974) Synthesis of all the methyl undecynoates and the methyl cis-undecenoates and a study of their gas-liquid chromatographic properties. J. Chromatogr., 97, 165–173. Lie Ken Jie, M.S.F. and Lam, C.C. (1995a) 1H-Nuclear magnetic resonance spectroscopic studies of saturated, acetylenic and ethylenic triacylglycerols. Chem. Phys. Lipids, 77, 155–171. Lie Ken Jie, M.S.F. and Lam, C.C. (1995b) 13C-Nuclear magnetic resonance spectroscopic studies of triacylglycerols of type AAA containing (Z)- and (E)-monoethylenic acyl groups. Chem. Phys. Lipids, 78, 15–27. Lie Ken Jie, M.S.F. and Lam, C.C. (1995c) 13C-NMR studies of polyunsaturated triacylglycerols of type AAA and mixed triacylglycerols containing saturated, acetylenic and ethylenic acyl groups. Chem. Phys. Lipids, 78, 1–13. Lie Ken Jie, M.S.F. et al. (1996a) C-13 nuclear magnetic resonance spectroscopy analysis of the triacylglycerol composition of some margarines. J. Am. Oil Chem. Soc., 73, 1011–1017. Lie Ken Jie, M.S.F. et al. (1996b) C-13 nuclear magnetic resonance spectroscopic analysis of the triacylglycerol composition of Biota orientalis and carrot seed oil. J. Am. Oil Chem. Soc., 73, 557–562. Lie Ken Jie, M.S.F. et al. (1997) Fatty acids, fatty acid analogues and their derivatives. Nat. Prod. Report, 14, 163–190. Lindlar, H. (1952) Ein neuer Katalysator fur selektive Hydrierungen. Helv. Chim. Acta, 35, 446–450. Liu, Y. et al. (2005) Spectrophotometric study of fluorescence sensing and selective binding of biochemical substrates by 2,2'-bridged bis(b-cyclodextrin) and its water-soluble fullerene conjugate. J. Phys. Chem. B., 109, 23739–23744. Lok, C.M. (1978) Versatile methods for the synthesis of mixedacid 1,2-diacylglycerols. Chem. Phys. Lipids, 22, 323–337. Lok, C.M. et al. (1985) Synthesis of glycidol esters and mono/ diacylglycerols from glycidol. Chem. Phys. Lipids, 36, 329–334.
383
5.5
Fullerene lipids
Lok, C.M. et al. (1976) Synthesis of chiral glycerides starting from D- and L-serine. Chem. Phys. Lipids, 16, 115–122. Loreau, O. et al. (2000) Large-scale preparation of (9Z,12E)[1-13C]-octadeca-9,12-dienoic acid, (9Z,12Z,15E)-[113C]-octadeca-9,12,15-trienoic acid and their [1-13C] all cis-isomers. Chem. Phys. Lipids, 106, 65–78. Marvell, E.N. and Li, T. (1973) Catalytic semihydrogenation of the triple bond. Synthesis, 457–468. Mattson, A.W. and Volpenhein, R.A. (1962) Synthesis and properties of glycerides. J. Lipid Res., 3, 281–295. Mattson, F.H. et al. (1964) Esterification of hydroxyl compounds by fatty acid anhydrides. J. Lipid Res., 5, 374–377. Mazur, A.W. et al. (1991) Regio- and stereoselective enzymatic esterification of glycerol and its derivatives. Chem. Phys. Lipids, 60, 189–199. Mchedlov-Petrossyan, N.O. et al. (1997) Colloidal dispersions of fullerene C60 in water: some properties and regularities of coagulation by electrolytes. J. Chem. Soc., Faraday Trans., 93, 4343–4346. McNeill, G.P. (1999) Enzymic processes in Lipid Synthesis and Manufacture, Ed. F.D. Gunstone, Sheffield Academic Press, Sheffield, pp. 288–320. McNeill, G.P. et al. (1999) Enzymatic enrichment of conjugated linoleic acid isomers and incorporation into triglycerides. J. Am. Oil Chem. Soc., 76, 1265–1268. Meier, M. S. and Kiegiel, J. (2001) Preparation and characterization of the fullerene diols 1,2-C60(OH)2, 1,2-C70(OH)2, and 5,6-C70(OH)2. Org. Lett., 3, 1717–1719. Mikkilineni, A.B. et al. (1988) The chemistry of L-ascorbic and D-isoascorbic acids. 2. R and S glyceraldehydes from a common intermediate. J. Org. Chem., 53, 6005–6009. Miura, S et al. (1999) A rapid method for enzymatic synthesis and purification of the structured triacylglycerol, 1,3dilauroyl-2-oleoyl-glycerol. J. Am. Oil Chem. Soc., 76, 927–931. Moss, R.A. et al. (1987) A convenient preparation of 1,2-diacylglycerols; ortho-iodobenzoyl as a protecting group. Tetrahedron Lett., 28, 5005–5008. Murakami, H. et al. (1996) Fullerene lipid chemistry: self-organized multibilayer films of a C60- bearing lipid with main and subphase transition. J. Am. Chem. Soc., 118, 4484–4485. Murphy, P.J. and Lee, S.E. (1999) Recent synthetic applications of the non-classical Wittig reaction. J. Chem. Soc. Perkin Trans., 1, 3049–3066. Nagao, T. et al. (2002) Fractionation and enrichment of CLA isomers by selective esterification with Candida rugosa lipase. J. Am. Oil Chem. Soc., 79, 303–308. Nakanishi, T. et al. (2002) Structure and electrochemistry of selforganized fullerene-lipid bilayer films. Chem. Eur. J., 8, 1641–1648. Neises, B. and Steglich, W. (1978) Simple method for the esterification of carboxylic acids. Angew. Chem. Int. Ed. Engl., 17, 522–524. Noller, C.R. and Bannerot, R.A. (1934) The synthesis of unsaturated fatty acids. Synthesis of oleic and elaidic acids. J. Am. Chem. Soc., 56, 1563–1565. Normant, J.F. and Alexakis, A. (1981) Carbometallation (Cmetallation) of alkynes: stereospecific synthesis of alkenyl derivatives. Synthesis, 841–870.
Nunn, K. et al. (1992) Conjugate reduction vicinal to butadiene tricarbonyl iron complexes. Application to the synthesis of (+/-)-6,7-dihydro-LTB4 methyl ester. J. Org. Chem., 57, 3359–3364. Osbond, J.M. et al. (1961) Essential fatty acids. Synthesis of linoleic, γ-linolenic, arachidonic, and docosa-4,7,10,13,16pentaenoic acid. J. Chem. Soc., 2779–2787. Osborn, H.T. and Akoh, C.C. (2002) Structured lipids — novel fats with medical, nutraceutical, and food applications. Compr. Rev. Food Sci. Food Safety, 1, 93–103. Pabon, H.J.J. et al. (1965) Synthesis of all-cis methyl 3,6,9,12octadecatetraenoate, methyl 4,7,10,13-nonadecatetraenoate and methyl 5,8,11,14,17-eicosapentaenoate. Recl. Trav. Chim. Pays-Bas, 84, 1319–1326. Peng, S. et al. (2002) Synthesis of isotopically labelled arachidonic acids to probe the reaction mechanism of prostaglandin H synthase. J. Am. Chem. Soc., 124, 10786–10796. Pfeiffer, F.R. et al. (1968) Glycerolipids I. Synthesis of D and L mono- and polyunsaturated 1,2-diglycerides via glycerol carbonates. Tetrahredon Lett., 32, 3549–3552. Pichat, L. et al. (1969) Methylations of metalated acetylenic compounds in hexamethylphosphoric triamide solution and use of the Wittig reaction for synthesizing 14C-labelled fatty acids. I. Synthesis of oleic-18-14C acid. Bull. Soc. Chim Fr., 4, 1198–1200. Pomonis, J.G. and Hakk, H. (1990) Synthesis of ω9-tetracosynoic and ω9-octacosynoic acids as entries into tritiated metabolic precursors of cis-9-tricosene and cis-9-heptacosene in the housefly. Lipids, 25, 821–826. Prassides, K. Ed. (2004) Fullerene-based Materials — Structure and Properties, Springer-Verlag, Berlin. Quaranta, A. et al. (2003) Photophysical properties of a dendritic methano[60]fullerene octadeca acid and its tertbutyl ester: evidence for aggregation of the acid form in water. Phys. Chem. Chem. Phys., 5, 843–848. Quinn, J.G. et al. (1967) Synthesis of 100-gram quantities of highly purified mixed acid triglycerides. J. Am. Oil Chem. Soc., 44, 439–442. Rachlin, A.I. et al. (1961) Synthesis of arachidonic acid. J. Org. Chem., 26, 2688–2693. Rakoff, H. (1982) Preparation of fatty acids and esters containing deuterium. Prog. Lipid Res., 21, 225–254. Rancan, F. et al. (2005) Fullerene-pyropheophorbide a complexes as sensitizer for photodynamic therapy: uptake and photo-induced cytotoxicity on Jurkat cells. J. Photochem. Photobiol. B, 80, 1–7. Raphael, R.A. and Sondheimer, F. (1950) The synthesis of longchain aliphatic acids from acetylenic compounds. The synthesis of linoleic acid. J. Chem. Soc., 2100–2103. Reeves, L.W. et al. (1979) A deuterium and sodium nuclear magnetic resonance investigation into the hexadecylpyridinium/hexadecyltrimethyl ammonium liquid crystalline system. Can. J. Chem., 57, 747–753. Rietmeijer, F.J.M. et al. (2004) C60 and giant fullerenes in soot condensed in vapors with variable C/H 2 ratio. Fullerenes, Nanotubes, and Carbon Nanostruct., 12, 659–680. Robles Medina, A.R. et al. (1999) Lipase-catalyzed esterification of glycerol and polyunsaturated fatty acids from fish and microalgae oil. J. Biotechnol., 70, 379–391.
384
Synthesis
Rosu, R. et al. (1999) Enzymatic synthesis of symmetrical 1,3diacylglycerols by direct esterification of glycerol in solvent-free system. J. Am. Oil Chem. Soc., 76, 839–843. Sandri, J. and Viala, J. (1995) Direct preparation of (Z,Z)-1,4dienic units with a new C6 homologating agent: synthesis of α-linolenic acid. Synthesis, 271–275. Sauerwald, T. et al. (1996) Zur umbedenenklichkeit vo stabilen Isotopen in der padiatrischen Forschung und Diagnostick. Monatsschr. Kinderheilkd, 144, 636–644. Schlosser, M. and Christmann, K.F. (1966) Trans-selective olefin synthesis. Angew. Chem. Int. Ed. Engl., 5, 126. Schmitt-Rozieres, M. et al. (2000) Enrichment of polyunsaturated fatty acids from sardine cannery effluents by enzymatic selective esterification. J. Am. Oil Chem. Soc., 77, 329–332. Scrivens, W.A. et al. (1994) Synthesis of 14C-labelled C60, its suspension in water and its uptake by human keratinocytes. J. Am. Chem. Soc., 116, 4517–4518. Shimada, Y. et al. (1997a) Purification of docosahexaenoic acid by selective esterification of fatty acids from tuna oil with Rhizopus delemar lipase. J. Am. Oil Chem. Soc., 74, 97–101. Shimada, Y. et al. (1997b) Purification of γ-linolenic acid from borage oil by a two-step enzymatic method. J. Am. Oil Chem. Soc., 74, 1465–1470. Shimada, Y. et al. (1997c) Purification of docosahexaenoic acid from tuna oil by a two-step enzymatic method: hydrolysis and selective esterification. J. Am. Oil Chem. Soc., 74, 1441–1446. Sijbesma, R. et al. (1993) Synthesis of a fullerene derivative for the inhibition of HIV enzymes. J. Am. Chem. Soc., 115, 6510–6512. Smith, C.R. Jr. (1972) Glyceride chirality, in Topics in Lipid Chemistry, Vol. 3, Ed. F.D. Gunstone, Elek Science, London, pp. 89–124. Smith, R.E. et al. (1994) Overview of SALATRIM, a family of low-calorie fats. J. Agric. Food Chem., 42, 432–434. Snyder, J.M. and Scholfield, C.R. (1982) cis-trans Isomerization of unsaturated fatty acids with p-toluenesulfinic acid. J. Am. Oil Chem. Soc., 59, 469–470. Softly, B.J. et al. (1994) Composition of representative SALATRIM fat preparations. J. Agric. Food Chem., 42, 461–467. Sonnet, P.E. (1991) A short highly regio- and stereoselective synthesis of triacylglycerols. Chem. Phys. Lipids, 58, 35–39. Sonnet, P.E. (1999) Synthesis of triacylglycerols in Lipid Synthesis and Manufacture, Ed. Gunstone, F.D., Sheffield Academic Press, Sheffield, pp. 162–184. Sonnet, P.E. and Dudley, R.L. (1994) Stereospecific synthesis of selected triglycerides: comments on acyl migration and analysis of configuration. Chem. Phys. Lipids, 72, 185–191. Sprecher, H. (1977) The organic synthesis of polyunsaturated fatty acids, in Polyunsaturated Fatty Acids, Eds. W.H. Kunau and R.T. Holman, American Oil Chemists’ Society, Champaign, IL, pp. 6–103. Sprecher, H. (1979) The organic synthesis of unsaturated fatty acids. Progr. Chem. Fats Other Lipids, 15, 219–254. Subramanian, G.B.V. and Rastogi, A. (1991) A facile synthesis of 1-thiacontanol from threo-aleuritic acid. Chem. Ind., 436–440.
Svatos, A.A. et al. (1994) Synthesis of deuterium labelled polyunsaturated fatty acids. Tetrahedron Lett., 35, 9497–9500. Taber, D.F. et al. (1982) Preparation of deuterated arachidonic acids, in Methods in Enzymology, Academic Press, New York, pp. 366–369. Taber, D.F. and You, K. (1995) New synthon for the convergent construction of skipped conjugation polyenes: synthesis of ethyl docosa-4,7,10,13,16,19-hexaenoate. J. Org. Chem., 60, 139–142. Tamvakopoulos, C.S. and Anderson, V.E. (1990) Synthesis of ω-labelled fatty acids. J. Labelled Cpd. Radiopharm., 28, 187–191. Thilgen, C. et al. (1997) The covalent chemistry of higher fullerenes: C70 and beyond. Angew. Chem. Int. Ed. Engl., 36, 2268–2280. Thomas III, A.E. et al. (1965) Quantitative estimation of isomeric monoglycerides by thin-layer chromatography. J. Am. Oil Chem. Soc., 42, 789–792. Tucker, W.P. et al. (1971) Synthesis of 11,11-dideutero-linoleic acid. J. Labelled Cpd. Comp. Radiopharm., 7, 11–15. Tulloch, A.P. (1979) Synthesis of deuterium and carbon-13 labelled lipids. Chem. Phys. Lipids, 24, 391–406. Turner, D.L. et al. (1968) The total synthesis of phosphatidyl(dioleoyl)hydroxyl-L-proline and its activity in blood-clottting systems. Lipids, 3, 228–233. Udell, T. et al. (2005) The effect of α-linolenic acid and linoleic acid on the growth and development of formula-fed infants: a systematic review and meta-analysis of randomized controlled trials. Lipids, 40, 1–11. Valicenti, A.J. et al. (1985) Synthesis of octadecynoic acids and [1-14C] labelled isomers of octadecenoic acids. Lipids, 20, 234–242. Van der Steen, D. et al. (1963) Synthesis of arachidonic acid and related higher unsaturated compounds. Recl. Trav. Chim. Pays-Bas, 82, 1015–1025. Vatèle, J-M. (1999) Polyene acids in Lipid Synthesis and Manufacture, Ed. F.D. Gunstone, Sheffield Academic Press, Sheffield, pp. 1–45. Vedejs, E. and Peterson, M.J. (1994) Stereochemistrsy and mechanism in the Wittig reaction in Topics in Stereochemistry, Vol. 21, John Wiley & Sons, New York, pp. 1–56. Verkade, P.E. (1953) Synthesis of glycerides. Chim. Ind. (Paris), 69, 239–251. Viala, J. and Sandri, J. (1992) Total stereospecific synthesis of all cis-5,8,11,14,17-eicosapentaenoic acid (EPA). Tetrahedon Lett., 33, 4897–4900. Viala, J. and Santelli, M. (1988) Efficient stereoselective synthesis of methyl arachidonate via C3 homologation. J. Org. Chem., 53, 6121–6123. Villeneuve, P. et al. (1994) Chiral synthesis of a triglyceride: example of 1-butyroyl-2-oleoyl-3-palmitoyl-sn-glycerol. Chem. Phys. Lipids, 72, 135–141. Voss, A. et al. (1991) The metabolism of 7,10,13,16,19-docosapentaenoic acid to 4,7,10,13,16,19-docosahexaenoic acid in rat liver is independent of a 4-desaturase. J. Biol. Chem., 266, 19995–20000. Watanabe, Y. et al. (2005a) Production of FAME from acid oil model using immobilized Candida antarctica lipase. J. Am. Oil Chem. Soc., 82, 825–831.
385
5.6 Glycerophospholipids
thiophospholipids. Issacson et al. (1990) describe the synthesis of phospholipids with hydroxylated fatty acids at the sn-2 position, while Lebeau et al. (1988) have synthesised phospholipids linked to steroid hormone derivatives. General books on phospholipids are those of Hawthorne and Ansell (1982) and Ceve (1993).
Watanabe, Y. et al. (2005b) Production of MAG of CLA by esterification with dehydration at ordinary temperature using Penicillium camembertii lipase. J. Am. Oil Chem. Soc., 82, 619–623. Watanabe, Y. et al. (2000) Continuous production of biodiesel fuel from vegetable oil using immobilized Candida antarctica lipase. J. Am. Oil Chem. Soc., 77, 355–360. Westerman, P.W. and Ghrayeb, N. (1982) Synthesis of esters of tetradecanoic acid deuterated at the penultimate carbon: some general procedures for the synthesis of selectively deuterated fatty acids. Chem. Phys. Lipids, 29, 381–387. Willis, W.M. et al. (1998) Lipid modification strategies in the production of nutritionally functional fats and oils. Crit. Rev. Food Sci. Nutr., 38, 639–674. Wüst, F. et al. (2000) A new approach for the synthesis of [11C]labelled fatty acids. J. Labelled Cpd. Radiopharm. 43, 1289–1300. Xu, X. (2000) Production of specific structured triacylglycerols by lipase-catalyzed reactions: a review. Eur. J. Lipid Sci. Technol., 102, 287–303. Zakharian, T.Y. et al. (2005) A fullerene-paclitaxel chemotherapeutic: synthesis, characterization and study of biological activity in tissue culture. J. Am. Chem. Soc., 127, 12508–12509. Zhang, X. et al. (2004) Iodo-controlled selective formation of pyrrolidino[60]fullerene and aziridino[60]fullerene from the reaction between C60 and amino acid esters. J. Org. Chem., 69, 5800–5802. Zlatanos, S.N. et al. (1985) High yield monoglycerides preparation from glycidol and carboxylic acids. J. Am. Oil Chem. Soc., 62, 1575–1577.
5.6
Glycerophospholipids
5.6.1
General comments
5.6.2 5.6.2.1
General considerations Sources of the glycerol backbone
Protected glycerol derivatives are convenient starting materials. The topic has been reviewed (Bittman, 1999). 2,3-OIsopropylidene-sn-glycerol and its enantiomer 1,2-Oisopropylidene-sn-glycerol are available both commercially and from many natural sources. Analysis and precautions in their use are detailed in Paltauf and Hermetter (1994). Other well-known precursors of the glycerol backbone are vicinal diols, such as 1-O-benzyl-sn-glycerol or its enantiomer. Two other 3-O-protected glycerol derivatives are 3-O-(4-methoxyphenyl)-sn-glycerol and 3-O-(4-methoxybenzyl)-sn-glycerol (Bittman, 1999). Other potential C-3 building blocks are reported by Paltauf and Hermetter (1994) although their individual use has, up until now, been limited. An exception is the use of racemic and optically active glycidols (and glycidol tosylates), which are commercially available. Epoxy alcohols, such as glycidol, can be easily ring-opened with nucleophiles to give 1,2-diols as the products. Regioselectivity can be improved by Lewis acid catalysis. A typical procedure might be that of Vinogradov et al. (1995). Alternatively, ring opening can be accomplished regiospecifically with primary or secondary alcohols in the presence of diisobutylaluminum hydride (Erukulla et al., 1995a) or with a sulfur nucleotide (see Bittman, 1999). Ring opening of glycidol derivatives with fatty alcohols also generates 1-O-alkyl-sn-glycerols for making ether-linked phosphoglycerides. The advantages of using BF3 to catalyse the ring opening have been discussed by Bittman (1999).
Excellent reviews dealing with phospholipid synthesis are those of Rosenthal (1975), Slotboom and Bonsen (1970), and Paltauf and Hermetter (1994). Other references that deal with general aspects include a discussion of the effect of solvents and catalysts on acylation (Mangroo and Gerber, 1988), the synthesis of phospholipids using a phosphotriester approach (Woolley and Eibl, 1988), and uniformly cleavable β-hydrogenated protecting groups (Lemman et al., 1990). Stepanov and Slivets (1986) describe the formation of phosphoester bonds in phosphoglyceride synthesis and include discussion of various methods of phosphorylation used for phosphomono-, -di and –triester formation. Particular attention was paid to a comparison of the chlorophosphate versus silver salt methods. Several papers deal with the preparation of derivatives of a number of phospholipids. An example of the synthesis of spin-labelled phospholipids is that of Lai et al. (1988), while Reynolds et al. (1991) describe the preparation of synthetic phospholipase substrates with p-nitrophenol linked to the sn-2 moiety of the phospholipid. Yu and Dennis (1991) detail the preparation of chiral
5.6.2.2
Diradylglycerols
Diradylglycerols are common intermediates for the synthesis of glycerophospholipids, though care has to be taken with acyl migration if there is an acyl group at the sn-2 position. Racemic diacylglycerols with identical saturated chains can be prepared from racemic benzylglycerol or from tetrahydropyranylglycerol. For diacylglycerols with one saturated acyl group at position-1 and a saturated or unsaturated group at positon-2, methods using allylesters of fatty acids can be utilised (Paltauf and Hermetter, 1994). For optically active diacylglycerols, methods similar to those for racemic compounds can be used, but with chiral adducts, such as 1- or 3-O-benzyl-sn-glycerol. A general route leading to various optically active saturated or unsaturated diaclyglycerols, with two different acyl chains starts from isopropylideneglycerol (Kodali and 386
Synthesis
Duclos, 1992). For other alternative methods of preparing diacylglycerols see Section 5.4 and Paltauf and Hermetter (1994). 5.6.2.3
5.6.2.5
Plant or bacterial phospholipase Ds (Section 10.2.2) can catalyse transphosphatidylation under appropriate conditions. This allows relatively inexpensive phosphatidylcholines to be used to make equivalent glycerophospholipids. The subject has been reviewed by Ella et al. (1997). Phospholipase D-catalysed exchange has proven useful for the preparation of phosphatidylserine, phosphatidylethanolamine, phosphatidylglycerol as well as antitumor phospholipid-nucleoside conjugates (Bittman, 1999). The reaction proceeds in two steps with phosphatidic acid as a product of the first step. (In fact, phosphatidic acid would be formed naturally in aqueous systems.) The phosphatidic acid can then be used to prepare various glycerophospholipids by coupling it with the desired alcohol via a condensing agent such as trichloroacetonitride in dry pyridine (e.g., Ali and Bittman, 1989).
Alkylacylglycerols
Reviews of methods for the synthesis of alkylacylglycerols are those of Paltauf (1983) and Paltauf and Hermetter (1991). These compounds are needed for the synthesis of alkylacylglycerophospholipids (which themselves are biochemical precursors for plasmalogens and platelet activating factor) or for making various model phospholipids. For 1-O-alkenyl-2-acylglycerols, synthesis usually begins with racemic or optically active 1-O-alkylglycerols. Tritylation of the primary hydroxyl group of an alkylglycerol, acylation at position 2 and ditritylation with BF3 or with bromomethylborane yields products of high isomeric purity (Paltauf and Hermetter, 1994). 1-Acyl-2-alkylglycerols can be produced by alkylation of benzylideneglycerol with alkylmethanesulphonates, followed by removal of the protecting group (acid-catalysed) and acylation of the 1-position. This gives racemic derivatives. For optically active compounds, 2-O-alkyl-1acylglycerol-3-arenesulphonates can be prepared and used for the subsequent synthesis of glycerophospholipids with a saturated alkyl chain (Ali and Bittman, 1990). For the synthesis of optically active 1-acyl-2-O-alkylglycerides with saturated or unsaturated acyl and alkyl groups, a convenient route begins with commercially available 1- or 3-O-benzyl-sn-glycerol (see Paltauf and Hermetter, 1994). Racemic di-O-alkylglycerols are usually made by alkylation of tetrahydropyranyl glycerol followed by acidcatalysed removal of the tetrahydropyranyl group. Those with saturated alkyl groups can also be made from 1- or 3-O-benzyl-sn-glycerols. For those compounds with different saturated or unsaturated alkyl chains, 1(3)-O-alkyl3(1)-O-trityl-sn-glycerols can be used as starting material (see Paltauf and Hermetter, 1994, where methods for making 1,3-di-O-alkylglycerols will also be found). 5.6.2.4
Transphosphatidylation
5.6.3
Phosphatidic acids
Phosphatidic acids can be synthesised by several routes (Figure 5.1). Paltauf and Hermetter (1994) consider that phosphorylation of diradylglycerols (see Section 5.6.2.2) is the most convenient. 5.6.3.1
Phosphorylation of diradylglycerols
Phosphorylation, using phosphorus oxychloride, is used to produce both saturated and unsaturated derivatives. A modification (Paltauf, 1976) of the method of Berecoechea et al. (1968) is recommended to give good yields. Several variants of the reaction conditions and phosphorylating agents are described by Paltauf and Hermetter (1994). One of these uses dimethylphosphorochloridate as the phosphorylating agent, in which case protecting methyl groups are removed with trimethylsilylbromide (Bittman et al. 1984). 5.6.3.2
Use of 1,2-diradylglycero-3-deoxy-3-iodo glycerols
Reaction of the above with silver salts of various phosphodiesters can be used and this should minimise the danger of acyl migration during phosphorylation. Silver salts of dibenzyl phosphate (Hessel et al. 1954), pnitrobentylphosphates (de Haas and Van Deenen, 1961), or di-p-xylxylphosphate (Molotkovsky et al. 1976) are examples. The latter allows preparation of phosphatidic acids containing unsaturated acyl or alkyl groups.
Glycerophosphoesters as starting materials
Depending on the final glycerophospholipid required, a number of glycerophosphoesters could be utilised. Thus, sn-3-glycerophosphocholine can be used to make phosphatidylcholines and, by subsequent acyl exchange (Section 5.6.4.2), other glycerophospholipids. Although sn-3-glycerophosphocholine can be synthesised, it is often prepared from phosphatidylcholine by deacylation (Paltauf and Hermetter, 1994). N-Trityl-glycerophosphoethanolamine has been used to prepare phosphatidylethanolamines (Hermetter et al., 1983). Details for other compounds, such as glycerol 3-phosphate or dihydroxyacetone phosphate that are useful substrates for phospholipid biosynthesis, are given in Paltauf and Hermetter (1994).
5.6.3.3
Acylation of glycerol 3-phosphate
In this procedure, α-glycerol phosphate is acylated directly using acyl anhydrides (Lapidot et al. 1969). The acyl anhydrides are prepared with dicyclohexyl-carbodiimide (Rosenthal, 1975). Optically active or racemic phosphatidic acids can be prepared and either saturated or unsaturated acyl residues added. The method depends on the ability of the fatty acid salt to suppress the formation of mixed phosphoric-carboxylic anhydrides with the
387
5.6 Glycerophospholipids
HC R2
O
P(OCH3)2
OH
O
1. POCl3, N(C2H5)3 or pyridine 1. Cl P(OPh)2 2. H2O/NaHCO3 2. H2/Pd 3. HCl
2. Trimethylsilylbromide H2C R2
O
O
CH H2C
H2C
O
P
OH
O
OH
1. AgO
H2C O
O
P
P(OpXyl)2
2. HCl in CHCl3
OH
CH H2C
R1 O
Acylanhydride, DMAP
HO
R1
C H2C
O 1. Cl
O
R2 OH
O
O
R1
CH H2C
I
OH R1. R2 = acyl or alkyl
FIGURE 5.1
Synthesis of phosphatidic acid. (Adapted from Paltauf, F. and Hermetter, A. (1994) Prog. Lipid Res., 33, 239-328.)
consequent phosphorylation of adjacent hydroxy groups. The use of acyl anhydrides rather than acyl chlorides probably prevents the formation of glycerol chlorhydrin esters (Aneja and Chadha, 1971a). Diacylation of glycerol 3-phosphate may be the preferred route where radioactive labelling of the glycerol backbone is required. An improved method for acylation of the pyridinium salt of glycerol 3-phosphate is that of Gupta et al. (1977). 5.6.3.4
5.6.4
Phosphatidylcholines
General methods for the chemical synthesis of phosphatidylcholine (lecithin) involve the use of phosphatidic acid, glycerylphosphorylcholine or diradylglycerols. 5.6.4.1
Synthesis of phosphatidylcholine from phosphatidic acid
Readily available phosphatidic acids (see Section 5.6.3) can be esterified with a suitable choline salt. Suitable condensing agents include carbodiimides, trichloroacetonitrile and sulfonyl chlorides (Rosenthal, 1975). Not only are these procedures relatively straightforward, but they are also suitable for the preparation of phosphatidylcholines radio-labelled in the choline moiety. The method of Aneja and Chadha (1971b) utilizes triisopropylbenzenesulfonyl chloride and is claimed to give products of better optical purity than other procedures. In this technique the phosphatidic acid is mixed and dried with choline acetate (in twice molar excess). The sulfonyl chloride derivative in dry pyridine is then added and, after reaction at 70°C for 1 h, the procedure is continued at room temperature for 4 h. Excess choline acetate can then be filtered off and the product purified, after passage through a mixed-bed resin column or by chromatography on a neutral alumina column.
Derivatives of phosphatidic acid
Monacyl (lyso) phosphatidic acid can be made chemically or through phospholipase A2 hydrolysis of phosphatidic acid (Van der Bend et al., 1992). Bis (diacylglycero) phosphoric acids (or bisphosphatidic acids) can be formed by condensation of 1,2-diacylglycerols with phenylphosphoryl dichloride followed by removal of the phenyl group via catalytic hydrogenolysis (Dang et al., 1982). Various other derivatives, such as ether-containing and glycerol analogues of phosphatidic acid have been synthesised (Rosenthal, 1975). Other methods for synthesis of alkyl analogues of phosphatidic acid, by techniques similar to those for the preparation of phosphatidic acid, are given in Paltauf and Hermetter (1994). Methods for the synthesis of chiral vinylic phosphonolipid (Schwartz et al., 1989a) and α-hydroxyphosphonolipid analogues of phosphatidic acid (Schwartz et al., 1989b) have been described.
388
Synthesis
5.6.4.2
an intermediate (see Section 5.6.4.1). An early procedure (Aneja and Chadha, 1971b) is simple, but has some drawbacks (Paltauf and Hermetter, 1994). An improved version, using the tetraphenylborate salt of choline, has been described (Harbison and Griffin, 1984). If the production of phosphatidate intermediates involves reaction of the diacylglycerols with phosphorus oxychloride then the dichlorophosphate intermediates can be condensed directly with choline chloride (Baer and Kindler, 1962). This method has been used very successfully by some (e.g., Brockerhoff and Ayengar, 1979), but may produce difficultto-remove by-products (Paltauf and Hermetter, 1994). Another method uses 2-bromoethyl dichlorophosphate (Hirt and Berchtold, 1958) and has been improved by Hansen et al. (1982) to reduce the reaction time and avoid side-reactions. Alternatively, the diradylglycerols can be condensed with 2-chloro-2-oxo-1,3,2-dioxaphospholane followed by treatment of the cyclic phosphates with trimethylamine (Chandrakumar and Hajdu, 1982). This type of synthesis, using dioxaphospholanes, tends to require long reaction times (for nucleophilic ring opening), but few byproducts are formed. Improvements have been reported, such as the use of chloro-oxazaphospholane as phosphorylating agent or methods that use phosphatidylethanolamine (or N-methyl-phosphatidylethanolamine) as intermediate (Patel et al., 1979; Stumpf and Lemmen, 1990). Phosphoamidite compling reactions, as utilised for nucleotides, have been used for phosphatidylcholine. For example, Bruzik et al. (1986) condensed dipalmitoylglycerol with chloro-(N-N-diisopropylamino) methoxyphosphine. However, subsequent steps are rather cumbersome and unsaturated derivatives are not possible. A different approach was taken by Hebert and Just (1990), who condensed 2-cyanoethyl (N-N-diisopropylamino) chlorophosphinite with bromoethanol to yield a dialkylphosphoramidite. The latter was then reacted with diacylglycerol in the presence of
Syntheses involving glycerophosphocholine
Acylation of glycerophosphocholine is the most straightforward method for phosphatidylcholine containing two identical saturated and unsaturated acyl groups. Gupta et al. (1977) used N, N-dimethyl-4-aminopyridine in chloroform for the acylation of a cadmium chloride complex of glycerophosphocholine with fatty acid anhydride. In contrast, acylation can occur with a twofold excess of acylimidazolide in the presence of methylsulfinylmethide (Warner and Benson, 1977). Modifications of reaction conditions and precautions are discussed by Paltauf and Hermetter (1994). For small scale preparation involving radioactive or other expensive fatty acids, a modification of the Gupta et al. (1977) method can be used (Mena and Djerassi, 1985). For mixed acid phosphatidylcholines, procedures usually involve phospholipase A2-removal of the acid from the sn-2 position followed by reacylation with a new fatty acid. Typically, fatty acyl anhydrides are used for the latter step (Gupta et al., 1977; Mason et al., 1981). Some improvements in the general methodology are discussed by Paltauf and Hermetter (1994). The acylation can also be carried out with 2-thiopyridylesters in the presence of silver ions (Nicholas et al., 1983). A three-step procedure for the preparation of mixed acid phosphatidylcholines with both saturated and unsaturated chains is depicted in Figure 5.2. The method by Hermetter et al. (1989) starts from a zinc chloride complex of glycerophosphocholine, which is converted to a 1-O-trityl derivative by reaction with triphenylmethylchloride, followed by acylation. The trityl group is then replaced by the chosen acyl moiety using acyl anhydride or acyl imidazolide. 5.6.4.3
Syntheses from diradylglycerols
Several ways of forming the phosphodiester bond are possible. A convenient method uses phosphatidic acid as
H2C
1. Tritylchloride, in DMF, ZnCl2
OH
O
H2C
O
Tr
2. Acylation HO
CH H2C
R1
O O
P
C
O
OCH2CH2N+ (CH3)3
O
CH H2C
O
P
OCH2CH2N+ (CH3)3
O–
O– O Acylanhydride or acylimidazolide BF3, Et2O, in CH2Cl2 0–5°C, 1h
O R1
C
H2C
O
H2C
R2
O
CH
O
C
O
P
OCH2CH2N+ (CH3)3
O–
FIGURE 5.2 Synthesis of mixed acid, 1,2-diacyl-sn-glycero-3-phosphocholines. (Adapted from Paltauf, F. and Hermetter, A. (1994) Prog. Lipid Res., 33, 239-328.)
389
5.6 Glycerophospholipids
TABLE 5.1
Phosphatidylcholine analogues (see Bonsen et al., 1972)
Type of Analogue
Compound
Stereoisomers of normal PC
1-sn-Phosphatidylcholine
Structural isomer of normal isomer
rac-1, 3-Diacylglycerol-2-phosphorylcholine
Modification of acyl chain at the 2 position of PC
2-Methyl derivatives of acyl chain 2,2-Dimethyl derivatives of acyl chain 3,3- Dimethyl derivatives of acyl chain rac-1, 2-Dibenzoylglycero-3-phosphorylcholine
Modification of ester linkages at the 2 position of PC
rac-1-Acyl-2-acylamido-2-deoxyglycero-3-phosphorylcholine 1-Acyl-2-alkanesulphonyl-sn-glycero-3-phosphorylcholine rac-1-Acyl-2-alkylglycerol-3-phosphorylcholine rac-1-Acyl-2-alkyl-2-deoxyglycero-3-phosphorylcholine
Compounds without an acyl group at the 2 position of PC
1-Acyl-sn-glycero-3-phosphorylcholine 1-Acyl-2-deoxyglycero-3-phosphorylcholine
Modification of glycerol backbone
1-Acylglycol-2-phosphorylcholine rac-1,2-Diacylbutanetriol-4-phosphorylcholine 2,2-(Diacylhydroxymethyl)propanol-1-phosphorylcholine 1,2-Diacyl-3-deoxy-sn-glycero-3-phosphorylcholine
Modification of the phosphate group
rac-1,2-Diacyl-3-deoxyglycero-3-sulfonic acid
tetrazole, followed by several steps to give phosphatidylglycerol in 65% yield. 5.6.4.4
antagonistic, or cytostatic activity or as phospholipid analogues for basic membrane biophysical studies. Details of their syntheses are given in Paltauf and Hermetter (1994), who also discuss other phosphatidylcholine derivatives, such as cyclopentano or sulfonium analogues or 1,2-dialkylmethylidene derivatives.
Other phosphatidylcholine derivatives
The synthesis of other phosphatidylcholines (other than plasmalogens; see Section 5.6.10) has been well reviewed (Rosenthal, 1975; Paltauf 1983; Zeisig et al., 1990). Experimental details are given in Paltauf and Hermetter (1991) and further discussion will be found in Paltauf and Hermetter (1994). A large variety of other phosphate analogues have also been synthesised, and these are reviewed by Bonsen et al. (1972). A summary of the various compounds for which methods have been developed is made in Table 5.1. For the preparation of mixed chain phosphatidylcholine analogues with bulky headgroups or with one acyl chain twice the length of the other, see Ali and Bittman (1989). Schwartz et al. (1989a) have described the synthesis of chiral vinylic phosphonolipid analogues, while Moschidis (1987) reported the preparation of 1-O-alkylethylene glycol phosphonic acid analogues of phosphatidylcholine. Thiol ester and thiol ether analogues are described by Hendrickson and Hendrickson (1990), while the formation of phosphatidylcholine labelled with 8-(2-anthroyl)octanoic acid, for use as a fluorescent probe, is detailed by Perochon and Tocanne (1991). General methods for fluorophore-labelled phospholipid substrates, including derivatives of phosphocholine, are given by Hendrickson (1991). Phosphatidylcholine analogues containing acylamino, alkylamino, or carbamyl groups have been made. They can have a variety of uses including potential inhibitors of phospholipase A2, as PAF analogues with PAF agonistic,
5.6.5
Phosphatidylethanolamines
Phosphatidylethanolamine is usually synthesised from phosphatidic acid or from iodohydrin diesters. The analogous route to that of phosphatidylcholine from glycerophosphorylcholine is not usually used because it is difficult to obtain the substrate in a sufficiently pure state. Care must be taken during the synthesis of phosphatidylethanolamines to protect the unsubstituted amino function. 5.6.5.1
Synthesis of phosphatidylethanolamine from phosphatidic acid
The first problem encountered in this route is the blocking of the primary amino group by a reagent that can be removed easily and under conditions that are not accompanied by ester hydrolysis or fatty acyl modification. The trityl group can be used for this purpose. In a typical procedure (Aneja et al., 1970) N-tritylethanolamine is condensed with a phosphatidic acid by means of 2,4,6-triisopropylbenzenesulphonyl chloride. The reaction is fast and, after heating for 1 min at 60°C to dissolve the phosphatidic acid, proceeds in 90% yield at room temperature in 1.5 h. The product is purified by triethylamine impregnated silicic acid chromatography. Removal of the trityl group can be accomplished by hydrogenolysis (not appropriate
390
Synthesis
H2C R
O
O
CH H2 C
R + HO
O O
P
CH2CH2NH
X
TPS Pyridine
OH
OH
H2C
O
R CF3COOH
R
O
CH
PtdEtn
O or HCl/CHCl3
H2C
O
P
OCH2CH2NH
X
O– X = Trityl or t
BOC
FIGURE 5.3 Synthesis of ethanolamine glycerophospholipids from phosphatidic acids. (Adapted from Paltauf, F. and Hermetter, A. (1994) Prog. Lipid Res., 33, 239–328.)
for unsaturated acyl groups) or by very mild acid hydrolysis (Rosenthal, 1975). Alternative condensing agents to sulphonyl chlorides include carbodiimides and trichloroacetonitrile. A depiction of the synthesis of phosphatidylethanolamines from phosphatidic acids is shown in Figure 5.3. Paltauf and Hermetter (1994) summarize alternative methodologies, including the use of phosphatidic acid dichloride for the condensation with ethanolamine (Eibl, 1978). 5.6.5.2
5.6.5.4
Reaction of diacylglycerols with a number of reagents can be used for the formulation of phosphatidylethanolamines. These include bromoethyl dichlorophosphate (Eibl and Nicksch, 1978; Figure 5.4), phosphorus oxychloride (see Figure 5.4) (via phosphatidic acid dichloride) as described by Eibl (1978), phosphorus acid-1,1,1,trichloro2-methyl-2-propylester-chloride-dimethylamide (Lemmen et al. 1990), phosphorus trichloride (Lindh and Stawinski, 1989), and chloro-(N,N-diisopropylamino) methoxylphosphine (Bruzik et al. 1986). Details of these methods and other pertinent discussion on the use of particular methods for specific purposes are given in the excellent review by Paltauf and Hermetter (1994).
Synthesis of phosphatidylethanolamine from α-iododeoxyglycerol diesters
This method is especially attractive for the synthesis of mixed-acid phosphatidylethanolamines because of the ease of preparation of specific diesters of α-iododeoxyglycerol. The phosphorylethanolamine function is introduced via the silver salt of a substituted derivative. As in the preparation of phosphatidylethanolamine from phosphatidic acid, it is essential that the protective groups can be easily removed. Silver t-butyl (N-t-butyloxycarbonyl-2aminoethyl) phosphate is the most useful reagent because it allows the simultaneous removal of both phosphoester and amino-protecting groups under conditions that do not affect unsaturation or ester groups (Rosenthal, 1975). Yields of about 50% for both the preparation of the silver salt of the substituted phosphorylethanolamine and for the condensation with the iododeoxyglycerol can be expected. 5.6.5.3
Synthesis of phosphatidylethanolamine from diacylglycerols
5.6.5.5
Other phosphatidylethanolamine derivatives
The synthesis of ether phosphatidylethanolamines is analogous to that for the formation of the corresponding phosphatidylcholines (Rosenthal, 1975). Relevant literature has been reviewed (Paltauf, 1983) and laboratory methods have been published (Paltauf and Hermetter, 1991). The N-methyl and N, N-dimethyl derivatives have been synthesised from their corresponding phosphatidic acids (Aneja et al., 1970). See also Calderon and Yague (1986) for the preparation of N, N-dimethylphosphatidylethanolamines. An often-used method utilises phospholipase D to catalyse transphosphatidylation of the corresponding phosphatidylcholine in the presence of N-methyl or dimethylethanolamine (e.g., Sisk and Huang, 1992). The formation of N-acylphosphatidylethanolamines and other analogues is detailed by Slotboom and Bonsen (1970) and that of ethanolamine plasmalogens in Section 5.6.10. Synthesis of 1-palmitoyl-2-diphenylhexatrienylpropanoylphosphatidylethanolamine for use as a fluorescent probe is given by Beck et al. (1990), while dialkylphosphatidylethanolamines are detailed by Ruocco and Makriyannis
Synthesis of phosphatidylethanolamine from glycerylphosphorylethanolamine
As mentioned above, this method is not often used. The amino nitrogen is blocked with phthaloyl or trityl groups before acylation to yield a phosphatidylethanolamine with two identical fatty acids. Mixed-acid phosphatidylethanolamines can be synthesised by deacylation (phospholipase A)reacylation (Slotboom and Bonsen, 1970). 391
5.6 Glycerophospholipids
1. Cl H2C R
O
O
R
O
CH2CH2Br
Cl 2. H2O
CH H2C
P
PtdEtn
3. NH3 OH (a)
H2C R
O
O
1. POCl3
CH H2C
2. HOCH2CH2NH2
R
O
O
CH H2C
OH
AcOH in i–propanol/H2O 2h
H2C
R
R O
O
O
CH2
N H
CH2
P
PtdEtn (b)
FIGURE 5.4 Synthesis of ethanolamine glycerophospholipids from diradylglycerols. (Adapted from Paltauf, F. and Hermetter, A. (1994) Prog. Lipid Res., 33, 239–328.)
be accomplished in a stepwise manner to give mixed-acid phosphatidylserines (Shvets et al., 1967).
(1989). Moschidis (1987) described the preparation of 1-Oalkylethylene glycol phosphonic acid analogues of phosphatidylethanolamine. Other derivatives include N-biotinyl phosphatidylethanolamine, carboxyacyl- and P-maleimidophenyl butyryl- compounds (Paltauf and Hermetter, 1994).
5.6.6
5.6.6.3
Woolley and Eibl (1988) have described the preparation of enantiomerically pure phosphatidylserines. They used a procedure that is applicable to the formation of many phospholipids. Their method was to make an appropriate diacylglycerol (or dialkylglycerol), to react this with phosphorus oxychloride, then an alkyl donor and finally use methanolysis to create a phosphate triester. The nature of the alkyl donor determined the headgroup of the final phospholipid, while demethylation was used at a later stage. In the case of phosphatidylserine the yield was 95%.
Phosphatidylserines
Three basic methods that can be used are discussed below. 5.6.6.1
Synthesis of phosphatidylserine from phosphatidic acid
One simple procedure allows the synthesis of phosphatidylserine by condensation of phosphatidic acid with an amino- and carboxy-protected serine. N-CarbobenzoxylDL-serine benzyl ester was condensed with the phosphatidic acid in the presence of triisopropylbenzenesulphonyl chloride. The protecting groups were then removed by hydrogenation, which limits the method to saturated phosphatidylserines. However, the use of different protecting groups should allow this method to be used for unsaturated compounds. An alternative procedure involves the introduction of the phosphate and serine functions via a complex silver salt to glycerol iodohydrin diesters (de Haas et al., 1964). 5.6.6.2
Use of phosphotriester intermediate
5.6.6.4
Compounds related to phosphatidylserine
Phosphatidylthreonine has been prepared by a method involving phosphorylation of isopropylideneglycerol with phosphorus oxychloride and a sulphonyl chlorideactivated condensation (Moore and Szelke, 1970).
5.6.7
Phosphatidylglycerols
Phosphatidylglycerol can be synthesised using 2,3-dibenzyl-sn-glycerol as one of the starting materials (Saunders and Schwartz, 1966). The blocking groups must be removed by catalytic hydrogenolysis, so the method can only be used for saturated molecular species. Furthermore, the use of phosphorus oxychloride also has disadvantages (Slotboom and Bonsen, 1970). Of the other reported syntheses of phosphatidylglycerol, the method of Bonsen et al. (1966) is the most widely used. It allows preparation of saturated or unsaturated mixed-acid and same-acid phosphatidylglycerols of any configuration. It should be borne in mind that
Synthesis of phosphatidylserine from substituted glycerophosphorylserine
1,2-Isopropylideneglycerol is treated successively with phosphorus oxychloride and N-phthalylserine. Hydrolysis of the ketal and the phosphomonochloride groups then yields glycero-3-phosphoryl-N-phthalylserine as the barium salt after neutralization with barium hydroxide. Acylation with an excess of acid chloride then furnishes the desired phosphatidylserine. In addition, the acylation can
392
Synthesis
H2 C O H
C O
C
CH3
H2C O
CH3
H C O
C
CH3
H2C O
CH3
H C O
H2C O PO OBz
H2C OH
H C O
C
CH3
H2C
CH3
OBz O CO C17H33
+ C15H31 CO O C H
CH3
H2C O PO
CH3
H2C O PO OBa0.5
OBz H2C O
C
OAg
H2 C
I
OBz H2C
O CO C17H33 H2C O
C15H31 CO O C H H2C
O
H C O PO
O
C
CH3 CH3
H2C
O CO C17H33 H2C OH
C15H31 CO O C H H2C
CH2
OBz
H O
PO
O
C
OH
CH2
ONa
FIGURE 5.5 Synthesis of a mixed-acid phosphatidylglycerol; Bz = benzyl. (Scheme adapted from Slotboom, A.J. and Bonsen, P.P.M. (1970). Chem. Phys. Lipids, 5, 301-397.)
2,3-diphytanyl-sn-glycerol-1-(mono-p-nitrobenzyl) phosphate. The latter was reacted with 1-benzyl-3-iododeoxy-2-t-butyl-sn-glycerol or its enantiomer, and the protecting groups were released by catalytic hydrogenolysis and then treatment with hydrogen chloride. Careful optical rotation measurements established that the natural compound was 1-sn-phosphatidyl-3-sn-glycerol (i.e., both glycerol moieties had the opposite configurations to the glycerol moieties of the diacyl phospholipid from other bacteria or from plants). For a fuller discussion of diether structural analogues of phoshatidylglycerol, phosphatidylglycerophosphate and phosphatidylglycerol sulphate, see Kates (1978). Methods have been reported for the synthesis of aminoacyl esters (Bonsen et al., 1965) and glucosaminyl esters (Verheij et al., 1971) of phosphatidylglycerol. Phosphatidylglycerol phosphate was synthesised by Bonsen and de Haas (1967). 1,3-Diiodo-2-O-t-butylglycerol was reacted with the silver salt of 1,2-diacyl-sn-glycerol-3-(monobenzyl) phosphate to produce the mono condensation product. This was reacted with the silver salt of di-t-butyl phosphate, and then anionic debenzylation and treatment with hydrogen chloride yielded 3-sn-phosphatidyl-1-rac-glycero-3 phosphate. The diether analogue of phosphatidylglycerol phosphate, which occurs in halobacteria, was synthesised by Joo and Kates (1968, 1969).
stereochemistry is especially important for this phospholipid since, in the natural compound, the unacylated glycerol has the sn-1 configuration. Therefore, 2,3isopropylidene-sn-glycerol is a starting material (Figure 5.5). The critical stage is the coupling of a silver salt and an iodo derivative. For various reasons (Slotboom and Bonsen, 1970), 1,2-diacyl-3-iododeoxy-sn-glycerol and the silver salt of sn-glycerol-1-phosphate, containing appropriate blocking groups, were used. Dibenzylphosphoric acid and dicyclohexylcarbodiimide were used to generate tetrabenzyl phosphate. This could then be reacted with 2,3-isopropylideneglycerol in the presence of imidazole to yield 2,3-isopropylidene-snglycerol-1-dibenzyl phosphate. Debenzylation of the latter was carried out with barium iodide and the barium salt was converted to the silver salt, which was reacted with 3-iododeoxy-1-oleoyl-2-palmitoyl-sn-glycerol. The fully protected phosphatidylglycerol was debenzylated, followed by reaction with boric acid in trimethyl borate, which affected removal of the isopropylidene group (Rosenthal, 1975). Yields of 30% are obtained (Slotboom and Bonsen, 1970). Rosenthal (1975) provides notes on certain parts of the preparative procedure. This includes preparation of dibenzylphosphoric acid and of 2,3-isopropylidene-snglycerol. The preparation of enantiomerically pure phosphatidylglycerol is described by Woolley and Eibl (1988), while Calderon and Yague (1988) reported the preparation of phosphatidylglycerols from diacylglycerols. Halophilic bacteria contain a diphytanyl analogue of phosphatidylglycerol with the opposite stereo-chemical configuration of the diphytanyglycerol moiety. Using diphytanylglycerol (from a natural source) as starting material, Joo and Kates (1969) prepared the iodo derivative. Condensation with the silver salt of di-p-nitrobenzyl phosphate was followed by debenzylation to give the silver salt of
5.6.8
Diphosphatidylglycerols (cardiolipins)
1,3-Diphosphatidylglycerol appears to be the structure of natural cardiolipin, but this has only been proven in a few cases (Slotboom and Bonsen, 1970). Some derivatives of diphosphatidylglycerol have also been isolated in high yields from beef heart using chromatography (cf. Slotboom and Bonsen, 1970) or by recrystallization as the barium salt followed by chromatography on ammoniatreated silicic acid (Takahashi et al., 1967).
393
5.6 Glycerophospholipids
Several methods of synthesizing diphosphatidylglycerol have been described. De Haas and van Deenen (1965) reacted 1 mol of 1,3-diiodo-2-O-t-butyloxypropane with 2 mol of the silver salt of 1-stearoyl-2-oleoyl-sn-glycerol3-(monobenzyl) phosphate. Anionic debenzylation and then hydrogen chloride successively released the blocking groups (see Figure 5.6; Rosenthal, 1975). Inoue and Nojima (1968) used a similar method for the synthesis of racemic diphosphatidylglycerol. These workers used a benzyl-protecting group instead of the t-butyl group and, therefore, had to use catalytic hydrogenolysis for its removal. Consequently, their method can be used only for saturated molecular species. A semisynthetic approach for making diphosphatidylglycerol with two identical acyl chains at the sn-2 position of both phosphatidyl groups is that of Dale and Robinson (1988). They blocked the free hydroxyl group of bovine heart cardiolipin with a tetrahydropyranyl group and then used phospholipase A2 followed by acylation. Saunders and Schwartz (1966) used 1,2-distearoyl-snglycerol, phosphorus oxychloride and 2-benzyl-glycerol as starting materials. Because they also removed the benzyl group by catalytic hydrogenolysis, only saturated derivatives could be made. de Haas and van Deenen (1963) have prepared acyl derivatives of diphosphatidylglycerol. A series of diphosphatidylalkanediols have been synthesised by Inoue and Nojima (1968). Moschidis (1988) has described the preparation of phosphono analogues of diphosphatidylglycerol, while Fowler et al. (1988) reported the synthesis of various photoreactive analogues. H2C
O
C17H33 CO O C
H
H2C
O
CO
5.6.9
Phosphatidylinositols and other inositol lipids
The asymmetry of the inositol moiety poses a considerable problem for the synthetic chemist. A synthetic asymmetrically substituted inositol derivative must, therefore, be used, and this should contain an optically active inositol moiety (the natural isomer being myo-inositol). Unsaturated phosphatidylinositol has been synthesised by Molotkovsky and Bergelson (1971) using an optically active glycerol derivative, but a racemic myo-inositol moiety, which was added from an acetylated inositol phosphate benzyl ester. Klyashchitski et al. (1971) prepared a completely optically active dipalmitoylphosphatidylinositol. However, because they used benzyl and phenyl protecting groups, their method is not suitable for unsaturated phosphatidylinositols. In the procedure of Molotkovsky and Bergelson (1971) rac-2,3,4,5,6-penta-O-acetyl-myo-inositol was first converted to rac-1-dibenzylphosphoryl-2,3,4,5,6-penta-Oacetyl-myo-inositol-1-benzyl phosphate, which in turn was made into a silver salt. The latter was reacted with 3-iododeoxy-1-lauroyl-2-oleoyl-sn-glycerol and the resultant triester was debenzylated with sodium iodide. The product was purified, reacted with hydrazine hydrate and the phosphatidylinositol finally purified by chromatography. Care was necessary to avoid deacylation of the product, and the method adopted seemed to avoid the danger of unprotected myo-inositol phosphotriesters undergoing positional migration. For details, refer to Rosenthal (1975) and Bergelson (1980). The synthesis of enantiomerically pure lyso- and alkylphosphatidylinositols is described by Filthath and Eibl H2C
C17H35 +
I
CH OBut
PO OAg
H2C
I
OBz (2 mol) OBz H2C
O CO C17H35 H2C
C17H33 CO O C H H2C
PO
O
O
CH OBut O
PO
O
CH2
H
CH2
C O CO C17H33
H2C
O CO C17H35
OBz 1. Bal2 2. HCl OH H2C O C17H33
CO C17H35 H2C
CO O C H H2C
O
CH OH O
PO
O
CH2
PO
O
CH2
H
C O CO C17H33
H 2C O
CO C17H35
OH
FIGURE 5.6 Synthesis of mixed-acid unsaturated diphosphatidylglycerol by the method of de Haas and van Deenen (1965). (Adapted from Slotboom, A.J. and Bonsen, P.P.M. (1970). Chem. Phys. Lipids, 5, 301–397.)
394
Synthesis
making a whole series of phosphatidylinositol phosphates. The overall methods are depicted in Figure 5.7 and details will be found in Desai et al. (1996) and references therein. Of particular note were the 3-phosphorylated phosphatidylinositols since there is much current interest in the activity and function of phosphatidylinositol 3-kinase (Payrastre, 2004). Alternative ways of making phosphatidylinositol-3-4,5-trisphosphate are described by Watanabe et al. (1994) and by Gou and Chen (1994). Synthesis of phosphatidylinositol phosphates in general is described by Desai et al. (1996). Paltauf and Hermetter (1994) described other methods for the production of phosphatidylinositol (often with different fatty acid combinations) by the use of semisynthetic
(1992). A chromatogenic substrate for phosphatidylinositol-specific phospholipase C enzymes has been prepared by Shashidhar et al. (1991). A method for the synthesis of isopropylidene derivatives of phosphatidylinositol is given by Noda and Keenan (1990). The higher inositides, diphosphoinositide (1-(3-sn-phosphatidyl)-L-myo-inositol 4-phosphate) and triphosphoinositide (1-(3-sn-phosphatidyl)-L-myo-inositol 4,5-bisphosphate) can be isolated from natural sources. Great care must be taken at several stages in order to obtain reasonable yields of material (cf. Schacht, 1981). Following the synthesis of a readily available chiral protected myo-inositol derivative for the synthesis of inosital phosphates, the derivative can be used to form a basis for
b
BnO OAll BnO
Pl–4, 5P2 d BnO OH BnO
OBn a O O CMe2
BnO OAll BnO
a. allylation followed by H3O
BnO OAll BnO
OBn OH OH c
+
BnO OAll BnO
b. benzylation c. tin-mediated p-methoxybenzylation d. tin-mediated benzylation e. p-methoxybenzylation
BnO e
f. deallylation
BnO OH OBn f OBn BnO OBn
OBn OBn OBn
Pl
BnO OH OBn e, f OH BnO OBn
OBn OpMB OBn
Pl–5P
BnO OH OBn b, f OH BnO OpMB
OBn OBn OpMB
Pl–4P
BnO OH f OBn OpMB BnO OpMB
OBn OpMB OpMB
Pl–4, 5P2
OAll
BnO
b
BnO OAll
pMBO
Pl–3. 4, 5, P3 Pl–4, 5, P2 BnO pMBO
OH
c BnO
OBn a O pMBO O CMe 2
BnO OAll
pMBO
OAll OBn OH OH
a. allylation followed by H3O+ b. benzylation
d
BnO OAll
pMBO
BnO OH f OBn OBn pMBO OBn
BnO OBn OH OpMB
b, f
OH
pMBO
BnO OH OBn e, f OH pMBO OBn
OBn OBn OBn
Pl Pl–3P
OBn OBn OpMB
Pl–3, 4P2
OBn OpMB OBn
Pl–3, 5P2
c. tin-mediated p-methoxybenzylation d. tin-mediated benzylation e. p-methoxybenzylation f. deallylation
BnO e
BnO
OAll
pMBO
OH OBn OpMB pMBO OpMB f
OBn OpMB OpMB
Pl–3, 4, 5P3
FIGURE 5.7 Two overall methods for synthesizing phosphatidylinositol phosphates. (From Desai, T. et al. (1996) In Synthesis in Lipid Chemistry (Ed., J.H.P. Tyman), Royal Soc. of Chemistry, Cambridge, U.K., pp. 67–93.)
395
5.6 Glycerophospholipids
of sulphanilic acid gave a neutral plasmalogen, which was modified so that the palmitic acid at position-2 was replaced with oleic acid. (This was because natural plasmalogens usually contain an unsaturated moiety at position-2). The modified neutral plasmalogen was hydrolysed with pancreatic lipase, which yielded 1-hexadec-1′-enyl-2-oleoyl-glycerol. (The advantages of using pancreatic lipase, which does not attack vinyl ether linkages, is fully discussed by Slotboom and Bonsen (1970).) The above derivative was then converted to the choline plasmalogen by the method of Hirt and Berchtold (1958). Interestingly, although total chemical synthesis of choline plasmalogen is well described (see Paltauf and Hermetter, 1994), most researchers seem to use semisynthetic methods. This is especially true for plasmalogens with defined alkenyl/acyl chain composition or for radioor fluorescent-labelled compounds. Methods for the total chemical synthesis of ethanolamine plasmalogens have been reported (Paltauf, 1983). In addition, ethanolamine plasmalogens with special moieties (e.g., fluorescent or radioactive acyl chains at the sn-2 position) can be made from the appropriate phosphatidylcholine using phospholipase D (Loidl et al., 1990). Radioactive ethanolamine plasmalogen can be conveniently prepared and isolated from protozoan cell cultures (Achterberg et al., 1986a). Radioactive ethanolamine plasmalogen can also be used to prepare dimethylethanolamine and choline analogues by means of a phospholipase D-catalysed reaction (Achterberg et al., 1986b). A simple method for producing lysoethanolamine plasmalogen uses mild alkaline hydrolysis, if necessary of a mixture of natural phosphatidylethanolamine and its plasmalogen (Hannahan et al., 1990). Appropriate use of protective groups during acylation then allows formation of plasmalogens with defined acyl groups (Hermetter and Paltauf, 1982). A partial synthesis scheme has been used for the formation of 1-(alk-cis-1′-enyl)-2-acyl-sn-glycero3-phosphoric acid (Eibl and Lands, 1970). A phosphotriester derivative of 1-dodec-1′-enyl-2stearoylglycerol has been made, but this was not converted into a plasmalogen. However, the 1-alkenyl-2-acyl-3bromodeoxyglycerol can be used as a starting compound to form the N, N-dimethyl derivative of 1-hexadec-1′-enyl2-stearoylglycero-3-phosphorylethanolamine. (See Rosenthal (1975) for more details.) Chacko et al. (1967) have reported the preparation of a synthetic plasmalogen analogue; an acetal phosphatidylcholine. The synthesis of plasmalogens and a number of substituted or double-labelled derivatives has also been summarized by Horrocks and Sharma (1982). Dialkylphosphatidylcholines and –phosphatidylethanolamines can be made by a method detailed by Abdelmageed et al. (1989). General methods for the preparation of alkyl ether and vinyl ether substrates for phospholipases are given by Paltauf and Hermetter (1991). These authors
routes. In one method, total chemical synthesis from phosphatidic acid yielded phosphatidylinositol in a 40% yield (Ward and Young (1988), and the diisopropylidene derivative could be converted to phosphatidylinositol 4-phosphate (Jones et al., 1989). Other methods for these lipids are also given in Paltauf and Hermetter (1994) and a method for preparing phosphatidylinositol-4,5-bisphosphate in 32% overall yield is that of Dreef et al.(1988). Verheij et al. (1970) have synthesised phosphatidylglucose and Lukyanov et al. (1965) have made a phosphatidylinositol aminoethyl phosphotriester isomer. Synthesis of glycosylphosphatidylinositol anchors is detailed by Gigg and Gigg (1997).
5.6.10
Plasmalogens
Plasmalogens are a group of phosphoglycerides that are phosphorylated derivatives of 1-(alk-cis-1-enyl)-2-acyl-snglycerol. The similarity in properties between plasmalogens and the corresponding 1,2-diacyl phospholipids has made isolation procedures difficult. Moreover, chemical synthesis is made difficult by the presence of both alkalilabile fatty acid ester bonds and a reactive, acid-labile cisvinyl ether bond. The synthesis of 1-alkenylglycerols, 1-alkenyl-2,3-diacylglycerols (neutral plasmalogens) or 1-alkenyl-2-acyl-3halodeoxyglycerols is detailed by Slotboom and Bonsen (1970) and Paltauf (1983). These syntheses are carried out in five ways: (1) methods involving debromination of cyclic glycerol-(1-bromoalkyl) acetals; (2) syntheses involving partial reduction of a triple bond; (3) procedures involving thermal elimination of 1,2-diacylglycerol from bis(1,2diacylglycerol) acetals; (4) methods involving elimination of p-toluenesulphonic acid or hydrogen iodide; and (5) syntheses involving elimination of hydrogen from 1chloroethers. Methods for the synthesis of choline plasmalogens have been reviewed (Paltauf, 1983). Synthesis of a choline plasmalogen has been reported by Slotboom et al. (1967). An addition reaction between hexadec-1-enyl ethyl ether and rac-1,2-dipalmitoylglycerol was catalysed by p-toluenesulphonic acid to yield a mixed acetal (Figure 5.8). Thermal elimination of ethanol in the presence CH2OH CHOCOR2 + R1CH CHOEt
(i)
CH2OCOR2
CHOCOR2
(ii)
CH2OCOR2
CH2OCH CHR1 CHOCOR2
CH2OCH(OEt)CH2R1
(iii)
CH2OCH CHR1 CHOCOR2
plasmalogen
2
CH2OCOR
CH2OH Reagents: I, p-CH3C6H4SO3H; ii, sulphanilic acid (thermal elimination); iii, pancreatic lipase.
FIGURE 5.8
Synthesis of choline plasmalogen.
396
Synthesis
method is that of de Haas and van Deenen (1965), who prepared four out of the six possible monoacylphosphatidylcholines. Billimoria and Lewis (1968) used trityl groups for blocking, but showed that removing them by refluxing with 90% aqueous acetic acid for 10 min rather than by hydrogenolysis gave 100% of the 1-monoacyl product. Although this complete migration demonstrated the importance of neutral solutions in the final step, it does suggest that their method could be used for the preparation of unsaturated 1-monoacylphosphoglycerides (cf. Rosenthal, 1975). Eibl et al. (1970) formed a D-α-monoacylphosphatidylcholine using a 1,2,5,6-D mannitol tetrabenzyl ether intermediate. Arnold et al. (1967) described a method for the preparation of DL-α and β-ether monoacylphosphatidylcholines, while O-methylated and acetylated monoacylphosphatidylcholines were synthesised by Weltzein and Westphal (1967). Enantiomerically pure ester and other lysophospholipids were prepared by Eibl and Woolley (1988), while Filthuth and Eibl (1992) reported the synthesis of lysophosphatidylinositol.
also describe the preparation of the biologically active platelet-activating factor (PAF) (1-alkyl-2-acetyl-sn-glycero-3-phosphocholine) (see Section 5.6.12), while Wang and Tai (1990) reported the synthesis of an aldehydic analogue of PAF for use in the preparation of antibodies.
5.6.11
Monoacylphosphoglycerides
Since 1- and 2-monoacylphosphoglycerides are important metabolic intermediates, there has been some interest in their chemical synthesis. In addition, specific phospholipases can be used to prepare either the 1- or 2-monoacyl compounds. Thus, pancreatic lipase attacks the 1-position of a phosphoglyceride to yield the 2-monoacyl product, while snake venom phospholipase removes only the 2-acyl moiety (see Section 10.2.2). For the chemical methods, a major problem is acyl migration. This is particularly rapid for the 2-monoacylphosphoglyceride, since the primary ester is more stable. Therefore, the primary hydroxyl must be protected throughout the synthesis and the blocking group only released at the final stage. So far, methods are available for saturated monoacylphosphoglycerides. These are discussed by Slotboom and Bonsen (1970). For production of a 1-monoacyl compound, the 2 position of glycerol can be protected with a benzyl ether group. Slotboom et al. (1963, 1967) prepared rac-1-stearoyllyso-phosphoglycerides starting from rac-1-stearoyl-2benzyl-3-iododeoxyglycerol (cf. Rosenthal. 1975). Other methods are detailed in Slotboom and Bonsen (1970). In addition, Fujiwara et al. (1967) have reported that 80 to 90% of the 2-acyl ester of phosphatidylcholine can be cleaved with laurylamine or certain alkoxides. Several methods for the synthesis of 2-monoacyl-phosphoglycerides are described in Slotboom and Bonsen (1970). These use benzyl or trityl groups to protect the free hydroxyl group and the blocking group is removed by hydrogenolysis. If this is carried out in neutral medium, then acyl migration is minimized. A representative O R
C
H2C O
O
CH H2 C
CH
CH
R
P
Platelet activating factor (PAF) and analogues
Platelet activating factor (PAF) is most easily made by semisynthetic procedures beginning from choline plasmalogen. The acyl group is removed from the sn-2 position and acetylation then uses acetic anhydride (Figure 5.9; see Demopoulos et al., 1979). Alternative natural sources of ether lipids that can be used as starting materials include ethanolamine plasmalogen (Kumar et al., 1984) or 1-Oalkylglycerols from ratfish liver oil (Muramatsu et al., 1981). Complete chemical synthesis of PAF has been accomplished with either saturated or unsaturated alkyl groups (Figure 5.10; see Paltauf and Hermetter, 1994). An alternative method (Surles et al., 1985) allows the synthesis O
H2 C
1. H2/Pd 2. OH–
O O
5.6.12
CH
HO
OCH2CH2N+(CH3)3
H2C
R
O
P
OCH2CH2N+(CH3)3
O– H2 C
O H3C
CH2
O
O–
Ac2O
CH2
C
O
O
CH2
R
O
CH H2C
CH2
O
P
OCH2CH2N+(CH3)3
O–
FIGURE 5.9 Semisynthetic preparation of PAF from natural choline plasmalogen. (Adapted from Paltauf, F. and Hermetter, A. (1994) Prog. Lipid Res., 33, 239–328.)
397
5.6 Glycerophospholipids
O
H2C HO
1. Tritylchloride
CH H2C
O
O
CH H2C
Bn
H3C
CH2
O
CH
O
CH2
C
H3C
2. POCl3, Cholinetosylate OTr
1. Pd/H2
P
PAF
2. Ac2O, DMAP
+
OCH2CH2N (CH3)3 (a)
R 1. Alkyl-O-Ts, NaH Ph
2. HCl, dioxam, H2O 3. (CH3O)TrCl 4. PhCOCl, pyridine
Ph
O
CH2
C
O
CH H2C
O O
CH2
CO
CH H 2C
R O
O
Tr(OCH3)
O O
P
OCH2CH2N+(CH3)3
O–
1. TBA 2. Ac2O, DMAP
1. H+
R
1. HCl/dioxan–H2O 2. POCl3 3. Cholinetosylate
R
CH
O O
O
H2C
O–
HO
O
2. Benzylchloride, NaOH
OH H2C
Bn
H2C
R
PAF (b)
FIGURE 5.10 Total chemical synthesis of PAF. (Adapted from Paltauf, F. and Hermetter, A. (1994) Prog. Lipid Res., 33, 239–328.) R = 9(Z)octadecenyl; Tr(OCH3) = 3-Methoxytrityl; TBA = Tetrabutylammoniumhydroxide.
and ciliates (see Section 1.2.3.4). Methods have been reported for the synthesis of: (1) phosphonolipids with the C-P bond between the phosphorus moiety and the polar headgroup, i.e., diacyl-sn-glycerol-3(X)-phosphonates; (2) phosphonolipids derived from 1,2-propanediol with a phosphonic acid attached to C-3, i.e., 1,2-diacyloxy- (or alkoxy-) propylphosphonic acid (or ester); (3) the phosphono analogue of sphingolipids, i.e., ceramide-(X)phosphonate; and (4) phosphinic acid analogues with two C-P bonds, i.e., 1,2-diacyloxy- (or alkoxy-)propyl-(X)phosphinate. These structures are indicated in Figure 5.11. For details of the synthesis of such compounds, see Slotboom and Bonsen (1970) and Rosenthal (1975). Hori and Nozawa (1982) have reviewed phosphonolipids in detail. A number of phosphonolipid analogues of phospholipids have been prepared. These include 1-O-alkylethylene glycol phosphonic acid analogues of phosphatidylcholine and phosphatidylethanolamine (Moschidis, 1987), a phosphono analogue of diphosphatidylglycerol (Moschidis, 1988), chiral vinylic analogues of phosphatidic acid and phosphatidylcholine (Schwartz et al., 1988a) and α-hydroxyphosphonolipid analogues of phosphatidic acid (Schwartz et al., 1989b).
of racemic and optically active PAF containing an unsaturated alkyl chain in a rather short reaction series. A number of chiral adducts other than glycerol derivatives, including D- mannitol and D-or L-tartaric acids, have been used as starting materials for the synthesis of PAF (Paltauf and Hermetter, 1994). These authors also describe the synthesis of a whole series of PAF analogues. Such compounds are of great current interest as PAF agonists or antagonists where they can have clinical value, e.g., for the treatment of bronchial asthma. Analogues include PAF modified in the alkyl chain, those containing N- or S- substitutes, those with modifications of the glycerol backbone and those with alterations in the head group. In addition, over a hundred derivatives of PAF with nonhydrolyzable substituents on the sn-2 position of glycerol have been synthesised (Zeisig et al., 1990). Many have potential as anticancer compounds and some have been investigated in clinical trials.
5.6.13
Phosphonolipids
These compounds occur naturally in a large number of organisms, particularly sea anemones, aquatic molluscs
398
Synthesis
CH2OR1
CH2OR1 R2OCH
R2OCH O
O
CH2POCH2CH2X
CH2OPCH2CH2X OH
OH
(1)
(2) CH2OR1 O
2
R OCH O
H3C[CH2]12CH=CHCH(OH)CHCH2OPCH2CH2X NH2
OH
(3)
CH2PCH2CH2X OH (4)
FIGURE 5.11 General structures of types of phosphonolipids that have been synthesised. X represents a polar head-group in each case. Note the two C-P bonds for compounds in Group 4.
5.6.14
choline analogue via transphosphatidylation by phospholipase D from cabbage. Chem. Phys. Lipids, 41, 349–353. Ali, S. and Bittman, R, (1989) Mixed-chain phosphatidylcholine analogues modified in the choline moiety: preparation of isomerically pure phospholipids with bulky head groups and one acyl chain twice as long as the other. Chem. Phys. Lipids., 50, 11–21. Ali, S., and Bittman, R. (1990). Synthesis of optically active 1acyl-2-O-alkyl-sn-glycero-3-phosphocholine via 1-O-benzyl-sn-glycerol 3-arenesulfonate. Biochem. Cell Biol., 68, 360–365. Aneja, R. and Chadha, J.S. (1971a). Acyl-chloro-deoxyglycerophosphorylcholines: structure of the so-called cyclic lysolecithins. Biochim. Biophys, Acta, 239, 84–91. Aneja, R. and Chadha, J.S. (1971b). A total synthesis of phosphatidylcholines. Biochim. Biophys. Acta, 248, 455–457. Aneja, R., Chadha, J.S. and Davies, A.P. (1970). A general synthesis of glycerophospholipids. Biochim. Biophys. Acta, 218, 102–111. Arnold, D. et al. (1967). Concerning the synthesis of lysolecithin and its ether analogs. Justus Liebigs Ann. Chem., 40, 234–239. Arthur, G. and Bittman, R. (1998). The inhibition of cell signalling pathways by antitumor ether lipids. Biochim. Biophys, Acta, 1390, 85–102. Baer, E. and Kindler, A. (1962). L- α -(Dioleoyl)lecithin. Alternate route to its synthesis. Biochemistry, 1, 518–521. Barlow, P.N. et al. (1988). Synthesis and some properties of constrained short-chain phosphatidylcholine analogues: (+)- abd (–)-(1,3/2)-1-O-(phosphocholine)-2,3,-O-dihexyanoylcyclopentane-1,2,3-triol. Chem. Phys. Lipids, 46, 157–164. Beck, A. et al. (1990). 1-Palmitoyl-2-[3-(diphenylhexatrienyl) propanoyl]-sn-glycero-3-phosphoethanolamine as a fluroscent membrane probe. Synthesis and partitioning properties. Chem. Phys. Lipids, 55, 13–24. Berecoechea, J. et al. (1968). Phospholipid ethers. Synthesis of dipalmityl-1,2-oxy-3-glyceryl-phosphoric acid and dipyro (dipalmityl-1,2-oxy-3-glyceryl phosphoric) acid. J. Bull. Soc. Chim. Biol., 50, 1561–1567. Bergelson, L.D. (1980) Lipid Biochemical Preparations, Elsevier, Amsterdam.
Phospholipid analogues
Phospholipid analogues are unnatural compounds with high metabolic stability, but which are readily incorporated into cell membranes (Geilen et al., 1994). Disruption of plasma membrane signalling pathways by such membraneactive compounds could represent a new approach to chemotherapy (Arthur and Bittman, 1988). Most currently used analogues are based on lysophosphatidylcholine (lyso PC) or on lyso-PAF. Structures synthesised include glycerol-containing sn-2 substituted analogues, long-chain, glycerol-free phosphobase agents, sugar-containing analogues of lysoPC or lysoPAF and phosphocholinecontaining analogues of sphingomyelin with truncated acyl side chains. References to their synthesis, and details of their anticancer effects and possible mechanisms of action are given in Wieder et al. (1999). Many analogues of glycerophospholipids are useful in basic science. These include fluorescent, photoactivate, and isotopically or spin-labelled compounds. There are some extremely useful compendia for such analogues and references to their synthesis in Paltauf and Hermetter (1994). In addition, these authors give information about the synthesis of polymerisable phospholipids, for use in stable liposomes (Ringsdorf et al., 1988) or in the ultrathin membranes of sensors or biomedical devices (Nakaya et al., 1990).
References Abdelmageed, O.H. et al.(1989). Syntheses of 2H- and 13C-labelled 1,2-di-O-hexadecyl-sn-glycero- 3-phosphoethanolamines and 1,2-di-O-hexadecyl-sn-glycero-3-phosphocholines. Chem. Phys. Lipids, 50, 163–169. Achterberg, V. et al. (1986a). Biosynthetic preparation of radioactively labeled ethanolamine plasmalogen (1-O-[114C]octadec-1′-enyl-2-acyl-sn-glycero-3-phosphoethanolamine) using a protozoan cell culture. Chem. Phys. Lipids, 41, 137–145. Achterberg, V. et al. (1986b). Conversion of radiolabelled ethanolamine plasmalogen into the dimethylethanolamine and
399
5.6 Glycerophospholipids
Billimoria, J.D. and Lewis, K.O. (1968). The synthesis of phospholipids. I. Phosphatidyl and lysophosphatidylethanolamines. J. Chem. Soc., [Perkin 1], 12, 1404–1412. Bittman, R. (1999). Chemical synthesis of glycerophospholipids and their analogs. In Lipid Synthesis and Manufcture, Ed., F.D. Gunstone, Sheffield Academic Press, Sheffield, U.K., pp. 185–207. Bittman, R. et al. (1984). Synthesis of phospholipids via dimethylphosphoryl chloride. Chem. Phys. Lipids, 34, 201–205. Bonsen, P.P.M. and de Haas, G. H. (1967). Silver di-t-butyl phosphate, a useful reagent in the synthesis of phosphonolipids. Synthesis of mixed-acid phosphatidic acid and phosphatidyl glycerophosphate. Chem. Phys. Lipids, 1, 100–109. Bonsen, P.P.M. et al. (1965). Synthesis and enzymic hydrolysis of an O-alanyl ester of phosphatidyl glycerol. Biochim. Biophys. Acta. 106, 93–105. Bonsen, P.P.M. et al. (1966). The synthesis of 3-phosphatidyl-1glycerol. Chem. Phys. Lipids,. 1, 33–40. Bonsen, P.P.M. et al. (1972). Chemical synthesis of some lecithin analogues—potential inhibitors of phospholipase A. Chem. Phys. Lipids, 8, 199–220. Brockerhoff, H. and Ayengar, N.K.N. (1979). Improved synthesis of choline phospholipids. Lipids, 14, 88–89. Brockerhoff, H. and Yurkowski, M. (1965). Simplified preparation of L-α-glyceryl phosphoryl choline. Can. J. Biochem., 43, 1777. Bruzik, K.S. et al. (1986). A general method for the synthesis of glycerophospholipids. J. Org. Chem., 51, 2368–2370. Caldern, J. and Yague, P. (1986). Phosphoric acid derivatives. XV. Phosphatidylethanolamines and N, N-dimethyl-phosphatidylethanolamines from diacylglycerol (O-acetoxyphenyl hydrogen phosphates). Chem. Phys. Lipids, 41, 147–157. Caldern, J. and Yague, P. (1988). Phosphoric acid derivatives. XVI. Phosphatidylglycerols from diacylglycerol (Oacetoxyphenyl hydrogen phosphates). Chem. Phys. Lipids, 46, 121–125. Ceve, G. Ed. (1993) Phospholipid Handbook, Marcel Dekker, New York. Chacko, G.K. et al. (1967). Plasmalogen synthesis. Use of lalkynylglycerols and the production of allenic ethers. J. Org. Chem., 32, 3698–3700. Chandrakumar, N.S. and Hajdu, J. (1982). A new method for the stereospecific synthesis of ether phospholipids. Preparation of the amide analog of platelet-activating factor and related derivatives. Tertrahedron Lett., 23, 1043–1046. Cubero Robles, E. and Van Den Berg, D. (1969). Synthesis of lecithins by acylation of O-(sn-glycero-3-phosphoryl) choline with fatty acid anhydrides, Biochim. Biophys. Acta, 187, 520–526. Dale, M. P., and Robinson, N.C., (1988). Synthesis of cardiolipin derivatives with protection of the free hydroxyl: its application to the study of cardiolipin stimulation of cytochrome c oxidase. Biochemistry, 27, 8270–8275. Dang, Q.Q. et al. (1982). Synthesis and identification of bis (diacylglycero) phosphoric acid and bis (monoacylglycero) phosphoric acid. Lipids, 17, 798–802. de Haas, G.H. and van Deenen, L.L.M. (1961). Synthesis of enantiomeric mixed-acid phosphatides. Rec. Trav. Chim., 80, 951–970.
de Haas, G.H. and van Deenen, L.L.M. (1963). Cardiolipin and derivatives. 1. Synthesis of an acyl derivative of diphosphatidylglycerol. Rec. Trav. Chim., 82, 1163–1172. de Haas, G.H. and van Deenen, L.L.M. (1965). Structural identification of isomeric lysolecithins, Biochim. Biophys. Acta, 106, 315–325. de Haas, G.H. et al. (1964). Synthesis of mixed-acid phosphatidylserine containing unsaturated fatty acids. Rec. Trav. Chim., 83, 99–1168. Demopoulos, C.A. et al. (1979). Platelet-activating factor. Evidence for 1-O-alkyl-2-acetyl-sn-glyceryl-3-phosphorylcholine as the active component (a new class of lipid chemical mediators). J. Biol.Chem., 254, 9355–9358. Desai, T. et al. (1996) Synthesis of phosphatidylinositol phosphates. In Synthesis in Lipid Chemistry, Ed., J.H.P. Tyman, Royal Soc. of Chemistry, Cambridge, U.K., pp. 67–93 . Dreef, C.E. et al. (1988). Synthesis of 1-O-(1,2-di-O-palmitoylsn-glycero-3-phospho)-D-myo-inositol 4,5-bisphosphate an analog of naturally occurring PtdIns(4,5) P2. Tetrahedron Lett., 29, 6513–6516. Eibl, H. (1978). Phospholipid synthesis: oxozaphospholanes and dioxaphospholanes as intermediates. Proc. Natl. Acad. Sci. U.S.A., 75, 4074–4077. Eibl, H. and Lands, W.E.M. (1970). Phosphorylation of 1-alkenyl-2-acylglycerol and preparation of 2-acylphosphoglycerides. Biochemistry, 9, 423–428. Eibl, H. and Nicksch, A. (1978). The synthesis of phospholipids by direct amination. Chem. Phys. Lipids, 22, 1–8. Eibl, H. and Woolley, P. (1988). A general synthetic method for enantiomerically pure ester and ether lysophospholipids. Chem. Phys. Lipids, 47, 63–68. Eibl, H. et al. (1970). Synthesis of choline phosphatides, 6.3palmitoyl-glycerol-1-phosphorylcholine (D- α-lysolecithin) and its inactivity as substrate for acyltransferases. Liebig’s Ann. Chem., 738, 161–169. Ella, K.M. et al. (1997). Utilization of alcohols by plant and microbial phospholipase D. Biochem. Mol. Biol. Intl., 41, 715–724. Erukella, R.K. et al. (1995). Stereospecific and regiospecific ring opening of glycidol with primary and secondary alcohols mediated by diisobutyl aluminium hydride. J. Chem. Soc. Perkin Trans. I., 18, 2199–2200. Filthuth, E. and Eibl, H. (1992). Synthesis of enantiomerically pure lysophosphatidylinositols and alkylphosphoinositols. Chem. Phys. Lipids, 60, 253–261. Fowler, W.T. et al. (1988). Photoreactive cardiolipin analogues. Chem. Phys. Lipids, 47, 261–271. Fujiwara, Y. et al. (1967). The preferential cleavage of the β-ester linkage of egg yolk lecithin by treatment with alkoxides and laurylamine. Bull. Chem. Soc. Japan, 40, 230–231. Geilen, C.C. et al. (1994). Uptake, subcellular distribution and metabolism of the phospholipid analog hexadecylphosphocholine in MDCK cells. Biochim. Biophys. Acta, 1211, 14–22. Gigg, R. and Gigg, J. (1997). Synthesis of glycosylphosphatidylinositol anchors. In Glycopeptides and Related Compounds, Eds., D.G. Large and C.D. Warren, Marcel Dekker, New York, pp. 327–392. Gou, D-M. and Chen, C-S. (1994). Synthesis of L-phosphatidylD-myo-inositol 3,4,5-trisphosphate, an important
400
Synthesis
intracellular signalling molecule. J. Chem. Soc. Chem. Commun., 2125–2126. Gupta, C.M. et al. (1977). Glycerophospholipid synthesis: Improved general method and new analogs containing photoactivable groups. Proc. Natl. Acad. Sci. U.S.A., 74, 4315–4319. Hanahan, D.J. et al. (1990). Novel route to preparation of high purity lysoplasmenylethanolamine. J. Lipid Res., 31, 2113–2117. Hansen, W.J. et al. (1982). An improved procedure for the synthesis of choline phospholipids via 2-bromoethyl dichlorophosphate. Lipids, 17, 453–459. Harbison, G.S. and Griffin, R.G. (1984). Improved method for the synthesis of phosphatidylcholines. J. Lipid. Res., 25, 1140–1142. Hebert, N. and Just, G. (1990). Synthesis of phospholipids using an inverse phosphate triester approach. J. Chem. Soc. Chem. Commun., 1497–1498. Hendrickson, H.S. (1991). Phospholipase A2 assays with fluorphore-labelled lipid substrates. Methods in Enzymology, New York, 197, 90–94. Hendrickson, H.S. and Hendrickson, E.K. (1990). A facile asymmetric synthesis of glycerol phospholipids via tritylglycidol prepared by the asymmetric epoxidation of allyl alcohol. Thiolester and thioether analogs of phosphatidylcholine. Chem. Phys. Lipids, 53, 115–120. Hermetter, A. et al. (1983). An improved method for the preparation of “mixed-chain” phosphatidylethanolamines. Chem. Phys, Lipids, 32, 145–152. Hermetter, A. et al. (1987). 1-O, N-ditritylglycerophosphoethanolamine, a novel intermediate for the facile preparation of mixed-acid phosphatidylethanolamines. Chem. Phys. Lipids, 43, 69–77. Hermetter, A. et al. (1989). 1-O-trityl-sn-glycero-3-phosphocholine: a new intermediate for the facile preparation of mixed-acid 1,2-diacylglycerophosphocholines. Chem. Phys Lipids., 50. 57–62. Hessel, L.W. et al. (1954). Phosphatidic acids and derivatives. IV. Synthesis of iodohydrin. Recueil. Trav. Chim. Pays-Bas Belgique, 73, 150–156. Hirt, R. and Berchtold, R. (1958). Zur synthese der phosphatide. 2. Eine neue synthese der lecithine. Pharm. Acta Helv., 33, 349–356. Hori, T. and Nozawa, Y. (1982). Phosphonolipids. In Phospholipids, Eds., J.N. Hawthorne and G.B. Ansell, Elsevier, Amsterdam, pp. 95–128. Horrocks, L.A. and Sharma, M. (1982). Plasmalogens and Oalkyl glycerophospholipids. In Phospholipids, Eds., J.N. Hawthorne and G.B. Ansell, Elsevier, Amsterdam, pp. 51–93. Inoue, K. and Nojima, S. (1968). Immunochemical studies of phospholipids. II. Syntheses of cardiolipin and its analogues. Chem. Pharm. Bull. (Tokyo), 16, 76–81. Isaacson, Y. et al. (1990). The synthesis and molecular dynamics of phospholipids having hydroxylated fatty acids at the sn-2 position. Chem. Phys, Lipids, 52, 217–226. Jones, M. et al. (1989). Improved syntheses of inositol phospholipid analogs. Tetrahedron Lett., 30, 5353–5356. Joo, C.N. and Kates, M. (1968). Synthesis of the phytanyl diether analogs of phosphatidyl glycerophosphate and phosphatidyl
glycerol present in Halobacterium cutirubrum. Biochim. Biophys. Acta, 152, 800–803. Joo, C.N. and Kates, M. (1969). Synthesis of the naturally occurring phytanyl diether analogs of phosphatidyl glycerophosphate and phosphatidyl glycerol. Biochim. Biophys. Acta, 176, 278–297. Kates, M. (1978). The phytanyl ether-linked polar lipids and isoprenoid neutral lipids of extremely halophilic bacteria. Prog. Chem. Fats Other Lipids, 15, 301–342. Klyashchitski, B.A. et al. (1971). Derivatives of asymmetrically substituted myoinositol. VI. Synthesis of 2,3,4,5,6- and 1,2,4,5,6-penta-0-benzyl-sn-myo-inositols. Total synthesis of 1-0-(1, 2-dipalmitoyl-sn-glyceryl-phosphoryl)sn-myo-inositol. Zh. Obshch. Khim., 41, 1386–1393. Kodali, D.R. and Duclos, R.I. Jr. (1992) Debenzylation and detritylation by bromodiemethylborane: synthesis of mono-acid or mixed-acid, 1,2- or 2,3-diacyl-sn-glycerols. Chem. Phys. Lipids, 61, 169–173. Kumar, R. et al. (1984). A facile route to semi-synthesis of acetyl glycerylether phosphoethanolamine and its choline analogue. J. Lipid Res., 25, 198–208. Lai, C. et al. (1988). Spin-label assay for phospholipase A2. Anal. Biochem., 172, 397–402. Lapidot, Y. et al. (1969). The synthesis of diacyl-DL-(and –L-)αglycerol phosphates. Chem. Phys. Lipids, 3, 125–134. Lebeau, L. et al. (1988). Synthesis of phospholipids linked to steroid hormone derivatives. Chem. Phys. Lipids, 46, 57–62. Lemmen, P. et al. (1990). A synthesis of phospholipids using a phosphite-triester approach and uniformly cleavablehalogenaed protecting groups. Chem. Phys. Lipids, 53, 65–75. Lindh, I. and Stawinsky, J. (1989). A general method for the synthesis of glycerophospholipids and their analogs via H-phosphonate intermediates. J. Org. Chem., 54, 1338–1342. Loidl, J. et al. (1990). Uptake of fluorescent plasmalogen analogs by cultured human skin fibroblasts deficient in plasmalogen. Biochim. Biophys. Acta, 1049, 75–84. Lukyanov, A.V. et al. (1965). Complex lipids. Synthesis of α-(α', α-distearoyl)glycerylphosphoryl-(1-myo-inositol)ethanolamine. Zh. Or.Khim., 36, 1029–1031. Mangroo, D. and Gerber, G.E. (1988). Phospholipid synthesis: effects of solvents and catalysts on acylation. Chem. Phys. Lipids, 48, 99–108. Mason, J.T. Broccoli, A.V. and Huang, C-H. (1981). Method for the synthesis of isomerically pure saturated mixed-chain phosphatidylcholines. Anal. Biochem., 113, 96–101. Mena, P.L. and Djerassi, C. (1985). Synthesis of 5,9-hexacosadienoic acid phospholipids. 11. Phospholipid studies of marine organisms. Chem. Phys. Lipids, 37, 257–270. Molotkovsky, J.G. and Bergelson, L.D. (1971). The resolution of racemic alcohols through the acid oxalates. A new synthesis of optically active myo-inositol 1-phosphates. Tetrahedron Lett. 50, 4791–4794. Molotkovsky, J.G. et al. (1976). A new protective group in the synthesis of phospholipids. Chem. Phys. Lipids, 17, 108–110. Moore, J.W. and Szelke, M. (1970). Synthesis of threonine phosphoglycerides. Tetrahedron Lett., 50, 4423–4426.
401
5.6 Glycerophospholipids
Moschidis, M.C. (1987). Synthesis of 1-O-alkylethylene glycol phosphonic acid analogues of lecithin and cephalin. Chem. Phys. Lipids, 43, 39–45. Moschidis, M.C. (1988). Synthesis of a phosphono analogue of cardiolipin and study of its silicic acid column chromatographic behaviour. Chem. Phys. Lipids, 46, 253–257. Muramatsu, T, et al. (1981). A facile method for the preparation of 1-O-alkyl-2-O-acetoyl-sn-glycero-3-phosphocholines (platelet activating factor). Chem, Phys, Lipids, 29, 121–127. Nakaya, T. et al. (1990). Polymerized phospholipid LangmuirBlodgett multilayer films. Langmuir, 6, 291–293. Nicholas, A.W. et al. (1983). Synthesis of mixed-acid phosphatidylcholines and high-pressure liquid chromatographic analysis of isomeric lysophosphatidylcholines. Lipids, 18, 434–438. Noda, N. and Keenan, R.W. (1990). Phosphatidylinositol isopropylidene derivatives: synthesis, characterization and use as synthetic intermediates. Chem. Phys. Lipids, 53, 53–63. Paltauf, F. (1976). An improved synthesis of 1-O-[3H]-alkyl2-acyl-sn-glycerol-3-phosphorylethanolamine with an unsaturated acyl chain. Chem. Phys. Lipids, 17, 148–154. Paltauf, F. (1983). Biosynthesis of 1-O-(1-alkenyl) glycerolipids (plamalogens). In Ether Lipids, Eds., H. K. Mangold and F. Paltauf, Academic Press, New York, pp. 107–128. Paltauf, F. and Hermetter, A. (1991). Preparation of alkyl ether and vinyl ether substrates for phospholipases. Meth.Enzymol., 197,134–148. Paltauf, F. and Hermetter, A. (1994) Strategies for the synthesis of glycerophospholipids. Prog. Lipid Res., 33, 239–328. Patel, K.M., Morrisett, J.D. and Sparrow, J.T. (1979). The conversion of phosphatidylethanolamine into phosphatidycholine labeled in the choline group using methyl iodide, 18-crown-6 and potassium carbonate. Lipids, 14, 596–597. Payrastre, B. (2004) Phosphoinositides. In Bioactive Lipids (Eds., A. Nicolaou and G. Kokotos), The Oily Press, Bridgewater, U.K., pp. 63–84. Perochon, E. and Tocanne, J.-F. (1991). Synthesis and phase properties of phosphatidylcholine labelled with 8-(2anthroyl)-octanoic acid, a solvatochromic fluorescent probe. Chem. Phys, Lipids, 58, 7–17. Reynolds, L.H., Washburn, W.N., Deems, R.A. and Dennis, E.A. (1991). Assay strategies and methods for phospholipases. Meth. Enzymol., 197, 3–23. Ringsdorf, H., Schlarb, B. and Venzmer, J. (1988). Molecular architecture and function in polymeric oriented systems. Models for the study of organization, surface recognition, and dynamics in biomembranes. Angew. Chem. Int. Ed., Engl., 27, 113–158. Roberts, M.F. (1991). Nuclear magnetic resonance spectroscopy to follow phospholipase kinetics and products. Meth. Enzymol., 197, 31–48. Rosenthal. A. F. (1975). Chemical synthesis of phospholipids and analogues of phospholipids containing carbon-phosphorus bonds. Meth. Enzymol., 35, 429–529. Saunders, R.M. and Schwartz, H.P. (1966). Synthesis of phosphatidylglycerol and diphosphatidylglycerol. J. Am. Chem. Soc., 88, 3844–3847.
Schacht, J. (1981). Extraction and purification of polyphosphoinositides. Meth. Enzymol., 72, 626–631. Schwartz, P.W. et al. (1989a). Phosphonolipids. The synthesis of chiral vinylic phosphonolipid analogues of phosphatidic acid and phosphatidyl choline. Chem. Phys. Lipids, 49, 1–7. Schwartz, P.W. et al. (1989b). Phosphonolipids. 2. α-Hydroxyphosphonolipid analogues of phosphatidic acid. Chem. Phys. Lipids, 49, 131–134. Shashidhar, M.S. et al. (1991). A chromogenic substrate for phosphatidylinositol-specific phospholipase C: 4-nitrophenyl myo-inositol-1-phosphate. Chem. Phys. Lipids, 60, 101–110. Shvets, V.I. et al. (1967). Complex lipids. XXVIII. Synthesis of highly unsaturated phosphatidylserines. Zh. Obschch Khim., 37, 1454–1458. Sisk, R. B. and Huang, C-H. (1992). Calorimetric studies on the influence of N-methylated headgroups on the mixing behavior of diheptadecanoyl phosphatidylcholine with 1behenoyl-2-lauroylphosphatidylcholine. Biophys. J., 61, 593–603. Slotboom, A.J. and Bonsen, P.P.M. (1970). Recent developments in the chemistry of phospholipids. Chem. Phys. Lipids, 5, 301–397. Slotboom, A.J. et al. (1963). On the synthesis of isomeric lysophosphatides. Rec. Trav. Chim., 82, 469–486. Slotboom, A.J. et al. (1967). On the synthesis of plasmalogens. Chem. Phys. Lipids, 1, 192–208. Stepanov, A.E. and Shvets, V.I. (1986). Formation of phosphoester bonds in phosphoglyceride synthesis. Chem. Phys. Lipids, 41, 1–51. Stumpf, R. and Lemmen, P. (1990). Syntheses of phospholipids via oxazaphospholanes. Z.Naturforsch B., 45, 1729–1731. Surles, J.R. et al. (1985). Facile synthesis of platelet-activating factor and racemic analogs containing unsaturation in the sn-1-alkyl chain. J. Med. Chem., 28, 73–78. Takahashi, T. et al. (1967). Rapid method for the preparation of cardiolipid by ion exchange. Bull. Soc. Chim. Biol., 49, 65–70. Van der Bend, R.L. et al. (1992). Metabolic conversion of the biologically active phospholipid, lysophosphatidic acid, in fibroblasts. Biochim. Biophys. Acta, 1125, 110–112. Verheij, H.M., Smith, P.F., Bonsen, P.P.M. and van Deenen, L.L.M. (1970). The chemical synthesis of a phosphatidylglucose. Biochim. Biophys. Acta, 218, 97–101. Verheij, H.M. et al. (1971). The chemical synthesis of two isomers of glucosaminylphosphatidylglycerol. Chem. Phys. Lipids., 6, 46–57. Vinogradov, S.V. et al. (1995) Synthesis of phospholipid-inhibitor conjugates by enzymatic transphosphoatidylation with phospholipase. D. J. Am. Chem. Soc., 115, 10487–10491. Wang, C.-J. and Tai, H.-H. (1990). A facile synthesis of an aldehydic analog of platelet activating factor and its use in the production of specific antibodies. Chem. Phys. Lipids, 55, 265–273. Ward, J.G. and Young, R.C. (1988). Synthesis of 1,2-dipalmitoylsn-glycer-3-yl-D-myo-inositol 1-phosphate. Tetrahedron Lett., 29, 6013–6016.
402
Synthesis
Warner, T.G. and Benson, A.A. (1977). An improved method for the preparation of unsaturated phosphatidylcholines: acylation of sn-glycero-3-phosphorylcholine in the presence of sodium methylsulfinylmethide. J. Lipid. Res., 18, 548–552. Watanabe, Y., Hirofuji, H. and Ozaki, S. (1994). Synthesis of a phosphatidylinositol 3,4,5-trisphosphate. Tetrahedron Lett., 35, 123–124. Weltzein, H.A. and Westphal, O. (1967). O-Methylated and Oacetylated lysolecithin. Justus Liebig’s Ann. Chem., 709, 240–243. Wieder, T., Reutter, W., Orfanos, C.E. and Geilen, C.C. (1999) Mechanism of action of phospholipid analogs as anticancer compounds. Prog. Lipid. Res., 38, 249–259. Woolley, P. and Eibl, H. (1988). Synthesis of enantiomerically pure phospholipids including phosphatidylserine and phosphatidylglycerol. Chem. Phys. Lipids, 47, 55–62. Yu, L. and Dennis, E.A (1991). Thio-based phospholipase assay. Meth. Enzymol, 197, 65–75. TABLE 5.2
Zeisig, R., Arndt, D. and Brachitz, H. (1990). Ether lipids — synthesis and application in tumor therapy. Die Pharmazie, 45, 809–818
5.7
Sphingolipids
Two general reviews of the chemistry and synthesis of sphingolipids are those of Hakomori (1983) and of Shapiro (1969). Table 5.2 indicates general methods that have been used for the formation of sphingosines or glycosphingolipids. Specific references for the various methods are shown in the table and in the text below. Gigg and Gigg (1966) and Gigg et al. (1966) have described two ways of synthesizing 4-D-hydroxysphinganine from D-galactosamine and 4-galactose. Gaver and Sweeley (1966) and Mendershausen and Sweeley (1969) prepared 3-oxo derivatives of sphinganine, N-acetyl-4trans-sphingenine, and N-carbobenzosphinganine. Stoffel
Summary of methods for the chemical synthesis of sphingosines and glycosphingolipids
Compound
Method
Sphingosines DL-Sphingosine Synthesis and separation of D- and L-sphingosines Differential synthesis of trans, erythro and threo; cis, erythro and threo
Reduction of DL-erythro-2-acetamido-3-keto-4-transoctadecanoic acid Resolution of DL-erythro-2-acetamido-3-hydroxy-4-transoctadecanoate through L(+)-acetylmandeloyl ester Condensation of 2-hexadecynal and 2-nitroethanol, followed by reduction
Dihydrosphingosine Dihydrosphingosine, DL-threo-, erythro-mixture Dihydrosphingosine D-Erythro- dihydrosphingosine D-Erythro- or D-threo-dihydrosphingosine D-Erythro- dihydrosphingosine
14
C-labelled DL-dihydrosphingosine followed by racemic separation
Oxamination of 3-keto-octadecanoate, followed by reduction with LiAlH4 Condensation of palmital and β-nitroethanol LiAlH4 reaction of β-benzamido-β-hydroxyoctadecanoic acid methyl ester Ammonolysis of trans or cis epoxide Reduction of β-N-acetylamino-β-keto-octanoate by LiAlH4 followed by treatment with methyl dichloroacetate and differential crystallization Separation of DL-sphingosine into D and L diastereoisomers
Phytosphingosine D-Ribo-D-amino-1,3,4-trihydroxysphinganine
Cerobrosides (Cerasine, phrenosine and glucocerebroside) Lactosylceramide (cytolipin H) Gangliotriaosylceramide (Tay-Sachs’ globoside) Sulfatide (3-O-sulphate galactosylceramide)
Glucosamine derivative (methyl-α-2-benzamido-2-deoxyglucofuranoside) is converted to the oxazoline, the 2allosamine derivative; compound is converted by oxidation to aldehyde, which is then coupled to olefin by Wittig reagenttriphenyltridecylphosphonium bromide; resulting product is reduced by catalytic hydrogenation N-Fatty acyl-3-O-benzoylsphingosine condensed with acetobromo sugars by Koenig-Knor type reaction; Hg(CN)2 as a catalyst N-Fatty acyl-3-O-benzoylsphingosine condensed with heptaacetyllactosyl bromide; Hg(CN)2 as a catalyst Condensation of decaacetylgangliotriaosyl chloride and Nstearoyl-3-O-benzoylsphingosine Reaction of 2,4,6-tri-O-acetyl-α-D-galactopyranosyl bromide with N-octadecanoyl-3-O-benzoyl-DL-dihydrosphingosine
Glycolipid analogues Gal-β(1→6) Gal-β(1→152)ceramide Gal-β(1→3) Gal-β(1→1)ceramide
Reaction analogous as for synthesis of lactosylceramide (see above)
See Hakomori, (1983) for details.
403
5.7 Sphingolipids
sulphatides and gangliosides. After reprecipitating the cerebrosides, these are benzoylated and the benzoate separated by silicic acid chromatography into fractions containing nonhydroxy and hydroxy fatty acids. Finally, the benzoyl groups are hydrolysed with methanolic sodium methoxide (Bergelson, 1980). Glucosylceramide can be synthesised by condensation of acetobromoglucose with 3-O-benzoylceramide. Benzoylceramide can be prepared by the method of Hay and Gray (1969) and is condensed with acetobromoglucose in the presence of mercuric cyanide catalyst (Bergelson, 1980). The benzoyl group is removed with methanolic sodium methoxide and the O-β-D-glycosyl (1→1) ceramide purified by silicic acid column chromatography. Fluharty et al. (1974) have reported the isolation of 35S-labelled cerebroside sulphate. They injected the brains of young rats intracerebrally with [35S]-sulfate and sacrificed the animals 3 days later. Total lipids were extracted and glycerolipids destroyed by alkaline hydrolysis. TEAEcellulose column chromatography was then used to purify [35S]-sulphatide. Jatzkewitz and Nowoczek (1967) have synthesised D-galactose 3-sulfate and shown it to be identical to the galactose sulgate present in brain sulfatides. Sulfatide can also be made by sulfation of 2,4,6-tri-Oacetylgalactopyranosylceramide followed deacetylation (Ishizuku et al., 1980). In addition, cerebroside 6-sulfate has been prepared by sulfation of cerebroside (Holmgren et al., 1980). Psychosine (1-O-galactosylsphingosine) was prepared by Radin (1974) starting from O-β-D-galactosyl (1→1) ceramide. The latter was deacylated by refluxing with aqueous potassium hydroxide in butanol. The potassium hydroxide was removed by precipitation as potassium perchlorate and fatty acids with hexane extraction. Finally, a chloroform/ methanol/water mixture allows partition of excess perchloric acid into the upper phase, the chloroform phase containing the psychosine product (cf. Bergelson, 1980). Sphingomyelin can be obtained from bovine brain following acetone treatment, by ethanol and petroleum ether extraction. It is separated from cerebrosides by aluminum oxide column chromatography and phosphoglycerides are removed by mild alkaline hydrolysis. Silicic acid column chromatography allows final purification (Hanahan, 1961). DL-2-N-Stearoyldihydrosphingomyelin was synthesised by Zvonkova et al. (1974). They started with DL-N-stearoylsphinganine and tritylated the primary hydroxyl group on C-1. The hydroxyl on C-3 was then benzoylated, when the trityl group could be removed with hydrogen chloride. The unmasked hydroxyl group was then phosphorylated with βchloroethylphosphodichloride to yield 1-O-(β-chloroethylphospho)-3-O-benzochloride. The latter was treated with trimethylamine to generate the quaternary ammonium of choline. Removal of the benzoyl group then yielded DL-2N-stearoyl-dihydrosphingomyelin. Stoffel (1975) has reported a method for the chemical synthesis of choline-labelled sphingomyelins. These
and Sticht (1967) have described a simple method for the synthesis of 3-dehydrosphinganine hydrochloride. For the preparation of radiolabelled sphinganines and sphingenines, refer to Stoffel and Sticht (1967) and to Stoffel et al. (1971) for the synthesis of radiolabelled sphingomyelin. For a summary of these syntheses and other chemical preparations of sphingolipids, see Shapiro (1969). A very thorough review of the synthesis of racemic sphingosines, chiral pool syntheses, and enantiomeric syntheses is given by Jung and Schmidt (1999). In addition to sphingosine, methods for the preparation of sphinganine, L-erythro-sphingosine and threo-sphingosine are also detailed. Furthermore, racemic and enantioselective syntheses of phytosphingosine are also described (Jung and Schmidt, 1999). The authors recommend the use of azido derivatives of sphingosine and phytosphingosine for the production of glycosphingolipids (see also CastroPolomino et al., 1997; Kratzer and Schmidt, 1995). This is also described in detail in Schmidt (1996). Barenholz and Gatt (1975) have described techniques for the separation of sphingosine bases by silicic acid chromatography. Bovine spinal cord lipids are used as starting material and hydrolysed in acid-methanol. Sphingosine and dihydrosphingosine can then be isolated from the hydrolysate by chromatography. For the isolation of phytosphingosine and dihydrosphingosine, a fraction obtained from the growth medium of the yeast Hansenula ciferri was used. Ceramides are often prepared from sphingomyelins by phospholipase C digestion (cf. Bergelson, 1980). Hay and Gray (1969) prepared ceramides by such methods. These were then tritylated and the 3-O-benzoyl ester was formed. After detritylation, the 3-O-benzoylceramide was condensed with the appropriate acetobromo sugar to form various neutral ceramides. Flowers (1967) described a twostep method for lactosylceramide preparation. Individual species of ceramides can be made by N-acylation of DL-sphinganine or sphingenine with stearic acid or with DL-2-acetoxypalmitic acid in the presence of a mixed carbodiimide. The free hydroxyl acid ceramide can then be generated by alkaline hydrolysis from the acetate and separation of diastereoisomers by preparative silica gel thin layer chromatography (TLC) (Hammarstrom, 1971). The ceramides prepared by one of the above methods can be conveniently analysed by TLC and characterized by enzymic hydrolysis. Particular points to pay attention to during these syntheses are summarized by Bergelson (1980). Galactosylceramide can be conveniently isolated from bovine brain tissue. If this organ is extracted successively with acetone, diethyl ether and then hot aqueous ethanol, sphingolipids are obtained in the final fraction. These precipitate on cooling and the residue is washed with acetone and dissolved in hot acetic acid. On cooling they can again be precipitated, dissolved in pyridine and passed through an aluminum oxide column to separate cerebrosides from 404
Synthesis
borohydride. After dialysis, the product is passed through a Dowex-50 column and freeze-dried (Bergelson, 1980). Successful synthesis of ganglio-N-triosylceramide (TaySachs’ globoside) was based on the successful synthesis of ganglio-N-triose. After the trisaccharide had been made, it was converted to octaacetyl bromide and, subsequently, condensed with 3-O-benzoyl-N-stearoylsphingosine (see Hakomori, 1983). Analogous procedures can be used to synthesise various neutral ceramides (Flowers, 1966). Gangliosides with fluorescent or paramagnetic probes on the lipid moiety have also been synthesised chemically (Acquotti et al., 1986). The Tay-Sachs’ ganglioside (GM2) has been prepared, which is specifically labelled in either the N-acetylneuraminosyl or N-acetylgalactosaminyl portions (Tallman et al., 1975). For the first type of labelling, GH1 is an intermediate and, for the second, GH3 is isolated from erythrocytes first and then converted to GM2. Methods for the preparation of inositol-containing sphingolipids are given in Schmidt (1996). These include D-erythro-ceramide-1-phosphoinositol and ceramidecontaining GPI (glycosyl phosphatidylinositol) anchors. Two useful review articles covering the isolation and purification of sphingolipids are those of Rouser et al. (1969) and Kundu (1981). Rouser et al. cover the use of DEAE- and TEAE-cellulose column chromatography, while Kundu gives a thorough summary of TLC methods. The latter includes solvent systems, detection methods, analytical and preparative methods, and radio-autography (fluorography). Skipski (1975) and Christie (2003) also discuss separation methods. Skipski (1975) recommends the use of primuline as a nondestructive spray for sphingolipids because it is sensitive and does not interfere with the subsequent analysis of carbohydrate, fatty acids, and sphingosines by gas-liquid chromatography (GLC) or with the colorimetric determination of sphingosines by the method of Yamamoto and Rouser (1970). Various analogues of sphingolipids have been synthesised, particularly those in which the glycosidic linkages are different from natural compounds (see Hakomori, 1983).
compounds are demethylated to the corresponding ceramide-1-phosphoryl-N,N-dimethylethanolamines. The purified compounds are then quaternized in high yield with [14C]- or [13C]- methyl iodide to yield the respective sphingomyelins, which are labelled in their choline moieties. The technique allows the preparation of sphingomyelins of high specific radioactivity regardless of their degree of unsaturation. Because many of the previous methods to make sphingomyelin were either lengthy or resulted in racemic mixtures, the use of azidosphingosine as starting material has been employed (Dong and Butcher, 1993; Schmidt, 1996). The method could be used to make ceramide-1-phosphate and sphingomyelin. Using 3-O-protected azidosphingosine, an analogous sequence could be used for sphingosine-1-phosphate and lysosphingomyelin (Kratzer and Schmidt, 1995). The isolation of gangliosides from brain has been reviewed by Kanfer (1969) and by Bergelson (1980). Other specific references are Suzuki (1965), Svennerholm et al. (1972) and Laine et al. (1974), and general comments on ganglioside separations will be found in Kates (1986) and Christie (2003). A typical isolation procedure makes use of the great hydrophilicity of gangliosides. Thus, a chloroform-methanol extract, when subjected to Folch partition, has most of the gangliosides in the upper phase. Phospholipid contaminants of this phase can be removed by alkaline hydrolysis and the gangliosides precipitate as their barium salts. From these the free acids can be generated and then the gangliosides can be crystallized from solution. Other methods of isolating total gangliosides use gel filtration and DEAE-cellulose chromatography for purification (cf. Bergelson, 1980). Some specific isolations are as follows. Haematoside (GM3) can be obtained from rat liver by Folch partition, dialysis, Sephadex G-50 filtration and preparative TLC on silica gel G. Chloroform/methanol/2.5 M NH4OH (60/35/ 8, by volume) is used as solvent and the zones are detected with iodine vapour or, better, with a nondestructive spray, such as primuline (Skipski, 1975). Haematoside is the least polar of the gangliosides and has a Rf of about 0.5. It can be eluted from the silica gel with chloroform/ methanol/ water (60/35/8, by volume). Monosialyltetrahexosylceramide (GH1) and disialyltetrahexosylceramide (GD1α) were isolated from bovine brain total gangliosides by silicic acid column chromatography (Svennerholm et al., 1972). Ganglioside GD1α has also been tritiated. The method involves catalytic addition of tritium gas to the unsaturated centres of the gangliosides isolated from beef brain. Because this results in a partial destruction of the ganglioside, the tritiated GD1α has to be purified by Folch partition, dialysis, and preparative TLC (Bergelson, 1980). Alternatively, purified beef brain ganglioside GD1α can be oxidized selectively in its N-acetylneuraminic acid residue with sodium metaperiodate. The oxidized ganglioside is then dialysed and reduced with tritiated
References Acquotti, D. et al. (1986). A new chemical procedure for the preparation of gangliosides carrying fluorescent or paramagnetic probes on the lipid moiety. Chem. Phys. Lipids, 40, 71–86. Barenholz, Y. and Gatt, S. (1975). Separation of sphingosine bases by chromatography on columns of silica gel. Meth. Enzymol., 35, 529–533. Bergelson, L.D. (1980) Lipid Biochemical Preparations, Elsevier, Amsterdam. Castro-Palomino, J.C. et al. (1997) Efficient synthesis of ganglioside GM2 for use in cancer vaccines. Angew. Chem. Int. Ed. Engl., 36, 1998–2001. Christie, W.W. (2003). Lipid Analysis, 3rd ed., The Oily Press, Bridgwater, U.K.
405
5.8
Glycosylglycerides
Dong, Z., and Butcher, J.A. (1993). An efficient route to Npalmitoyl-D-erythro-sphingomyelin and its carbon-13labelled derivatives. Chem. Phys. Lipids, 66, 41–46. Flowers, H.M. (1966). Substituted cerebrosides. II. Synthetic dihydrosulfatides. Carbo. Res., 2, 371-379. Flowers, H.M. (1967). Substituted cerebrosides. III. 1-O-(4-Oβ-D-hexopyranosyl--D-glucopyranosyl) dihydroceramides. Carb.. Res., 4, 42–48. Fluharty, A.L. et al. (1974). Simplified procedure for preparation of 35S-labelled brain sulfatide. Lipids, 9, 865–869. Gaver, R.C. and Sweeley, C.C. (1966). Chemistry and metabolism of sphingolipids. 3-Oxo derivatives of N-acetylsphingosine and N-acetyldihydrosphingosine. J. Am. Chem. Soc., 88, 3643–3647. Gigg, J. and Gigg, R. (1966). A synthesis of phytosphingosines from D-galactose. J. Chem. Soc., 20, 1876–1879. Gigg, J. et al. (1966). A synthesis of phytosphingosines from Dglucosamine. J. Chem. Soc., 20, 1872–1876. Hakomori, S.-I. (1983). Chemistry of glycosphingolipids. In Sphingolipid Biochemistry (Eds., J. N. Kanfer and S I. Hakomori), Plenum, New York, pp. 1–64. Hammarstrom, S. (1971). A convenient procedure for the synthesis of ceramides. J. Lipid Res., 12, 760–765. Hanahan, D.G. (1961). Sphingomyelin. Biochem. Prep., 8, 121–124. Hay, J.B. and Gray, G.M. (1969). The partial synthesis of some naturally occurring glycosphingolipids with special reference to O-β-D-galactosyl-(1→4)-O-β-D-galactosyl(1→1)-ceramide. Chem. Phys. Lipids, 3, 59–69. Holmgren, J. et al. (1980). Sendai virus receptor: proposed recognition structure based on binding to plastic-absorbed gangliosides. Proc. Natl. Acad. Sci. USA, 77, 1947–1950. Ishizuku, I. et al. (1978). Sulfated glyceroglycolipids in rat brain. Structure, sulfation in vivo, and accumulation in whole brain during development. J. Biol. Chem., 253, 898–907. Jatzkewitz, H. and Nowoczek, G. (1967). Synthesis of sulphur35-labelled D-galactose sulphate and ceramide D-galactose sulphate (cerebroside sulphate). Chem. Ber., 100, 1667–1674. Jung, K.H., and Schmidt, R.R. (1999). Sphingolipids. In Lipid Synthesis and Manufacture, Ed., F.D. Gunstone, Academic Press, Sheffield, U.K., pp. 208–249. Kanfer, J.W. (1969). Preparation of gangliosides. Meth. Enzymol., 14, 660–664. Kates, M. (1986) Techniques in Lipidology, 2nd ed., Elsevier, Amsterdam. Kratzer, B. and Schmidt, R. R. (1995). Efficient synthesis of sphingosine-1-phosphate, ceramide-1-phosphate, lysosphingomyelin and sphingomyelin, Liebig’s Ann., 957–963. Kundu, S.K. (1981). Thin-layer chromatography of neutral glycosphingolipids and gangliosides. Meth. Enzymol., 72, 185–204. Laine, R.A. et al. (1974). Isolation and characterization of membrane glycosphingolipids. Meth. Membr. Biol., 2, 205–244. Mendershausen, P.B. and Sweeley, C.C. (1969). Chemistry and metabolism of sphingolipids. Chemical synthesis of 2amino-1-hydroxyoctadecan-3-one(3-ketosphinganine). Biochemistry, 8, 2633–2635. Radin, N.S. (1974). Preparative isolation of cerebrosides (galactosyl and glucosyl ceramide). J. Lipid Res., 17, 290–293.
Rouser, G. et al. (1969). Diethylaminoethyl and triethylaminoethyl cellulose column chromatographic procedures for phospholipids, glycolipids and pigments. Meth. Enzymol., 4, 272–317. Schmidt, R.R. (1996). Sphingosines, phosphosphingolipids and glycosphingolipids. In Synthesis in Lipid Chemistry, Ed., J.H.P. Tyman, Royal Soc. Chem., Cambridge, pp. 93–118. Shapiro, D. (1969). Chemistry of Sphingolipids, Hermann, Paris. Skipski, V.P. (1975). Thin-layer chromatography of neutral glycosphingolipids. Meth. Enzymol., 35, 396–425. Stoffel, W. (1975). Chemical synthesis of choline-labelled lecithins and sphingomyelins. Meth. Enzymol., 35, 533–541. Stoffel, W. and Sticht, G. (1967). Metabolism of sphingosine bases, II. Studies on the degradation and transformation of [3-14C]erythro-DL-dihydrosphingosine, [7-3H]erythroDL-dihydrosphingosine in rat liver. Hoppe-Seyler’s Z. Physiol.Chem., 348, 1345–1351. Stoffel, W. et al. (1971). A simple chemical method for labelling phosphatidylcholine and sphingomyelin in the choline moiety. Hoppe-Seyler’s Z. Physiol.Chem., 352, 1058–1064. Suzuki, K. (1965). The pattern of mammalian brain gangliosides. II. Evaluation of the extraction procedures, postmortem changes and the effect of formalin preservation. J. Neurochem., 12, 629–638. Svennerholm, L. et al. (1972). Gangliosides, isolation. Methods Carb. Chem. 6, 464–474. Tallman, J.F. et al. (1975). The preparation of Tay-Sachs ganglioside specifically labelled in either the N-acetylneuraminosyl or N-acetylgalactosaminyl portion of the molecule. Meth. Enzymol., 38, 541–548. Yamamoto, A. and Rouser, G. (1970). Spectrophotometric determination of molar amounts of glycosphingolipids and ceramide by hydrolysis and reaction with trinitrobenzenesulfonic acid. Lipids, 5, 442–444. Zvonkova, E.N. et al. (1974). Method for synthesis of sphingomyelins from ceramides. Khim Prir. Sojed, 5, 553–558.
5.8
Glycosylglycerides
Analytical separation and small-scale isolation of galactosylglycerides can be carried out conveniently by TLC on silica gel G (cf. Hitchcock and Nichols, 1971; Harwood, 1980). Sastry (1974) has discussed the various solvent systems in some detail, and the method of Khan and Williams (1977) is particularly useful. For larger-scale preparations, column chromatography by the method of O'Brien and Benson (1964) will be found to be good for the isolation of all three major plant glycosylglycerides. Yields of the two galactosylglycerides may be improved by the use of repeated chromatography on silica gel or on carbon-Celite columns (van der Veen et al., 1967). A rapid method of batch elutions from silica gel has been reported by de Stefanis and Ponte (1969). Most of the fractions obtained by these methods are slightly contaminated, but can be conveniently cleaned up by TLC or, in the case of phospholipid contaminants, by passage through Amberlite MB-3 (Sastry and Kates, 1964). Different methods of isolation and good advice on various techniques for the separation of glycosylglycerides are described by Heinz (1996). 406
Synthesis
Seminolipids (sulfated galactosylacylalkylglycerols occurring in testes and brain of animals) (see Murray and Narasimhan, 1990), α-glucosyldiacylglycerols (occurring in bacteria) (see Kates, 1990b), and monoglycosyldialkylglycerols have all been synthesised chemically (see Gigg and Gigg, 1990, for more details). Glycosylglycerols have been synthesised by several workers. Wickberg (1958) reported the synthesis of O-αD-galactopyranosyl- and O-β-D-galactopyranosyl (1→1)D-glycerols using a condensation reaction between tetra-O-acetyl-α-D-galactopyranosyl bromide and methylene-bis-2-O-(3-O-benzoyl-D-glycerol). He also prepared the corresponding L-glycerol galactosides. A second method was used by Charlson et al. (1957) for the formation of O-α-D-galactopyranosyl- and O-β-D-galactopyranosyl(1→2)glycerols. Brundish and Baddiley (1968) have reported the formation of O-α-D-glucopyranosyl- and O-β-D-glucopyranosyl(1→1)-D-glycerols and their anomeric L-glycerol galactosides. They have also synthesised four positional isomers of diglucosylglycerols. The major glycosylglyceride of Pneumococcus, O-α-D-galactopyranosyl(1→2)-O-α-D-glucopyranosyl(1→1)-D-glycerol, was synthesised by Brundish et al. (1967). Sulfoquinovosylglycerol has been synthesised by Miyano and Benson (1962). Wehrli and Pomeranz (1969) reported the synthesis of 1,2-dipalmitoyl- and 1-palmitoyl-2-linoleoyl-3-O-(-β-Dgalactopyranosyl)-sn-glycerols. They acylated the primary hydroxyl groups of 2,5-methylene-D-mannitol and then cleaved the mannitol moiety between C-3 and C-4 with lead tetraacetate. The resultant aldehyde was reduced and galactose attached by the Koenigs–Knorr reaction. The acetal was then hydrolysed and the hydroxyl group acylated.
A useful reference for general chemical procedures to synthesise glycoglycerolipids is that of Gigg and Gigg (1990). They cover the synthesis of mono-, di- and triglycosyldiacylglycerols, glycosylalkylacylglycerols, glucosyldialkylglycerols, and some phosphatidylglucosyldiacylglycerols. Wehrli and Pomeranz (1969) prepared the chiral β-galactosyldiacylglycerol by a method shown in Figure 5.12. Condensation of acetobromogalactose with the glycerol derivative, bis(1-acyl-2-methylene-glycerol), was followed by acid hydrolysis to give the acetylated galactosylmonoacylglycerol. Reaction with acid chloride then gave acetylated galactosyl-diacylglycerol. Hydrazinolysis was used to remove acetyl groups and yield monogalactosyldiacylglycerol stereochernically identical to the naturally occurring plant lipid. The same route could be used to prepare galactosyldiacylglycerols with two different fatty acids (Gigg and Gigg, 1990). Various β-glycosyldiacylglycerols have been prepared by an orthoacetate glycosidation procedure to diacylglycerols (see Shvets et al., 1973). Glycosides of 2-deoxy-Dglucose with racemic diacylglycerols and glucosyl- or mannosyldiacylglycerols containing fluorescent labels in the lipid and sugar portions have been made (see Gigg and Gigg, 1990). A saturated β-galactosyidiacylglycerol was prepared via an intermediate formed by condensation of aceto-bromogalactose with 1,2-di-O-benzyl-sn-glycerol (Gent and Gigg, 1975). The latter compound was also used by Batrakov et al. (1976) for the preparation of βgalactosyl-, β-glucosyl-, and α-mannosyldiacylglycerols. Improvements in their procedure resulting in high yields (70 to 75%) have been reported (Mannock et al., 1987; van Boeckel et al., 1985).
CH2OCOR1
CH2OH
AcO
H C O CH2 O C H CH2OCOR1
OAc
CH2OH
Br
OAc Acetobromogalactose
Bis(1-acyl-2-methyleneglycerol)
Condensation
CH2OAc O O AcO OAc
CH2OAc O
CH2OCOR1
CH2 H C
OAc
O CH2 O C
CH2
OCOR1
H O
CH2
O AcOH2C OAc
OAc
OAc Acid hydrolysis
CH2OAc O O AcO OAc
H OAc
FIGURE 5.12
CH2 C OH CH2OCOR1
OH group acylated
Hydrazinolysis
with fatty acid chloride
to remove acetyl groups
HO
CH2OH O O OH
CH2 H
OH
C OCOR2 CH2OCOR1
Synthesis of galactosyldiacylglycerol. Ac = CH3CO. See Wehrli and Pomeranz (1969).
407
5.8
Glycosylglycerides
involving condensation of perbenzylated glucosylbromide with the α-glucosyldiacylglycerol derivative. The important plant glycolipid, digalactosyldiacylglycerol (α-D-galactopyranosyl(1→6)-β-D-galactopyranosyl (1→3)-1,2-di-O-acylglycerol) was synthesised using a route beginning with the condensation of acetobromogalactose with 1,2-di-O-benzyl-sn-glycerol as shown in Figure 5.13. The β-galactosyldiacylglycerol was also prepared. Other diglycosyldiacylglycerols, such as those important in bacteria and for serological activity, have been synthesised (see Gigg and Gigg, 1990, for more details). The plant sulfolipid (sulfoquinovosyldiacylglycerol) was synthesised (Gigg, 1978; Gigg et al., 1980) by a route shown in Figure 5.14. A glucosyl chloride derivative was condensed with 1,2-di-O-(but-2-enyl)-sn-glycerol to give α-glucoside, which was converted to its acetonide. The toluene-β-sulfonate was converted to the thioacetate, which was then hydrolysed and oxidised (with iodine) to give the disulfide. The isopropylidene group was removed by acid hydrolysis and the alcohol product was acetylated with hexadecanoyl chloride. The resultant ester was oxidized with 3-chloroperbenzoic acid and subsequent debenzylation yielded sulfoquinovosyldiacylglycerol. A comprehensive recent review on sulfoquinovosyldiacylglycerol is that by Harwood and Okanenko (2003). Methods for the chemical synthesis of 6-sn-phosphatidylglucosyldiacylglycerol (found in streptococci), diglycosyldialkylglycerols, triglycosyldiacylglycerols (occurring in plants, bacteria and gastric secretions) (see Kates,
Hydrazinolysis of the acetylated glycolipid completed the procedure. A simpler method, but which involved loss of optical activity at C-2 of the glycerol, was reported by Bashkatova et al. (1971a). They formed 1,2-dipalmitoyl3-O-(β-D-glucopyranosyl)-sn-glycerol as well as the mannose equivalent. Their procedure has also been used to synthesise diglycosyldiacylglycerols (cf. Bashkatova et al., 1971b; Shvets et al., 1973). A useful technique described by Heinz (1971) has been used to prepare different molecular species of monogalactosyl- and digalactosyldiacylglycerols. He used a semisynthetic method starting from compounds isolated from natural sources. The hydroxyl hydrogens of the galactose residues of the substrates were substituted with O-(1methoxyethyl) groups and the acyl groups were then removed with sodium methoxide. Acylation was then carried out with the desired fatty acyl chloride, and the protecting groups were removed with boric acid. Heisig and Heinz (1972) also used a similar method for preparing galactosylmonoacylglycerols. The first synthetic diglycosyldiacylglycerol was prepared by Wehrli and Pomeranz (1969) using a condensation reaction between racemic isopropylideneglycerol and acetobromocellobiose. The isopropylidene group was removed by acid hydrolysis. Acylation with fatty acids and deacetylation with hydrazine yielded β-D-glucopyranosyl(1→4)-D-glucosyldiacylglycerol. Another diglucosyldiacylglycerol (derived from kojibiose, α-D-glucopyranosyl(1→2)-D-glucose) was prepared by van Boeckel and van Boom (1985). They used a route
AcO
CH2OAc O O OAc
H C
CH2OR O O
BnO
CH2
CH2 O O
BnO O
OBn
CH2 H
OBn OBn R = CH2CH CHCH3 converted to R = H then converted to R = Bn CH2OR1 O OR1
CMe2
OBn R = CPh3 converted to R=H
OAc R = Bn converted to R=H CH2OR O BnO
R1O
O
CH2O
CH2OR
OBn
H C
OBn
OR
CH2
O
C
CMe2
CH2O
CH2 O O
R1O O
OR1
CH2 H C
OR1 OR1
OR2
CH2OR2
R1 = Bn, R2 = H converted to R1 = H, R2 = CO(CH2)14CH3
FIGURE 5.13
Synthesis of digalactosyldiacylglycerol. Ac = CH3CO, Bn = CH2Ph. See Gent and Gigg (1975).
408
Synthesis
CH2OAc O
CH2OAc O
OBn BnO
OBn
Cl
OBn
SCH2
BnO
OBn
O
O
CH2O
CMe2
OBn
BnO
SCH2
BnO
OBn
H C OBn
OR RO
CH2OR
OBn
CHCH3
CH2 O
O
CH2O
CMe2
CH2 H
OR
O OR
2
R = H converted to R = CO(CH2)14CH3
FIGURE 5.14
CH2OCH2CH
H C
CH2SO3H O
CH2 O
CHCH3
OBn R = OH converted to R = OSO2Ph(pMe) then converted to R = SAc
2
O
OCH2CH
CH2R O
CH2 H C
C
O
BnO
OBn
O
CH2 H
COCO(CH2)14CH3 CH2OCO(CH2)14CH3
R = Bn converted to R=H
Synthesis of sulfoquinovosyldiacylglycerol. Ac = CH3CO, Bn = CH2Ph. Adapted from Gigg and Gigg, (1990).
1990a) and triglycosyl-O-phytanylglycerol (found in halobacteria) (see Kates, 1990b) are discussed by Gigg and Gigg (1990). Heinz (1996) has made a very thorough review of the structure, isolation and analysis of glycosylglycerides. The latter includes structural analytical methods such as mass and nuclear magnetic resonance (NMR) spectrometry.
canoyl-β-D-galactopyranosyl)-L-glycerol. J. Chem. Soc., Perkin Trans., 4, 364–370. Gigg, R. (1978). Studies on the synthesis of sulphur-containing glycolipids (“sulfoglycolipids”). Am. Chem. Soc. Symp. Ser., 77, 44–66. Gigg, J. and Gigg, R. (1990). Synthesis of glycoglycerolipids. In Handbook of Lipid Research, vol.6, Glycolipids, Phosphoglycolipids and Sulfoglycolipids, Ed., M. Kates, Plenum, New York, pp. 489–506. Gigg, R. et al. (1980). Synthesis of 3-O-(6-deoxy-6-sulpho-α-dglucopyranosyl)-1,2-di-O-hexadecanoyl-l-glycerol, sulphoquinovosyl diglyceride. J. Chem. Soc., Perkin Trans., 1, 2490–2493. Harwood, J.L. (1980). Plant acyl lipids: structure, distribution and analysis. In Biochemistry of Plants, vol. 4, Eds., P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 1–55. Harwood, J.L. and Okanenko, A.A. (2003). Sulphoquinovosyldiacylglycerol — the sulpholipid of higher plants. In Sulphur in Plants, Eds., Y.P. Abrol and A. Ahmad, Kluwer, Dordrecht, pp. 189–220. Heinz, E. (1971). Semisynthetic galactolipids of plant origin. Biochim. Biophys. Acta, 231, 567–544. Heinz, E. (1996). Plant glycolipids: structure, isolation and analysis. In Advances in Lipid Methodology, Ed. W.W. Christie, vol.3. The Oily Press, Dundee, U.K., pp. 211–332. Heisig, O.M.R.A. and Heinz, E. (1972). Semisynthetic lysogalactolipids of plant origin. Phytochemistry, 11, 815–818. Hitchcock, C. and Nichols, B.W. (1971). Plant Lipid Biochemistry, Academic Press, London. Kates, M. Ed. (1990a). Handbook of Lipid Research, vol. 6, Glycolipids, Phosphoglycolipids and Sulfoglycolipids, Plenum, New York. Kates, M. (1990b). Glyco-, phosphoglyco- and sulfoglycoglycerolipids of bacteria. In Handbook of Lipid Research, vol. 6,
References Bashkatova, A.I. et al. (1971a). Glycosyldiglycerides. 1. Synthesis of mannopyranosyldiglyceride and glucopyranosyldiglyceride. Zh. Org. Khim., 7, 1644–1652. Bashkatova, A.I. et al. (1971b). Diglycosyldiglyceride synthesis. Zh. Org. Khim., 7, 2627. Batrakov, S.G. et al. (1976). Synthesis of monoglycosyl diglycerides and their derivatives. Bull. Acad. Sci. USSR, Chem., 25, 626–632. Brundish, D.E. and Baddiley, J. (1968). Synthesis of glucosylglycerols and diglucosylglycerols, and their identification in small amounts. Carbohydr. Res., 8, 308–316. Brundish, D.E. et al. (1967). The structure and possible function of the glycolipid from Staphyloccus lactis 13. Biochem. J., 105, 885–889. Charlson, A.J. et al. (1957). The configuration of glycosidic linkages in oligosaccharides. IV. Further degradations of reducing disaccharides to 2-O-glycosyl-glycerol. Can. J.Chem., 35, 365–373. De Stefanis, V.A. and Ponte, J.G. (1969). Simple method for isolation of crude mono-and digalactosyl diglycerides from wheat flour. Biochim. Biophys. Acta, 176, 198–201. Gent, P.A. and Gigg, R. (1975). Synthesis of 1,2-di-O-hexadecanoyl-3-O-(β-D-galactopyranosyl)-L-glycerol (a ‘galactosyl diglyceride’) and 1,2-di-O-octadecanoyl-3-O-(6-O-octade-
409
5.9 Bulk separation procedures
refined high-oleic sunflower oil with 80 to 90% of oleic acid may be used in place of triolein. A similar situation exists with phospholipids. A crude mixture (lecithin) may suffice for some purposes, but materials of increasing purity (and cost) frequently display interesting and valuable properties. Procedures, therefore, have been developed to isolate individual phospholipid classes and then to manipulate the fatty acids within these.
Glycolipids, Phosphoglycolipids and Sulfoglycolipids, Ed., M. Kates, Plenum, New York, pp. 1–122. Khan, M. and Williams, J.P. (1977). Improved thin-layer chromatographic method for the separation of major phospholipids and glycolipids from plant lipid extracts and phosphatidyl glycerol and bis(monoacylglyceryl) phosphate from animal lipid extracts. J. Chromatogr., 140, 179–185. Mannock, D.A. et al. (1987). An improved procedure for the preparation of 1,2-di-O-acyl-3-O-(β-D-glucopyranosyl)sn-glycerols. Chem. Phys. Lipids, 43, 113–127. Miyano, M. and Benson, A.A. (1962). The plant sulfolipid, VI: configuration of the glycerol moiety. J. Am. Chem. Soc., 84, 59–62. Murray, R.K. and Narasimhan, R. (1990). Glycoglycerolipids of animal tissues. In Handbook of Lipid Research, vol. 6, Glycolipids, Phosphoglycolipids and Sulfoglycolipids (Ed., M. Kates), Plenum, New York, pp. 321–361. O’Brien, J.S. and Benson, A.A. (1964). Isolation and fatty acid composition of the plant sulfolipid and galactolipids. J. Lipid Res., 5, 432–436. Sastry, P.S. (1974). Glycosyl glycerides. Adv. Lipid Res., 12, 251–310. Sastry, P.S. and Kates, M. (1964). Lipid components of leaves. V. Galactolipids, cerebrosides, and lecithin of runner-bean leaves. Biochemistry, 3, 1271–1280. Shvets, V.I. et al. (1973). Synthesis of glycosyl diglycerides. Chem. Phys. Lipids, 10, 267–285. Van Boeckel, C.A.A. and van Boom, J.H. (1985). Synthesis of phosphatidyl-α-glucosyl glycerol containing a dioleoyl diglyceride moiety-application of the tetraisopropyldisiloxane-1,3-diyl (tips) protecting group in sugar chemistry, 3. Tetrahedron, 41, 4545–4555. Van Boeckel, C.A.A. et al. (1985). Synthesis of phosphatidyl-βglucosyl glycerol containing a dioleoyl diglyceride moietyapplication of the tetraisopropyldisiloxane-1,3-diyl (tips) protecting group in sugar chemistry, 4. Tetrahedron, 41, 4557–4565. Van der Veen, J. et al. (1967). Comparison of column chromatographic methods for the quantitative determination of mono- and digalactosyldiglycerides in fresh alfalfa. Lipids, 2, 406-410. Wehrli, H.P. and Pomeranz, Y. (1969). Synthesis of galactosyl glycerides and related lipids. Chem. Phys. Lipids, 3, 357–370. Wickerberg, B. (1958). Synthesis of 1-glycerol D-galactopyranosides. Acta Chem. Scand.,12, 1187–1201.
5.9
5.9.1
Urea fractionation
Urea normally crystallises in tetragonal form, but in the presence of certain aliphatic molecules it forms hexagonal prisms within which some of the aliphatic compound is trapped. These prisms are built up from urea: six molecules form a unit cell 11.1 × 10−10 m long and 8.2 × 10−10 m in diameter containing a channel in which an open chain molecule may be held so long as it is fulfils certain dimensional qualifications. It must not be too short or it will not be held in the channel, and it must not be too wide if it is to fit into the free space, estimated at around 6 × 10−10 m diameter. Many straight chain acids and alkyl esters satisfy these conditions and readily form complexes with urea. These are also called adducts, inclusion compounds, or clathrates. Urea fractionation cannot be applied to triacylglycerols. Saturated acids form stable complexes more readily than do unsaturated acids, and oleic acid forms inclusion compounds more readily than do the polyunsaturated acids. In practice, urea and mixed acids are dissolved in hot methanol or urea and methyl esters are dissolved in a hot methanol–ethanol mixture. The solution is crystallised at temperatures between 0 and 4ºC. After separation, the adduct and mother liquor each furnish acids or esters when mixed with water and extracted with ether or petroleum ether in the usual way. This procedure is used to separate mixtures of fatty acids according to their degree of unsaturation or to separate straight-chain from branched-chain acids. It has been applied recently, in particular, to separate conjugated linoleic acid isomers and to concentrate the highly unsaturated omega-3 acids in fish oils. Some examples are cited below. Urea fractionation is simple, does not damage PUFA, can be repeated one or more times to enhance the enrichment of a particular acid, and can be affected on a gram, kg, or tonne scale. It is considered by many to be environmentally benign, but can require large volumes of solvents and a considerable aqueous reject. Hayes (2002a and b) has discussed the physical processes involved in urea fractionation and has given some examples of its use. He lists the main selectivities in descending order of importance as:
Bulk separation procedures
Syntheses of molecules related to fatty acids have been reported earlier in this chapter. For the most part, these are based on procedures developed and carried out by organic chemists. They have their place in the research laboratory, but there are many other demands for fatty materials that are met by purification or concentration of compounds already present in natural fats. This applies over a range of scales for the research laboratory and for commercial operations and the most commonly employed procedures are described in the following sections. On an industrial scale there has to be a balance between the purity of the product and its cost. As a simple example,
• Increasing discrimination against inclusion as the number of double bonds per molecule increases. • Preference for molecules of longer chain length. • Preference for trans rather than cis double bonds. • Sensitivity for double bond position. 410
Synthesis
of this acid on LDL-receptor activity. Conditions are reported through which the concentration of pinolenic acid rose from 14% in the oil to 45% in the mother liquor after urea fractionation (Lee 2004).
He further reviews the use of this separation procedure on a large scale suggesting that the urea (and solvent) can be recovered for re-use or used directly as fertiliser. Urea fractionation was one step in the enrichment of γ-linolenic acid and of stearidonic acid from blackcurrant seed oil carried out on the tonne scale. This included an interesting separation of the α- and γ-isomers of linolenic acid. Under appropriate conditions γ-linolenic acid and stearidonic acid concentrated in the mother liquor fractions while α-linolenic acid was mainly in the adduct (Wille et al., 1988, 1991). Commercially produced CLA contains two major isomers (9c11t and 10t12c) and attempts have been made to separate these by urea fractionation with the 9,11 isomer enriched in the adduct and the 10t12c isomer enriched in the mother liquor (Ma et al., 2002). In attempts to isolate pure or enriched fractions of highly unsaturated PUFA (EPA and/or DHA) from fish oils urea fractionation on a large scale is frequently employed as the first step to produce concentrates that can be further purified by other procedures, such as molecular distillation or selective enzymatic reaction. Urea fractionation is not only effective for this purpose, it is relatively simple to carry out on kg or tonne scale under mild conditions that do not damage the highly unsaturated acids. For example, the EPA and DHA levels of squid ethyl esters, submitted to urea fractionation, were raised from 12 to 28% and from 15 to 36%, respectively (Hwang et al., 2001). This leads to a significant reduction in the mass of material that has to be submitted to the later processes. At the same time there is a marked reduction in the levels of 20:1 and 22:1 acids that is particularly helpful if the acids are to be subject to molecular distillation to isolate EPA and/or DHA. A second example involves the concentration of PUFA from mackerel processing waste (Zuta et al., 2003). As a more esoteric example, urea crystallisation was used to prepare a concentrate of pinolenic acid (all-cis5,9,12-18:3) from Korean pine nut oil to study the effect TABLE 5.3
5.9.2
Distillation under reduced pressure
Distillation of unsaturated methyl or ethyl esters is not generally advised because they may be exposed to high temperatures for a long time. Under these conditions double bond isomerisation (both positional and configurational) and/or cyclisation may occur even though oxidation can be minimised by working under vacuum. Exposure to heat is reduced in short path or molecular distillation and this finds many applications (see below for examples). Reduced pressure distillation depends on boiling point and can be used to separate acids or alkyl esters according to their chain length but not by their degree of unsaturation. On a commercial scale C8 and C10 acids can be isolated from hydrolysed lauric oils and the acids from fish oils may be distilled to give separate cuts rich in C14, C16, C18, C20, and C22 acids. Since the effectiveness of distillation depends on differences of boiling point, chain length is more important than degree of unsaturation. It is possible to separate the esters of C12 to C22 acids from each other and distillation is commonly used for this purpose. On the other hand, stearate, oleate, linoleate, and linolenate are not usefully separable by distillation. Distillation of fatty acids is important in industry where it is used primarily to separate saturated fatty acids such as (1) 16:0 and 18:0 from hydrogenated tallow or hydrogenated seed oils; (2) 18:0, 20:0, 22:0, and 24:0 from hydrogenated fish oil or hydrogenated high-erucic rapeseed oil; and (3) 8:0, 10:0, 12:0, and 14:0 from lauric oils. The temperatures range from 160 to 230°C at pressures below 20 mmHg in specially constructed stills (up to 2m in diameter) that run continuously. Some boiling points are collected together in Table 5.3.
Names and boiling points of some alkanoic acids and their methyl esters
Acid
Acid
Methyl Ester
Chain
Systematic
Trivial
Mol
BP
BP
Length
Name
Name
wt
(°C)a
(°C)a
88.1 116.2 144.2 172.3 200.3 228.4 256.4 284.5 312.5 340.6 368.6
164 206 240 271 1301 1491 1671 1841 2041 – –
103 151 195 228 262 1141 1361 1561 1882 2062 2222
4 6 8 10 12 14 16 18 20 22 24 a
butanoic hexanoic octanoic decanoic dodecanoic tetradecanoic hexadecanoic octadecanoic eicosanoic docosanoic tetracosanoic
butyric caproic caprylic capric lauric myristic palmitic stearic arachidic behenic lignoceric
BP at 760 mm Hg or at 1 or 2 mm Hg as indicated by the superscript.
Source: Gunstone, F.D., Harwood, J.L., and Padley, F.E. (1994) The Lipid Handbook, 2nd ed., Chapman & Hall, London.
411
5.9 Bulk separation procedures
Soybean deodoriser distillate is an important by-product containing valuable sterols and tocopherols among other less valuable components. The isolation and recovery of the sterols and tocopherols can be achieved in a number of ways that generally include molecular distillation to recover and separate the sterols (sometimes as esters) and tocopherols. Appropriate fractions contain tocopherols (up to 76%), sterols (96%), and sterol esters (98%) (Shimada et al., 2000, Watanabe, 2004). Many oleochemical processes yield glycerol (in the form of a dilute aqueous solution also containing salts and other impurities) as a commercially important by-product. Evaporation and distillation are important processes in converting this so-called “sweetwater” to glycerol at concentrations of 88%, 95%, and 99.5%. When heated (to around 230°C for 4 to 8 hours) with a catalyst, such as montmorillonite (4%), unsaturated acids from tall oil or other sources are converted to dimers and trimers that can be separated by distillation. The dimer fraction (~ 80% dimer and 20% trimer) is a mixture of C36 dibasic acids used mainly as polyamides. Glycerolysis is employed on a commercial scale to convert triacylglycerols to mixtures of monoacylglycerols and diacylglycerols by reaction with glycerol in the presence of a basic catalyst. Concentrates of monoacylglycerol (90 to 95%) are produced by molecular distillation and are widely used as emulsifiers.
The major uses of distillation at laboratory to industrial scale include: • Separation of saturated acids from C6 to C22 (and occasionally of monounsaturated acids) using appropriate starting materials (lauric oils, C16/C18 seed oils or animal fats, fish oils). This is the major route to saturated acids on a commercial scale. These are obtained at purities up to 99% though lower grades are available (cheaper) and find many uses. • Purification of glycerol from aqueous solutions. • Isolation of dimer acid concentrates. • Tall oil furnishes fractions rich in fatty acids and rich in resin acids (Hubers and Fritz, 1988). • Isolation of products from enzymatic processes. • Separation of volatile materials — both desirable and undesirable — such as tocopherols, squalene, sterols, and persistent organic pollutants. • Concentration of monoacylglycerols from mixed glycerolysis products. Molecular distillation is used to separate squalene, diacylglycerol ethers, and concentrates of n-3 acids from fish oils on a scale of 200k/24 hours (Anon 2004). The level of organic environmental pollutants, such as PAH and halogenated organic compounds in fish oils, can be reduced by short path distillation. This procedure is improved by addition of 3 to 6% of an ester mixture that co-distills with the pollutants. The procedure can be applied to fish oils for human consumption or to those that are to be fed to farmed fish thereby reducing their levels in the final marketed product (Breivik and Thorstad 2005). The products of enzymatic hydrolysis, alcoholysis, or interesterification frequently contain free acids, alkyl esters, free alcohols, and glycerol esters. To recover the desired fraction it may be useful to remove one or more volatile fractions by molecular distillation. Depending on its nature the desired material is sometimes present in a distillate and sometimes in the residue. (See Breivik et al. [1997], Xu et al. [2001, 2002], Yang et al. [2004], Haraldsson et al., [2006].) Molecular distillation can be used to separate C20 and C22 acids/esters and is frequently a stage in the isolation of concentrated 20:5 and 22:6 after these have been separated from less unsaturated acids/esters by urea fractionation, e.g., the level of EPA was raised from 9 to 15% and that of DHA from 13 to 25% (Hwang et al., 2001, Liang and Hwang, 2000). The best samples of commercial CLA contain only two CLA isomers (9c11t and 10t12c), but for some purposes it has been desirable to separate these or, at least, to obtain concentrates. This has been achieved through enzymecatalysed (Candida rugosa lipase) esterification with lauric acid. The 9c11t isomer is the more reactive and concentrates in the ester fraction while the 10t12c isomer remains mainly as unreacted acid. The esters, free acids, and unreacted lauryl alcohol are separated by molecular distillation (Nagao, 2002).
5.9.3
Crystallisation, fractionation, and hydrophilisation
Crystallisation from an appropriate solvent is a classical method for the isolation and purification of solids, but since most of the interesting fatty acids are liquid this method has to be modified. With only a slight change in procedure and equipment, liquid acids can be crystallised conveniently at temperatures down to –78ºC using solid carbon dioxide as refrigerant. Alternatively, the conventional procedure can be applied to the higher melting salts. Lead salts crystallised from acetic acid permit the separation of saturated from unsaturated acids and lithium salts crystallised from acetone provide a separation of saturated and monoene acids from polyunsaturated fatty acids. More details of these classical procedures are given in the second edition of The Lipid Handbook (Gunstone et al., 1994). The separation of natural fats by crystallisation of the complex mixtures of triacylglycerols present is generally referred to as fractionation. This is now an industrial process of considerable importance applied particularly to palm oil, but also to palm kernel oil, milk fat, and to animal body fats. It has been discussed in detail in Section 4.4. Crystallisation is frequently a step in the production of pure synthetic acids and triacylglycerols discussed in the earlier sections of this chapter. Hydrophilisation is a process for obtaining concentrates of oleic acid from tallow or palm fatty acids. After 412
Synthesis
crystallisation of the mixed acids at about 20°C, the crystals are treated with a wetting agent and form an aqueous suspension that can be separated from the liquid fraction by centrifugation. The latter is mainly oleic acid (70 to 75%) along with palmitoleic, linoleic, and 10% or less of saturated acids. Hydrophilisation is sometimes described as the Henkel or Lipofrac process (Fritz 1988).
5.9.4
Fish oil triacylglycerols (20–30 % EPA+DHA) Ethanolysis or hydrolysis Ethyl esters or free fatty acids (20–30 % EPA+DHA)
Lipase
Short-path distillation
Enzymatic enhancement
Under appropriate conditions lipases promote not only the hydrolysis of esters but also their formation from acids and alcohols or from other esters serving as a source of acids and alcohols. Because of specificities shown by certain lipases they can distinguish between acids according to their chain length or, more important in the present discussion, according to the position of the double bonds with respect to the acyl function. Reactions are slower with acids or esters that have a double bond close to the carboxyl or ester function. This has been exploited in preparing concentrates of γ-linolenic acid (6,9,12-18:3), EPA (5,8,11,14,17-20:3), and DHA (4,7,10,13,16,19-22:6) all of which have a double bond closer to the carboxyl group than the common ∆-9 acids, such as oleic and linoleic. Of particular value in this context are the lipases from Candida cylindricea, C. rugosa, C. antarctica, Rhizopus miehei, R. delemar, Pseudomonas fluorescens, and Geotrichum candidum, which distinguish unsaturated acids according to double bond position. These possibilities are illustrated in Figure 5.15 showing how it is possible to prepare EPA and DHA each of at least 95% purity. CLA has been identified at low levels in milk fat (3 to 6 mg/g of total fat), butter fat (12 to 14), cheeses (2 to 20), and in lamb and beef meat (4 to 5). Several isomers may be present and the major component (the 9c11t isomer) is designated rumenic acid (linked to its formation in the rumen of the cow). This acid is believed to be a metabolic product resulting from linoleic acid by two linked pathways: isomerisation of linoleic acid (9c12c-18:2) and ∆9-desaturation of vaccenic acid (11t18:1). The 7t9c and 10t12c dienes are also present at lower levels along with many other isomers. CLA preparations can be made in larger volumes and higher concentrations by alkali isomerisation of linoleic-rich vegetable oils, such as safflower. Early preparations were mixtures of the 9c11t and 10t12c dienes and other isomers formed by further isomerisation. Now that the reaction is better understood and better controlled in terms of alkali selected, choice of solvent, and restriction of reaction temperature, the product contains these two isomers in similar proportions as virtually the only CLA present along with unreacted palmitic, stearic, and oleic acids from the starting material. These two CLA isomers show different physiological properties and procedures have been devised to concentrate them individually in
Lipase
Ethyl esters or free fatty acids (50 % EPA+DHA) Urea inclusion Ethyl esters or free fatty acids (80–85 % EPA+DHA) HPLC
> 95 % EPA
> 95 % DHA
FIGURE 5.15 Possible routes to EPA and DHA from fish oil triacylglycerols by enzymatic enhancement and traditional separation techniques. (Adapted from Haraldsson, G.G., Hjaltason, B., in Modifying Lipids for Use in Food, Gunstone, F.D. (Ed.), Woodhead Publishing, Cambridge, England, 2006, pp. 347–352.)
separate fractions. This is best achieved using enzymes (such as that from the fungus Geotrichum candidum), which distinguish between these two isomers with the 9c11t isomer being the more reactive. The products of the isomerisation process are free acids that are generally converted to triacylglycerols before being used in human or animal diets. This can be done enzymatically with lipases, such as those from Mucor miehei or Candida antarctica,, since esterification then proceeds under mild conditions without modification of the double bond systems in the CLA. Further examples and references are cited in Section 5.4.4.
5.9.5
Chemical methods
Chemical methods of isolating fatty acids according to their level of unsaturation depends on reactions that distinguish one acid or one group of acids from others present in a mixture and which, after appropriate separation, can be reversed to give the isolated acid(s). The reaction of olefinic acids with silver ions might be considered to fall into this category though the complex is never isolated. Olefinic compounds also react with mercuric acetate to give sadducts that are easily converted to methoxy bromomercuric compounds – CH(HgBr)CH(OMe) — by reaction with NaBr-MeOH. It is possible by chromatographic procedures to separate saturated, from monoene (mono adduct) from polyene acids (di-adduct, tri-adduct, etc). The olefinic compounds can be regenerated by reaction with methanolic HCl (Section 8.5). 413
5.9 Bulk separation procedures
RCH=CH(CH2)n COOH
(CH2)nCOO–
R
or
R
Gaiday, N.V., et al., Separation of natural polyunsaturated fatty acids by means of iodolactonization, J. Am. Oil Chem. Soc., 68, 230–233, 1991. Gunstone, F.D. (1994) Crystallisation, in The Lipid Handbook, Gunstone, F.D., Harwood, J.L., and Padley, F.B., Eds., 2nd ed., Chapman and Hall, London, pp. 225–226. Haraldsson, G.G. and Hjaltason, B. (2006) PUFA Production from Marine Sources, in Modifying Lipids for Use in Food, Gunstone, F.D., Ed., Woodhead Publishing, Cambridge, England, 2006, 347–352. Hayes, D.G. (2002a) Urea inclusion compound formation, Inform, 13, 781–783. Hayes, D.G. (2002b) Free fatty acid fractionation via urea inclusion compounds, Inform, 13, 832–834. Huibers, D.T.A. and Fritz, E. (1988) Distillation of fatty acids, in Fatty Acids in Industry, R.W. Johnson and E. Fritz, Eds., Marcel Dekker, New York, pp. 85–112. Hwang, L-S. and Liang, J-H. (2001) Fractionation of ureapretreated squid visceral oil ethyl esters, J. Am. Oil Chem. Soc., 78, 473–476. Lee, J-W. et al., (2004) Selective increase in pinolenic acid (allcis-5,9,12-18:3) in Korean pine nut oil by crystallisation and its effect on LDL-receptor activity, Lipids, 39, 383–387. Liang, J-H. and Hwang, L.S. (2000) Fractionation of squid visceral oil ethyl esters by short-path distillation, J. Am. Oil Chem. Soc., 77, 773–777. Ma, D.W.L. et al. (2002) Countercurrent approach to the enrichment of ∆9c11t- and ∆10t12c-18:2 isomers by urea complexation, J. Am. Oil Chem. Soc., 79, 755–758. Nagao, T. (2002) Fractionation and enrichment of CLA isomers by selective esterification with Candida rugosa lipase, J. Am. Oil Chem. Soc., 79, 303–308. Shimada, Y. et al., (2000) Facile purification of tocopherols from soybean oil deodorizer distillate in high yield using lipase, J. Am. Oil Chem. Soc., 77, 1009–1013. Watanabe, Y. (2004) Purification of tocopherols and phytosterols by a two-step in situ enzymatic reaction, J. Am. Oil Chem. Soc., 81, 339–345. Wille, H-J. and Traitler, H. (1988) Continuous process for the concentration of polyunsaturated fatty acids, Fat Sci. Technol., 90, 476–481. Wille, H-J. et al., (1991) Preparation of stearidonic acid concentrate, Fat Sci. Technol., 93, 363–368. Xu, X. et al., Purification of specific structured lipids by distillation: effects on acyl migration, J. Am. Oil Chem. Soc., 78, 715–718, 2001. Xu, X. et al., (2002) Purification and deodorisation of structured lipids by short path distillation, Eur. J. Lipid Sci. Technol., 104, 745–755. Yang, T. et al., (2004) Diacylglycerols from butterfat: production by glycerolysis and short-path distillation and analysis of physical properties, J. Am. Oil Chem. Soc., 81, 979–987. Zuta, C.P. et al. (2003) Concentrating PUFA from mackerel processing waste, J. Am. Oil Chem. Soc., 80, 933–936.
I+
(i)
O
O
I
R
O
O I
(ii) delta-4-acid
(iii) delta-5-acid
FIGURE 5.16 Iodolactonisation of acids with unsaturation at ∆4 (as in DHA) or ∆5 (as in EPA or AA). (i) I2, KI, KHCO3; EtOH at 25°C or tetrahydrofuran at −2 to +6°C; (ii) and (iii) TMSCl and NaI in CH3CN.
Appropriate olefinic acids react with KHCO3, I2, and KI to give iodolactones (Gaidey et al., 1991). Acids with a ∆4 double bond (such as DHA) form an iodo γ-lactone and acids with a ∆5 double bond (such as AA or EPA) form an iodo δ-lactone. This scheme is formulated in Figure 5.16. There are several interesting features about these reactions: • The iodolactones are neutral molecules and are easily separated from unreacted acids. This provides the basis of a method for separating and isolating ∆4 and ∆5 acids from acids that do not have unsaturation at these positions. • The original unsaturated acids can be recovered from the iodolactones by reaction with trimethylsilyl iodide (or Me3SiCl and NaI). • The iodolactonisation reaction is dependent on solvent, reaction temperature, and the ratio of iodine to iodide. For example, with ethanol as solvent and a reaction temperature of 25°, DHA gives a maximum yield of iodo-γ-lactone in 10min, but EPA requires 90 min to produce the iodo-δ-lactone.
References Anon., (2004) Squalene, ethers and n-3 fatty acids from fish oil, Lipid Technol., 16, 26. Breivik, H., et al., (1997) Preparation of highly purified concentrates of eicosapentaenoic acid and docosahexaenoic acid, J. Am. Oil Chem. Soc., 74, 1425–1429. Breivik, H. and Thorstad, O. (2005) Removal of organic environmental pollutants from fish oil by short-path distillation, Lipid Technology, 17, 55–58. Fritz, E. (1988) Separation of fatty acids, in Fatty Acids in Industry, R.W. Johnson and E. Fritz, Eds., Marcel Dekker, New York, pp. 73–84.
414
6 ANALYSIS
A. J. Dijkstra, W. W. Christie and G. Knothe
their relevance and facilitate the understanding of old, but often surprisingly relevant literature; it also describes a simple method to determine the oil loss during bleaching that, for some unknown reason, was never developed into an official method.
Lipid analysis comprises a vast range of methods, approaches and analyses that serve totally different purposes. Oils and fats are analysed on the shop floor by a plant operator titrating their free fatty acid content so that he/she can calculate how much lye to use in the chemical refining process. A homemaker may analyse the contents of a partially used bottle of walnut oil organoleptically for rancidity to decide whether or not to throw it away. An analyst in a clinical laboratory uses instrumental methods to determine fatty acid compositions and may even distinguish between various fatty acid isomers. A plant breeder uses an automated, high throughput system of analysis to screen individual seeds in a nondestructive manner to decide which seeds to plant next season. Several of these analytical methods have been developed and validated by a close cooperation between the parties that are directly involved. These methods have been published by international or national bodies that issue standards or by national learned societies. The reader is referred to these publications for details about their execution, but in being referred, he/she faces an embarrassment of choice. Rather than repeat details of selected methods, we have considered it more useful to the reader to list official methods and then to provide a commentary on some aspects relating to lipid analysis (Sections 6.1 to 6.3). Accordingly, the first of these sections (Section 6.1) lists a selection of available, official methods and aims at putting them somewhat in perspective. The second section discusses the various possibilities facing an industrial operator so that hopefully, he/she can make a sensible choice from what is offered. The third section (Section 6.3) discusses several methods in some detail to highlight
6.1
Introduction
The importance of the analysis of oils, fats, and related products is well illustrated by the fact that the American Oil Chemists’ Society (AOCS) originates from the Society of Cotton Products Analysts (Blank, 1942). The reason why analysis is so important is that these products are traded and that trading contracts may contain penalties if a certain property does not meet a certain, specified limit. Then buyer and seller must apply the same standards and use the same analytical methods. Accordingly, in several countries, such as for instance France and The Netherlands, Official Methods are issued by the respective national Bureaus of Standards. In this respect there is extensive international collaboration, as illustrated, for example, by the fact that some French norms (AFNOR) and Dutch norms (NEN) are identical to ISO (International Standards Organisation) norms and/or have been published in English as well. AOCS Official Methods may also mention that they are identical to ISO norms. In fact, organisations like FOSFA (Federation of Oils, Seeds and Fats Associations) rely more and more on ISO methods, now that the International Union of Pure and Applied Chemistry (IUPAC) has stopped issuing agreed methods. Accordingly, the first column in Table 6.1 provides the ISO method numbers even when the method is still under development. In other countries like Germany and the U.S., the learned societies in the 415
6.1
Introduction
field of edible oils and fats have taken the initiative to develop, validate, and distribute such Official Methods; they also organise proficiency contests, etc., not just for their members but as a service to society. Not surprisingly, there isn’t a lot of bureaucracy involved before an analytical method becomes an Official Method. Moreover, chemists concerned in this “beatification” process may be more worried about the reproducibility and the repeatability of the method than about its pertinence (for definitions, see Section 6.2). This has led, for instance, to the situation whereby a method to determine the Oil Stability Index (OSI) became official in 1997 (AOCS Official Method Cd 12b), although it was already known (Dijkstra et al., 1995; Dijkstra et al., 1996) that, although it is quite reproducible, the method does not predict shelf life (Lacoste et al., 1999). On the other hand, subsequent authors (Broadbent and Pike, 2003) claim that the OSI correlates so well with a sensory induction period to warrant the use of the OSI in industrial applications. However, they stored their samples at 60°C and what happens at that temperature may not be the same as what happens during storage at ambient temperatures. Methods may also seem to promise more than they actually deliver:
• Acidity is not the same as free fatty acid content; phosphatides present in crude oil are also titrated with alkali so that water degumming an oil apparently causes its FFA to decrease. • Alkalinity is not the same as soap content because there may be additional alkaline compounds present. • Besides, calcium and magnesium soaps are not determined by the normal titration method (Sen Gupta, 1988). • Phosphorus may originate from phosphatides, but also from residual phosphoric acid used as a degumming acid. • A palmitic acid content as determined by GLC analysis of FAME is not the same as the palmitic acid content of the sample, since the GLC analysis does not take dimeric, polymeric, and oxidised compounds into account. For crude oils, the difference will be only marginal, but for oils that have been used in deep fat frying the “increase in palmitic acid content” can be quite considerable. On paper it is; in reality it does not exist. • The iodine value is a very useful parameter to follow the progress of a hydrogenation reaction, but what about conjugated double bonds?
• The oil content of oil seeds, for instance, is a parameter that is of vital interest to the crusher. Oil content is determined by grinding the seeds and extracting the ground seeds in a prescribed manner. However, the method does not say anything about the extent to which seed cells are opened up in the industrial crushing plant and, therefore, does not predict potential oil yield. The amount of oil extracted and especially its phosphatide content also vary widely according to the solvent used (Desnuelle et al., 1951). • Quite correctly, the French method refers to the extract obtained with hexane, as the “so-called oil content”1 and this is only one example of the fact that the analytical chemists who write specifications are fully aware of their limitations. Laboratory technicians carrying out the analyses will also be aware of this, but higher up in the hierarchy, problems may arise if people interpret “oil content” at face value. • Similarly, weight loss on heating a sample at atmospheric pressure or in vacuo is not the same as its water content. Calling this weight loss “moisture and volatiles” may be nearer the truth, but even so, heating the sample may lead to chemical dehydration and/or thermal degradation reactions and there is no guarantee that all free water will leave the sample under test conditions.
The above list is far from exhaustive and the instances given are only examples of the need to be aware of the limitations of analytical methods. However, the limitations of a method are always clearly indicated and within its limitations each method can still be quite useful. Far more serious situations arise when there is no Official Method as is, for example, the case for the determination of oil retention by bleaching earth. Again, knowing how much oil he/she stands to lose when bleaching oil constitutes vital information for the refiner. However, extracting spent earth with a solvent in a Soxhlet — and this is what most people do in the absence of an Official Method — underestimates the oil loss since this extraction doesn’t extract oil that has been polymerised (Morton, 1995). Treating the spent earth in a muffle furnace (see Section 6.3) is a far more accurate way of providing the data required to calculate the yield loss on bleaching (Dijkstra, 1993), but this is not an Official Method. In fact, there are quite a few parameters in lipid technology for which there is no Official Method to quantify them; neutral, triglyceride oil in soapstock is one of them, but this is not the time or place to examine this in detail. Readers can do so for themselves and then, hopefully, do something about it. The standards and methods listed in Table 6.1 are a selection of the methods published by the issuing bodies concerned. For source materials, meal and oils and fats alone, the AOCS has published almost 250 different Official Methods and this figure excludes so-called “surplus methods,” methods that often have become obsolete
1
Extrait à hexane, dit “teneur en huile.”
416
Analysis
TABLE 6.1
Numbers (and latest year of issue) of Official Analytical Methods or norms issued in various countries
Subject matter
ISOa
AFNORb
NENc
AOCSd
DGFe
General Nomenclature, oilseeds, oils, fats Execution and evaluation of ring test Application of repeatability limit and reproducibility limit Detection limits, significance tests Validation of analytical methods Hazardous materials
5507 (02)
V00-300
C-I 1 (99) A-II 1 (03)E 6303 (88)
A-II 2a (97)E A-II 2b (97) A-II 3 A-III 1 (97)
Vegetable oil source materials Sampling Sample preparation Impurities Moisture and volatile matter Oil and water content (pulse NMR) Oil and water content (NIR) Oil content (reference method) Oil content (continuous wave NMR) Acidity of oil in oilseeds Glucosinolate in rapeseed (HPLC) Glucosinolate (x-ray fluorescence) Glucosinolate HPLC rapid method Chlorophyll (spectrophotometrically) Bulk density Pressure loss on air flow through grain
542 (95) 664 (90) 658 (02) 665 (00) 10565 (98)
V03-900 V03-902 V03-904 V03-909 V03-916
659 (98) 5511 (92) 729 (88) 9167-1 (92) 9167-2 (94) 10519
V03-905 V03-907 V03-906 V03-918-1 V03-918-2 V03-918-3 V03-911
4174
V03-743
5500 (86) 5502 (92) 771 (77)
V03-920 V03-926 V03-921
734-1 (98) 734-2 (98) 10632 (00) 749 (77) 735 (77)
V03-924-1 V03-924-2
Aa 1 etc.f Am 4 (02) Aa 3 etc.f Ak 4 (99) Am 1 (01) Am 2 (93) Ak 3 (00) Aa 6 etc. f Ak 1 (97)
B-I B-I B-I B-I B-I B-I
1 (87) 2 (87) 4 (87) 5a (87) 5a (89) 5b
B-I 7a (97) B-1 7 (97)
Ak 2 (92) B-I 8 (89)
Oilseed residues Sampling Preparation meal test sample Moisture and volatile matter Residual oil content (petroleum ether) Residual oil (petroleum ether/hexane) Residual oil content (rapid extraction) Oil and water content (pulse NMR) Total ash content Hydrochloric acid insoluble ash Protein content (ammonia titration) Crude fibre in meal Urease activity in soya products Free residual hexane Total residual hexane (GLC) Screen test
Ba 1 (97) Ba 2 Ba 3 (97)
Ak 5 (01) Ba 5a (97) Ba 5b (97) Ba 4d (03) Ba 6 (97)f Ba 9 (97) Ba 13 (97) Ba 14 (97) Bc 7 (97)
V03-922 V03-923 V18-100 V03-040 V03-942 V03-930 V03-927
5506 (88) 9289 (95) 8892 (87)
B-II 1 (87) B-II 2 (87) B-II 3 (87) B-II 4a (03) B-II 4b (87) B-II 4c B-II 5 (89) B-II 6 (89) B-II 7 (87) B-II 8 (89) B-II 8a (87) B-II 9 (89)
Oils and fats Sampling Insoluble impurities Sediment Preparation of test sample Unsaponifiable (diethyl ether) Unsaponifiable (pet ether/hexane) Moisture and volatiles Moisture (distillation) Moisture (Karl Fischer) Free fatty acids/acid value
5555 (02) 663 (00) 15301 (01) 661 (05) 3596 (00) 18609 (01) 662 (98) 934 (80) 8534 (96) 660 (96)
T60-280 T60-202 T60-200 T60-205-1 T60-205-2 T60-201 T60-218 T60-225 T60-204
Polar compounds in deep frying fats Polymerised triglycerides (HPSEC) Butyric acid (GLC) Residual hexane Residual hexane (low values) Volatile org. contaminants by GC-MS
8420 (02) 16931 (02)
T60-248 T60-247
9832 (02)
T60-253 T60-257
15303 (01)
6300 (98)
6328 (84)
6323 (87)
6354 (91)
C 1 (00) Ca 3a (97) Ca 3d (02) Ca 6b (01) Ca 6a (97) Ca 2f (97) Ca 2a (87) Ca 2e (97) Ca 5d (03) Cd 3d (03) Cd 20 (01) Cd 22 (00) Ca 5c (97) Ca 3b (02)
C-III 11a (84)
C-III C-III C-III C-III C-III C-III
1a (77) 1b (77) 12 ((97) 13 (97) 13a (97) 4 (97)
C-III 3b (84) C-III 3c (02) C-III 8 (97)
Ca 3c (01)
(Continued)
417
6.1
Introduction
TABLE 6.1
Continued
Subject matter Low-boiling halogenated hydrocarbons Polycyclic aromatic hydrocarbons (GC-MS) Benzo(a)pyrene (RP HPLC) Ash content Polythene Soaps/alkalinity Phosphorus (colorimetric) Phosphorus (graphite furnace AAS) Phosphorus (ICP) Trace metals (ICP) Trace metals (graphite furnace AAS) Lead (graphite furnace AAS) Cadmium (graphite furnace AAS) Slip melting point (Amendment for palm oil) Titre Cloud point Mass per volume (pycnometer) Cold test Dropping point Solid fat content (pulse NMR) Refractive index Flash point (Pensky–Martens) Smoke point (open cup) Colour (Lovibond) Colour (Iodine scale) Colour (Gardner) Colour (spectrophotometric) PUFA with cis,cis-1,4-diene Iodine value (Wijs) Saponification value Ester value Hydroxyl value Polybromide reaction for trienes 1-Monoglycerides and free glycerol Monoglycerides (GLC) Antioxidants (TLC) Antioxidants (HPLC) BHA and BHT (GLC) Gallates (spectrophotometric) Individual and total sterols (GLC) Total sterols (enzymatic) Sterol composition (GLC) Sterol composition (HPLC plus GLC) Tocopherols and tocotrienols (HPLC) Stigmadienes (HPLC) Total trans isomers (IR) Isolated trans content (FTIR) Trans isomers (capillary GLC) For olive oil Preparation of FAME (BF3) FAME (alkaline ester interchange) Fatty acid composition (GLC) (capillary GLC) Fatty acid composition 2-position (GLC) Peroxide value (chloroform) Peroxide value (isooctane) p-Anisidine value
ISOa
AFNORb
16035 (05) 15753 (05) 15302 (98) 6884 (85) 6656 (02) 10539 (02) 10540-1 (03) 10540-2 (03) 10540-3 (03) 21033 8294 (99) 12193 (04) 15774 (00) 6321 (02) 6321 A1 935 (88)
T60-255 T60-290 T60-240 T60-217 T60-228
NENc
6327 6355 6330 6349
(77) (84) (77) (91)
T60-251 T60-252 T60-226
6313 (77)
T60-208
6316 (77) 6315 (76) 6311 (93)
6883 (00)
T60-214
8292 (95) 6320 (00) 15267 (98)
T60-250 T60-212 T60-256
15305 (98)
T60-259
6308 (76) 6309 (76) 6310 (76)
T60-224 7847 (87) 3961 (99) 3657 (03)
T60-203 T60-206 T60-213
7366 (87)
T60-245
5558 (82)
T60-235
6463 (82) 6464 (83) 12228 (99)
T60-237
6337 6339 6342 6369 6338 6340
(82) (76) (80) (76) (91) (91)
AOCSd
Cd 21 (97) Ca 11 (03) Ca 16 (02) Cc17 (97) Ca 12a (9) Ca 12b (02) Ca 20 (01) Ca 17 (01) Ca 18 (03) Ca 18c (03) Ca 18d (01) Cc 3b (02) Cc 12 (97) Cc 6 (97) Cc 10c (02) Cc 11 (03) Cc 18 (01) Cd 16 (99) Cc 7 (02) Cc 9c (97) Cc 9a (97) Cc 13e (02)
Cc 13c (97) Cd 15 (97) Cd 1 (93) Cd 3 (02) Cd 13 (97) Cd 11 (03) Cd 11b (03)
DGFe C-III 17b C-III 17c C-III 10 (97) C-III 15 (97) C-III 16a (03)
C-III 18
C-IV 3a (03) C-IV 3 (81) C-IV 2c (02) C-IV 3d (02)E C-IV C-IV C-IV C-IV C-IV C-IV C-IV
3g (03)E 5 (02) 8 (02)E 9 (02)E 4b (98) 4a (98) 4c (02)
C-V 11d (02)E C-V 3 (02)E C-V 4 (53) C-V 17a (98)E C-V 16 (57) C-VI 5 (02) C-VI 5a (02) C-VI 9 (00)E
Ce 6 (97)
T60-258 T60-243 T60-232 T60-254
6799
6350 (77)
9936 (04) 15788 (019)
Ce 8 (97) Cd 26 (03) 6336 (81)
13884 (03) 15304 (02) 5509 (00)
T60-233
5508 (90)
T60-234
6800 (98)
T60-241 T60-220
3960 (01) 6885 (98)
T60-246
Cd 14d (99) Ce 1f (02) Ch 2a (02) Ce 2 (97) Ce1 (97) Ce 1e (01) Cd 8 (03) Cd 8b (03) Cd 18 (97)
C-VI 8b (99)E C-IV 11 (98) C-IV 11a
C-VI 11a (98) C-VI 11d (98) C-VI 10a (00)
C-VI 6a (02)E C-VI 6e (84)
(Continued)
418
Analysis
TABLE 6.1
Continued ISOa
Subject matter ®
Oxidative stability (Rancimat ) CBE in chocolate (qualitative) CBE in chocolate (quantitative)
AFNORb
6886 (96) 23275-1 23275-2
T60-219
NENc
AOCSd
DGFe
Cd 12b (97)
C-VI 6f
a
http://www/iso.org http://www.afnor.fr c http://normen.nen.nl d http://onlinemethods.aocs.org e http://www.dgfett.de f The Official Methods and Recommended Practices of the AOCS provide these methods per vegetable source material. Abbreviations: CBE, cocoa butter equivalents; FTIR, Fourier transform infrared; HPSEC, high-performance size exclusion chromatography; RP HPLC, reversed phase high performance liquid chromatography. b
together, e.g., phosphorus (colorimetric), phosphorus (ICP: inductively coupled plasma). To indicate to what extent a method will be applicable in a given laboratory, instruments to be used such as GLC, HPLC, ICP, etc. have been listed between brackets behind the subject matter. This does not mean that no dedicated equipment is required if nothing has been listed, but that, in general, these requirements will form less of a financial hurdle.
because a more recent analytical method was found to be more accurate, cheaper, or faster.2 The selection of these methods is somewhat arbitrary, but intends to provide an overview of the most commonly used methods. Methods that reflect national idiosyncrasies, such as the “Break Test (crude soya bean oil)” and the test to prove the presence of sesame seed oil have also been omitted, now that margarine in various European countries no longer has to distinguish itself from butter by the incorporation of 2% sesame seed oil that could then be detected by the Baudouin test involving shaking an oil sample that had been washed with strong aqueous hydrochloric acid with an alcoholic solution of furfural (Boekenoogen, 1948). Italy was the last country to abolish this regulation. Subjective methods involving taste panels have also been omitted. In general, the list comprises methods that an industrial laboratory should know about and might want to use. If the various issuing bodies each provide their own method for a certain property, these methods, as listed in Table 6.1, are similar; some may be identical, but there may also be slight variations. These variations may refer to sample weight, the composition of solvent mixtures, duration of treatment, temperature, etc. For practical purposes, they should be regarded as equivalent, but when a contract prescribes a specific method, that is the method that should be used. A common feature of the tables of contents provided by all the various issuing bodies for their methods is that the order in which the various analytical methods have been listed is to some extent historically determined, but otherwise arbitrary. Each issuing body has its own order. Consequently, Table 6.1 is also organised in a somewhat arbitrary manner, except that different methods for the determination of the same property have been grouped 2
References AOCS (2003a) Official Methods and Recommended Practices of the AOCS, 5th ed., 2nd printing, AOCS Press, Champaign, IL. Blank, E.W. (1942) Chronological list of important dates in the history of the fats and waxes, Oil & Soap, 19, 110–113. Boekenoogen, H.A. (1948), De Scheikunde der Oliën en Vetten, A. Oosthoek’s Uitgevers Mij., Utrecht. Broadbent, C.J. and Pike, O.A. (2003) Oil Stability Index correlated with sensory determination of oxidative stability in canola oil, J. Am. Oil Chem. Soc., 80, 59–63. Desnuelle, P. et al. (1951) Traitement thermique des graines oléagineuses et teneur en lécithine des huiles, Oléagineux, 6, 264–267. Deutsche Gesellschaft für Fettwissenschaft (2004) DGFEinheitsmethoden. Dijkstra, A.J. (1993) Degumming, refining, washing and drying fats and oils, in Proceedings of the World Conference on Oilseed Technology and Utilization, Applewhite, T.H., Ed., American Oil Chemists’ Society, Champaign, IL, 138–151. Dijkstra, A.J. et al. (1995) Interpreting oil stability data, in OilsFats-Lipids 1995, Proceedings of the 21st ISF Congress, Castenmiller, W.A.M., Ed., P.J. Barnes & Associates, Bridgwater, U.K., 629–637. Dijkstra, A.J. et al. (1996) Interpreting the oil stability index, Oléagineux Corps Gras Lipides/OCL, 3, 378–386. Lacoste, F. et al. (1999) Comparison of Rancimat stability test and ambient storage of edible oil, Paper presented at the 23rd ISF World Congress, Brighton, U.K. Morton, R.B. (1995) Oil content of filter cakes, Paper presented at the 86th AOCS Annual Meeting and Expo, San Antonio, TX. Sen Gupta, A.K. (1988) Micellar structures and their implication in the chemistry and technology of fats and other lipids, Fat Sci. Technol., 90, 251–256.
This also holds for the “Detection of sulfur by the coin test” (AOCS Official Method Ca 8a). The author was taught to prove the presence of sulfur by his late father, also a chemist, who referred to the method as the “hepar-method” (hepar = liver) because the silver sulfide on the silver coin has a colour reminiscent of boiled liver.
419
6.2
Requirements stemming from quality control and process investigation
6.2
spectroscopy (Dijkstra and Meert, 1982) or even IPC-MS (Wiedemann et al., 2004), which also requires a sophisticated instrument, is preferably carried out in-house for the simple reason that the analytical data are required rapidly since they form the basis of urgent decisions and subcontracting may take too long. On the other hand, if the property concerned can be determined by NIR (near infrared spectroscopy), a service called “Quality Trait Analysis” (QTA) has recently been established (Anon., 2005), which allows a crusher to relay the NIR-probe signal via the Internet to a central QTA® processor that sends the estimated values back within minutes. This service obviates the need for analytically skilled personnel and the development and maintenance of the database used to arrive at the NIR-based estimates. It only requires investment in NIR equipment and training of plant personnel in how to use it and relay the data. The system has the potential disadvantage that NIRbased estimates cannot be used as evidence during disputes; these stipulate classical, often wet methods. Of course, the QTA® service uses these methods to arrive at its database, but it may not be in a position to act as an expert witness in such disputes because of conflict of interests. Moreover, although the use of the database by several different crushers saves money, the QTA® service has to make a profit to survive. Accordingly, a situation may well arise whereby relatively small, single site crushers make use of an external QTA® service, whereas large, multisite operators prefer an in-house service that may operate along similar lines. They may also operate a central processor and a single database serving several users throughout the company. In-house services have the advantage that an additional analysis hardly causes additional costs, whereas the operator relying on an external QTA® service probably has to pay a standard amount for each additional analysis. Accordingly, a crusher (especially if MBA controlled) may want to save real money by decreasing the number of samples being analysed and, in doing so, lose control over his process and product; this may well be more costly in the long run.
Requirements stemming from quality control and process investigation
Before discussing ways and means to meet regular analytical requirements, a reminder of the definition of a few analytical concepts may be opportune: • Accuracy: the accuracy of a measurement is given by its bias, which is the systematic difference between the true value and the measured value. Such a bias may, for instance, originate from an inappropriate choice of indicator during a titration that visualises the endpoint too early (or too late). Incomplete cell opening when determining the oil content of a seed by solvent extraction will also lead to a bias. • Reproducibility: the degree to which a measurement can be repeated within the same analytical laboratory. The reproducibility of an analytical determination is commonly quantified by its standard deviation, a measure of the spread of the individual data points around their average. • Repeatability: the extent to which the same analytical method yields the same results when applied by different analytical laboratories. This means that the reproducibility of a measurement is always better (smaller standard deviation) than its repeatability. The latter can also be quantified by a standard deviation, although mathematically, this is not correct since the individual data points are distributed around individual averages that have a different bias with respect to the true value and, thus, do not constitute a normal distribution themselves.
6.2.1
In-house vs. external analyses
Table 6.1 in Section 6.1 lists a large number of analytical methods, but this list is only a selection of the large number of methods available. Many of them are rarely used and most of them are hardly ever needed by a plant manager operating a plant and selling its products. He can limit the methods to be used to those that permit him to make sure that his products are within specification and those that enable him to quantify and, subsequently, optimise the performance of his plant or plants. However, he must decide first of all whether he should have the analyses carried out in-house, ask an external laboratory to carry them out, or opt for a combination of both approaches. There may be occasions when analyses are called for that require sophisticated instrumentation. If these occasions are incidental, it is most convenient to subcontract these analyses to a specialised outside laboratory. Phosphatide composition by 31P NMR can be an example of such an analysis. On the other hand, the determination of residual phosphorus by inductively coupled plasma (ICP)
6.2.2
Labour intensive vs. automated analyses
The analytical methods required for product quality control follow from the sales specifications of the various products. Thus, if urease activity is specified, the crusher has to make sure that he can measure it. Occasionally, the analytical method is also specified. This was, for example, the case when an external company recuperating oil from spent bleaching earth by extraction with hexane, specified that the oil content of the earth had to be determined by Soxhlet extraction with hexane. If an oil colour is specified, the refiner must provide the data required. If the colour is specified in Lovibond 420
Analysis
(Verleyen et al., 2001) that is accurate and reproducible, but has the disadvantage of requiring an initial calibration that is oil-type specific. Subsequently, another FTIR method was developed (Al-Alawi et al., 2004b) that has the advantage that it is specific for fatty acids by measuring them as soaps extracted from the oil by a basic methanolic solution. This specificity is also an advantage over the titration method that just measures acidity. The FTIR method is fast and can be automated so that 60 samples can be measured per hour and, above all, it is far more reproducible, having a standard deviation that is about half the standard deviation of the titration method (Al-Alawi et al., 2004a). Using this method requires the purchase of an FTIR spectrophotometer, but then, this is a multipurpose instrument that can also be used for other determinations. Moreover, since the labour requirement of this FTIR method is almost negligible, the availability of the instrument will encourage duplicate measurements to be made to verify possible rogue values and outliers. Consequently, its sheer availability may lead to more soundly based decisionmaking. The FTIR instrument can also be used for the simultaneous determination of trans-content and the iodine value (Sedman et al., 1998; Sedman et al., 2000) or just the iodine value (Cox et al., 2000). Soap can also be determined by FTIR spectroscopy with an accuracy that is comparable to that of AOCS Method Cc17-95 and an improved repeatability. Because the reproducibility of instrumental methods is usually quite close to its repeatability, the reproducibility of the FTIR-based method will be an even greater improvement compared with the titration method. Finally, water can be determined with high accuracy (in the order of ±10 ppm) by extracting the water from the oil with acetonitrile and measuring the extract against dry acetonitrile (Al-Alawi et al., 2005). It can be concluded that even for relatively small laboratories the investment in an FTIR instrument looks likely to be fully justified. Gas-liquid chromatography (GLC) is a similar, widely applicable analytical technique. Its main use is for the determination of fatty acid compositions, but it can also be used for triglyceride analysis according to carbon number, the determination of the sterol composition, which can be useful for the detection of adulterations and many more incidental applications. Moreover, determining fatty acid compositions makes it possible to calculate trans content, saponification value, and iodine value. An instrument that can operate with capillary columns should be the object of choice. Pulse-NMR as an analytical technique, which is also referred to as “time-domain NMR” (Todt et al., 2001), has quite a number of applications, but some of them require dedicated equipment to ensure sufficient sample homogeneity. The technique was developed for the measurement of the solid fat content (SFC) of fat blends (Van den Enden et al., 1978, 1982; Duynhoven et al.,
units — and this is invariably the case — he has a number of options: • He can purchase a cheap, manual instrument that was developed for the brewing industry (Stillman, 1955) and rely upon the skill of his laboratory assistants in comparing the colour of a set of built-in glasses with the oil colour to come up with proper readings. • He can also purchase a more expensive automated instrument that immediately provides readings in these units and readings that are more reproducible because they eliminate operator influence. • He can stay away from Lovibond and purchase a spectrometer (UV-visible) and calculate on the basis of the visible light transmission spectra to which Lovibond-values the oil colour would correspond (Maes et al., 1997). Since a spectrometer can serve several other purposes, such as the determination of chlorophyll content, which equipment to purchase should be an obvious choice. Investment in the simplest analytical equipment that does the specified job is, of course, easy to justify by being unavoidable. However, it can be difficult to justify less simple and, thus, more expensive equipment. Labour saving in comparison with the simple equipment is becoming a very sound reason as well as the actual analyses it is to perform. In practice however, once the equipment has been purchased and commissioned, it can often be used for other, unforeseen analyses and, thus, yield unexpected positive results that would in hindsight have easily justified its purchase. Laboratory equipment should not be regarded like plant equipment as far as investment decisions are concerned. Plant equipment should be in almost constant use, but most laboratory equipment is only used sporadically. Nobody wants to peer down a microscope all day, but on the occasions that only a microscopic examination can provide the answer, its presence is essential. Not all equipment in a plant laboratory needs to be highly sophisticated and require highly skilled staff to keep it running. Drying ovens, for instance, were never very sophisticated and that is still the case today; they just do what they have to do. Analytical balances, on the other hand, are regularly improved and instead of purchasing a rather elementary model, a plant laboratory might be better served by a model that can be linked to a laboratory information management system (LIMS). Perhaps it does not have to be linked up straightaway, but it would be a pity if this were impossible when such a system were to be installed at a later date. The FFA (free (nonesterified) fatty acid) content of an oil is specified for both crude and refined oils. It is commonly determined by titration and this method of analysis can be carried out on the shop floor. Recently, an FTIR (Fourier transform infrared) method has been developed 421
6.2
Requirements stemming from quality control and process investigation
parameter to be determined, so if a method prescribing it is to be used, the equipment has to be purchased. Again, a versatile and multipurpose instrument is recommended and if some measurements look like becoming routine, autoinjection and connection to a LIMS merit consideration. Investment in equipment to measure the OSI cannot be avoided if this property has been incorporated in sales specifications. Since the OSI does not predict oil shelf life (Lacoste et al., 1999), its inclusion in a specification preferably should be resisted. This saves money, analysis time and interminable discussions.
1999). The ISO 8292 standard and the AOCS Official method Cd 16b-93 describe what thermal treatment to give to the sample before measuring its SFC (see also Section 7.2.5). When using the method for the endpoint control of a hydrogenation batch (Rutledge et al., 1988) a different thermal treatment is used to save time. The equipment most frequently used for this application operates with a Larmor frequency of 20 MHz and a magnetic field strength of 0.47 T. It can accommodate sample tubes with a diameter of up to 25 mm, but also provides accurate data with sample tubes of 18 and even 10 mm diameter. Consequently, this piece of equipment can also be used in the confectionery industry to determine the oil/fat and moisture in chocolate, cocoa powder and milk powder. In addition, it can be used to determine the droplet size distribution in both water-in-oil emulsions like margarine and butter and oil-in-water emulsions like mayonnaise, dressings, and soft cheese. For routine SFC measurements that require the sample to follow a prescribed temperature profile, a laboratory robot that transfers the sample tubes from one thermostat to another and finally to the measuring cell, makes the thermal history more reproducible, saves labour, and avoids human errors. It can also be easily integrated into a LIMS (laboratory information management system). The instrument used for this kind of determination has been programmed to average the results of three consecutive measurements and present this average as the reading. In principle, the repeatability of this reading could be improved by calculating the average of more than three consecutive measurements, but in practice, the magnet is kept thermostatically at a temperature (e.g., 40°C) that causes crystals present in the sample to melt. Accordingly, the sample gets less and less representative with time so that the time span available for representative measurements is limited. Thus, three consecutive measurements is a compromise between representability and repeatability. For less homogenous products like oilseeds, nuts, and especially olives, a larger sample tube diameter is required to ensure sample homogeneity and a model, therefore, must be selected that can accommodate tube diameters up to 52 mm; it operates with a Larmor frequency of 7.5 MHz and a magnetic field strength of 0.17 T in accordance with ISO 10565, ISO 10632, and AOCS Ak 4-99. Ak 5-01 and permits simultaneous measurement of oil content and moisture content in oilseeds and their meal. In comparison with NIR reflectance methods, the pulseNMR method’s advantage is that it analyses the entire sample instead of just investigating its surface and determines the property concerned in a direct manner leading to full proportionality between the NMR signal and the property. NIR reflectance methods are indirect and arrive at estimates based upon multivariant correlations. Several Official Methods mention the use of HPLC, reversed phase HPLC or gel permeation HPLC. Like GLC, the equipment is indispensable for the property or
References Al-Alawi, A. et al. (2004a) A new FTIR method for the analysis of low levels of FFA in edible oils, Paper presented at the 3rd EuroFedLipid Congress, Edinburgh, p. 40 of the Abstracts. Al-Alawi, A. et al. (2004b) New FTIR method for the determination of FFA in oils, J. Am. Oil Chem. Soc., 81, 441–446. Al-Alawi, A. et al. (2005) A new Fourier transform infrared method for the determination of moisture in edible oils, Appl. Spectrosc., 59, 1295–1299. Anon. (2005) A user-friendly process monitoring system, INFORM, 16, 654; see also www.qta.com. Cox, R. et al. (2000) Determination of iodine value with a Fourier transform-near-infrared based global calibration using disposable vials: an international collaborative study, J. Am. Oil Chem. Soc., 77, 1229–1234. Dijkstra, A.J. and Meert, D. (1982) Determination of trace elements in oils by plasma emission spectroscopy, J. Am. Oil Chem. Soc., 59, 199–204. Duynhoven, J. van et al. (1999) Solid fat determination by NMR, INFORM, 10, 479–484. Enden, J.C. van den et al. (1978) A method for the determination of the solid phase content of fats using pulse Nuclear Magnetic Resonance, Fette Seifen Anstrichm., 80, 180–186. Enden, J.C. van den et al. (1982) Determination of the solid fat content of hard confectionary butters, J. Am. Oil Chem. Soc., 59, 433–439. Lacoste, F. et al. (1999) Comparison of Rancimat stability test and ambient storage of edible oil, Paper presented at the 23rd ISF World Congress, Brighton, U.K. Maes, P.J. et al. (1997) Converting spectra into color indices, INFORM, 8, 1245–1252. Rutledge, D.N. et al. (1988) Contribution à l’étude d’une méthode de contrôle rapide de la qualité des margarines par RMN-IBR, Rev. Franç. Corps Gras, 35, 157–162. Sedman, J. et al. (2000) Simultaneous determination of iodine value and trans content of fats and oils by single-bounce horizontal attenuated tot reflectance Fourier transform infrared spectroscopy, J. Am. Oil Chem. Soc., 77, 399–403. Sedman, J. et al. (1998) Industrial validation of Fourier transform infrared trans and iodine value analyses of fats and oils, J. Am. Oil Chem. Soc., 75, 33–39. Stillman, R.C. (1955) Bleach and color methods, J. Am. Oil Chem. Soc., 32, 587–593.
422
Analysis
fatty acid content by oxidising the unsaturated fatty acids in a controlled manner, followed by a separation based on the insolubility of the magnesium soaps of saturated fatty acids. If no linolenic acid is present, using the iodine value (IV) and the saturated fatty acid content allows us to calculate the oleic acid and the linoleic acid contents in terms of weight percentage, because the two unknowns left (oleic acid content and linoleic acid content) can be calculated from the iodine value with the sum of the three fatty acids content being 100. Instead of determining the saturated fatty acid content, the rhodanic acid (thiocyanic acid) value could have been determined. Whereas both double bonds in linoleic acid add an iodine molecule during the determination of the iodine value, only one of them adds a rhodanic acid molecule. Accordingly, the difference between the iodine value and the rhodanic acid value is indicative of the linoleic acid content and as shown by Boekenoogen (1948, p. 160), the two methods yield quite close results. Rhodanic acid also adds to linolenic acid and, since apparently two methylene interrupted bonds are too close to accommodate two rhodanic acid molecules, it was felt that linolenic acid would be able to accommodate two of them. However, if the first one were to be added to the central double bond in the linolenic acid, the two outer ones would be too close for further addition. This would mean the addition of a single molecule, just as for linoleic acid. In practice (Kass et al., 1940), linolenic acid is found to add on average 1.83 mole equivalents of rhodanic acid, indicating some addition to the central bond. So, in the past, when the analytical tools were more limited, ingenuity was called for. When UV-spectrophotometers became available, this provided more direct information since linoleic acid could be determined by conjugating the acid. But until the advent of chromatographic separation methods, the number of components that could be determined was limited to the number of independent values that were available.
Todt, H. et al. (2001) Quality control with time-domain NMR, Eur. J. Lipid Sci. Technol., 103, 835–840. Verleyen, T. et al. (2001) Influence of triacylglycerol characteristics on the determination of free fatty acids in vegetable oils by Fourier transform infrared spectroscopy, J. Am. Oil Chem. Soc., 78, 981–984. Wiedemann, S.C.C. et al. (2004) Direct analysis of Ca, P, and Fe in oleochemicals by inductively coupled plasma MS, J. Am. Oil Chem. Soc., 81, 437–440.
6.3
Some selected analytical methods
Because the edible oils and fats industry has a long and respectable history, the analytical methods developed within the industry reflect this time span. Some methods date from when analytical chemistry was limited to gravimetric and volumetric measurements. Examples of purely gravimetric methods are the measurements of oil content in oilseeds, meal, etc., of their water content, the determination of the unsaponifiable content of oils and of the oil content of spent bleaching earth. A combination of gravimetric and volumetric methods is used to determine the FFA content of oils, their saponification value, hydroxyl value, peroxide value, iodine value, and the determination of the water content of substrates by azeotropic distillation or by Karl Fischer’s method. As will be shown below, several “values” that were formerly determined by these volumetric methods can now be determined more quickly and often more accurately by instrumental methods. That is to say, the instrumental method provides the data with which the “value”, such as the iodine value, can be calculated. This calculation demands an understanding of the basic principles of the old method. Accordingly, some of these methods will be explained here; more detailed descriptions can be found in the respective Official Methods. In general, the gravimetric and/or volumetric methods are well proven. The equipment used in these methods is straightforward and comprises an analytical balance and standard glassware. Accordingly, the initial outlay is quite low, but the methods tend to take a long time and to be labour intensive. Besides, the repeatability of these methods may be unacceptably poor. Because of these disadvantages, several of these ancient methods have been replaced by modern, often instrumental methods, such as the FTIR-based method for the determination of FFA content (Al-Alawi et al., 2005b), soap and water (Al-Alawi et al., 2005a). In addition, novel methods are constantly being developed to measure properties that were rarely measured in the past. Take, for instance, the fatty acid composition. The average molecular weight of the fatty acids has been known for a long time, since it follows from the saponification value as obtained by titration, but this value does not tell us anything about the fatty acid composition. The determination of this fatty acid composition comprises as the first step the determination of the saturated
6.3.1
Acidity
For an oil refiner, the acidity of his oil is an important parameter since it affects the yield loss on refining. The acidity of oil is commonly expressed as either the acid number or the FFA content. The acid number is defined as the amount [mg] of KOH (potassium hydroxide) needed to neutralise 1 g of oil; being a ratio of two weights (mg/g), it has no dimension. The FFA content is defined as a mass percentage of these acids while assuming them to be oleic acid with a relative molecular mass of 282 (weight % expressed as oleic acid). So an oil with an FFA of 1.0 % contains 0.01 g oleic acid per g of oil, which amount of oleic acid equals 0.01:282 = 3.546 × 10−5 mol = 3.546 × 10−2 mmol. Given the relative molecular mass of KOH of 56.11, the amount of potassium hydroxide required to neutralise this amount of oleic acid 423
6.3 Some selected analytical methods
equals 3.546 × 56.11 × 10−2 = 1.9897. For practical purposes, a ratio of 2 can be assumed between the acid number and the FFA content of oil, provided the latter is expressed as oleic acid.
6.3.2
TABLE 6.2 Literature values for the average relative molecular mass of constituent fatty acids (MWFA), relative molecular mass of the oil (MWoil) and saponification values (SV) for various oils Oil type Canola oil Coconut oil Corn oil Cottonseed oil Fish oil (anchovy) Lard Milk fat Palm oil Palm olein Palm stearin Palm kernel oil Rapeseed oil (HEAR) Soybean oil Sunflower seed oil Tallow
Saponification value
Because the FFA is expressed as oleic acid, the theoretical loss on refining has to be adjusted for the average relative molecular mass of the constituent fatty acids. This average can be calculated from the saponification value of the oil, which has been defined as the amount (in mg) of KOH required to neutralise all fatty acids resulting from complete saponification of a 1 g oil sample. This means that lauric oils with an abundance of medium chain fatty acids contain more fatty acids per unit of mass than for instance soybean oil and, thus, have a higher saponification value. Triolein, for instance, has a relative molecular mass of 884 and the KOH equivalent mass of the three oleic acids in this species (its saponification value or SV) equals:
SV = (1000 : 884) × 3 × 56.11 = 190.42
3 × 56.11 × 1000 SV
SV
890 674 887 860 818 872 893 851 854 833 704 983 878 878 860
189 248 190 196 206 193 188 198 197 202 239 171 192 192 196
If the chain lengths of the fatty acids do not differ very much, the above method of calculation can be simplified by omitting the first step. Using the weight percent data as if they represented molar fractions multiplied by 100 will not introduce noticeable errors for oils like soya bean oil, canola and sunflower seed oil. For lauric oils and fish oils, the first step is preferably maintained. Acid oils resulting from soap stock acidulation contain both FFA and glycerides. The former can be quantified by the acid value and the saponification value is indicative of the average relative molecular mass of the fatty acid moieties, both free and bound, present in the sample. The difference between the saponification value and the acid value is referred to as the ester value and is indicative of the ester bonds still present in the acid oil sample. It is also expressed as the amount (in mg) of KOH required to saponify these ester bonds present in a 1 g oil sample.
(6.1)
(6.2)
which allows the average relative molecular mass of the fatty acids (MWFA) to be calculated according to:
MWFA = (MWoil − 38) : 3
MWoil
284 212 283 274 260 278 245 271 272 265 222 315 280 280 274
HEAR, high erucic acid rapeseed
Working back to the average relative molecular mass of the fatty acids first of all requires the relative molecular mass of the oil (MWoil) to be calculated according to:
MWoil =
MWFA
(6.3)
In the above equation, the term 38 is the difference in relative molecular mass between a triglyceride and its constituent fatty acids. The average relative molecular mass of the fatty acids (MWFA) can then be used to calculate the actual fatty acid content of an oil sample with a given FFA content expressed as oleic acid by multiplying this FFA content with the MWFA value and dividing by 282, which is the relative molecular mass of oleic acid. Often literature values for the MWFA are used as listed in Table 6.2. The saponification value can also be calculated from the fatty acid composition as determined by GLC. The results are expressed in weight percent of FAME (fatty acid methyl esters). Consequently, the first step in the calculation entails the calculation of the molar composition by dividing each weight percent by the corresponding relative molecular mass of the FAME and normalising to a total of unity. The molar fractions obtained are then multiplied by the corresponding relative molecular mass of the fatty acid and these products are then added to give the average relative molecular mass of the fatty acids (MWFA) from which the saponification value can be calculated.
6.3.3
Iodine value
The iodine value (IV) is indicative of the degree of unsaturation of a triglyceride (triacylglycerol) oil. It is defined as the amount of iodine (in g) added to 100 g of oil; the IV, therefore, has no dimension. Accordingly, triolein with a relative molecular mass of 884 and three double bonds per molecule has an iodine value of (100:884) × 3 × 2 × 126.9 = 86.13. Less unsaturated oils, such as the lauric oils, have much lower iodine values (9 to 18) and vegetable oils with a high linoleic acid content, such as sunflower seed oil, corn oil, and soybean oil, have higher values (120 to 132). Very high values are displayed by linseed oil (185) because of its high linolenic acid content and fish oils containing fatty acids with five or even six double bonds. Iodine only adds quantitatively to non-conjugated double bonds and when the oil is oxidised, there may be some substitution. IV values should be interpreted with some 424
Analysis
caution, but for following the overall progress of a hydrogenation reaction for example, the IV is a most useful parameter indeed. It can also be calculated from the fatty acid composition as determined by GLC. Like the calculation of the saponification value from GLC data, the calculation of the IV also entails the calculation of the molar fatty acid composition as a first step. The molar fractions are then multiplied by the number of double bonds in the fatty acid and the products, thus obtained, are added. If this total (sum of double bonds) is called SDB, the IV can be calculated according to:
IV = S DB
884 86.13 MWoil
6.3.5
The titre of a fatty material is defined as the solidification point (in °C) of its water-insoluble fatty acids. If the fatty material is an oil or fat, it has to be saponified before its titre can be determined. This determination requires a special, but simple apparatus comprising a beaker, a wide-necked flask, and a test tube fitted with a manual stirrer and a thermometer. Although the test itself may have become obsolete, awareness of its existence is recommended since it facilitates understanding the literature. For example, a patent describing the frac-tionation of fatty acids (Myers and Muckerheide, 1942) indicates the quality of the stearic acid by quoting its low iodine value and illustrates the absence of saturated fatty acids in the oleic acid fraction by quoting its titre of 1.0°C.
(6.4)
Again, the first step can be omitted for oils with mainly C18 fatty acids. For these oils, the weight fractions of the fatty acids can be immediately multiplied by their number of double bonds, followed by totalling the products obtained, and multiplying this total with the factor 86.13 (the IV of triolein) immediately yields the IV. For fish oils, this simplification would be an oversimplification.
6.3.4
Titre
6.3.6
Hydroxyl value
The hydroxyl value of a fat is defined as “the amount (in mg) of KOH required to neutralise the amount of acetic acid capable of combining by acetylation with 1 g of oil.” This definition is based on the way the hydroxyl value is determined; an oil sample is acetylated with a known excess of acetic acid anhydride. This excess is then allowed to hydrolyse and the amount of acetic acid formed during both the acetylation and the hydrolysis reactions is then determined by titration; the hydroxyl value follows from the difference between this measurement and a blank. During the acetylation reaction, free hydroxyl groups in hydroxy acids and in partial glycerides are acetylated without distinction; presumably, any free hydroxyl groups in lysophosphatides present are also acetylated. This means that the hydroxy acid content has to be corrected for these partial glycerides and/or lysocompounds when this content is arrived at on the basis of the hydroxyl value. The monoglyceride monostearate has a relative molecular mass of 358 so 1 g contains (1 000:358) × 2 = 5.5866 mmol free hydroxyl groups; this amount corresponds to 5.5866 × 56.11 = 313.5 mg KOH. Accordingly the hydroxyl value of monostearate is 313.5. The hydroxyl value of the diglyceride distearate is much lower since this compound has only one free hydroxyl group per molecule and a much higher relative molecular mass than monostearate so that there are fewer mmol per gram. Its hydroxyl value equals 56.11 × (1 000:624) = 89.9, which is just over a quarter of the value of monostearate. Although the hydroxyl value is primarily used for the characterization of oils with hydroxy acids, such as castor oil, it can also be used to give an impression of the partial glyceride content. In fact, partial hydrolysis of an oil leads to free fatty acids and an equivalent amount of free hydroxyl groups in partial glycerides and, since both the acid value and the hydroxyl value are expressed as an
Fatty acid composition
The determination of the fatty acid composition of oil by GLC invariably uses the FAME. Various methods for obtaining a representative sample of FAME have been described, for example, by BF3-catalysed esterification of free fatty acids. In practice, the method by Jáky (1971) is most convenient. Oil (0.2 g) is dissolved in diethyl ether (2 ml) and some (2 ml) methanolic solution of KOH (3%) is added. After shaking and a reaction time of some 3 min, pentane (10 to 15 ml) and water (2 to 3 ml) are added and shaken. The upper layer is decanted and washed twice with water and dried by adding some sodium sulfate. The pentane solution is then ready for injection. The only glassware required is a couple of standard test tubes. Normally, the gas chromatogram peak area can be used to calculate the content of a certain fatty acid; calibration is only required for very demanding analyses and for verification. Nevertheless, FAME gas chromatograms must be interpreted with care since they only refer to standard fatty acids and exclude oxidised, oligomerised, etc., breakdown products. Concluding that the palmitic acid content of a used frying fat equals the chromatogram reading may seriously overestimate its content. The increase in palmitic acid that coincides with a decrease in linolenic acid in oil from damaged soybeans (Robertson et al., 1973) is also indicative of the presence of non-eluting material. Using an internal standard (e.g., C17:0 as provided by triheptadecanoate to ensure that it undergoes the same FAME preparation procedure) overcomes this problem. 425
6.4 Chromatographic analysis of lipids
The amount Mf′ of fresh bleaching earth used to arrive at the amount Ms of spent earth then equals:
amount (in mg) of KOH per gram of oil, it follows that the acid value and the hydroxyl value of this partially hydrolysed oil should be quite close.
6.3.7
M 'f =
Oil content of spent bleaching earth
As pointed out in Section 3.7.5 and shown in Table 3.10 (Morton, 1995), Soxhlet extraction of spent bleaching earth as recommended by Boring (1995) only partially removes residual oil from the earth. Oxidised and polymerised oil is not extracted. Consequently, Soxhlet extraction only provides an answer to how much oil can be recuperated by solvent extraction of the spent earth, but does not allow the yield loss of the bleaching step to be worked out. For the determination of the yield loss of the bleaching step, a procedure involving the determination of the water content and the weight loss during calcination in a muffle furnace is recommended (Dijkstra, 1993). The fresh bleaching earth is thereby assumed to consist of a calcination residue (R), water (W), and other, nonaqueous volatile constituents (V) that are evaporated in the muffle furnace. The spent bleaching earth also contains these components and in addition a certain amount of oil (O). By using the subscripts f and s for the fresh earth and the spent earth, respectively, and introducing M to denote the mass of the samples, the mass (Mf) of the sample fresh earth can be written as:
OL =
Al-Alawi, A. et al. (2005a) A new Fourier transform infrared method for the determination of moisture in edible oils, Appl. Spectrosc., 59, 1295–1299. Al-Alawi, A. et al. (2005b) A new FTIR method for the analysis of low levels of FFA in refined edible oils, Spectrosc. Lett., 38, 389–403. Boekenoogen, H.A. (1948) De Scheikunde der Oliën en Vetten, A. Oosthoek’s Uitgevers Mij., Utrecht. Boring, S.A. (1995) Soybean processing quality control, in Practical Handbook of Soybean Processing and Utilization, Erickson, D.R., Ed., AOCS Press, Champaign, IL, 483–503. Dijkstra, A.J. (1993) Degumming, refining, washing and drying fats and oils, in Proceedings of the World Conference on Oilseed Technology and Utilization, Applewhite, T.H., (Ed.), American Oil Chemists’ Society, Champaign, IL, 138–151. Jáky, M. (1971) Über den Reaktionsmechanismus der Bildung von Fettsäure-methylestern, Fette Seifen Anstrichm., 73, 216–220. Kass, J.P. et al. (1940) The thiocyanogen value of linolenic acid, Oil & Soap, 17, 118–119. Morton, R.B. (1995) Oil content of filter cakes, Paper presented at the 86th AOCS Annual Meeting & Expo, San Antonio, TX. Myers, L.D. and Muckerheide, V.J. (1942), Method of separating fatty acids, US Patent 2,293,676, assigned to Emery Industries Inc. Robertson, J.A. et al. (1973) Chemical evaluation of oil from field- and storage-damaged soybeans, J. Am. Oil Chem. Soc., 50, 443–445.
(6.5)
Similarly, the mass (Ms) of the sample spent earth can be written as: (6.6)
Since the nonaqueous volatiles V will be a fixed fraction of the calcination residue R:
Vs V f = Rs R f
or Vs = V f
Rs Rf
Chromatographic analysis of lipids
6.4.1
Introduction
The study of lipids has assumed considerable importance in recent years with the recognition that they are involved in many vital biological processes in animals, plants, and microorganisms. Methods for the analysis of lipids, therefore, are essential for many research, clinical, and quality control applications, and chromatographic techniques are especially important. Indeed, lipid analysts were at the forefront in the development of gas chromatography (GC) and thin-layer chromatography (TLC), and these techniques provided the springboard
Rs Rf
or
Os = M s − Rs − Ws − (M f − Rf − Wf )
6.4 (6.7)
the amount of oil in the spent earth can then be expressed as:
Os = M s − Rs − Ws − V f
R f (M s − Rs − Ws ) − Rs (M f − R f − W f ) 100% (6.10) RsM f
References
so that
M s = Rs + Ws + Vs + Os
(6.9)
so that the oil loss (OL) can be expressed as a percentage of the amount of fresh earth according to:
M f = Rf +Wf +Vf
V f = M f − R f −W f
Rs Mf Rf
Rs Rf
(6.8)
426
Analysis
liquid chromatography techniques) presents particular problems, and these are discussed below (Section 6.4.5).
for the explosive growth of knowledge in the fields of lipid chemistry and biochemistry that has occurred over the past 30 years. High-performance liquid chromatography (HPLC) was taken up less quickly, but many separations of lipids by means of this technique have now been described that cannot be rivaled by other methods. When such techniques are coupled to mass spectrometry and related techniques, substantial additional information is obtainable. Indeed mass spectrometry without a chromatographic interface is being used increasingly. It is, however, worth noting how much can still be done with relatively simple equipment. The analysis of lipids presents many problems, mainly because of their diverse nature. Some are almost hydrocarbon-like in their physical properties, while others are slightly soluble in water and there is a spectrum of lipids with properties between these extremes. Also, lipids lack spectroscopic chromophores that permit straightforward identification or quantification. Lipid analysts must first decide exactly what information is needed from a sample, and then decide which of a wide range of techniques available is most suited to the problem. Such techniques have been reviewed in much greater detail than is possible here (Christie, 2003). In the analysis of lipids, the use of organic solvents and potentially toxic reagents is essential and appropriate safety precautions must be taken at all times.
6.4.2 6.4.2.1
6.4.2.2
Gas-liquid partition chromatography
GC with capillary or wall-coated, open-tubular columns may be used for any lipid that can be volatilized. The columns consist of narrow bore tubing (0.1 to 0.3 mm in internal diameter and 25 or 50 m in length commonly) of fused silica, the inner wall of which is coated with the liquid phase. Packed columns, i.e., up to 5 mm wide and containing inert support materials coated with stationary phase, are little used for routine analysis, but may be required for some specialist applications, e.g., combined mass and radioactivity measurements. The function of the column is to allow partitioning of the constituents of a sample between the stationary and mobile phases and this is aided by having the liquid phase as a thin film with a large surface area accessible to the flow of the gas phase. As the sample passes down the column, the molecules of each component partition between the liquid and gas phases according to their individual equilibrium constants, and so travel at different rates and are separated. The efficiency of a given system is dependent on a number of factors, including the nature and flow-rate of the carrier gas, column dimensions, liquid-phase thickness, and column temperature, and these must be optimized to maximize the resolution. Hydrogen is the most efficient carrier gas, but helium may be preferred for safety reasons. Nitrogen is sometimes used for cost and safety reasons, but is much less efficient. Gas chromatography with flame-ionization detection has become the definitive technique for analysis of the fatty acid components of lipids, when separation is both by chain-length and by degree of unsaturation. The detector is robust and linear over several orders of magnitude in sample size, so accurate and reproducible quantification is possible with a high degree of automation. While the technique can be utilized for the separation of any lipid molecule capable of volatilization, or which can be converted to nonpolar volatile derivatives, up to and including triacylglycerols, separation worsens and quantification becomes more problematic as the molecular weights of the analytes increase.
Chromatographic modes of separation Adsorption chromatography
Silica gel is by far the most widely used adsorbent in lowpressure column chromatography and in TLC, and now has innumerable uses in HPLC, but especially for classes of lipids separated according to the number and nature of their functional groups. The adsorptive properties of silica gel are due to silanol groups, i.e., hydroxyl groups linked to silicon, which are attached to the surface and can be free or hydrogen-bonded. In addition, there is water of hydration, which exists first in a strongly bound layer and then in one or more loosely bound layers on the surface. In the early days of the technique, different brands of silica gel tended to vary greatly in their properties, but modern precoated TLC plates and HPLC adsorbents from reputable suppliers can be relatively uniform and consistent in their properties. In the most common application of this separation mode, in which components are separated according to the number and type of polar functional groups (ester, phosphate, hydroxyl and amine groups, for example), isocratic elution with a mobile phase of constant composition may be practical for restricted types of lipid class, and this is the usual mode for TLC applications. On the other hand, gradient elution in which the polarity of the mobile phase is increased at a controlled rate affords greater versatility for a wide range of lipid types. Detection of lipids separated by adsorption chromatography (and other
6.4.2.3
Reversed-phase liquid-partition chromatography
In “reversed-phase” chromatography, the separation is based on the selective interactions of solute molecules with a relatively nonpolar liquid stationary phase and a relatively polar liquid mobile phase. The mechanism of this interaction has been reviewed in relation to lipid separations by Nikolova-Damyanova (1997). HPLC in the reversed-phase mode has been much used by lipid analysts for the separation of molecular species of lipids within a single lipid class, i.e., they are separated principally according to the sum of the chain-lengths of the fatty acyl or alkyl moieties, together with an appreciable dependence 427
6.4 Chromatographic analysis of lipids
residues bonded chemically to a silica matrix. These afford much better and reproducible resolution. The technique has been reviewed comprehensively by NikolovaDamyanova (1992). The main value is for the separation of simple fatty acid fractions, depending on the number, position and geometry of double bonds, from complex mixtures for further analysis, e.g., for trans fatty acids. It is also very useful for molecular species of simple lipids, such as triacylglycerols, where it complements separations by reversed-phase HPLC.
on the number and configuration of any double bonds. Analogous TLC methods are available, but have been little used because of technical difficulties. Although there are many nonpolar stationary phases in use in reversed-phase HPLC, the most widely employed and most important for lipid analysis are those consisting of relatively long-chain hydrocarbons, bonded chemically to the surface of spherical silica (with particle sizes of 3 to 10 µm). Of these, by far the most widely used stationary phase consists of octadecylsilyl (“C18” or “ODS”) groups. For most lipid separations, either acetonitrile or methanol is a major component of the mobile phase, accompanied by a modifier solvent chosen according to the nature of the lipid. Under optimum conditions, separations are according to the number of carbons in the aliphatic chains of the lipids and by the number of double bonds, each double bond reducing the retention time by the equivalent of about two methylene groups. 6.4.2.4
6.4.3 6.4.3.1
Introduction and preliminaries
As a first step, it is necessary to extract the lipids from their tissue matrices and free them of any nonlipid contaminants before analysis can be attempted. Ideally, this should be done immediately after removal of the tissue from the living organism, but if this is not possible, the tissue should be stored in such a way that it does not deteriorate significantly. For example, samples should be frozen as rapidly as possible, ideally with dry ice or liquid nitrogen, and stored in sealed glass containers at −20°C (or lower) in an atmosphere of nitrogen. Subsequent extraction should be undertaken without thawing. During extraction, as in many other aspects of analysis, it is possible to inadvertently introduce contaminants or to bring about some unwanted change in the composition of the lipids. Autoxidation of double bonds in fatty acids, for example, is particularly troublesome and care must be taken at all steps in the analysis of lipids, not just during storage and extraction, to eliminate the problem. Natural tissue antioxidants, such as the tocopherols, afford some protection to lipid extracts, but it is usually advisable to add further synthetic antioxidants, such as 2,6-di-tertbutyl-p-cresol (BHT) to storage solvents at a level of 50 to 100 mg/l. The presence of large amounts of unesterified fatty acids, diacylglycerols, phosphatidic acid, or lysophospholipids in lipid extracts must be an indication that some permanent damage to the tissues and, hence, to the lipids has occurred. Similarly, lipoxygenases can cause artefactual formation of oxygenated fatty acids, compounding the effects of autoxidation. Such changes can make a crucial difference to the concentrations of some important lipid metabolites, such as free fatty acid and 1,2-diacylglycerol concentrations of tissues. For example, very low free fatty acid concentrations were observed when heart tissue was frozen rapidly and pulverized at dry ice temperatures before extraction (Kramer and Hulan, 1978) in comparison to more conventional approaches. All containers made of plastic materials should be avoided for storage purposes, as plasticisers will leach out and contaminate extracts. Extraction procedures in general have been reviewed in some detail elsewhere (Christie, 1993b).
Ion-exchange chromatography
In essence, the process of ion exchange can be considered as a competition between the solute ions and counter ions present in the mobile phase for fixed sites of opposite charge on a support. The quality and extent of a given separation can be manipulated by varying the nature and concentration of the counter ion or by varying the pH of the mobile phase. Silica-based HPLC phases with a chemically bonded primary amine group (anion exchanger) or sulfonic acid (a strong cation exchanger) as the active agent have been applied to phospho- and glycolipid analyses with some success and afford distinctive separations. Earlier, the cellulose-based ion-exchange media, diethylaminoethyl (DEAE)- and triethylaminoethyl (TEAE)-cellulose (anion exchangers) and carboxymethyl (CM)-cellulose (cation exchanger), were much used by lipid analysts in lowpressure column applications for the separation of polar complex lipids. Although this methodology works well in semi-preparative applications, it now appears to be much less used than formerly. 6.4.2.5
Lipid extraction, storage, and sample handling
Silver ion chromatography
One specialized type of chromatography of particular importance in lipid analysis is a form of complexation chromatography in which silver ions are associated in some manner with the stationary phase. The technique is termed silver ion or argentation chromatography. The principle of the method is that silver ions interact reversibly with the pi electrons of double bonds (cis more strongly than trans) to form transient polar complexes; the greater the number of double bonds in a molecule, the stronger the complex formation and the longer it is retained. Silica gel impregnated with silver nitrate has been used in TLC applications and even in HPLC columns, but a more practical HPLC technique consists in binding silver ions to an ion exchange phase, e.g., phenylsulfonate 428
Analysis
The “Folch” procedure is the standard by which other extraction methods are judged. However, it is timeconsuming and uses large volumes of solvents. In contrast, a method described by Bligh and Dyer (1959), which was devised originally for the extraction of phospholipids from fish muscle tissue in a relatively economical manner, is often recommended for large samples with a high proportion of endogenous water. Quantitative recovery of triacylglycerols is not always achieved when these are major components of a tissue and a second extraction with chloroform alone is then recommended. With green plant tissues, it is necessary to extract the lipids first with isopropanol, in order to deactivate the enzymes, especially lipases, before partitioning with chloroform, methanol, and saline solution as in the Folch method. Special procedures are also required for extraction of such highly polar lipids as polyphosphoinositides and gangliosides (Christie, 1993b; 2003).
Lipid extracts or lipid standards are still susceptible to autoxidation, though not to enzymic attack. They should be dissolved in a small volume of a relatively nonpolar solvent, such as hexane, and stored at -20°C in glass (never plastic) containers, with screw-caps lined with TeflonTM (PTFE), from which air is excluded by flushing with a stream of nitrogen and in the presence of antioxidants. 6.4.3.2
Extraction methods
Lipids occur in tissues in a variety of physical forms. For example, simple lipids are often part of large aggregates in storage tissues, from which they are extracted with relative ease by solvents such as hexane or diethyl ether. On the other hand, complex lipids are usually constituents of membranes (where they occur in a close association with such compounds as proteins and polysaccharides) and they are not extracted so readily. In this instance, the solvents must not only dissolve the lipids readily, but also overcome the interactions between the lipids and the tissue matrix. Most lipid analysts use chloroform-methanol (2:1 by volume), with the endogenous water in the tissue as a ternary component of the system, to extract lipids from animal, plant, and bacterial tissues. The tissue can be homogenized in the presence of both solvents, but better results may be obtained if the tissue is first extracted with methanol alone before the chloroform is added to the mixture. With difficult samples, more than one extraction may be needed and, with lyophilized tissues, it may be necessary to rehydrate prior to carrying out the extraction. The homogenization and extraction should be performed in equipment in which the drive to the blades is from above, so that the solvent does not come into contact with any lubricated bearings. For safety reasons, there is some interest in isopropanol-hexane (3:2 by volume) as an extractant because its toxicity is relatively low, while watersaturated butanol has been recommended for some difficult samples. Lipid extracts from tissues, obtained in this way, will contain significant amounts of nonlipid contaminants, such as sugars, amino acids, urea, and salts, which must be removed before the lipids are analysed. Most analysts use a simple washing procedure, in which a chloroformmethanol (2:1 by volume) extract is shaken and equilibrated with one-fourth its volume of saline solution (Folch, Lees and Stanley, 1957). The mixture partitions into two layers, of which the lower phase is composed of chloroform–methanol–water in the proportions 85:14:1 (by volume) and contains virtually all of the lipids, while the upper phase consists of the same solvents in the proportions of 3:49:48, and contains much of the nonlipid contaminants. It is important that the proportions of chloroform, methanol, and water in the combined phases should be as close as possible to 8:4:3 (by volume), otherwise selective losses of lipids may occur.
6.4.4 6.4.4.1
Fatty acid analysis Introduction
Gas chromatography (GC) with flame-ionization detection is undoubtedly the technique that would be chosen in most circumstances for determination of fatty acid compositions, after conversion to simple ester derivatives. Here, the technique is discussed in terms of capillary columns only, as packed columns are now virtually obsolete, other than for combined mass and radioactivity measurements, where greater sample loads are necessary. The technology is robust and mature, and high precision is possible. Individual fatty acids can usually be identified by GC with reasonable certainty from their relative retention times, especially if the analysis is carried out with a variety of stationary phases, and taking into account the large body of knowledge that now exists on the compositions of specific tissues or organisms. On the other hand, when definitive confirmation of fatty acid structures is required, chemical-degradative and spectroscopic procedures are available. The only technique to compare with GC for the analysis of fatty acid derivatives is HPLC in the reversed-phase mode with UV-absorbing derivatives, although both the capital and running costs of this technique are appreciably higher. HPLC has advantages for the isolation of specific components on a small scale for structural analysis or for radioactivity measurements (Christie, 2003). 6.4.4.2
Preparation of fatty acid derivatives for chromatography
Before fatty acid components of lipids are analysed, it is usually necessary to prepare nonpolar derivatives of various kinds, but usually the methyl esters. Because of the high sensitivity of GC procedures, small amounts of material (usually less than 1 mg and certainly less than 10 mg) are all that is required. Methods for preparation of 429
6.4 Chromatographic analysis of lipids
While short-chain acids, as in milk fat or coconut oil, are completely esterified by all of the procedures described above, quantitative recovery of the esters from the reaction medium can be very difficult because of their high volatility and partial solubility in water. The best methods are those in which there are no aqueous extraction or solvent removal steps and in which the reagents are not heated; these criteria are met by the alkaline transesterification procedure of Christopherson and Glass (1969). Butyl rather than methyl esters may be preferred in this instance. Sphingolipids, which contain fatty acids linked by N-acyl bonds, are not easily transesterified under acidic or basic conditions, but methylation can be achieved with methanol containing concentrated hydrochloric acid (5:1 v/v) under reflux for 5 hours.
the methyl ester derivatives have been reviewed (Christie, 1993a). Other derivatives are preferred for mass spectrometric identification. Free fatty acids are esterified and O-acyl lipids transesterified by heating them with a large excess of anhydrous methanol in the presence of an acidic catalyst, such as 5% (w/v) anhydrous hydrogen chloride in methanol. It is most often prepared by bubbling hydrogen chloride gas into dry methanol, but a simpler procedure is to add acetyl chloride (5 ml) slowly to cooled dry methanol (50 ml). It is usual to heat the lipid sample in the reagent under reflux for about 2 hours, but equally effective esterification is obtained if the reaction mixture is heated in a stoppered tube at 50°C overnight. Boron trifluoride in methanol (12 to 14% w/v) has also been much used as a transesterification catalyst and, in particular as a rapid means of esterifying free fatty acids, and it is recommended in some internationally approved methods. However, there are a number of well-documented disadvantages, especially as the reagent ages. A solution of 1 to 2% (v/v) concentrated sulfuric acid in methanol transesterifies lipids in the same manner and at much the same rate and the reagent can readily be prepared fresh whenever it is required. If any of the acidic reagents is used carelessly, for example, if they are permitted to super-heat in air or are stored too long, they can cause some decomposition of polyunsaturated fatty acids. Acidic reagents will also release aldehydes from plasmalogens, converting them to dimethyl acetals. O-acyl lipids are transesterified very rapidly in anhydrous methanol in the presence of a basic catalyst. Free fatty acids are not esterified, however, and care must be taken to exclude water from the reaction medium to prevent their formation by hydrolysis of lipids. 0.5M sodium methoxide in anhydrous methanol, prepared simply by dissolving fresh clean sodium in dry methanol, is the most popular reagent, but potassium methoxide (or occasionally the hydroxide) has also been used as catalyst. The reagent is stable for some months at refrigeration temperature, especially if oxygen-free methanol is used in its preparation. The reaction with lipids is very rapid; phosphoglycerides, for example, are completely transesterified in a few minutes at room temperature. Diazomethane was used as a rapid means of specific methylation of free fatty acids in the past, but is now banned in many countries as both it and the reagents required for its preparation are toxic and potentially carcinogenic. It should be noted that nonpolar lipids, such as cholesterol esters or triacylglycerols, are not soluble in reagents composed predominantly of methanol and will not react in a reasonable time, unless a further solvent is added to effect solution. Toluene is usually recommended for the purpose. Special care may be required if samples are believed to contain fatty acids with cyclopropene, cyclopropane, and epoxyl groups, as they are disrupted by acidic conditions.
6.4.4.3
GC analysis of fatty acid derivatives
The liquid phases in use for the GC analysis of methyl ester derivatives are almost exclusively polar polyesters, with which unsaturated components elute after the analogous saturated fatty acids. They permit clear separations of esters of the same chain-length and with zero or up to six double bonds. Polyesters can be classified according to their degree of polarity and, in current practice, only two main types need be considered, i.e., those of low to medium polarity, such as those of the CarbowaxTM type (polyethylene glycol — under various trade names), and those of high polarity, such as CP-Sil 88TM (Chrompack, Middleburg, The Netherlands), BPX70TM (SGE Melbourne, Australia), or SP-2340TM (Sigma-Aldrich, Poole, U.K.). Nonpolar phases may be required in mass spectrometry because of their low bleed characteristics, but afford poorer resolution (unsaturated elute before analogous saturated fatty acids in this instance). Fused silica capillary columns are robust and, with modern cross-linked and chemically bonded phases, will last for 2 or more years, while affording high resolution and low adsorptivity (quantitative recovery). Excellent results can be obtained with columns of 10 or 25 m in length with most samples, and longer columns are only rarely required. For analytical purposes, those columns used most often have internal diameters of 0.2 to 0.3 mm. Changing the polarity of a polyester phase does not change the order of elution of components within a given chain-length group, but it can affect the elution order relative to components of other chain lengths. With the phases of low to medium polarity, all the unsaturated C18 fatty acids, i.e., with one to four double bonds, emerge from the column before any of the C20 components. Therefore, columns of fused silica coated with phases of the Carbowax type can be recommended for the analysis of the common fatty acids of animal and plant origin, as they give satisfactory resolution of all the important polyunsaturated fatty acids with no interference by components differing in chain-length by two carbon atoms, at least until the C22 to C24 regions. Thus, a 25-metre column (0.25 430
Analysis
16:0 5
20:4(n-6)
18:1(n – 9) 10
15
24:1
22:5(n – 3)
22:6(n – 3)
24:0
22:4(n – 6) 22:5(n – 6)
20:4(n – 3) 20:5(n – 3) 22:0
20:3(n – 6)
18:3(n – 6) 18:3(n – 3) 18:4 20:0 20:1
C16 acetal 14:0
} 16:1
{
6000 4000
C18 acetals
10000
18:2(n – 6)
18:0
12000
BHT
Detector response
14000
8000
Their principal disadvantage is that there is some overlap of fatty acids of different chain lengths and the nature and extent of the problem can be rather sensitive to column temperature or the temperature gradient (see below). The inherent resolution is such that there may be few problems of actual overlap of major components, but a multiplicity of peaks can be revealed, so compounding the identification problems. Columns of the more polar phases are essential for analysis of lipids containing trans fatty acids or for conjugated linoleic acid isomers. Then, 100 metre columns are often recommended and, indeed, may be essential. In Europe, CP-Sil 88TM seems to be preferred, while the SP-2340TM stationary phase is favoured in the U.S. (Ratnayake, 1998). However, more accurate results are obtained if such GC techniques are used after a preliminary separation by silver ion chromatography. Standards are available to permit identification of the common range of fatty acids likely to be encountered in samples, and analysts soon acquire an intuitive understanding of the relationship between the retention times of peaks on a GC trace and their identity. Secondary external reference standards consisting of a natural fatty acid mixture of known composition can also be useful; for example, cod liver oil has been used in this way to identify the fatty acids of other marine species (Ackman and Burgher, 1965). Another useful means of identification is to use equivalent chain-lengths (ECLs) (Miwa et al., 1960). These are obtained by reference to the straight line obtained by plotting the logarithms of the retention times of a homologous series of straight-chain
Erythrocytes
18:1(n – 7)
16000
solvent
mm i.d.) coated with Carbowax 20MTM is suitable for all routine analyses of clinical, seed oil, and fish oil samples. The nature of the injection system for capillary GC is obviously important, and the easiest to use is one of the split/splitless type. It is capable of high accuracy in fatty acid analysis and it should probably be the first choice for this purpose, especially as it is more easily coupled to an automatic injection facility than are those of the on-column type. It can be assumed that all modern GC instruments will be supplied with a temperature-programming facility. A separation of methyl esters of fatty acids from erythrocyte lipids on a Carbowax 20MTM column is illustrated in Figure 6.1. Each of the main chain-length groups is reasonably well resolved. For example, three minor 16:1 isomers are seen and they are distinct from the trace amounts of C17 fatty acids. Similarly the important C18 components are clearly separated and they are in a different region of the chromatogram from the C20 unsaturated constituents. With the last, the only serious overlap problem is with 20:3(n-3), which co-chromatographs with 20:4(n-6). Finally, all the biologically important C22 fatty acids are cleanly resolved, although care is necessary to distinguish them from the 24:0 and 24:1 fatty acids. In this instance, the methyl esters were prepared by acid-catalysed-methylation so that vinyl-ether bound chains of plasmalogens were converted to dimethyl acetals, which elute just ahead of the corresponding fatty acid derivatives. With columns of the highest polarity (e.g., CP-Sil 88TM, BPX70 TM , or SP-2340 TM ), excellent resolution is obtained, especially of positional or geometrical isomers.
20
Time (min)
FIGURE 6.1 Separation of fatty acids of human erythrocytes as methyl esters by GC on a column (25 m × 0.25 mm × 0.2 µm), coated with CP-Wax 52CBTM (Chrompack BV, The Netherlands). The oven temperature was held at 170°C for 3 min, then was raised by 4°C/min to 220°C. Hydrogen was the carrier gas at a flow rate of 1 ml/min. (From Christie, W.W. (2003) Lipid Analysis, 3rd ed., Oily Press, Bridgwater, U.K. With permission.)
431
6.4 Chromatographic analysis of lipids
1
H- and 13C-nuclear magnetic resonance spectroscopy are extremely powerful techniques for structural analysis of fatty acids, but may require more of a pure fatty acid component than is easily attainable (Gunstone, 1992; 1993). On the other hand, mass spectrometry coupled to GC requires very little material and prefractionation is only occasionally necessary (Christie, 1997). Methyl esters are not particularly good derivatives for mass spectrometry and a better approach is to prepare derivatives designed for mass spectrometry, such as the pyrrolidides, picolinyl esters or 4,4´-dimethyloxazolines (Christie, 1997; 1998). Methyl esters are by far the most widely used derivatives for the analysis of fatty acids in general, as they are simple in structure and have good chromatographic properties. In addition, there is a wealth of data on relative retention times and how they vary with the type and position of various structural features. Location of methyl branch points or the position and type of oxygen atom (hydroxyl, epoxyl, keto) in the chain is usually possible by GC-MS, especially when standards are available for comparison. On the other hand, the most common methyl-branched fatty acids, i.e., with iso- or anteiso-methyl branches, can only be recognized from subtle changes in the intensity of some minor ions (Apon and Nicolaides, 1975). Ring structures, other than cyclopropane rings, attached to the aliphatic chain can usually be recognized, although it is never easy to define the positions with certainty. It is usually considered that methyl ester derivatives are not suitable for locating double bonds or other centres of unsaturation, and this is certainly true for monoenoic fatty acids and most dienes, but polyenoic fatty acids often give distinctive mass spectrometric fingerprints with specific ions related to double bond position. This has not always been recognized. As examples, Figure 6.2 shows the mass spectra of methyl esters of three octadecatrienoic fatty acids (molecular weight 292). It is apparent that each spectrum is different and distinctive. The best known of these are methyl γ- and α-linolenate, and Holman and Rahm (1971) showed that the mass spectrum of each contained several ions that were indicative of double bond position. In particular, the spectrum of methyl 6,9,12-octadecatrienoate contains an ion at m/z = 150, presumed to be formed from the terminal part of the molecule by cleavage between carbons 7 and 8. The analogous ion in the mass spectrum of methyl 9,12,15-octadecatrienoate is at m/z = 108, and is formed following cleavage between carbons 10 and 11. It is now recognized that ions at m/z = 150 and 108 may be in general characteristic for methyl esters of polyunsaturated fatty acids of the (n-6) and (n-3) families, respectively (Fellenberg et al., 1987). On the other hand, while they are useful guides, it is not certain that they are always present or unique to the mass spectra of such fatty acids.
saturated fatty acid methyl esters against the number of carbon atoms in the aliphatic chain of component under isothermal operating conditions. The retention times of the unknown acids are measured under the same conditions and the ECL values are read directly from the graph, for comparison with data obtained with a range of standards or with those published in the literature. With modern flame ionisation detectors, the areas under the peaks on the GC traces are within limits linearly proportional to the amount (by weight) of material eluting from the columns. Manual and electronic methods are available, but there is no doubt that the latter are best (Christie, 2003). Problems of measuring peak areas arise mainly when components are not completely separated, and there is no way of overcoming this difficulty entirely. When overlapping peaks have distinct maxima, computer analysis of peak shapes may improve the accuracy of the estimation. Where one component is visible only as a minor shoulder or broadening of a major peak, no manual or computer method is likely to give very precise results for the individual components, although electronic integration can at least give an accurate measure of the total amount of material present in a multiple peak. With all column types, it is necessary to check at frequent regular intervals whether losses are occurring by running standard mixtures of accurately known composition through the columns. Such checks also give an estimate of the precision and reproducibility of the method of measurement. Small correction factors may be applied, when high precision is required, to compensate for the fact that the carboxyl carbon atom in each ester is not ionized appreciably during combustion (Ackman, 1972). If analysts find further correction factors to be necessary, it is probable that some aspect of the chromatographic technique, instrument settings, or column installation has not been carried out correctly. In general, the proper approach to the generation of results of high accuracy is to optimize the equipment parameters and operational technique (sample preparation and injection) so that the true answer is obtained with a primary standard, rather than to employ empirical correction factors to correct for faulty practice (Craske, 1993). It is often of value to add an internal standard, usually an odd-chain fatty acid that does not occur in the sample, as an aid to quantification. This is especially useful if the total amount of each fatty acid, as opposed to the relative proportion, is required. 6.4.4.4
Definitive identification of fatty acid structure by mass spectrometry
GC analysis alone can be an excellent guide to the identity of fatty acids, but there are times when this may not be sufficient. Then, it may be necessary to isolate particular components by silver ion or reversed-phase chromatography (Nikolova-Damyanova, 1992; 1997) for structural analysis by chemical degradative techniques, reviewed by Sébédio (1995), or by spectroscopic means. 432
Analysis
79 90
CH3OOC
80
67
150
Abundance %
70 93
60 50 40
107
55
121
30
163
135
20
M+ 292
150 175
194
10
235 243
208 60
80
100
120
140
160
180 m/z
200
220
264
240
260
280
79 CH3OOC
90 80
108
Abundance %
70 ∗ 95 108
67
60 50 40
55
30
121
20
M+ 292
135 149 163 173
10 60
80
100
120
140
160
180 m/z
191 203 200
236 223 220
249 240
261 260
280
67 90
81
141
80
150
CH3OOC
Abundance %
70 60 50 40
95 55
109 136
30 121
20
∗ 150
∗ 141
10 60
80
100
120
140
161 177 191 160
180 m/z
207
200
221 220
∗ 243 240
M+ 292
261 260
280
FIGURE 6.2 Mass spectra of methyl 6,9,12-octadecatrienoate or γ-linolenate (top), 9,12,15-octadecatrienoate or α-linolenate (middle), and 5,9,12-octadecatrienoate (bottom).
addition, it has a small ion at m/z = 141 that is presumably the corresponding fragment from the carboxyl end of the molecule. We have found an analogous ion in the mass spectra of many different methyl esters of fatty acids with
cis,cis,cis-Octadeca-5,9,12-trienoic acid is a common constituent of many conifer species, especially pines. The mass spectrum of its methyl ester has the characteristic ion at m/z = 150 for the n-6 family of fatty acids. In 433
6.4 Chromatographic analysis of lipids
boxyl group, other characteristic features are evident. Thus, DMOX and pyrrolidides derivatives with double bonds in positions 4, 5, and 6 have characteristic oddnumbered ions at m/z = 139, 153, and 167, respectively, while at pos-ition 3, m/z = 152 is the base peak and at position 2, m/z = 110 is a prominent ion. The mass spectra can become a little more difficult to interpret with increasing degree of unsaturation, and often the DMOX spectra are more informative. Taking α-linolenate as an example, for the picolinyl ester (Figure 6.3), gaps of 26 amu between m/z = 234 and 260, 274 and 300, and 314 and 340 locate the double bonds in positions 9, 12, and 15, respectively. For the DMOX derivative, gaps of 12 amu between m/z = 196 and 208, 236, and 248 and between 276 and 298 locate double bonds in positions 9, 12, and 15, respectively. The mass spectrum of the pyrrolidide derivative is very similar to that of the DMOX in this instance. In polyunsaturated acids it is sometimes useful to look for gaps of 40 amu between ions corresponding to fragments containing the double bond and the alphamethylene group. When the position of all double bonds are not located unambiguously, the positions of the other double bonds can be inferred with reasonable confidence because most polyunsaturated double bonds are methylene-interrupted (the total number of carbons and degree of unsaturation can be determined from the molecular ion). DMOX derivatives are particularly informative for locating double bond position in conjugated dienes, such as conjugated linoleic acid (CLA). Double bond positions can be confirmed unambiguously by deuterating the double bonds and looking for gaps of 15 amu between adjacent carbons. However, the reaction is best carried out on simplified mixtures as deuterated components with the same number of carbons but different degrees of deuteration (from fatty acids of different degrees of unsaturation) will co-elute. In general, problems of interpretation tend to arise when functional groups are adjacent to either end of the molecule. When they are close to the carboxyl group, interpretation can be assisted by reference to spectra of similar types of compound. When the functional group is near the terminal part of the molecule, picolinyl esters give the best results usually, while pyrrolidides may also be suitable. However, DMOX derivatives often give confusing results. This can be seen from Figure 6.4, which illustrates spectra of anteiso-methyl-hexadecanoate in the form of various derivatives. The picolinyl ester gives a clear and unambiguous spectrum in which the gap of 28 amu between m/z = 304 and 332 defines the position of the methyl group. The appropriate part of the spectrum of the pyrrolidide has to be magnified for clarity, but again a gap of 28 amu (m/z = 266 to 294) locates the methyl group. However, there is no such gap in the spectrum of the DMOX derivative and it is presumed that a loss of one or more methyl groups from the derivatizing moiety confounds the expected pattern.
a 5,9-double bond structure from sponges, including dienes as well as polyenes (unpublished). In addition, there is a distinctive ion at m/z = 243 or [M-49]+, and this is present in the spectra of other 5,9-dienoic or polyenoic esters. It is hard to see how this could arise other than by sequential loss of a methoxyl ion and water (not necessarily in this order). However, it may also occur in some other spectra. Many different alternative derivatives have been utilized for identification of fatty acids by mass spectrometry, but the most useful contain nitrogen atoms because the ester/ amide moiety rather than the alkyl chain carries the charge when the molecule is ionized. Radical-induced cleavage occurs in a relatively simple way, giving mass spectra that are usually easy to interpret in terms of the positions of functional groups, including double bonds, in the chain. Of the many derivatives that have been tested, pyrrolidides (Andersson, 1978), 4,4-dimethyloxazoline (DMOX) (Spitzer, 1997) and picolinyl ester (Harvey, 1992) derivatives have been widely used and a wealth of information is available on their mass spectrometric properties (see www.lipidlibrary.co.uk). DMOX derivatives are only slightly less volatile than methyl esters on gas chromatography, and their resolution is comparable. On the other hand, picolinyl esters require a temperature of about 50°C higher to elute and usually require specialist hightemperature cross-linked polar GC phases. Pyrrolidides have intermediate properties. In the mass spectra of these derivatives, a saturated chain is indicated by regular gaps of 14 amu between ions due to cleavage of adjacent methylene groups and functional groups in the chain lead to a change in this pattern. For example, methyl branches are located by a gap of 28 rather than 14 amu (see below). By coincidence, pyrrolidides and DMOX derivatives of a given fatty acid have the same molecular weight, although the structures are very different. With electron-impact ionization, they give very similar mass spectra with often identical types of fragmentation although the relative abundances of the ions may vary and DMOX derivatives tend to give somewhat more intense diagnostic ions. For example, in both, the McLafferty ion is at m/z =113 and this is usually the base peak. In general, exactly the same type of fragmentation is seen (with some important exceptions described below). The most common application for these derivatives is in the location of double bonds in unsaturated fatty acids. If a double bond is positioned between carbons n and n + 1 then a gap of 12 amu between ions corresponding to fragments containing n-1 and n carbons is usually observed for pyrrolidide and DMOX derivatives. A gap of 26 amu between ions corresponding to fragments containing n – 1 and n + 1 carbons is more reliable for picolinyl esters. Monoenes (C18) with double bonds between positions 7 and 5 follow these rules, and also give intense allylic ions corresponding to fragments containing n + 1 and n + 2 carbons. When the double bond is closer to the car434
Analysis
92
260
220
90
300
340
CH2OOC
80
M+ 369
N
Abundance %
70 108 60 164
50 40
300 67
30 20
151
55 135
10 60
80
100
120 140
260
206 220
178
160 180 200 220 240 260 m/z
314
280 300
340 354
320 340
126 90 80 70 Abundance %
208
N
248
288
113 O
60
196
276
236
50 M+
40
331 30
55
79 182
140
93
10
152 60
80
262
222
20
100
120
140
168
160
180
276
236 196
208
200 m/z
288
248 220
240
260
280
302 316
300
320
m/z
FIGURE 6.3 Mass spectra of 9,12,15-octadecatrienoate or α-linolenate in the form of the picolinyl ester (upper) and dimethyloxazoline derivative (lower).
Small proprietary solid-phase extraction cartridges of silica gel are convenient for the purpose, nonpolar lipids being recovered by elution with hexane-diethyl ether (1:1 by volume), while the complex lipids (phospholipids and glycolipids) are recovered by elution first with methanol and then with chloroform–methanol–water (3:5:2 by volume) (Bitman et al., 1984). Alternatively, aminopropyl-bonded phase cartridges (Bond-ElutTM) have been much used for a more comprehensive fractionation, though particular care is required to recover the acidic lipids quantitatively. Separate fractions containing simple lipids, free fatty acids, glycolipids, zwitterionic phospholipids, and acidic phospholipids can be recovered by a simple elution sequence (Christie et al., 1998) (Table 6.3). Sphingolipids, and especially the complex glycosphingolipids, are best analysed separately and can be isolated by a similar method to the last (Bodennec et al., 2000). More comprehensive separations of glycosphingolipids
There are many examples of characterizing the fatty acids in natural mixtures using GC-MS in the literature. However, often the mixtures are complex and it should be noted that complementary techniques, e.g., reversed-phase and silver-ion HPLC, are extremely useful for fractionating samples into simpler mixtures prior to GC-MS analysis to allow a full characterization of all fatty acids (Christie, 1997; 1978). It is always helpful when at least two derivatization techniques are employed as an aid to identification.
6.4.5 6.4.5.1
Separation and analysis of lipid classes Preliminaries
While it is often possible to analyse a number of different lipid classes in a single analytical procedure, it is often easier technically to isolate small amounts of pure lipid classes after a preliminary fractionation kind has been carried out. 435
6.4 Chromatographic analysis of lipids
92 108
332
90 CH2OOC
80 Abundance %
70
N 304
60
164 151
50 40 30
57
304
69 220 206
178
10 60
M+ 361
332
20
80
234
262
290
346
100 120 140 160 180 200 220 240 260 280 300 320 340 m/z 113
90 294
80 Abundance %
70
N OC
60 50
266
40
×5
M+
30 20 10
126 55
60
70
80
100
182
154
85 98 120
140
160
323
294
168
140
196
180 200 m/z
224 238 252 220
240
266
260
280 280
300
320
113 90 80
Abundance %
70
N
60
O
50 40 30
126
20 10
55 60
72 83 80
98 100
140 120
140
168 182 196 210 224 160
180 200 m/z
220
280 252 266 240
260
280
294
308 M+ 323
300
FIGURE 6.4 Mass spectra of 14-methyl-hexadecanoate (anteiso-methyl) as the picolinyl ester (top), pyrrolidide (middle), and dimethyloxazoline derivatives (bottom).
ing O-acyl fatty acids are converted to methyl esters and water-soluble products, while the glycosphingolipids and sphingomyelin (which contain amide-bound fatty acids) are not affected (Wells and Ditmer, 1965). Hakomori (1983) has reviewed such procedures.
can be obtained by chromatography following derivatization or by larger-scale, ion-exchange chromatography. It is also possible to eliminate any nonsphingolipid components from a concentrate of these compounds by mild alkaline transesterification, by means of which any lipids contain436
Analysis
TABLE 6.3
frequently present in lipids subjected to chemical or enzymatic hydroperoxidation. In addition to mere detection, second-derivative UV spectroscopy has proved of value for the identification of configurational isomers of conjugated dienes. In some instances, it has proved possible to convert lipids to derivatives that absorb strongly in the UV range. For example, fatty acids have been converted to aromatic esters (e.g., phenacyl or naphthacyl), the sugar moieties of glycolipids have been benzoylated and diacylglycerols derived from phospholipids have been esterified with aromatic acid derivatives for analysis. Good quantification is often possible, as there is then a linear molar response to the derivatizing moiety. Most lipids exhibit a weak absorbance in the range 200 to 210 nm and that is the result of the presence of isolated double bonds predominantly, although other functional groups, such as carbonyl, carboxyl, phosphate, amino, and quaternary ammonium, have some effect; it is sometimes termed “end absorption.” There are a number of disadvantages to using UV detection at such wavelengths, however. Many of the solvents of proven value in the chromatography of lipids, such as chloroform, acetone, ethyl acetate, or toluene, absorb strongly between 200 and 220 nm and so cannot be used. Because small differences in the degree of unsaturation of each component can make a big difference to the response, quantification is not at all easy, and saturated lipids might be overlooked. However, direct quantification has been achieved by determining the apparent extinction coefficient for each component in standard mixtures very similar to those to be analysed. If the samples are likely to be variable in composition or indeed if the composition is not determined, it is necessary to collect the peaks for estimation by appropriate micromethods. Many analysts have followed this approach, using phosphorus analysis for phospholipids, for example. GC analysis of the methyl ester derivatives of fatty acid constituents, prepared from fractions, with an added internal standard permits identification and quantification simultaneously and has wider applicability.
Preliminary separation of lipid classes
An IsoluteTM NH2 (500 mg) (or equivalent) cartridge is conditioned by elution with iso-hexane (2 ml) and the lipid extract (up to 4 mg lipid or 2 mg complex lipids) is applied to the column in the minimum volume of chloroform. All solvents are allowed to run under gravity in the following sequence. Simple lipids are eluted with diethyl ether (10 ml). Free fatty acids can be eluted with diethyl ether-acetic acid (98:2, v/v; 5 ml) (optional, as small amounts of glycolipids may co-elute). Sterol glycosides, and mono- and digalactosyldiacylglycerols are eluted with acetone-pyridine (1:1, 10 ml). Phosphatidylethanolamine and the choline-containing lipids are eluted with methanol (20 ml). Acidic phospholipids (phosphatidylglycerol, phosphatidylinositol and phosphatidylserine) together with sulfoquinovosyldiacylglycerol are recovered by elution with chloroform–methanol–28% aqueous ammonia containing 0.05M ammonium acetate (4:1:0.1 by volume; 10 ml). In this instance, the lipids are recovered by adding methanol (2 ml), water (3 ml), and acetic acid (0.12 ml). Two phases form. The upper aqueous phase is removed by means of a Pasteur pipette and discarded; the lower contains the lipids. The solvents from each fraction are evaporated by means of a rotary evaporator or in a gentle stream of nitrogen at a temperature no greater than 50°C. Lipid components are stored in iso-hexane (with BHT) at 4°C.
6.4.5.2
Detectors for HPLC
Before considering specific methods for separating lipid classes by HPLC, it is advisable to consider the properties of the various detection systems available (Table 6.4.), since these determine to a large extent the nature of the solvent systems that can be used. They have been reviewed by Christie (1992, 2003). Spectrophotometric detectors in the UV-visible range are probably by far the most widely used detectors for HPLC in general. Much the best response is given with compounds containing conjugated double bond systems and aromatic rings, but such substituent groups are found only rarely in natural lipids. Some seed oils contain fatty acids with conjugated double bond systems and these are TABLE 6.4
Properties of different detectors
Detector
Advantages
Disadvantages
UV
Relatively inexpensive and widely available
RI
Linear response to most lipids
Few natural lipids have suitable chromophores Detection at ~210 nm where isolated double bonds absorb is sensitive to the nature of the fatty acids Gradient elution rarely possible Many solvents not suitable as they absorb at the wavelengths of interest. Not very sensitive Gradient elution not possible Sensitive to minor fluctuations in ambient temperature
ELSD
Sensitive response to most lipids Most solvents can be used, and with complex gradients
Response is not rectilinear Requires careful calibration for each lipid class Inorganic ions cannot be used in mobile phase
437
6.4 Chromatographic analysis of lipids
and reproducibility as those from most other analytical methods in use with intact lipids.
Refractive index (RI) detectors, which function by monitoring continuously the difference in refractive index between the eluent and the pure mobile phase, are universal in their scope and can be used with any solute for which the refractive index is different from that of the mobile phase. They are at their best with isocratic elution and, unfortunately are very sensitive to fluctuations in ambient temperature. For quantitative analysis of lipids, the consensus appears to be that acceptable results can be obtained by equating detector response directly with the mass of components, on the assumption that differences in the chain-length or degree of unsaturation of the fatty acyl moieties will have little or no effect on the refractive indices of molecular species. Most analysts have used this simple approach. On the other hand, a few systematic studies have demonstrated that careful calibration with pure standards can improve the accuracy. Only for size-exclusion chromatography do RI detectors appear to be especially favoured, although they are also well suited to the isolation of particular lipid components by preparative HPLC. With evaporative light-scattering detectors, the solvent emerging from the end of the HPLC column is evaporated in a stream of air or nitrogen in a heating chamber; the solute does not evaporate, but is nebulized and passes in the form of minute droplets through a light beam, which is reflected and refracted. The amount of scattered light is measured and this bears a relationship to the amount of material in the eluent. These detectors can be considered as universal in their applicability, in that they will respond to any solute that does not evaporate before passing through the light beam. Most solvents, including ketones, esters, and chlorinated and aromatic compounds, can be used in complex gradients; up to 20% of water and small amounts of ionic species are also permissible. They give excellent results under gradient elution conditions and are simple and rugged in use, with sensitivity better than that of refractive index detection. The detector is destructive in that the sample is lost, but it is possible to insert a stream splitter between the end of the column and the detector to divert much of the sample to a collection device. Sadly, the evaporative light-scattering detector does not give a rectilinear response to analyte concentrations. In fact, detector response increases sigmoidally with increasing sample concentration in a manner that can be predicted by changes in the size distribution of particles in the aerosol. Thus, at low solute concentrations, the solute particles scatter light to a proportionately lesser extent. As the diameters of the droplets begin to approach the wavelength of light, they no longer affect its passage and the response falls off rapidly (Mourey and Oppenheimer, 1984). Also, different lipid classes give very different responses with evaporative light-scattering detection, although chain-length and degree of unsaturation of the acyl constituents do not have a significant effect. With careful calibration, data obtained in lipid analyses of this kind should be at least as reliable in terms of accuracy
6.4.5.3
Separation of simple lipid classes
Although TLC is sometimes considered old-fashioned or “low-tech,” it is effective and gives excellent separations of simple lipid classes. Many solvent systems have been used for the purpose with silica gel G as adsorbent in a single dimension. Those used most frequently contain hexane, diethyl ether and acetic (or formic) acid in various proportions. For example, with these solvents in the ratio 80:20:2 (by vol.), the separation illustrated schematically in Figure 6.5 is achieved in which most of the common simple lipids are separated — cholesterol esters near the solvent front followed by triacylglycerols, free fatty acids, cholesterol, diacylglycerols (1,3- and 1,2-) and monoacylglycerols. Phospholipids and other complex lipids remain at the origin and are determined as a single class — a useful property of TLC methods. A 0.1% (w/v) solution of 2′,7′-dichlorofluorescein in 95% methanol is most frequently used for detection and causes lipids to show up as yellow spots under UV light. The plates are sprayed in a uniform manner until just visibly moist. In recent years, a 0.05% solution of primulin dye (a 1:100 dilution of a 5% aqueous stock into acetone–water, 8:2, v/v) appears to be gaining popularity for the purpose (Skipski, 1975). These sprays are nondestructive and the lipids can be recovered from the plates for further analysis. As all lipids contain fatty acids, it is possible to determine the amounts of lipid classes by quantifying the amounts of the fatty acids that they contain. Typically, the fatty acid components of each lipid are converted to methyl esters in the presence of a known amount of the ester of an acid that does not occur naturally in the sample (e.g., an oddchain compound, such as methyl heptadecanoate (17:0)), Cholesterol esters Triacylglycerols
Free fatty acids
Cholesterol 1, 3-diacylglycerols 1, 2-diacylglycerols Monoacylglycerols Phospholipids
FIGURE 6.5 Schematic TLC separation of simple lipids on silica gel G. Developing solvent: hexane-diethyl ether-formic acid (80:20:2 by vol.).
438
Analysis
peaks tended to interfere and some variation in response with fatty acid composition was observed. However, others showed that this elution system can give acceptable accuracy with relatively simple mixtures, such as those obtained in commerce by glycerolysis of seed oils, if an internal standard (ricinoleic acid) is used and a careful calibration is performed (Ritchie and Jee, 1985). The detector response was found to be rectilinear up to as much as 1 mg of glyceride. Relatively few applications of UV spectrophotometry at 200 to 210 nm in the separation of simple lipids have been described, possibly because this form of detection is of limited value for quantification purposes. Some analysts have used columns of silica gel, but better results appear to have been obtained with cyanopropyl-bonded stationary phases. One potential advantage is that unsaturated terpenoid “lipids,” such as retinol, vitamin E, dolichol, ubiquinone, and their esterified forms show up prominently (Palmer et al., 1984). Evaporative light-scattering detection has been used extensively as part of methods to separate both simple and complex lipids in a single step (see Section 6.5.5 below). However, it has been used also for simple lipids alone. Both silica gel and diol stationary phases have been used for the purpose, usually with a gradient of isopropanol (or with 0.1% acetic acid added) into hexane as the mobile phase. However, some of the more convincing separations have been with bonded nitrile phases. For example, Foglia and Jones (1997) demonstrated a comprehensive separation of cholesterol esters, methyl esters, tri-, di- and monoacylglycerols, cholesterol, and free fatty acids. In this instance, a bonded nitrile column (PhenomenexTM) was employed with a gradient of methyl tert-butyl ether and acetic acid in hexane. The authors calibrated the detector carefully for each of the analytes with suitable standards and obtained the typical curvilinear weight-response curves for evaporative light-scattering detection.
and this serves as an internal standard. Transesterification can be performed on pure lipids recovered from TLC plates, or it can be carried out in the presence of the adsorbent. By means of gas chromatography, the total amount of the fatty acids relative to that of the standard is obtained by dividing the sum of the areas of the relevant peaks on the recorder trace by that of the internal standard. It has the additional merit that both the fatty acid compositions and the amounts of the lipid classes in a given mixture are determined in a single analysis. However, it is necessary to allow for the weight of nonfatty acid material (e.g., glycerol or cholesterol) in each lipid class by multiplying each result by appropriate arithmetic factors. Problems arise also with complex lipids that contain variable amounts of etherlinked alkyl moieties. To avoid such problems, many analysts simply record the total amount of fatty acids in each lipid class. Alternatively, the plates may be sprayed with a solution of 50% sulfuric acid, or similar reagent and the lipids made visible as a black deposit of carbon by heating the plates at 180°C for an hour or so. Although such charring procedures have the obvious disadvantage that they completely destroy the lipids, they are very sensitive, and as little as 1 µg of lipid can be detected by this means. Sterols give a red-purple colour in a few minutes with charring reagents before blackening and this is a useful diagnostic guide. Of course, charring procedures cannot be used with commercial plastic-backed plates. Procedures of this kind are often used to quantify components separated by TLC, when the amount of charred material is measured by means of a scanning photodensitometer, which is capable of reasonable precision after suitable calibration. The procedure has a number of disadvantages; the sample is destroyed, the yield of carbon is variable and affected by degree of unsaturation, authentic standards with a similar fatty acid composition to the analytes are necessary for calibration, but are not always available, and constant updating of the calibration is necessary. While a number of HPLC methods have been described for the separation of simple lipid classes, they do not appear to be in widespread use. One reason is that it is not possible to apply total lipid extracts to columns, as complex lipids would accumulate, causing a build-up of pressure and eventually blocking them. Simple lipid fractions prepared as in Section 6.4.5.1 can be used. Another reason why HPLC is not employed more often is because the detectors available are not ideal for the purpose. A refractive index (RI) detector was utilized with a column of silica gel and isocratic elution with iso-octanetetrahydrofuran-formic acid (90:10:0.5 by volume) to separate most of the common simple lipid classes encountered in animal tissue extracts, such as those of liver (Greenspan and Schroeder, 1982). Cholesterol esters, triacylglycerols and cholesterol all gave symmetrical peaks. Although an attempt was made to use the technique quantitatively, the results were not convincing, as negative solvent
6.4.5.4 Separation of phospholipid classes Preparative-scale procedures involving adsorption and ionexchange chromatography for isolation of phospholipid fractions have been described (Christie, 2003), but only the analytical separations are described here. Phospholipid fractions prepared as described in section 6.4.5.1 are usually analysed, but total extracts may also be used, as simple lipids tend to migrate with the solvent front, so do not interfere with analyses. For many years, TLC was by far the most widely used method for the analytical scale separation of individual complex lipid classes, and with little modification and a minimum loss of resolution it can be used for small-scale preparative purposes. Although HPLC has now made considerable inroads, TLC retains many advantages. For example, TLC offers considerable versatility and precision in lipid analysis with relatively low capital costs. In the analysis of complex lipids, it is a simple matter to use specific spray reagents to detect particular functional 439
6.4 Chromatographic analysis of lipids
groups in lipids separated by TLC, e.g. ninhydrin for free amino groups as in phosphatidylethanolamine and phosphatidylserine, but this is not possible with HPLC. While highly polar lipids, such as the phosphoinositides, are not always well resolved with TLC, they are at least detected; this cannot be guaranteed with HPLC. Commercial precoated TLC plates, especially, give reproducible results in analyses of complex lipids. Phospholipids and glycolipids separated by TLC can be recovered, after they have been detected by an appropriate nondestructive method, by scraping the adsorbent band into a small chromatographic column or sintered disc funnel and eluting with chloroform-methanol-water (5:5:1, by volume). One-dimensional TLC procedures are preferred for rapid group separations or for smallscale preparative purposes; two-dimensional TLC procedures will resolve the maximum number of distinct components. Innumerable combinations of mobile-phase solvents have been described for the separation of phospholipids over the years, many of which should be re-evaluated with modern adsorbents. However, methyl acetate–isopropanol–chloroform–methanol–0.25% aqueous potassium chloride (25:25:25:10:9, by volume) enables separation of the important phospholipids in animal tissues, including all the choline-containing phospholipids, phosphatidylserine, phosphatidylinositol and phosphatidylethanolamine can be separated from each other (Vitiello and Zanetta, 1978). Phosphatidic acid and phosphatidylglycerol tend to run together just behind the phosphatidylethanolamine spot. Similarly, many two-dimensional systems have been devised that are suited to the analysis of animal or plant lipids. The most successful separations are achieved when contrasting solvents are used for development in each direction; for example, a neutral or basic solvent mixture in the first direction may be followed by development with an acidic solvent mixture in the second direction, or the second system may contain acetone to retard the migration of phospholipids relative to glycolipids, with which they might otherwise overlap. As with one-dimensional TLC, many published systems represent minor adjustments only of earlier ones to suit local conditions (e.g., of temperature or humidity). As an example, plant phospholipids and glycosyldiacylglycerols can be separated by first developing the plate in chloroform–methanol–water (75:25:2.5, by volume) in the first direction. After allowing sufficient time for drying, the plate is developed, at right angles to the first development, in chloroform–methanol–acetic acid–water (80:9:12:2, by volume), as illustrated in Figure 6.6. (Rouser et al., 1967). Complex lipids can be located on TLC plates by either the nonspecific destructive or the nondestructive reagents described in Section 6.4.5.3. If the latter are used, care must be taken to remove polar solvents, such as water or glacial acetic acid, otherwise lipid spots will be obscured.
Simple lipids
First direction
MGDG
DGDG
PE
SQDG PG PI
DPG
PC PS
Second direction
FIGURE 6.6 Schematic two-dimensional TLC separations of complex lipids from plant tissues on silica gel. Solvent systems: first direction, chloroform–methanol–water (75:25:2.5, by volume), and second direction, chloroform–methanol–acetic acid–water (80:9:12:2, by volume). Abbreviations: DPG, diphosphatidylglycerol; PE, phosphatidylethanolamine; PC, phosphatidylcholine; PS, phosphatidylserine; PI, phosphatidylinositol; PG, phosphatidylglycerol; MGDG, monogalactosyldiacylglycerol; DGDG, digalactosyldiacylglycerol; SQDG, sulfoquinovosyldiacyglycerol.
This is particularly important if the fatty acid components are required for further analysis and, in this instance, solvents should be evaporated in a vacuum oven at close to room temperature or by blowing nitrogen on to the surface of the plate. Excess acetic acid can be removed by neutralization with an ammonia spray. Such precautions minimize the risk of autoxidation. Phospholipids can be quantified by phosphorus determination, or by analysis of the fatty acid components with an internal standard as discussed in the previous section. HPLC methods are now used routinely in large numbers of laboratories for the analysis of phospholipids, although there is room still for improvement, especially in the analysis of acidic phospholipids. The topic has been comprehensively reviewed (Christie, 1996). HPLC is much more expensive than TLC in terms of both equipment and running costs, but it can be automated to a considerable degree and gives much cleaner fractions in micro-preparative applications. There is no simple recipe that can be recommended unequivocally, as the method of choice will depend to a large degree on the nature of the equipment available to the analyst. For example, when UV spectrophotometry is the only type of detector available, isocratic elution may be essential. Gradient elution techniques afford possibilities for improved resolution, but an evaporative light-scattering detector is then required. HPLC in the adsorption mode must be used for class separations of phospholipids and, in most published work, silica gel has been the adsorbent although chemically bonded stationary phases are being used increasingly. The evaporative 440
Analysis
light-scattering detector can also be employed for analyses of this kind with complicated gradients in the mobile phase to improve resolution and this is discussed in the next section. Two solvent systems transparent at UV wavelengths in the range of 200 to 210 nm were developed first for phospholipid separations, and these still form the basis of most published methods today, i.e., mixtures of hexane–isopropanol–water (Geurts van Kessel et al., 1977) and acetonitrile–water (sometimes with added methanol) (Jungalwala et al., 1976) (Table 6.5). With the latter system, phosphatidylethanolamine elutes before phosphatidylcholine and then sphingomyelin and, indeed all the choline-containing phospholipids tend to be well resolved. A special virtue is that acidic lipids, like phosphatidylserine and phosphatidylinositol, are eluted with relative ease ahead of phosphatidylethanolamine. With mobile phases based on hexane–isopropanol–water, phosphatidylethanolamine elutes before phosphatidylcholine, but the other choline-containing lipids, such as sphingomyelin and lysophosphatidylcholine, tend to be less well resolved. The acidic lipids, such as phosphatidylinositol, phosphatidylserine, and phosphatidic acid, are separated from each other, but in this instance they emerge between phosphatidylethanolamine and phosphatidylcholine. This system has proved easier to adapt to simultaneous separation of simple lipids and glycolipids than has that based on acetonitrile. In addition, it is necessary to be aware of the fact that phospholipids are ionic molecules and may require an ionic species in solution if they are to elute as sharp peaks. Sulfuric and phosphoric acids have often been used, but apart from dissolving HPLC equipment they will bring about complete destruction of any
TABLE 6.5 Phospholipid separations with two different solvent systems The order of elution of phospholipids in mobile phases based on acetonitrile and propan-2-ol.* Acetonitrile-based
Propan-2-ol-based
phosphatidic acid cardiolipin phosphatidylinositol phosphatidylserine phosphatidylethanolamine phosphatidylcholine sphingomyelin lysophosphatidylcholine
cardiolipin phosphatidylethanolamine phosphatidylinositol phosphatidylserine phosphatidic acid phosphatidylcholine sphingomyelin lysophosphatidylcholine
*Note there may be some modification to the order given (especially of cardiolipin), depending on the nature of other solvents and of any ionic species in the mobile phase.
plasmalogens present. Similarly, ammonia can have deleterious effects on packing materials by dissolving silica or hydrolysing chemically bonded phases. Organic buffers are often preferred as counter ions in the mobile phase, as it is still possible to use evaporative lightscattering detection. Isocratic elution methods have the merit of employing simple pumping systems, so reducing the requirements in terms of costly equipment. UV spectrophotometric detectors are by far the most common in laboratories, and they are usually set at 205 nm where isolated double bonds absorb. Of the large number of published procedures for phospholipid analysis of this kind, that of Patton et al. (1982) appears particularly convincing, and has been adopted by many others (Figure 6.7). Their mobile phase is based on hexane–isopropanol–water, but contains a phosphate buffer, ethanol and acetic acid also, while silica gel is the stationary phase. In the
NL
Recorder response
PE
PC
PI
PS
x2 x1
DPG
x3
x4
SPH
PA 0
20
40
60
80
100
LPC 120
Time (min)
FIGURE 6.7 Isocratic elution of rat liver phospholipids from a column of silica gel with hexane–isopropanol–25 mM phosphate buffer–ethanol–acetic acid (367:490:62:100:0.6 by volume) as mobile phase at a flow rate of 0.5 mL/min for the first 60 min, then of 1 mL/min, and with spectrophotometric detection at 205 nm. X1, X2, X3, and X4 are unknown lipids. Abbreviations: NL, neutral lipids; DPG, diphosphatidylglycerol; PE, phosphatidylethanolamine; PA, phosphatidic acid; PI, phosphatidylinositol; PS, phosphatidylserine; PC, phosphatidylcholine; SPH, sphingomyelin; LPC, lysophosphatidylcholine. (Adapted from Patton, G.M. et al. (1982) J. Lipid Res., 23, 190. With permission.)
441
6.4 Chromatographic analysis of lipids
ends with a solvent containing water to elute the phospholipids; a solvent of medium polarity is then needed to mediate the transfer from one extreme to the other and mixtures based on isopropanol (with added chloroform to improve the selectivity of the separation) give satisfactory results. Finally, a gradient is generated in the reverse direction to remove most of the bound water and to re-equilibrate the column prior to the next analysis. Plant lipids are a special challenge because of their content of glycolipids, but it has proved possible to resolve all the important glycolipid classes as well as the simple lipids and phospholipids in a single chromatographic run. Instead of silica, a new stationary phase that is proving especially useful is chemically similar to diol and silica gel in its properties and is manufactured by polymerizing and cross-linking vinyl alcohol to silica gel, i.e., PVA SilTM (YMC Co., Japan). The whole surface is covered and deactivated so that the mobile phase and analytes interact with a uniform layer of hydroxyl groups only. Again, a ternary elution gradient scheme is necessary with isohexane-methyltertbutyl ether (98:2, v/v), isopro-panol–acetonitrile–chloroform–acetic acid (84:8:8:0.025 by volume) and isopropanol–water–triethylamine (50:50:0.2 by volume) as the three components, and simple lipids, glycolipids and phospholipids are eluted sequentially under mild conditions (Christie et al., 1998). The separation is illustrated in Figure 6.8. When evaporative light-scattering detectors are used directly in quantitative analysis, it is necessary to work out the optimum conditions for the desired separations first and then carry out a calibration with lipid standards that are as close as possible in composition to the
original paper, as each component was eluted, it was collected, washed to remove the buffer and determined by phosphorus assay. In addition, the fatty acid composition of each lipid class was obtained with relative ease by gas chromatographic analysis after trans methylation, though it should also have been possible to add an internal standard for quantification purposes. Direct quantification from the response of the UV detector has been attempted by a number of analysts, but this is dependent on the degree of unsaturation of each lipid class. Very careful calibration is essential with standards identical to the lipids in the samples, but rectilinear responses can be obtained for each lipid. Such a method may be of value for routine analysis of very similar samples. 6.4.5.5
Simultaneous separation of simple and complex lipid classes Although it is important to be able to analyse phospholipid classes per se, it is even more of a technical challenge to accomplish simultaneous separation of both simple and complex lipid classes in a single chromatographic run. At the extremes of polarity, cholesterol esters are almost hydrocarbon-like while lysophospholipids are partly soluble in water and there is a broad spectrum of lipids with differing properties in between. However, useful separations are possible with the aid of an evaporative light-scattering detector (ELSD) and a ternary gradient pumping system (Christie, 1985). When solvents are selected for the mobile phase, the choice is constrained by the need for sufficient volatility for evaporation in the detector under conditions that do not cause evaporation of the solute. Usually, a gradient is selected that starts with hexane to separate the lipids of low polarity and SE WE
S
C-b
MGDG
DGDG
C-a
mV
600
400
PC
PG PE SQDG
200 CER TG 0
PI 5
10
15
20
25
30
35
Time (min)
FIGURE 6.8 Separation of lipid classes from leaves of a single plant of Arabidopsis thaliana by HPLC on a YMC-PVA-SilTM column (chromatography conditions as described in the text). Abbreviations: WE, wax esters; SE, sterol esters; TG, triacylglycerol; C-a, chlorophyll a; C-b, chlorophyll b; SG, sterylglycosides; MGDG, monogalactosyldiacylglycerol; DGDG, digalactosyldiacylglycerol; CER, cerebroside; PG, phosphatidylglycerol; SQDG, sulfoquinovosyldiacylglycerol; PE, phosphatidylethanolamine; PI, phosphatidylinositol, PS, phosphatidylserine; PC, phosphatidylcholine.
442
Analysis
partially resolved lipid profile may often contain sufficient information, provided that enough data are available on normal and diseased states to enable significant comparisons to be made. In historical terms, the development of the methodology has followed closely behind that for the separation of intact triacylglycerols since these inevitably must be resolved if the procedure is to be of value. Capillary columns of fused silica with polarizable stationary phases are preferred. Improvements in computerised data handling have contributed greatly to applications involving routine screening, but careful calibration is necessary, especially for the components of highest molecular weight, the triacylglycerols.
samples to be analysed. The operational parameters for the instrument, such as gas pressure and flow rate, evaporator temperature and attenuation, must also be standardised rigorously. Of course, very different results are obtained with different commercial instruments. If the elution conditions or detector settings have to be changed later for any reason, a tedious re-calibration may be necessary. The response differs for each lipid class and is rectilinear over a limited concentration range only. A plot of detector response vs. amount of lipid is sigmoidal, but a straight line is obtained when the logarithm of detector response is plotted against the logarithm of the amount. With careful calibration, the data obtained in lipid analyses of this kind appear to be at least as accurate and reproducible as those from any other analytical method in use for lipids. An internal standard, such as phosphatidyl-N,N-dimethylethanolamine, can be utilized to improve direct quantification with the ELSD. Many lipid analysts have used an ELSD and columns of silica gel to separate phospholipid classes in the absence of simple lipids. One practical system for routine analysis uses a diol column and a gradient of hexane–isopropanol–acetic acid to water–isopropanol–acetic acid (Herslöf et al., 1990; Kaufmann, 1995). With this, it is interesting that acidic lipids, such as phosphatidylinositol and phosphatidylserine, elute after phosphatidylcholine. This procedure is under consideration as a standard approved method for “plant lecithins.” The PVA-SilTM column is also being used increasingly for phospholipid separations per se (cf. Deschamps et al., 2001). A great deal has been accomplished by separating natural lipid mixtures according to molecular weight by high-temperature GC, principally in the laboratory of Kuksis (Kuksis et al., 1990; 1993). In brief, the lipids are first digested with phospholipase C, which converts phosphatidylcholine, lysophosphatidylcholine, and sphingomyelin to diacylglycerols, monoacylglycerols, and ceramides, respectively. The hydrolysis products are converted to the trimethylsilyl (TMS) ether (or related) derivatives, while the cholesterol and free fatty acids also react to form TMS ether and ester derivatives, respectively. Tridecanoin is added as an internal standard for quantification purposes, and the mixture is subjected to GC separation over a large temperature range so that as many as possible of the components are separated. The absolute amounts of the various lipid classes are easily determined, while the proportions of the molecular species give an indication of the chain-length distributions of the fatty acid constituents. Although such methodology could in theory be applied to any tissue, most work has been done on human plasma lipids and related body fluids such as lymph, as rapid screening procedures here can lead to the diagnosis of disorders of lipid metabolism and can assist in monitoring the effects of clinical therapy. For these purposes, a
6.4.5.6
Separation of sphingolipid classes
Sphingomyelin is best analysed with other phospholipids. Analyses of the glycosphingolipids are considered here. These lipids are essential participants in a number of vital processes in living tissues, as they perform a definitive function in the immunogenicity and antigenicity of cells, and they are involved in cellular interactions, differentiation and oncogenesis. The nature of the carbohydrate residue is especially important in this regard, but a further distinctive structural feature of sphingolipids, including both the glycosphingolipids and sphingomyelin, is the presence of long-chain aliphatic bases (sphingoid bases) linked by amide bonds to long-chain fatty acids, which are rather different in composition from those found in the glycerolipids; these are also important for biological activity. Because of the complexity of the carbohydrate residue, analysis of such compounds has become a distinct discipline in its own right. TLC has been used most often for the purpose and HP-TLC especially could be considered to have set the standard. It has the advantage that immunological tests can be applied directly to the TLC plate. However, HPLC procedures are becoming more important. Procedures for TLC separation of neutral glycosphingolipids into classes differing in the number and type of hexose units have been reviewed elsewhere (Schnaar and Needham, 1994; van Echten-Deckert, 2000). Commercial precoated HP-TLC plates, with finer grades of adsorbents, are preferred for most analytical applications. Before use, they should be pre-run in the solvent to be used for the separation and then dried at 125°C for 15 min and cooled and stored in a sealed dry box. Chloroform–methanol–water mixtures varying in proportions from 70:30:4 to 50:40:10 (by volume), depending on the complexity and polarity of the sample, are usually recommended for neutral glycolipids. As a good practical compromise, these solvents in the ratio 60:35:8 can be tried. Ionic species must be added to the mobile phase when gangliosides are present, however. In general, glycosphingolipids with one to four hexose units are clearly separated, as are species differing in the nature of the hexose unit (glucose vs. galactose). Some bands may split into two, according to the presence of normal and 2-hydroxy fatty acyl constituents. 443
6.4 Chromatographic analysis of lipids
Under ideal circumstances, the analyst would wish to separate a lipid into individual molecular species in its native form, in order that the biosynthesis or metabolism of every part of the molecule can potentially be studied, or so that the physical properties of each species can be assessed in relation to those of the whole. With polar complex lipids, the analysis can be simplified in a technical sense by converting the compounds to nonpolar forms by removing the polar head group by enzymatic or chemical hydrolysis. The approach of the analyst will depend on the nature and amount of information required, or on the availability of specific equipment. The chromatographic procedures utilized for the separation of molecular species of glycerolipids resemble in kind those used for the separation of fatty acids (Section 6.4.4), modified according the higher molecular weight of the former. With most lipids, the separations achieved depend on the combined physical properties of all the aliphatic residues in each molecule. Considering triacylglycerols for illustrative purposes, reversed-phase HPLC will separate molecules according to the sum of the chain lengths of the fatty acids, with the retention times being reduced by the equivalent of approximately two carbon atoms for each double bond in the three fatty acid constituents. Silver ion chromatography will separate those molecules containing three saturated fatty acids from those with one monoenoic and two saturated fatty acids, and these are, in turn, separated from further fractions with an increasing degree of unsaturation. Adsorption chromatography can be used to separate molecular species containing three normal fatty acids from those containing two normal fatty acids and one with a polar substituent, such as a hydroxyl group. Often no single method will give the required degree of fractionation, but if two of the above separation modes are used in sequence, a high degree of molecular simplification may be possible. GC and HPLC coupled to mass spectrometry are increasingly proving their value for identification of molecular species.
There have been two general approaches to the separation and analysis of glycosphingolipids by means of HPLC, each having its devotees and being suited to particular purposes, i.e., to subject them to chromatography in the native form or to convert them to nonpolar UV-absorbing derivatives, such as the benzoates, prior to fractionation. These have also been reviewed elsewhere (Christie, 2003; McCluer et al., 1989). Improved methods are increasingly being sought for the chromatographic isolation of glycosphingolipids in the native state, in part to reduce the number of steps and to minimize possible losses or alterations during derivatization, but mainly to permit direct investigations of the antigenicity of specific components, or for physical chemical studies of their interactions with cellular membranes. Adsorption chromatography with silica gel has generally been favoured, but bonded phases have found some applications. As with other aspects of HPLC analysis of lipids, the choice of solvents for the mobile phase has frequently been dependent upon the availability of a particular detector in a laboratory. After conversion to perbenzoyl derivatives, glycosphingolipids are much less polar and can be eluted with nonaqueous mobile phases. More importantly, they can be detected and quantified spectrophotometrically with high sensitivity and specificity by their absorbance at 230 nm. If an appropriate benzoylation method is used, they can be restored to their native state by alkali-catalysed transesterification.
6.4.6 6.4.6.1
Separation and analysis of molecular species of lipids The nature of the problem
Each lipid class in a tissue exists in nature as a complex mixture of related components in which the composition of the aliphatic residues varies from one molecule to the next. Sometimes, as in cholesterol esters, only the single fatty acid component will change. On the other hand in triacylglycerols, each of the three positions in the molecules may contain a different fatty acid. Both the long-chain base and fatty acid constituents of sphingolipids can be variable and can exist in distinctive combinations. For a complete structural analysis of a lipid, it, therefore, is necessary to separate it into molecular species, i.e., into groups of molecules with single specific alkyl or acyl moieties (fatty acids, alcohols, alkyl ethers, etc.) in all the relevant portions of the molecule. With lipids that contain only one aliphatic residue, this can frequently be accomplished without difficulty. When there are two aliphatic residues, the task is much more difficult, but is not impossible. While triacylglycerols can be subjected to some considerable molecular simplifica-tion, it is not yet feasible technically to obtain single species, especially if differing enantiomeric forms are considered. For example, a triacylglycerol with only five different fatty acid constituents may consist of 75 different molecular species (not including enantiomers).
6.4.6.2
Molecular species of triacylglycerols
The most useful single procedure for separation of molecular species of triacylglycerols is arguably reversedphase HPLC (Nikolova-Damyanova, 1997). However, silver ion chromatography (HPLC or TLC) followed by re-chromatography of fractions by reversed-phase HPLC or high-temperature GC will always give more information. As mentioned above, when reversed-phase HPLC is applied to intact triacylglycerols, the separation is in ascending order of the total number of carbon atoms in the aliphatic chains of the three fatty acids, with a double bond in any of the fatty acids reducing the retention time to roughly that of a component with two fewer carbon atoms in total. The relative retention time of a given component has been defined in terms of an “equivalent carbon number” (ECN) or “partition number” value, defined as the actual number of carbon atoms in the aliphatic residues 444
Analysis
third of log α of the corresponding simple triacylglycerol. The retention time of any triacylglycerol could be calculated from such data, or from the graphical relationships. Unlike relative retention times, the numerical values can be seen to have some immediate relevance. However, such data are not easily applied to unknowns. All of such methods present the analyst with lengthy preparatory work, not only in standardizing the chromatographic conditions, but also in running a variety of natural samples and standards to obtain retention data. Graphical methods can be tedious and can obviously only be used if sufficient data points are available. Perrin and Naudet (1985) have simplified the task by tabulating retention data relative to that of triolein for 120 different triacylglycerol species commonly encountered in seed oils. The absolute values listed may not be reproduced directly by other workers, but the data are certainly important for predicting relative orders of elution. The same detector and quantification problems arise in triacylglycerol analysis as in many other aspects of the HPLC of lipids (see Section 6.4.5.2 above). When UV detection at 200 to 210 nm (for isolated double bonds) must be employed, isopropanol, methyltertbutyl ether, or tetrahydrofuran are the most appropriate modifiers of the mobile phase. Alternatively, propionitrile has been used as the sole solvent, but this is costly and highly toxic. Herslöf (1981) showed that by using a wavelength of 215 to 230 nm, i.e., the region where the carbonyl group exhibits a weak absorbance, but away from that where isolated double bonds absorb, sufficiently sensitive detection and good quantification could be obtained with acetonitrile-tetrahydrofuran mobile phases. There are fewer constraints in the choice of solvents when an ELSD is used, and Figure 6.9 illustrates the separation of a low-erucic rapeseed oil, with a gradient of acetonitrile-acetone (author, unpublished). Of course, careful calibration is required if the technique is to be used quantitatively. Silver ion complexation chromatography, used in conjunction with TLC, revolutionized the study of triacylglycerol structures during the 1960s (Nikolova-Damyanova, 1992). With relatively simple equipment, it is possible to obtain distinct molecular fractions, separated on the basis of a single well-defined property, i.e., degree of unsaturation. More recently, silver-ion HPLC has come to the fore and affords distinct advantages. When fractions obtained in this way are subsequently separated by reversed-phase TLC or by high-temperature GLC, additional information is obtained on the chain-length distributions of the fatty acid constituents. For example, silver nitrate TLC has proved immensely useful in separating triacylglycerols, containing a normal range of fatty acids with zero to three cis-double bonds, into simpler molecular species. A triacylglycerol of this type can contain species with up to nine double bonds in the fatty acid moieties per mole of glycerol. Components migrate in the order:
(CN) less twice the number of double bonds (n) per molecule (the carbons of the glycerol moiety are not counted for this purpose), i.e., ECN = CN – 2n Two components having the same ECN value are said to be “critical pairs.” For example, triacylglycerol species containing the fatty acid combinations 16:0-16:0-16:0, 16:0-16:0-18:1, 16:0-18:1-18:1, and 18:1-18:1-18:1 have the same ECN value and tend to elute close together. However, with the best modern equipment, all four components should be separable. The positions of the fatty acids within the triacylglycerol molecules have no effect on the nature of the separations. The ECN concept was useful in the early days of the technique, when the resolving power was relatively limited. On the other hand, the formula is now only of utility as a rough guide to what may elute in a given area of a chromatogram, since the greatly increased resolving power of modern HPLC phases means that the factor for each double bond is not simply 2.0, but has to be defined much more precisely. Also, this factor can no longer be treated as a constant, as a second double bond in a molecule has a slightly different effect from the first. Accordingly, more complex formulae are necessary to define the order of elution of triacylglycerols from modern reversed-phase columns, which in essence means from octadecylsilyl (ODS) stationary phases, as these have been used almost exclusively for the purpose. Acetonitrile is used as the main component of the mobile phase, but it must be modified by a proportion of an additional solvent, such as chloroform, tetrahydrofuran, (or better methyltert-butyl ether), dichloromethane, isopropanol, or acetone. The choice has often been governed by the availability of a specific type of detector rather than by objective separation criteria. Control of column temperature, usually in the range 20 to 25°C, is desirable for reproducible retention times and optimum quantification. Various methods of describing the elution characteristics of triacylglycerols in quantitative terms have been published of which the most successful is probably that of Goiffon et al. (1981a,b). They presented a scheme for identifying triacylglycerol species in which the principle of additivity of solution free energies of saturated and unsaturated acyl residues (up to C18) was utilized. The results were presented as a plot of log k (or retention volume) against the number of double bonds in the triacylglycerol. In practical terms, identifications were accomplished most readily by plotting the number of double bonds in the molecule against the logarithm of the retention time for each component relative to that of triolein, expressed as log α. Parallel straight lines were obtained for all the homologous series. For a given triacylglycerol species, log α was the sum of the equivalent values for each of the three constituent fatty acids, the latter being equal to one 445
6.4 Chromatographic analysis of lipids
600 LOO
OOO
500 TOO LLL LLO
mV (span = 600)
400
300
200 OTT
LOP
100 LLT LOT LPT 0
10
20
OOP
TOP LLP STP
30
40
50
OPP 60
OOS 70
Minutes (span = 75)
FIGURE 6.9 Reversed-phase HPLC of triacylglycerols from low-erucate rapeseed oil, with a gradient of acetonitrile-acetone, an ODS stationary phase, and evaporative light-scattering detection. Abbreviations: P, palmitate; S, stearate; O, oleate; L, linoleate; T, linolenate.
of heptane-acetonitrile, although base-line drift can be a problem and quantification is difficult if not impossible. Evaporative light-scattering detectors are those most often employed, although as in other applications the nonlinearity of the response, and the difficulties of calibration present problems. By including a stream-splitter just before the detector, fractions can be collected, an internal standard added and, following trans-methylation, fractions can be identified by GC of the component fatty acids. The order of elution of triacylglycerol species is very similar if not identical to that by silver ion TLC, and is easy to understand, intuitively, unlike HPLC in the reversed-phase mode. A typical separation is illustrated in Figure 6.10 (Christie, 1988). Although useful separations of intact triacylglycerols by high-temperature GC can be achieved routinely, the technique can still be fraught with difficulties (Buchgraber et al., 2004). The conditions necessary to elute lipids of such high molecular weight from the columns approach the limits of thermal stability both of the stationary phases and of the compounds themselves. Columns constructed from fused silica are preferred, the length being a compromise between the optimum in terms of resolution with a need to limit the exposure time of the analyte to high temperatures to the minimum; commonly the length is 5 to 25 m with an internal diameter of 0.2 to 0.32 mm. Initially, nonpolar stationary phases only (of the methyl silicone type) were used in high-temperature GC, and cross-linking and chemical bonding improved the properties of the columns appreciably. More polar (or polarizable) bonded phases, consisting of phenylmethyl silicones, came into use later and are increasingly being used.
SSS > SSM > SMM > SSD > MMM > SMD > MMD > SDD > SST > MDD > SMT > MMT > DDD > SDT > MDT > DDT > STT > MTT > DTT > TTT where S, M, D, and T denote saturated, mono-, di- and trienoic acids, respectively (they do not indicate the positions of the fatty acids on the glycerol moiety). Silver nitrate (up to 10%) is incorporated into the layers and the solvent systems generally employed consist of hexane–diethyl ether, toluene–diethyl ether, or chloroform (alcohol-free)–methanol mixtures. As all the fractions listed above cannot be separated on one plate, it is common practice to separate the least polar fractions first with hexane–diethyl ether (80:20, v/v) or chloro-form–methanol (197:3, v/v) and then to separate the remaining fractions with more polar solvents such as diethyl ether alone or chloroform–methanol (96:4, v/v). Bands are detected under UV light after spraying with 2´,7´-dichlorofluorescein solution, components are recovered and they are identified and determined by gas chromatography of the fatty acid constituents with an added internal standard (the concentration depending on the scale of the separation). Columns with silver ions attached to cation exchangers (silica based) are used exclusively now for silver ion HPLC of lipids and prepacked columns are available from commercial sources or they are easy to prepare from a commercial ion-exchange HPLC column (Christie, 1987). While chlorinated solvents together with gradients of acetone and/or acetonitrile, which complexes strongly with silver ions displacing unsaturated lipids, were the first practical systems to be described, heptane-acetonitrile is often preferred nowadays. Ultraviolet detectors at wavelengths of about 205 nm can be used with gradients 446
Analysis
SMD
alternative techniques (reviewed by Blomberg et al., 1998). Here the advantages are that elution temperatures can be employed that are low in comparison to gas chromatography, while flame-ionization detection permits more convenient and accurate quantification than is possible by HPLC. Similar chromatographic columns are used as with gas chromatography. The technique is being used in industry in process control of triacylglycerol composition. It is also possible to use the technique in conjunction with mass spectrometry, as the relative volatility of the mobile phase is an advantage in comparison to conventional HPLC-MS (Laakso and Manninen, 1998).
MMM 0
10
20 Time (min)
MDD + SMT
DDD + SDT
SSS
SMM SSD
MMD SSD + SST
SSM
Rat adipose tissue
30
6.4.6.3
Molecular species of complex glycerolipids
Molecular species of polar lipids, such as glycerophosphatides or glycosyldiacylglycerols, can be separated in native form (see below), but analysis is simplified if the polar head group can be removed. While this is self-evident for GC, it can also be true for HPLC as the polar moiety can interfere and limit the quality of the separation. Also, its removal means that molecular species of each complex lipid are analysed under exactly the same conditions. Gas chromatography on a polarizable stationary phase is probably the most appropriate method for analysis of diacylglycerol derivatives when suitable equipment is available. High resolution is possible and the sensitivity, linearity and convenience of the flame ionization detector means that quantification is a relatively simple task. GC can also be linked to mass spectrometry to simplify identification of the separated components. Reversed-phase HPLC after conversion of diacylglycerols to UV-absorbing derivatives is a useful alternative that permits good resolution and simplifies quantification with increased sensitivity. Also, it can be used as a micro-preparative technique, so that fractions can be collected for more detailed analysis or for radioactivity assays in biological experiments, for example. Some years ago, silver ion TLC would have been the first method most analysts would have adopted, but the convenience and accuracy of modern instrumental methods have tended to displace it. Nonetheless, it should not be forgotten as it permits a great deal of useful information to be obtained with relatively simple equipment. A variety of enzymes capable of releasing 1,2-diacyl-snglycerols from phosphoglycerides or of ceramide from sphingomyelin, and termed phospholipase C, have been isolated from microorganisms, but especially from Clostridium perfringens and Bacillus cereus. Each organism produces an enzyme with distinctive properties for particular applications and these are available commercially. Bell (1997) has reviewed the methodology and recommends the use of phospholipase C from B. cereus for the preparation of diacylglycerols from phosphatidylcholine, phosphatidylethanolamine and phosphatidylserine. An enzyme from B. thuringiensis is recommended for phosphatidylinositol. The diacylglycerols produced should be derivatized at once
40
FIGURE 6.10 Separation of triacylglycerols from rat adipose tissue by HPLC on a NucleosilTM 5SA column in the silver ion form with a gradient of acetone into dichloromethane-dichlorethane (1:1, v/v). Abbreviations: S, saturated; M, monoenoic; D, dienoic; T, trienoic fatty acyl residues. (From Christie, W.W. (1988) J. Chromatogr., 454, 273. With permission.)
When nonpolar phases are used in capillary columns, triacylglycerols are separated according to molecular weight only and there is no useful resolution by degree of unsaturation, although some partial separations may be seen on occasion. Components varying in carbon number from 44 to 56 can be clearly resolved. Some remarkably effective separations of triacylglycerols, both by chainlength and degree of unsaturation, have been achieved on capillary columns coated with more polar (or “polarizable”) silicone phases containing a high proportion of phenyl groups, but the recovery of polyunsaturated species is poor. The linearity and robustness of the flame-ionization detector is a considerable advantage, but the key to a wider acceptance of high-temperature GC of triacylglycerols on capillary columns is the precision that can be attained in quantification. Good injection technique and clean samples can eliminate some of the losses, but there is little that can be done to prevent thermal degradation entirely. By careful optimization of the operating conditions, reproducible if not quantitative recoveries can be attained and careful calibration can give meaningful results, at least for the more saturated species as in cocoa butter (Buchgraber et al., 2003; 2004). Supercritical fluid chromatography is a hybrid between GC and HPLC, using much of the instrumentation of the former while the mobile phase is in effect a liquefied gas, commonly carbon dioxide. Much effort is being expended in developing the instrumentation and applications, but only the published separations of molecular species of triacylglycerols appear to offer real competition to 447
6.4 Chromatographic analysis of lipids
reference standards or on the basis of relative retention times calculated from the additive contributions of the fatty acyl chains. While good separations are possible on nonpolar columns, such as SE-54TM, much better results are obtained in separations of unsaturated species on polar phases, such as RTx-2330TM or SP-2380TM, with which unsaturated species follow saturated. Indeed, molecular species with positional isomers of double bonds in the fatty acyl residues were found to be separable. As an example, a separation of the diacylglycerol moieties derived from the phosphatidylcholine of rat liver is illustrated in Figure 6.11 (Myher and Kuksis, 1989). As the molecular weights of diacylglycerol derivatives are much lower than those of triacylglycerols, there appear to be no difficulties with quantification and uncorrected detector responses of TMS ether derivatives should give comparable results to those obtained by other means, provided that all the instrumental parameters have been properly optimized. Excellent resolution of molecular species of diacylglycerols derived from phospholipids has been achieved by means of reversed-phase HPLC. Although intact phospholipids can also be separated by this technique, the diacylglycerol approach means that there is no requirement for inorganic ions in the mobile phase, while UVabsorbing or fluorescent derivatives of diacylglycerols may be employed, simplifying detection and quantification. In addition, complementary chromatographic techniques can more easily be brought to bear for the further resolution of fractions and individual components can be collected for further analysis (reviewed by Bell, 1996; Christie, 2003). The principle of the separation is the same as that described above (Section 6.4.6.2) for triacylglycerols,
to TMS ether, acetates, or UV-absorbing derivatives, for example, as these can be stored for long periods in an inert atmosphere at low temperatures without coming to harm. Ceramides can be prepared from sphingomyelin by digestion with phospholipase C in this manner. In addition, diacylglycerols can be released from the glycosyldiacylglycerols of plants by a series of simple chemical reactions, and then analysed by the methods below (Heinz, 1996; Heinze et al., 1984). The procedure involves oxidation with periodic acid in methanol; after decomposition of excess reagent with ethyleneglycol, the oxidized lipid is extracted and treated with 1,1-dimethylhydrazine to release the diacylglycerols by beta-elimination. With many phospholipid classes (and phosphatidylethanolamine and phosphatidylserine especially), it is advisable to separate alkenylacyl, alkylacyl and diacyl forms (the “diradyl” forms) as the acetate or other derivatives before proceeding to more detailed analysis. The compounds migrate in the order stated in the form of various derivatives and can be adequately resolved on silica gel TLC layers, with a first development to half way up the plate with hexane–diethyl ether (1:1, v/v) followed by a full development in the same direction in toluene (Renkonen and Luukkonen, 1976) (HP-TLC plates may be preferred nowadays). Nakagawa and Horrocks (1983) achieved similar separations with HPLC. The greater resolving power of capillary columns has been put to good use for the resolution of diacylglycerol species derived from phospholipids, first with apolar and, more recently, with polarizable stationary phases in columns of fused silica. GC-MS will greatly simplify identification of peaks, but in the absence of this facility excellent results can be obtained by comparison with relative retention times of 13
2425
RTx2330
Detector response
11
4
5 8910
33
12
39 20 1519 21 22
26 2930
36 37 38 40
44
Time
FIGURE 6.11 GC separation of TMS ethers of diacylglycerols derived from the phosphatidylcholine of rat liver on a fused silica capillary column coated with cross-bonded RTx-2330 (Restek, Port Matilda, PA) with hydrogen as carrier gas at an isothermal temperature of 250°C. Peak identifications: 4, 16:0-16:0; 5, 16:0-16:1; 8, 16:0-17:0; 10, 16:0-18:0; 11, 16:0-18:1(n-9); 12, 16:0-18:1(n-7); 13, 16:0-18:2; 19, 17:0-18:2; 20, 18:0-18:1(n-9); 21, 18:0-18:1(n-7); 22, 18:1(n-9)-18:1(n-9); 24, 18:0-18:2; 25, 16:0-20:4, 16:0-20:3, 18:1(n9)-18:2; 26, 18:1(n-7)-18:2; 29, 18:2-18:2, 16:0-20:5; 30, 17:0-20:4; 33, 18:0-20:4(n-6), 18:0-23(n-6); 36, 18:1(n-9)-20:4; 38, 18:0-20:5; 39, 16:0-22:5, 16:0-22:6; 44, 18:0-22:5, 18:0-22:6. (From Myher, J.J. and Kuksis, A. (1989) J. Chromatogr., 471, 187. With permission.)
448
Analysis
following preparation by the chemical procedure described above; the p-anisoyl derivatives were then prepared and fractionated on an ODS column with a gradient of 30 to 0% water in acetonitrile as the mobile phase with specific detection and quantification at 250 nm (Kesselmeier and Heinz, 1985). HPLC in the reversed-phase mode is virtually the only technique that needs be considered for the separation of molecular species of intact polar lipids nowadays. Again, columns of the ODS type with a high carbon loading are used mainly, though octyl phases offer alternative properties for some applications. Mobile phases are based on methanol or acetonitrile with additional modifier solvents (usually including water) and it is essential to add an ionic species to the mobile phase to counter unwanted interactions of the ionic head group with the stationary phase that would otherwise cause peak broadening. Patton, Fasulo and Robins published a seminal paper on the subject of separations of phospholipid molecular species in 1982. They utilized a column of the ODS type with methanol–water–acetonitrile as the components of the mobile phase with choline chloride added as the ion suppressant. The separation can be considered as bimodal, with those molecular species containing a 16:0 fatty acyl group eluting before those containing 18:0. As with triacylglycerols, the position of the acyl group within the molecule has no effect on separation by reversed-phase chromatography. All of the major components in rat liver phosphatidylcholine are clearly resolved, although inevitably some minor fractions co-elute with the main ones. With plant lipids, the molecular species differ markedly from those found in the corresponding lipids of animal tissues and they tend to be much simpler in composition. Essentially the same elution conditions as those of Patton et al. have now been used in a number of laboratories to effect similar separations of phosphatidylcholine species, always a good recommendation. However, there may be times when an alternative elution scheme is required, either to suit the availability of particular HPLC equipment, especially the detection system, or to change the selectivity of the separation to permit resolution of specific molecular species. Phosphatidylethanolamine and phosphatidylinositol from rat liver were separated into molecular species and quantified under exactly the same conditions employed for phosphatidylcholine and, by increasing the polarity of the mobile phase, phosphatidylserine could be fractionated. Indeed, fractions identical in composition were obtained, although the relative proportions were rather different, as expected. In order to use evaporative light-scattering detection, molecular species of phosphatidylethanolamine from animal tissues were separated by reversed phase HPLC on an ODS phase with methanol–acetonitrile (7:3, v/v) containing 5 µM triethylamine as the mobile phase (Brouwers et al., 1999). After careful calibration, components were quantified directly from the detector responses of an
except that only two long-chain fatty acids need be considered, i.e., the separation is in ascending order of the total number of carbon atoms in the aliphatic chains of the two fatty acids, with a double bond in any of the fatty acids reducing the retention time to roughly that of a component with two fewer carbon atoms. Components considered formerly as critical pairs, e.g., dipalmitoyl and palmitoyloleoyl species, are easily separated on modern columns. However, only in exceptional circumstances is it possible to separate isomers in which the positions of the fatty acids on the glycerol moiety differ. The choice of stationary and mobile phases tends to be the same as for triacylglycerols and many of the practical points raised in the section dealing with these lipids are appropriate here also. ODS columns with a high carbon content are usually favoured, but octyl phases are a little different in selectivity and this can be useful in some applications. Various types of diacylglycerol derivatives have been employed. Acetates are the simplest, but UV detection at low wavelengths must then be used, leading to difficulties in quantification. However, this approach may be useful if it is intended that fractions be subjected to identification directly by MS, or indirectly following collection by GC or GC-MS. Various UV-absorbing and fluorescent derivatives have been utilized and there appears to be no information on which of these is best in chromatographic terms. Benzoates are perhaps the simplest UV-absorbing derivatives, but 3,5-dinitrobenzoates offer greater sensitivity (to picomole levels). Anthroyl and naphthoyl are the most common fluorescent derivatives to have been used and these can increase the sensitivity of detection by a further three orders of magnitude. With dinitrobenzoyl derivatives of diacylglycerols, impressive separations of molecular species derived from phosphatidylcholine from a variety of tissues have been achieved. For example, with a column of UltrasphereTM ODS and elution with acetonitrile–isopropanol (4:1 by volume) as the mobile phase, 29 distinct fractions were detected, identified, and quantified (Takamura et al., 1986). When methanol–isopropanol (19:1, v/v) was the mobile phase, only 17 fractions were seen, but some components were resolved that were not separated by the previous system. By collecting fractions containing more than one component from the first eluent, a more comprehensive analysis can be obtained by rerunning with the second eluent. In this way, up to 36 distinct molecular species can be obtained from each lipid class. This method used in sequence with silver ion chromatography might prove to be an even more thorough approach. These methods have the advantages that isocratic elution and, therefore, simpler HPLC equipment can be used, while the detector response is directly proportional to the molar amount of each species. Molecular species of plant galactolipids (mono- and digalactosyldiacylglycerols and sulfoquinovosyldiacylglycerol) have been analysed in the form of diacylglycerols, 449
6.4 Chromatographic analysis of lipids
of ODS groups relative to the inert support (20% by weight) and eluted with methanol UV detection at 210 nm (Hirabayashi et al., 1986). A large number of components were resolved and most were provisionally identified by their relative retention times as containing homologous series of saturated hydroxy and nonhydroxy fatty acids linked to sphingosine. These predictions were confirmed for the most abundant components by fast atom bombardment MS, a technique that is extremely useful for the structural identification of glycolipids. A further series of peaks was identified as containing phytosphingosine linked to nonhydroxy saturated fatty acids. Gangliosides are more difficult to analyse by HPLC because of their high polarity, but good results can be achieved by a suitable choice of mobile phase (Sonnino et al., 1985). As with the separation of classes of glycolipids (see Section 6.4.5.6), benzoylation has proved useful for the analysis of molecular species compositions as it enables sensitive and quantitative determination by UV detectors. In general, much better separations are achieved than is possible with the underivatized glycolipids, and with less polar mobile phases. Among the first applications of derivatization of sphingolipids for separations of molecular species was one of perbenzoylated glucosylceramide from a patient with Gaucher’s disease. This was fractionated on a column of an ODS phase, with methanol as the mobile phase and UV detection at 254 nm (Suzuki et al., 1976). Nine fractions were obtained, differing in their fatty acid compositions, but apparently not in the compositions of the long-chain bases. Perbenzoylation has also proved a useful approach to resolution of molecular species of highly polar monosialogangliosides and glycolipid sulfates.
evaporative light-scattering detector, although peaks could also be collected for analysis (see previous section). The galactosyldiacylglycerols of higher plants and algae in the intact form have been successfully separated into molecular species by reversed-phase HPLC by several research groups, and again this seems to be the only technique that need be considered nowadays. The methodology has been reviewed by Heinz (1996). For example, mono- and digalactosyldiacylglycerols from leaves were fractionated by this technique with a gradient of 50 to 0% water in acetonitrile as the mobile phase, and with spectrophotometric detection at 200 nm (Kesselmeier and Heinz, 1985). Response factors were determined for each molecular species so that the technique could be used quantitatively. The procedure developed by Patton et al. (1982) described above has also been adapted for galactolipids by Demandre et al. (1985). In an alternative approach, digalactosyldiacylglycerols from oats were peracetylated with acetic anhydride and pyridine to reduce their polarity prior to analysis by reversed-phase HPLC (Bergqvist and Holmback, 2000), although, perhaps surprisingly, benzoylation does not appear to have been attempted. 6.4.6.4
Molecular species of sphingolipids
In sphingomyelin and the glycosphingolipids, the molecular species contain different combinations of a fatty acid moiety with a long-chain base and reversedphase HPLC is the preferred method of analysis. In only a few studies have attempts been made to separate intact sphingomyelins into simpler molecular fractions and, as they are relatively saturated lipids, UV detection at low wavelengths is not ideal. An HPLC system that makes use of an evaporative light-scattering detector may afford better results, and Olsson et al. (1992) have described suitable conditions. Molecular species of sphingomyelin can also be analysed in the form of ceramides after removal of the polar head group (as described in the previous section). It is also possible to generate ceramides from glycosphingolipids by means of ceramide glycanase, which removes the carbohydrate moiety, but I am not aware of it being used for analysis of molecular species of the ceramide as opposed to the oligosaccharide portion of such lipids (though a chemical procedure has been used) (see Christie, 2003). Ceramides are less polar and more volatile than the parent sphingolipids and can be analysed by both high-temperature GC and HPLC methods, which can be linked to mass spectrometry. Methods for the analysis of ceramides by HPLC (Cremesti and Fischl, 2000; Gaudin et al., 2002) and GC (Raith et al., 2000) were recently reviewed. There are advantages in being able to resolve underivatized glycosphingolipid species, as these can then be subjected to immunogenicity tests. Monoglycosylceramides obtained from the intestines of Japanese quail were subjected to HPLC on phase comprising a very high proportion
6.4.7
Mass spectrometry of lipids
Mass spectrometry is proving to be an extremely important tool for the analysis of intact lipids and this has produced a revolution in the approach to the problem for those with access to the required equipment. Indeed, this is frequently described in the literature as a new branch of science — “lipidomics” — in which the kinetics of lipid metabolism and the interactions of lipids with cellular proteins is studied via a detailed quantification of a cell’s lipidome (lipid classes and individual molecular species), with the hope of obtaining new insights into health and disease. The topic has been the subject of reviews (Murphy et al., 2001; Han and Gross, 2003; 2005) and a monograph (Byrdwell, 2005). Fast atom bombardment (FAB), atmospheric pressure-chemical ionization (APCI), and matrix-assisted laser desorption/ ionization linked to time-of-flight (MALDI-TOF) mass spectrometry can all have great value for specific lipid classes but electrospray ionization (ESI) is the most sensitive and appears to have the greatest general utility. The technique is developing so rapidly that these notes may soon be out of date. 450
Analysis
HPLC-MS with APCI has been widely used for the analysis of triacylglycerols. The technique results in relatively limited fragmentation, but with distinct diagnostic ions related to the nature of the acyl moieties. In particular, there is usually a protonated molecular ion ([M+H]+) and a series of ions derived from diacylglycerol fragments ([M-RCOO] + or [DG] + ). The degree of unsaturation is the main factor that governs the intensities of the various ions and, in particular, the proportions of the diacylglycerol ions relative to the protonated molecular ion; the higher the degree of unsaturation the more intense is the protonated molecular ion, which becomes the base peak with four or more double bonds in the molecule. As an example, Figure 6.12 illustrates the mass spectrum of the triacylglycerol, 20:4 – 18:3 – 16:0. The protonated molecular ion is at m/z = 879, and ions at m/z = 574, 600, and 622 represent the diacylglycerols formed by loss of arachidonic, linolenic, and palmitic acids, respectively. Note that the spectra do not give information on the positional distributions of fatty acids on the glycerol moiety. The technique is especially compatible with reversed-phase chromatography, but it has also been used in conjunction with silver ion chromatography. The large differences in the response factors for each molecular species (depending on the degree of unsaturation of the acyl moieties) and instrumental factors must place a limit on the accuracy of quantification of triacylglycerols by APCI-MS. However, by careful calibration and the use of internal standards, acceptable results can be obtained, especially in routine analyses of similar samples. While other ionization techniques continue to make a significant contribution, the development of electrospray ionization mass spectrometry (ESI-MS) has greatly simplified the task of analysis of triacylglycerols, phospholipids and sphingolipids. As this ionization technique does not cause extensive fragmentation, it has proved to be highly sensitive, accurate and reproducible without a need for complicated chromatographic steps, i.e., samples can be introduced to the instrument via a probe rather than via an HPLC system. With ESI, a high electric field is applied to nebulize a solution as it emerges from a needle. The field imparts a charge to the droplets and this builds up as the solvent evaporates until a point is reached where ions must be ejected from the surface. The ions are focused by an electronic lens and pass via a skimmer into the mass spectrometer. The correct pH and solvent composition are important, and lithium, sodium, or ammonium salts may be introduced to form [M+Li]+, [M+Na]+, or [M+NH4]+ ions, respectively. Both positive and negative ion spectra can be obtained and each may give useful structural information with specific lipid molecules. The technique works best at low concentrations of lipid in the infusion solution, e.g., fmol to pmol of total lipid per µl, when lipid-lipid interactions and ion suppression are not relevant and the response is highly linear.
Electron-impact ionization is still of value for GC-MS, when the lipids are sufficiently volatile for this technique to be suitable. For example, phospholipids can be analysed after conversion to simple diacylglycerol derivatives (see Section 6.4.6.3) and GC-MS of intact triacylglycerols is feasible but with some technical difficulty. In a typical electron-impact mass spectrum of a triacylglycerol, there is a rather small molecular ion, followed by a unique peak for an ester at [M-18]+ (or loss of water). More importantly, there are intense ions that are characteristic of the various fatty acyl residues and these fall into two classes, i.e., those containing two acyl residues and those with only one. In the more important first class, there is an ion equivalent to the loss of an acyloxy group, i.e., [M-RCOO]+, together with a related ion, but minus a further hydrogen atom. For example, the loss of a palmitoyl acyloxy group (equivalent to 255 amu) gives ions equivalent to [M-255]+ and [M-256]+, while the loss of an oleoyl moiety gives ions at [M-281]+ and [M-282]+. The relative intensity within each pair is dependent on whether the ion fragment contains an unsaturated residue (when the smaller ion is more intense). The other important class of diagnostic ions that contains the individual fatty acid moieties is of the form RCO+, though if the fatty acid group is unsaturated an additional hydrogen atom is lost. Thus, an oleoyl moiety produces an ion at m/z = 264, while that from palmitate is at m/z = 239. Related ions with an additional 74 amu corresponding to the glycerol backbone, are found at m/z = 339 and 313, respectively; in this instance, the presence of a double bond has no effect. The electron-impact mass spectra of acetate derivatives of diacylglycerols are equivalent to those of triacylglycerols, except that one of the acyl moieties is an acetyl residue. Published mass spectral data are sparse, but it is apparent that the molecular ion tends to be rather small or nonexistent, although an ion representing loss of water ([M-18]+) can usually be seen. Loss of the acetyl group gives ions at [M-59]+ or [M-60]+, depending on the degree of unsaturation of the residual ion, and this is probably the best marker for determining the molecular weight. In addition, ions are seen for the loss of one or both of the other acyloxy moieties. TMS ether derivatives of 1,2-diacylglycerols tend to give much better spectra. Although the molecular ion is rarely seen, ions equivalent to [M-15]+ (loss of a methyl group) and [M-90]+ (loss of the TMS ether moiety) can be used to determine the molecular weight and, hence, the total carbon number and degree of unsaturation of the acyl moieties. An important diagnostic ion results from the loss of an acyloxy residue, i.e., [M−RCOO]+ or [M-RCOOH]+ if the residual acyl group is unsaturated. Other useful ions are equivalent to [RCO + 74]+, [RCO + 90]+, [RCO]+, and [RCO − 1]+. Characteristic ions at m/z = 145 and 129 contain the TMS group and parts of the glycerol backbone. 451
6.4 Chromatographic analysis of lipids
879
100
(M + H)+
20:4 18:3 = 878 16:0 80
20:4 18:3 = 622 304
20:4
60
= 600
%
16:0 18:3 = 574 16:0
40
278 574 256
20 600 622 200
400
600 m/z
800
1000
FIGURE 6.12 Mass spectrum (LC-MS with APCI detection) of the triacylglycerol, 20:4-18:3-16:0. (From Christie, W.W. (2003) Lipid Analysis, 3rd ed., Oily Press, Bridgwater, U.K. With permission.)
For example, Han and Gross (2003) favour the use of lithium ions in solution and positive ion mass spectrometry to analyse neutral or zwitterionic lipids, such as phosphatidylcholine, sphingomyelin, triacylglycerols and cerebrosides; the same technique with negative ion mass spectrometry is favoured for free acids, ceramides and phosphatidylethanolamine. Anionic lipids, such as cardiolipin, phosphatidylglycerol, phosphatidylinositols, phosphatidylserine, phosphatidic acid and sulfatides are analysed by negative-ion ESI-MS in the diluted chloroform extracts without added ions, and they are quantified by comparisons of the individual ion peak intensity with an internal standard. After identification of the main molecular species, tandem mass spectrometry is performed to estimate the molar ratios of any isobaric molecular forms. Variations in the methodology permit distinction of those fatty acids esterified to the sn-1 and sn-2 positions. Similar principles apply to the analysis of the complex glycosphingolipids, although in a recent review (Merrill et al., 2005) HPLC, in both the normal- and reverse-phase modes, linked to tandem mass spectrometry is recommended over the direct inlet approach as this enables separation of isometric and isobaric species (such as glucosylceramides and galactosylceramides). The main problem now lies in evaluating the vast amounts of data resulting from such measurements, especially when tandem mass spectrometry techniques are applied simultaneously. This has required the development and application of complex computational algorithms, most of which are not published in detail.
References Ackman, R.G. (1972) The analysis of fatty acids and related materials by gas-liquid chromatography. Prog. Chem. Fats Other Lipids, 12, 165–284. Ackman, R.G. and Burgher, R.D. (1965) Cod liver oil fatty acids as secondary reference standards in GLC of polyunsaturated fatty acids of animal origin — analysis of a dermal oil of Atlantic leatherback turtle. J. Am. Oil Chem. Soc., 42, 38–42. Andersson, B.A. (1978) Mass spectrometry of fatty acid pyrrolidides. Prog. Chem. Fats Other Lipids, 16, 279–308. Apon, J.M.B. and Nicolaides, N. (1975) The determination of the position isomers of the methyl branched FA esters by capillary gas chromatography/mass spectrometry. J Chromatogr. Sci., 13, 467–473. Bell, M.V. (1997) Separations of molecular species of phospholipids by high-performance liquid chromatography. In Advances in Lipid Methodology — Four, Ed., W.W. Christie, Oily Press, Dundee, U.K., pp. 45–82. Bergqvist, M. and Holmback, J. (2000) Nuclear magnetic resonance spectroscopy and reversed-phase high-performance liquid chromatography of peracetylated digalactosyldiacylglycerols. J. Am. Oil Chem. Soc., 77, 757–761. Bitman, J. et al. (1984) Comparison of the phospholipid composition of breast milk from mothers of term and preterm infants. Am. J. Clin. Nutr., 40, 1103–1119. Bligh, E.G. and Dyer, W.J. (1959) A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol., 37, 911–917. Blomberg, L.G. et al. (1998) Characterization of lipids by supercritical fluid chromatography and supercritical fluid extraction. In Lipid Analysis in Oils and Fats, Ed., R.J. Hamilton, Blackie, London, pp. 34–58.
452
Analysis
Bodennec, J. et al. (2000) A procedure for fractionation of sphingolipid classes by solid-phase extraction on aminopropyl cartridges. J. Lipid Res., 41, 1524–1531. Brouwers, J.F.H.M. et al. (1999) Rapid separation and identification of phosphatidylethanolamine molecular species. J. Lipid Res., 40, 164–169. Buchgraber, M. et al. (2003) Capillary GLC: a robust method to characterise the triglyceride profile of cocoa butter — results of an intercomparison study. Eur. J. Lipid Sci. Technol., 105, 754–760. Buchgraber, M. et al. (2004) Triacylglycerol profiling by using chromatographic techniques. Eur. J. Lipid Sci. Technol., 106, 621–648. Byrdwell, W.C. Ed. (2005) Modern Methods for Lipid Analysis by Liquid Chromatography/Mass Spectrometry and Related Techniques, AOCS Press, Champaign, IL. Christie, W.W. (1985) Rapid separation and quantification of lipid classes by high performance liquid chromatography and mass (light-scattering) detection. J. Lipid Res., 26, 507–512. Christie, W.W. (1987) A stable silver-loaded column for the separation of lipids by high-performance liquid chromatography. J. High Resol. Chromatogr. Chromatogr. Commun., 10, 148–150. Christie, W.W. (1988) Separation of molecular species of triacylglycerols by HPLC with a silver ion column. J. Chromatogr., 454, 273–284. Christie, W.W. Ed. (1992) Detectors for high-performance liquid chromatography of lipids with special reference to evaporative light-scattering detection. In Advances in Lipid Methodology — One, Oily Press, Bridgwater, U.K., pp. 239–271. Christie, W.W. Ed. (1993a) Preparation of ester derivatives of fatty acids for chromatographic analysis. In Advances in Lipid Methodology — Two, Oily Press, Bridgwater, U.K., pp. 69–111. Christie, W.W. Ed. (1993b) Preparation of lipid extracts from tissues. In Advances in Lipid Methodology — Two, Oily Press, Bridgwater, U.K., pp. 195–213. Christie, W.W. Ed. (1996) Separation of phospholipid classes by high-performance liquid chromatography. In Advances in Lipid Methodology —Three, Oily Press, Bridgwater, U.K., pp. 77–107. Christie, W.W. Ed. (1997) Structural analysis of fatty acids. In Advances in Lipid Methodology — Four, Oily Press, Bridgwater, U.K., p. 119–169. Christie, W.W. (1998) Gas chromatography mass spectrometry methods for structural analysis of fatty acids. Lipids, 33, 343–353. Christie, W.W. (2003) Lipid Analysis, 3rd ed., Oily Press, Bridgwater, U.K. Christie, W.W. et al. (1998) New procedures for rapid screening of leaf lipid components from Arabidopsis. Phytochem. Anal., 9, 53–57. Christopherson, S.W. and Glass, R.L. (1969) Preparation of milk fat methyl esters by alcoholysis in an essentially nonalcoholic solution. J. Dairy Sci., 52, 1289–1290. Craske, J.D. (1993) Separation of instrumental and chemical errors in the analysis of oils by GC — a collaborative evaluation. J. Am. Oil Chem. Soc., 70, 325–334.
Cremesti, A.E. and Fischl, A.S. (2000) Current methods for the identification and quantitation of ceramides: an overview. Lipids, 35, 937–945. Demandre, C. et al. (1985) Analysis of molecular species of plant polar lipids by high-performance and gas-liquid chromatography. Phytochemistry, 24, 481–485. Deschamps, F.S. et al. (2001) Assessment of the retention properties of poly(vinyl alcohol) stationary phase for lipid class profiling in liquid chromatography. J. Chromatogr. A, 928, 127. Fellenberg, A.J. et al. (1987) Simple mass spectrometric differentiation of the n–3, n–6 and n–9 series of methylene interrupted polyenoic acids. Biomed. Environ. Mass Spectrom., 14, 127–137. Foglia, T.A. and Jones, K.C. (1997) Quantitation of neutral lipid mixtures using high-performance liquid chromatography with light scattering detection. J. Liq. Chromatogr. Rel. Technol., 20, 1829–1838. Folch, J. et al. (1957) A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem., 226, 497–509. Gaudin, K. et al. (2002) Structure-retention diagrams of ceramides established for their identification. J.Chromatogr.A, 973, 69–83. Geurts van Kessel, W.S.M. et al. (1977) High performance liquid chromatographic separation and direct ultraviolet detection of phospholipids. Biochim. Biophys. Acta, 486, 524–530. Goiffon, J-P. et al. (1981a) High-pressure liquid-chromatography application for fatty triglycerides analysis. 2. Retention parameters of triglycerides. Rev. Franc. Corps Gras, 28, 199–207. Goiffon, J-P. et al. (1981b) High-pressure liquid-chromatography application for fatty triglycerides analysis. 1. A study of the best operating conditions for soyabean oil. Rev. Franc. Corps Gras, 28, 167–170. Greenspan, M.D. and Schroeder, E.A. (1982) Separation and detection of neutral lipids and free fatty acids in a liver extract by high-performance liquid chromatography. Anal. Biochem., 127, 441–448. Gunstone, F.D. (1992) High resolution 1H and 13C NMR. In Lipid Analysis. A Practical Approach, Eds., R.J. Hamilton and S. Hamilton, IRL Press, Oxford, pp. 243–262. Gunstone, F.D. (1993) High resolution 13C NMR spectroscopy of lipids. In Advances in Lipid Methodology — Two, Ed., W.W. Christie, Oily Press, Bridgwater, U.K., pp. 1–68. Hakomori, S.-I. (1983) Chemistry of glycosphingolipids. In Sphingolipid Biochemistry, Handbook of Lipid Research, Vol. 3, Eds., J.N. Kanfer and S-I. Hakomori, Plenum Press, New York, pp. 1–165. Han, X.L. and Gross, R.W. (2003) Global analyses of cellular lipidomes directly from crude extracts of biological samples by ESI mass spectrometry: a bridge to lipidomics. J. Lipid Res., 44, 1071–1079. Han, X.L. and Gross, R.W. (2005) Shotgun lipidomics: electrospray ionization mass spectrometric analysis and quantitation of cellular lipidomes directly from crude extracts of biological samples. Mass Spectrom. Rev., 24, 367–412. Harvey, D.J. (1992) Mass spectrometry of picolinyl and other nitrogen-containing derivatives of fatty acids. In Advances in Lipid Methodology — One, Ed., W.W. Christie, Oily Press, Dundee, U.K., pp. 19–80.
453
6.4 Chromatographic analysis of lipids
Heinz, E. (1996) Plant glycolipids: structure, isolation and analysis. In Advances in Lipid Methodology — Three, Ed., W.W. Christie, Oily Press, Dundee, U.K., pp. 211–332. Heinze, F.J. et al. (1984) Release of diacylglycerol moieties from various glycosyl diacylglycerols. Anal. Biochem., 139, 126–133. Herslöf, B. (1981) HPLC of triglycerides using UV detection. J. High Resolut. Chromatogr. Chromatogr. Commun, 4, 471–473. Herslöf, B. et al. (1990) Characterization of lecithins and phospholipids by HPLC with light scattering detection. In Phospholipids, Eds., I. Hanin and G. Pepeu, Plenum Press, New York, pp. 295–298. Hirabayashi, Y. et al. (1986) An improved method for the separation of molecular species of cerebrosides. Lipids, 21, 710–714. Holman, R.T. and Rahm, J.J. (1971) Analysis and characterization of polyunsaturated fatty acids. Prog. Chem. Fats Other Lipids, 9, 15–90. Jungalwala, F.B. et al. (1976) High-performance liquid-chromatography of phosphatidylcholine and sphingomyelin with detection in the range of 200 nm. Biochem. J., 155, 55–60. Kaufmann, P. (1995) Multivariate optimisation strategy for liquid chromatography. II. Exploring and finding optimal conditions in the search area of the multidimensional solvent space. Chemometr. Intell. Lab. Systems, 27, 105–114. Kesselmeier, J. and Heinz, E. (1985) Separation and quantitation of molecular species from plant lipids by high-performance liquid chromatography. Anal. Biochem., 144, 319–328. Kramer, J.K.G. and Hulan, H.W. (1978) A comparison of procedures to determine free fatty acids in rat heart. J. Lipid Res., 19, 103–106. Kuksis, A. et al. (1993) Quantitation of plasma lipids by gas-liquid chromatography on high-temperature polarizable capillary columns. J. Lipid Res., 34, 1029–1038. Kuksis, A. et al. (1990) Gas-liquid chromatographic profiling of plasma lipids using high-temperature polarizable capillary columns. J. Chromatogr., 500, 427–441. Laakso, P. and Manninen, P. (1998) Mass spectrometric techniques in the analysis of triacylglycerols. In Spectral Properties of Lipids, Eds., R.J. Hamilton and J. Cast, Sheffield Academic Press, Sheffield, U.K., pp. 141–190. McCluer, R.H. et al. (1989) High-performance liquid chromatography of membrane lipids, glycosphingolipids and phospholipids. Meth. Enzymol., 172, 538–575. Merrill, A.H. et al. (2005) Sphingolipidomics: high-throughput, structure-specific, and quantitative analysis of sphingolipids by liquid chromatography tandem mass spectrometry. Methods, 36, 207–224. Miwa, T.K. et al. (1960) Gas chromatographic characterization of fatty acids — identification constants for monocarboxylic and dicarboxylic methyl esters. Anal. Chem., 32, 1739–1742. Mourey, T.H. and Oppenheimer, L.E. (1984) Principles of operation of an evaporative light-scattering detector for liquid chromatography. Anal. Chem., 56, 2427–2434. Murphy, R.C. et al. (2001) Analysis of nonvolatile lipids by mass spectrometry. Chem. Rev., 101, 479–526.
Myher, J.J. and Kuksis, A. (1989) Relative gas-liquid chromatography retention factors of trimethylsilyl ethers of diradylglycerols on polar capillary columns. J. Chromatogr., 471, 187–204. Nakagawa, Y. and Horrocks, L.A. (1983) Separation of alkenylacyl, alkylacyl and diacyl analogues and their molecular species by high performance liquid chromatography. J. Lipid Res., 24, 1268–1275. Nikolova-Damyanova, B. (1992) Silver ion chromatography and lipids. In Advances in Lipid Methodology — One, Ed., W.W. Christie, Oily Press, Bridgwater, U.K., pp. 181–237. Nikolova-Damyanova, B. (1997) Reversed-phase highperformance liquid chromatography: general principles and application to the analysis of fatty acids and triacylglycerols. In Advances in Lipid Methodology — Four, Ed., W.W. Christie, Oily Press, Bridgwater, U.K., pp. 193–251. Olsson, N.U. et al. (1992) HPLC separation of molecular species of intact sphingomyelin, utilizing multivariate design and optimization. Chromatographia, 34, 529–534. Palmer, D.N. et al. (1984) Separation of some neutral lipids by normal-phase high-performance liquid chomatography on a cyanopropyl column: ubiquinone, dolichol and cholesterol levels in sheep liver. Anal. Biochem., 140, 315–319. Patton, G.M. et al. (1982) Separation of phospholipids and individual molecular species of phospholipids by high performance liquid chromatography. J. Lipid Res. 23, 190–196. Perrin, J-L. and Naudet, M. (1985) Identification and quantification of triglycerides from natural fats by HPLC. Rev. Franc. Corps Gras, 32, 301–303. Raith, K. et al. (2000) Ceramide analysis utilizing gas chromatography-mass spectrometry. J. Chromatogr. A, 876, 229–233. Ratnayake, W.M.N. (1998) Analysis of trans fatty acids. In Trans Fatty Acids in Human Nutrition, Eds., J.L. Sébédio and W.W. Christie, Oily Press, Bridgwater, U.K., pp. 115–161. Renkonen, O. and Luukkonen, A. (1976) Thin-layer chromatography analysis of complex lipids. In Lipid Chromatographic Analysis, 2nd ed., Vol.1, Ed., G.V. Marinetti, Marcel Dekker, New York, pp. 1–58. Ritchie, A.S. and Jee, M.H. (1985) High-performance liquid chromatographic technique for the separation of lipid classes. J. Chromatogr. A, 329, 273–280. Rouser, G. et al. (1976) Column chromatographic and associated procedures for separation and determination of phosphatides and glycolipids. In Lipid Chromatographic Analysis, 2nd ed., Vol. 3, Ed., G.V. Marinetti, Marcel Dekker, New York, pp. 713–776. Schnaar, R.L. and Needham, L.K. (1994) Thin-layer chromatography of glycosphingolipids. Meth. Enzymol., 230, 371–389. Sébédio, J.L. (1995) Classical chemical techniques for fatty acid analysis. In New Trends in Lipid and Lipoprotein Analyses, Eds., J.-L. Sebedio and E.G. Perkins, AOCS Press, Champaign, IL, pp. 277–289. Skipski, V.P. (1975) Thin-layer chromatography of neutral glycosphingolipids. Meth. Enzymol., 35, 396–425. Sonnino, S. et al. (1985) Preparation of GM1 ganglioside molecular species having homogeneous fatty acid and long chain base moieties. J. Lipid Res., 26, 248–257. Spitzer, V. (1997) Structure analysis of fatty acids by gas chromatography — low resolution electron impact mass
454
Analysis
Table 6.6 contains information on the most common functional groups in lipid chemistry and their 1H-NMR chemical shifts and is taken largely from the second edition of The Lipid Handbook (Gunstone et al., 1994). To illustrate these values, the 1H-NMR spectrum of methyl linoleate (Figure 6.13) is given as an example since it contains most of the common functional groups in a fatty acid chain. The assignments in Figure 6.13 are shown at the bottom of this page. The abbreviations in parentheses refer to the splitting of the signals, s = singlet, d = doublet, t = triplet, q4 = quartet, q5 = quintet, m = multiplet (usually broad; br). The H-3 qu in t e t i s o f t e n n o t w e l l r e s o l v e d . From the information in Table 6.6 and Figure 6.13, the spectra of other types of fatty acids can be deduced. For example, a saturated fatty acid chain (such as in methyl palmitate or methyl stearate) will show only five signals, namely those designated with the letters a, b, c, d, and h. In a monounsaturated fatty acid chain, such as in methyl oleate, the signals designated a, b, c, d, e, f, and h will be observed since the bis-allylic protons designated g are missing. In free fatty acids, of course, the singlet caused by the methyl ester protons disappears and is replaced by the OH of the acid. The nature of the ester does not affect the position of the signals of the protons in0 the fatty acid chain; however, signals of the ester moiety may overlap those of the fatty acid chain. For example, in ethyl esters the signal of the ethyl CH3 group is contained within the large signal of the CH2 groups of the fatty acid chain, in propyl esters the signal of H-3 in the fatty acid chain is overlapped by the second methylene group in the propyl moiety, and in butyl esters the signal of one methylene group is contained in the large methylene signal of the fatty acid methylene units. The exact position of the signals caused by the protons in the fatty acid chain depends on the proximity of other functional groups. For example, in monounsaturated fatty acids, the signal of the olefinic protons “migrates” downfield and is split when the double bond is at C2 (Gunstone and Ismail,1967; Frost and Gunstone, 1975). In the case of proximity to C1, shifts are moved downfield for acids vs. methyl esters and trans vs. cis double bond configuration. For example, in CCl4 the signals of 2(Z)-octadecenoic acid were observed at 6.285 (C3) and 5.735 (C2) ppm, while the corresponding ester showed peaks at 6.145 (C3) and 5.680 (C2) ppm, and 2(E)-octadecenoic acid displayed shifts at 7.01 (C3) and 5.75 (C2) ppm (Frost and Gunstone, 1975). With the signals toward the middle of the chain, shift values around 5.30 ppm were observed in CCl4 (Gunstone and Ismail, 1967; Frost and Gunstone, 1975). Coupling constants are a facile method to distinguish cis and trans double bonds in 1H-NMR, especially for monounsaturated compounds.
spectrometry of their 4,4-dimethyloxazoline derivatives — a review. Prog. Lipid Res., 35, 387–408. Suzuki, A. et al. (1976) Separation of molecular species of glucosyl ceramide by high-performance liquid chromatography of their benzoyl derivatives. J. Biochem. (Tokyo), 80, 1181–1183. Takamura, H. et al. (1986) Quantitative analysis of polyenoic phospholipid molecular species by high-performance liquid chromatography. Lipids, 21, 356–361. van Echten-Deckert, G. (2000) Sphingolipid extraction and analysis by thin-layer chromatography. Meth. Enzymol., 312, 64–79. Vitiello, F. and Zanetta, J-P. (1978) Thin-layer chromatography of phospholipids. J. Chromatogr. A, 166, 637–640. Wells, M.A. and Dittmer, J.C. (1965) A preparative method for the isolation of brain cerebroside, sulfatide and sphingomyelin. J. Chromatogr., 18, 503–511.
6.5
Nuclear Magnetic Resonance Spectroscopy
Nuclear magnetic resonance (NMR) spectroscopy is one of the most common analytical methods in chemistry and the related sciences due to the wealth of information available from the spectra. Accordingly, both 1H- and 13CNMR spectroscopy are routinely used in lipid chemistry. This section briefly discusses the basics of analyzing lipids by NMR but also touches on subjects such as quantification. The NMR phenomenon was discovered independently and simultaneously by Purcell and his associates at Harvard University and by Bloch and coworkers at Stanford University, for which they were jointly awarded the Nobel prize in physics in 1952. The theoretical background of NMR is described in numerous textbooks and websites and therefore will not be discussed here. A basic discussion is also available in the second edition of The Lipid Handbook (Gunstone et al., 1994). The website (www.lipidlibrary.co.uk) and the literature cited here (e.g., Gunstone, 1993) contain additional detailed information. For the purpose of obtaining NMR spectra, the lipid sample is usually dissolved in a deuterated solvent or one that does not contain hydrogen. Deuterated chloroform (CDCl3) is currently the most common solvent used for this purpose. In less recent literature, often spectra obtained in CCl4 or other solvents were reported. The data from less recent literature using the τ scale (TMS: τ = 10) have been recalculated to conform to the δ scale.
6.5.1
1
H-NMR spectroscopy
All values reported here use the δ scale, in which the signal of tetramethylsilane (TMS) has been assigned the value δ = 0. In practice, many spectra are now obtained using the known shift values of the solvent signal(s) as reference.
CH3–O–CO–CH2 – CH2 – (CH2)4 – CH2 – CH = CH – CH2 – CH = CH – CH2 – (CH2)3 – CH3 a b c d e f f g f f e d h (s) (t) (q5, br) (m) (q4) (m) (m) (t) (m) (m) (q4) (m) (t)
455
6.5 Nuclear Magnetic Resonance Spectroscopy
TABLE 6.6 Assignments of proton signals in the 1H-NMR spectra of fatty compounds; all values relative to tetramethylsilane (TMS) = 0 ppm Structure
Shift values
— CH2 — (cyclopropane)
(-0.3) – 0.6
a
— CH2 — (cyclopropene)
0.6 (singlet)
— CH3 (terminal methyl in alkyl chain)
0.85–0.90 (triplet)
— CH3 (branched, saturated isoprenoid)
0.85–0.90 (singlet or doublet)
— C(CH3)2 isopropyl methyl
1.2–1.3
(ω1)CH2, saturated alkyl chain
1.21.3
— CH2 — , acyl C-3, saturated chains
1.58
— CH2 — , acyl C-4 to C-(ω3). saturated chains; (ω2)CH2, saturated chain
1.2–1.3
RSH (sulfhydryl)
1.1–1.5b
RNH2 (amino)
1.1–1.5 (1.8)b
R2NH (imino)
0.4–1.6 (2.2)b
R3C-H (saturated)
1.4–1.7
— C=C — CH3 (allylic methyl)
1.6–1.9 (doublet)
— C=C—CH2 — (allylic methylene)
2.04 (doublet)
— C=C — CH2 — C=C — (diallylic methylene)
2.8 (triplet)
— CH2 — COOR, acyl C2
2.1–2.3 (triplet)
— CH2 — CO — (α-metbylene in ketone)
2.2–2.5
COOR-CH3 (methyl in acetoxy)
1.9–2.6 (singlet)
Ar — CH3
2.1–2.5
— C–C — H (terminal acetylene, nonconjugated )
2.5–2.7
— O — CH3 (methoxy ether, aliphatic)
3.3–3.8 (singlet)
— O — CH3 (methyl ester, aliphatic)
3.6–3.8 (singlet)
— CH — OH, sn-2 in glycerol
3.75 (multiplet)
— CH2 — OH, sn-1 or sn-3 in glycerol
3.6 (doublet)
— O — CH2 — (aliphatic saturated alcohol or ether)
3.4–3.7
— CH2 — O — CO — R (sn-1 or sn-3 esterified glycerol)
4.2–4.4
— CH — O — CO — R (sn-2 esterified glycerol)
5.1–5.2 (quintet)
— CH2 — O — R (sn-1- or sn-3-O-alkylglycerol)
3.5–3.6
— CH — O — R (sn-2-O-alkylated glycerol)
3.6–3.7
— CH — O — P (O-acylglycerol; sn-1 or sn-3)
3.9
— CH2 — O — P (O-alkylglycerol; sn-1 or sn-3)
3.9
— CH2 — O — P (choline or sulfocholine)
4.3–4.4
R — OH (hydroxyl proton)
3.0–5.3
R — CH = CH — O (vinyl ether)
5.8 (cis), 6.0 (trans)
C=CH2 (terminal vinyl, nonconjugated)
4.6–5.0
H — C = C — H (olefinic or cyclic; nonconjugated)
5.1–5.9 (multiplet)
— CH = CH — R, cis-∆2 in fatty acid chain
7.0 (), 5.8 ()
3
cis-∆ in fatty acid chain
5.6
cis-∆4 in fatty acid chain
5.5
5
cis-∆ in fatty acid chain
5.4
cis-∆6 in fatty acid chain
5.4
9
cis-∆ in fatty acid chain
5.3
cis-∆12 in fatty acid chain
5.3
(Continued)
456
Analysis
TABLE 6.6
Continued
Structure
Shift values
— (CH3)C=C — H (olefinic isoprenoid)
5.0–5.1
a
— C = CH2 (terminal vinyl, conjugated)
5.3–5.7 (6.2)
H — C = C — H (olefinic, conjugated; diene or triene)
5.8–6.5 (7.1)
— CO — N — H (amide NH and CO)
5.5–8.5
Ar - H (benzenoid)
7.3 - 8.5
RCHO (aldehyde proton aliphatic saturated)
(9.5) 9.7–9.8
aliphatic, α,β-unsaturated
9.5–9.7
R-COOH (carboxyl)
10.5–12.0
a
Values in parentheses apply to compounds that may absorb outside this range. Concentration-dependent; higher δ when diluted. Source: Gunstone, F.D., Harwood, J.L., and Padley, F.B. (1994) The Lipid Handbook, 2nd ed., Chapman & Hall, London. With permission.) b
a
d h
b e g f c
5.5
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0
Chemical shift (ppm)
FIGURE 6.13
1
H-NMR spectrum of methyl linoleate. (See text for explanation of letters.)
peaks are observed when the two double bonds have identical configuration and four signals when the configuration is not identical. Thus, in methyl 9(Z),11(Z)-octadecadienoate, a multiplet assigned to the “outer” positions
In contrast to methyl linoleate with methyleneinterrupted unsaturation, in conjugated linoleic acid (CLA) the peaks of the olefinic protons are split. The number of signals depends on the double bond configuration. Two 457
6.5 Nuclear Magnetic Resonance Spectroscopy
of the trans isomers. For methyl cis-2,3-epoxyoctadecanoate, the signals are observed at 3.30 and 2.99 ppm, while for the trans isomer a value of 2.99 ppm is observed (solvent CCl4; Gunstone and Jacobsberg, 1972). When the epoxy group is located towards the middle of the chain, the shift values in CCl4 are 2.68 to 2.70 ppm for the cisepoxy group and 2.45 to 2.49 for trans-epoxy. For the methyl ester of vernolic acid (cis-12,13-epoxy-9(Z)-octadecenoic acid), the protons at C-12 and C-13 give a signal at 2.82 ppm, while olefinic protons signal is split at 5.33 ppm (H-10) and 5.43 ppm (H-9) and the signal of the diastereotopic H-11 protons is split at 2.10 and 2.27 ppm for one proton each (Fürmeier and Metzger, 2003). When the epoxy group is located directly next to the double bond, the signal of the epoxy group is split; for example, in methyl trans-11,12-epoxy-9(Z)-octadecenoate, the shifts are observed as multiplets 3.31-3.35 ppm (H-11) and 2.79-2.83 ppm (H-12) and the olefinic protons at 5.64 to 5.73 ppm (H-9) and 5.00 to 5.08 ppm (H-10) (Lie Ken Jie et al., 2003). For methyl cis-9,10-epoxy-11(E)-octadecenoate, the values were 5.92 ppm (H-12), 5.25 to 5.34 ppm (m, H-11) for the olefinic protons and 3.36 ppm (q, H-10) as well as 3.01 to 3.09 (m, H-9) for the epoxy protons (Lie Ken Jie et al., 2003). In fatty compounds with oxo (keto) moieties, the protons α to this functionality are often observed around 2.50 ppm. Thus, in methyl 3-oxo-hexadecanoate, the signal of the protons at C2 was a singlet at 3.45 ppm (Lie Ken Jie and Lam, 1996). The protons to the oxo group at C4 were observed at 2.53 ppm. With an oxo group α to the terminal methyl group, the signal of the latter became a singlet downfield at 2.13 ppm in methyl 12-oxo-3(E)-truidecenoate (Lie Ken Jie and Lam, 1996). The protons to the oxo groups in fatty compounds with 1,2-dioxo groups gave triplets at 2.70 ppm (Knothe 2002). If an oxo group is close to a double bond the signal of the olefinic protons splits and is moved downfield considerably, for example, in 9(E)-11-oxooctadecenoic acid the olefinic signals were observed at 6.8 ppm (H-9) and 6.1 ppm (H-10) with the protons α to the oxo group again resonating at 2.5 ppm (Porter and Wujek, 1987). Similar results are available from other literature. For isostearic acid (16-methylheptadecanoic acid), the two equivalent terminal methyl groups show a doublet. If the branching is in the middle of the chain, for example, a lone methyl group, then the branched methyl group displays a doublet slightly upfield of the triplet generated by the terminal methyl group. If the methyl group is close to C1, such as at C3, the signal of the distereotopic protons at C2 is split. If the methyl grouup is close to a double bond, such as in 9-methyl-10(Z)octadecenoic acid, the signal of the olefinic protons is split again, as it is with other groups, in this case giving signals at 5.08 ppm (H-10) and 5.24 to 5.37 (broad multiplet of H-11) (Carballeira et al., 1999). In fatty acids with a cyclopropene unit in the chain, the protons of the ring methylene caused a singlet at 0.73 to
9 and 12 and doublet of doublets at 6.22 caused by the “inner” protons 10 and 11 is observed), while for the corresponding all-E isomer the shift values are 5.56 ppm (“outer” protons) and 5.96 ppm (“inner” protons) (Lie Ken Jie et al., 1997). The 9(Z),11(E) isomer displayed two multiplets at 5.32 and 5.56 ppm and two triplets at 5.82 ppm and 6.24 ppm attributable to H-9, H-12, H-10, and H-11, respectively (Lie Ken Jie et al., 1997). In fatty compounds with triple bonds, the signals of the protons adjacent to the triple bonds are those of major interest. These propargylic protons resonate at 2.05 to 2.3 ppm in monoacetylenic fatty acid chains with the downfield shifts occurring in case of greater proximity of the triple bond to C1 (Gunstone and Ismail, 1967; Frost and Gunstone, 1975). The signals of the propargylic protons can be split depending on the proximity of one of the triple bonds to C1 or the proximity of triple bonds to each other. When the triple bond is located at C3, the shift of C2 was observed at 3.205 ppm (acid) or 3.135 ppm (methyl ester) and for the triple bond at C4 the shifts were 2.445 ppm for the acid and 2.33 ppm for the ester (Gunstone and Ismail, 1967; Frost and Gunstone, 1975). With increasing distance from C1, the shift of C2 approximates its “usual” value of about 2.25 ppm. For bis-propargylic protons in compounds with two triple binds, shift values around 2.98 to 3.07 ppm (in CCl4) are observed (Frost and Gunstone, 1975). Results for chains with mixed unsaturation (mainly one triple and one double bond, etc.) were also reported (Frost and Gunstone, 1975). The introduction of other functional groups into the fatty acid chain induces changes to the NMR spectrum. Probably the most common additional moieties contain oxygen. For example, in saturated hydroxy fatty acids, an additional signal for the proton attached to the hydroxybearing carbon arises. This signal moves downfield, similar to the discussion above on olefinic protons when the OH group “moves” closer to either end of the molecule. In CCl4, the signal of the proton attached to the hydroxybearing carbon was observed at 3.93 ppm for 2-hydroxystearate, 3.80 ppm for 3-hydroxystearate, 3.48 ppm for 4-hydroxystearate, while for the stearates with the OH at C7-C14 this signal was observed at 3.38 to 3.40 ppm (Tulloch, 1966). For 15-, 16-, 17-, and 18-hydroxystearate, the signals were observed at 3.42, 3.34, 3.58, and 3.46 ppm, respectively (Tulloch, 1966). The introduction of the hydroxy group into unsaturated fatty acids leads to more additional peaks. For example, in methyl ricinoleate an additional signal (triplet) at around 2.1 ppm (in CDCl3), while the signal of the olefinic protons is split due to the proximity of the OH group into signals at about 5.45 (H-10) ppm and 5.60 ppm (H-9). In epoxy fatty acids, the shift values of the protons attached to the carbons carrying the epoxy group depend on the cis- or trans configuration of the epoxy group. The shifts of the cis epoxy isomers are downfield from those 458
Analysis
0.75 ppm (Gosalbo et al., 1993). Protons β to the ring overlapped with the C3 protons to give a signal at 1.57 ppm while a signal at 2.37 ppm was assigned to CH2 adjacent to the ring (Gosalbo et al., 1993). In fatty acids with a terminal cyclopentene ring, the olefinic protons in the ring gave a signal at 5.58 ppm and the methine proton of the carbon carrying the ring was observed at 2.60 ppm (Blaise et al., 1997). The signal of the methylene protons to the carbon in the cyclopentene ring gave two signals at 2.02 and 1.27 ppm, while the two protons on the carbon to the ring also gave a split signal at 1.25 and 1.35 ppm (Blaise et al., 1997). Signal integration yields information on the number of protons causing the signals. Thus, the number of double bonds, etc., can be determined in this fashion. In methyl linolenate, which contains an ω-3 double bond, the signal of the terminal methyl group at C18 is shifted slightly downfield and in mixtures can be integrated separately of the corresponding signal of other components. This forms the basis of some quantifications, such as determination of the fatty acid profile. As an example of a triacylglycerol, the 1H-NMR spectrum of tristearin is shown in Figure 6.14. The assignments in the figure are as follows: a
at around 3.95 ppm (C2 glycerol proton) with the two protons at C3 split at about 3.65 and 3.73 ppm.
6.5.2
HA
H – C – O – CO – R
HB – C – O – CO – R a
C-NMR spectroscopy
Similar to the discussion of 1H-NMR above, an unsaturated compound shall serve as starting point for the discussion. 13C-NMR data for linoleic acid have been given in the literature (Gunstone, 1993) and are shown at the bottom of this page. The 13C-NMR spectrum of the methyl ester of linoleic acid is shown in Figure 6.15. Major differences to the values for linoleic acid are the presence of the peak of the methyl ester moiety at 51.41 ppm and the upfield shift of the carbonyl carbon to 174.27 ppm. Generally, many effects observed in 1H-NMR are also found in 13C-NMR. For example, the methyl and methylene signals are upfield in the spectrum, while signals of olefinic carbons are farther downfield. The number and nature of double bonds affects the chemical shifts as does the proximity of multiple double bonds to each other and the presence of functional groups. In case of acylglycerols, it plays a role if the double bond is in the α or β chain with detailed data given by Lie Ken Jie and Lam (1995). Again, qualitatively the spectra of other fatty compounds can be reasonably deduced from the information given above for linoleic acid. For example, the spectrum of oleic acid (or methyl oleate) will not show the signal of the bis-allylic carbon and there will be only two olefinic carbon signals. The number of signals in the “methylene envelope” at around 29 to 30 ppm, however, will obviously increase. With the increasing number of closely spaced signals, correctly carrying out assignments of the signals in the methylene envelope to individual carbons becomes increasingly difficult and other more sophisticated experiments, such as those described briefly below, may not always help. Table 6.7 is a compilation of the chemical shifts of functional groups in 13C-NMR. While most authors report 13C-NMR data to two decimal places, these data can vary slightly from laboratory to laboratory and even when repeating the same experiment in a certain laboratory. Therefore, ranges are given in Table 6.7, similar to Table 6.6.
HB – C – O – CO – R
b
13
HA
R = CH2 – CH2 – (CH2)14 – CH3 c d e f
6.5.2.1
The 1H-NMR spectra of numerous triacylglycerols have been discussed in the literature (Lie Ken Jie and Lam, 1995). The 1H-NMR spectrum of the corresponding 1,3-diacylglycerol (1,3-distearin) also shows the resonance of the HA and HB protons at around 4.2 ppm, but in this case it almost overlaps the proton at the C2 carbon of the glycerol backbone. In the 1-monoacylglycerol, numerous peaks can be identified. The most upfield signals (doublet of doublets) are those of the two C1 glycerol protons at 4.18 and 4.25 ppm. The other protons give individual signals
Double bonds
Since there are numerous combinations of double bond positions and configurations possible in unsaturated fatty acids with more than one double bond and the relatively narrow range in which they are observed (with the exception of some functional groups inducing shift changes), these individual combinations are not listed separately in Table 6.7. The shifts of trans double bonds are slightly downfield from those of cis double bonds, for example, if a lone double bond is located towards the
HOOC – CH2 – CH2 – (CH2)4 – CH2 – CH = CH – CH2 – CH = CH – CH2 – CH2 – CH2 – CH2 – CH3 180.54 34.15 24.70 29.08-29.63 27.22 130.02 128.12 25.67 127.95 130.21 27.25
459
29.19
31.58
22.62 14.09
6.5 Nuclear Magnetic Resonance Spectroscopy
e
5.8
5.6
5.4
5.2
5.0
4.8
4.6
4.4
4.2
4.0
f c
5.5
FIGURE 6.14
1
d
a
b
5.0
4.5
4.0
3.5 3.0 2.5 Chemical shift (ppm)
2.0
1.5
1.0
0.5
0
H-NMR spectrum of tristearin. (See text for explanation of letters.)
CDCl3
192 184 176 168 160 152 144 136 128 120 112 104 96 88 80 72 64 56 48 40 32 24 16 Chemical shift (ppm)
FIGURE 6.15
13
C-NMR spectrum of methyl linoleate.
460
8
0
Analysis
TABLE 6.7
Assignments of carbon signals in the 13C-NMR spectra of fatty compounds (CDCl3)
Acids and Esters
Assignment
HOOC – (CH2)x
179–181
HOOC – CHCH3 – (CH2)x
183–184
CH3O – CO – ; other alkyl esters
172–175
CH3O – CO
51–52
Glycerol esters Triacylglycerols Glyc-O-CO-(CH2)x – CH3
173–173.2 (chain); 172.7-172.9 (chain)
Glyc-O-CO-(CH2)x – CH = CH - R
172.5–172.9 (chain); 172.1–172.5 (chain)
CH2 – O – CO – R | CH – O – CO – R | CH2 – O – CO – R
62–62.5 68.5–69 62–62.5
Diacylglycerols CH2 – O – CO – R | CH – O – CO – R | CH2OH CH2 – O – CO – R | CH2 | CH2 – O – CO – R
ca. 61.5–62 72–72.5 ca. 62 ca. 65 68–68.5 ca. 65
Monoacylglycerols CH2 – O – CO – R | CHOH | CH2OH CH2 OH | CH – O – CO – R | CH2OH
63–63.5 70-70.5 65.65.5 61.5–62 74.5–75 61.5–62
Methyl esters – (CH2)x – CH3 ; x > 1
13.5–14.5
Methylene in saturated chains (CH2)x
29–30
HOOC – CH2 – (CH2)x – or CH3O – CO – CH2 – (CH2)x –
34-35
HOOC – CH2 – CH2 – (CH2)x – or CH3 O – CO – CH2 – CH2 – (CH2)x –
23–26
– (CH2)x – CH2 – CH3
22–23
– (CH2)x – CH2 – CH2 – CH3
31.5–32.5
Double bonds – CH = CH –
125–135 (exceptions given below)
(CH2)x – CH = CH2
114–115
(CH2)x – CH = CH2
139–140
(CH2)x – CH2 – CH = CH2
33–35
(CH2)x – CH2 – C = C – CH2 – (CH2)y
27–28 (cis); 32–33 (trans)
HOOC – CH2 – C = C
33–34 (cis)
HOOC – CH2 – CH2 – C = C
22–23 (cis)
– CH=CH – CH2 – CH3
20–21(cis); 25–26 (trans)
(Continued)
461
6.5 Nuclear Magnetic Resonance Spectroscopy
TABLE 6.7
Continued Assignment
– CH=CH – CH2 – CH2 – CH3
29-30 (cis); 34–35 (trans)
– C = C – CH2 – CH = CH –
25–26 (all cis); 35–36 (all trans); 30–31 (cis, trans)
– CH = C = CH –
200–205
– CH = C = CH –
90–92
Triple bonds (CH2)x – C ≡ C – (CH2)y
79–81
(CH2)x – CH2 – C
18–19
≡ C – CH2 – (CH2)y – C ≡ C – CH2 – C ≡ C – – C ≡ C – CH2 – C ≡ C – – C ≡ C – CH2 – CH2 – C ≡ C – –C≡C–C≡C– – C ≡ C –C ≡ C – –CH2-C ≡ C –
74–75 9–10 19–20 77–78 65–66 18–20
Hydroxy groups (CH2)x – CHOH – (CH2)y
71–73
(CH2)x – CH2 – CHOH – CH2 – (CH2)y
37–38
(CH2)x – CH = CH – CHOH – (CH2)y
67–68 (cis); 73–74 (trans)
(CH2)x – CH = CH – CHOH – (CH2)y
131–134; (trans; CH adjacent to CHOH downfield)
(CH2)x – CH = CH – CH2 – CHOH – (CH2)y
133–134 (cis)
(CH2)x – CH = CH – CH2 – CHOH – (CH2)y
125–126 (cis)
(CH2)x – CH = CH – CH2 – CHOH – (CH2)y
35–36 (cis)
(CH2)x – CH = CH – CH2 – CH2 – CHOH – (CH2)y
130–131 (cis)
(CH2)x – CH = CH – CH2 – CH2 – CHOH – (CH2)y
129–130 (cis)
(CH2)x – CH = CH – CH2 – CH2 – CHOH – (CH2)y
23–24 (cis)
(CH2)x – CH = CH – CH2 – CH2 – CHOH – (CH2)y
37–38 (cis)
(CH2)x – CHOH – CHOH – (CH2)y
74–75 (erythro slightly > threo)
(CH2)x – CH2 – CHOH – CHOH – CH2 – (CH2)y
31–32 (erythro), 33–34 (threo)
(CH2)x – CHOH – CH2 – CH2 – CHOH – (CH2)y
32–33 (erythro), 33–34 (threo)
(CH2)x – CH2 – CHOH – CH2 – CH2 – CHOH – CH2 – (CH2)y
37–38 (threo slightly > erythro)
(CH2)x – CHOH – CH = CH – CHOH – (CH2)y
133–134; (trans; threo slightly > erythro)
Hydroperoxy compounds (CH2)x – CH2 – CH(OOH) – CH = CH – CH= CH – (CH2)y
86–88
(CH2)x – CH2 – CH(OOH) – CH = CH – CH = CH – (CH2)y
25–26
Oxo compounds (CH2)x – CO – (CH2)y
209–213
(CH2)x – CH2 – CO – (CH2)y
42–43
(CH2)x CH = CH – CH2 – CO – (CH2)y
41–42
(CH2)x CH = CH – CH2 – CO – (CH2)y
133–134
(CH2)x CH = CH – CH2 – CO – (CH2)y
121–122
(CH2)x – CO – CH2 – CH2 – CO – (CH2)y
36–37
(CH2)x – CO – CH = CH – CO – (CH2)y
203-204 cis; 200–201 trans
(CH2)x – CO – CH = CH – CO – (CH2)y
135–136 cis; 136–137 trans
Epoxy and furanoid compounds
O (CH2)x CH CH (CH2)y (CH2)x CH2
56–57 (cis); 58–59 (trans)
O CH CH CH2 (CH2)y
27–28 (cis); 31–32 (trans)
(Continued)
462
Analysis
Continued
TABLE 6.7
Assignment
O (CH2)x CH CH CH2 CH CH CH2 (CH2)y
25–26 (cis; cis-epoxy)
O (CH2)x CH CH CH2 CH CH CH2 (CH2)y (CH2)x CH CH CH2
132–133 (cis; cis-epoxy)
O CH CH CH2 (CH2)y
123–124 (cis; cis-epoxy)
O (CH2)2 C
104–105
C (CH2)
C C
154–155
Acetoxy (CH2)x – CO(COCH3) – (CH2)y
73–75
(CH2)x – CO(COCH3) – (CH2)y
170–171
(CH2)x – CO(COCH3) – (CH2)y
20–22
CH3 – O – CO – CO(COCH3) – (CH2)y
72–73
CH3 – O – CO – CO(COCH3) – (CH2)y
70–71
(CH2)x – CH2 – COCH3
64–65
Cyclic compounds (cyclopropene fatty acids) 7–8
CH2 (CH2)x
CH
CH
(CH2)y
109–110
CH2 (CH2)x
CH2
HOOC
(CH2)x
CH
CH2
CH
(CH2)y
25.5–26.5
CH2 CH2
CH
CH
CH2
(CH2)y; x = 2–3
107–109
CH2
(CH2)y; x = 2–3
110–112
CH2 HOOC
(CH2)x
CH2
CH
CH
Branched compounds (CH2)x – CH(CH3) – (CH2)y
19–20
(CH2)x – CH(CH3) – (CH2)y
32–33
(CH2)x – CH2 – CH(CH3) – CH2 – (CH2)y
3637
HOOC – CH(CH)3 – (CH2)x – or CH3 – O – CO – CH(CH)3 – (CH2)x
39–40
HOOC – CH(CH)3 – (CH2)x – or CH3 – O – CO – CH(CH)3 – (CH2)x
16.5–17.5
CH3 – O – CO – CH2 – CH(CH)3 – (CH2)x
30–31
CH3 – O – CO – CH2 – CH(CH)3 – (CH2)x
19–20
CH3 – O – CO – CH2 – CH(CH)3 – (CH2)x
41–42
CH3 – O – CO – CH2 – CH2 – CH(CH)3 – (CH2)x
31–32
CH3 – O – CO – CH2 – CH2 – CH(CH)3 – (CH2)x
19–20
case of signal overlap). The positions and configuration of double bonds may, in many cases, be better determined by the shifts of the allylic methylene carbons, although for lone double bonds the shifts of the olefinic carbons and their differences can be definitive, especially if the lone double bond approaches one end of the molecule. Extensive reports on the 13C-NMR spectra of unsaturated fatty acids and esters are Gunstone et al. (1976, 1977). Other extensive reports exist on fatty acids
middle of a long chain, the values for cis are about 129.9 ppm and trans 130.4 ppm. Similar to 1H-NMR spectroscopy, when the double bond approaches one end of the molecule, the signals of the olefinic protons become more strongly separated. Such differences are usually evaluated in terms of additive increments, but rational functions can be used (Knothe and Nelsen 1998). In practice, the number of double bonds can be determined by the number of signals and intensity (in 463
6.5 Nuclear Magnetic Resonance Spectroscopy
literature for those not discussed here. However, the application and possible results of some methods will be briefly outlined here, but without discussion of their physical background. DEPT (distortionless enhancement of polarization transfer) is an experiment that is useful for determining the number of protons attached to a carbon atom. In the usual DEPT experiment (DEPT-135; the number referring to the pulse angle of the experiment), the 13C signals appear either as positive or negative peaks or disappear. Positive peaks correlate with carbons carrying an odd number of protons, in other words methine or methyl carbons. Negative peaks correlate with carbons carrying an even number of protons, in other words methylene carbons. Carbons without an attached hydrogen do not give a visible peak. This experiment, therefore, is useful for distinguishing carbons in areas of the spectrum where there can be significant overlap of the type of signalcausing carbons. C1 carbons in fatty acids and esters are easily distinguishable because their signal disappears in comparison to the “normal” proton-decoupled 13C-NMR experiment. Figure 6.16 depicts the DEPT-135 spectrum of methyl linoleate. A comparison of this spectrum with the “normal” 13C-NMR spectrum in Figure 6.15 shows the differences between the two experiments as discussed above. In the probably less routinely performed DEPT-90, only the peaks of methine are determined, thus allowing to distinguish between these species and methyl, which both give positive peaks in DEPT-135. In DEPT-45, all protonated carbons give positive peaks. Therefore, DEPT-45 appears less useful because DEPT-135, the “routine” DEPT, has the same feature but distinguishes carbons based on odd or even numbers of protons attached. However, DEPT-45 is more sensitive than the “normal” 13C experiment, making it useful when less sample is available, but caution must be used in interpretation because of the missing peaks of nonprotonated carbons. Two-dimensional experiments provide a means of identifying nuclei that are mutually coupled. In the COSY (COrrelation SpectroscopY) experiment, homonuclearcoupled protons are identified. The 1H spectrum serves as x- and y-axis for a contour plot. The diagonal of the contour plot results from the cross peaks of an individual signal with itself. Besides the peaks in the diagonal, various cross peaks “scattered” across the contour plot, but symmetrical to the diagonal as mirror plane, signal coupling of specific protons. This is often very useful in identifying which kind of carbon a specific functional group is bound. For example, in a COSY experiment of a fatty acid, the signal of the terminal methyl group will show a contour outside the diagonal that can be correlated with the broad peak caused by the methylene protons, showing that the terminal group is connected to a CH2 group. Because of the low natural abundance of 13C (ca. 1.1%), correlation spectroscopy between carbons is of no significance. Instead, heteronuclear correlation (HETCOR)
with functional groups, for example, those with oxo (Tulloch, 1977), hydroxy and acetoxy groups (Tulloch, 1978) and epoxy groups (Gunstone and Schuler, 1975). Generally, the shifts of methylene carbons cover a greater range of ppm values and are very sensitive to the nature of the adjacent functional group(s). For this reason, they are dealt with more explicitly in Table 6.7, being arranged according to the functional group influencing the shift. Methylene carbons allylic to trans double bonds often show downfield shifts of 4 to 5 ppm compared to those methylenes allylic to cis double bonds. The fact that 13C-NMR spectra cover a wider ppm range (typically 0 to 200 ppm and even beyond) facilitates some aspects of evaluation compared to 1H-NMR. A major aspect is that some functional groups impart shift values in parts of the spectrum where they do not overlap with the peaks of other moieties. Salient examples in this respect are epoxy groups, which show peaks in the range of 53 to 60 ppm and oxo groups, which cause significant downfield shifts beyond 200 ppm, around 210 to 212 for an oxo group in the middle of a fatty acid chain (see also Table 6.7). To some extent, this effect can also be found in methylenes adjacent to a group. For example, when an oxo group is present, the adjacent methylenes are shifted downfield to about 42 ppm. The influence of other moieties with more upfield signals, however, can then shift the signal of these methylenes upfield. Free fatty acid and methyl esters can be distinguished in this fashion, too, with the signals of C1 in free fatty acids around 179 to 181 ppm and in alkyl esters around 172 to 175 ppm and the methyl carbon in methyl esters causing a peak at 51 to 52 ppm. Again, such shifts may be perturbed by the presence of other functional groups. Double and triple bonds can be recognized in this fashion, too. For triple bonds under the influence of other unsaturation or other groups, it may need to be considered that the signal of chloroform in 13C-NMR is close to their range. Diastereomers, such as those resulting from the presence of two hydroxy groups in the chain, can also be distinguished by 13C-NMR with some data given in Table 6.7. A full discussion of all possible changes to a spectrum for various combinations of functional groups is certainly beyond the scope of this section. Instead the reader is referred to Table 6.7 for some common combinations, including those not discussed in the text, and the literature cited in this chapter, which gives more details including numerous other functional group combinations and provides additional references.
6.5.3
Other NMR experiments
Besides acquiring the “normal” 1H- and 13C-NMR spectra, other NMR experiments can prove to be very helpful in assigning peaks and removing remaining ambiguities. Not all possible experiments in NMR are discussed, rather the reader is referred to the specialized 464
Analysis
192 184 176 168 160 152 144 136 128 120 112 104 96 88 80 72 64 56 48 40 32 24 16
8
0
Chemical shift (ppm)
FIGURE 6.16
DEPT-135 (13C) spectrum of methyl linoleate.
between 1H and 13C is now a common experiment yielding useful information. The common experiment is termed HMQC (heteronuclear multiple-quantum correlation). The experiment yields a contour map similar to the one described for COSY above, except that one axis is now the 13C spectrum instead of the 1H spectrum, and, of course, there is no diagonal of cross peaks. The contour plot indicates which 1H-NMR peaks correlate with which 13C peaks. An application is the following: in methyl linoleate shown above, several peaks are found in the region of about 24.5 to 27.27 ppm in 13C. Fortunately all of these peaks can be correlated with distinct peaks in the 1HNMR spectrum, enabling their assignment. One potential difficulty is, however, that peaks that are close together can often only be assigned to the type of carbon with this procedure, not necessarily the specific carbon. An example is the allylic carbons in linoleic acid with signals at 27.22 and 27.25 ppm. The exact assignment to a specific carbon is often the result of careful comparison with other known model compounds and/or of theoretical considerations. Another example is the terminal methyl group of longchain fatty acids usually exhibits its 13C peak at 14.00 to 14.10 ppm. In a saturated fatty acid chain with a methyl group in the middle of the chain, this methyl group will generate another signal separate from that of the terminal methyl group. The branched methyl group will generate a doublet instead of the triplet of the terminal methyl group and this doublet is slightly upfield of the terminal CH3 triplet. By heteronuclear correlation, a 13C peak at 19.7
ppm correlates with the doublet of the mid-chain methyl group. Two-dimensional (2D) methods are also useful for identifying diastereotopic protons and assigning the 13C signals of such species. Thus, the combination of “regular” spectra, DEPT and 2D experiments is a powerful tool for determining the structure of lipids and organic compounds in general.
6.5.4
31
P-NMR spectroscopy
Phosphorus NMR (31P-NMR) spectroscopy is of interest for the analysis of phospholipids, but can also be employed when quantifying lipid components by means of phosphorus-containing derivatives. An advantage of 31P-NMR is that phosphorus is monoisotopic, rendering it sensitive compared to other nuclei with less natural abundance. Signals in 31P-NMR spectroscopy are distributed over almost 700 ppm and sometimes beyond that, thus facilitating separation of signals based on different environments and subtle structural differences. However, most signals of phospholipids are found in the relatively narrow range of + 5 to − 1 ppm, with each phospholipid giving a characteristic signal. Triphenylphosphate (TPP) is the commonly used internal standard in quantitative 31P-NMR, an advantage being that its shift value of –17.8 ppm is beyond that of natural phospholipids, although artificial phospholipids, such as distearoylphosphatidylglycerol (DSPG) are also employed (Diehl 2002). 31P-NMR is a reference method of the 465
6.5 Nuclear Magnetic Resonance Spectroscopy
results correlate even better with GC (Miyake et al., 1998). When determining the fatty acid profile of oil, the integration values of the protons need to be divided by 3. Using the equations (Knothe and Kenar, 2004):
International Lecithin and Phospholipid Society and was preferred over HPLC methods (Diehl, 2002). Recent overviews of the applications of 31P-NMR to problems in phospholipid research were provided by Diehl (2001, 2002). For some earlier papers on the analysis of phospholipids by 31P-NMR, see Meneses and Glonck (1988) and London and Feigenson (1979). The applications of 31P-NMR as discussed in the articles by Diehl (2001, 2002) are summarized below. For quantification of phospholipids in a mixture, absolute quantities are obtained by adding an internal standard. This is of significance for the composition of lecithins and body fluids such as lung surfactant and human blood plasma. An example is the ratio of phosphatidylcholine to lysophosphatidylcholine in human plasma indicating the level of inflammation in rheumatoid arthritis (Fuchs et al., 2005). Determination of the fatty acid profile of phospholipids: 31P-NMR distinguishes double unsaturated and saturated/ unsaturated phospholipids. This determination takes advantage of signal splitting in ultra-high resolution 31P-NMR when phospholipids from bilayers and liposomes in aqueous solution, which yield broad signals. The splitting depends on the pH value of the solution with greatest splitting occurring at the pK of the investigated phospholipid. Thus, the composition of phosphatidic acid is 40% saturated/unsaturated and 60% saturated/unsaturated. For phospholipid structure determination, in combination with enzymatic reactions, 31P-NMR can be used to determine the chiral center of a racemic phospholipid. Also, signal splitting indicates diastereomeric phospholipids while lack of splitting indicates pure enantiomers. Besides analyzing phospholipids, 31P-NMR can be used for other purposes in lipid chemistry. For example, phosphorus-containing derivatives used for quantification purposes may exhibit signals in other spectral regions than phospholipids. Thus, the relevant peaks of mono- and diacylglycerols derivatized with 2-chloro-4,4,5,5-tetramethyldioxaphospholane are observed around 145 to 150 ppm (Spyros and Dais, 2000; Schiller et al., 2002). 31PNMR can be used accordingly to quantify the mono- and diacylglycerols in olive oils by phosphorylizing the free hydroxy groups with 2-chloro-4,4,5,5-tetramethyldioxaphospholane and integrating the appropriate peaks (Spyros and Dais, 2000). The method may be applicable to other oil constituents with labile protons. Thus, thermally stressed vegetable oils can be analyzed with 31P-NMR (Schiller et al., 2002).
6.5.5
AC18:3 = Iexper,methyl,C18:3/(Iexper,methyl,C18:3 + Iexper,methyl,rest) AC18:2 = 0.5 (Iexper,bisallylic - 4AC18:3) AC18:1 = (Iexper,allylic/4) - AC18:2 – AC18:3 in which A indicates the amounts of the subscripted fatty acids and I indicates the experimentally determined integration values of the terminal methyl, bisallylic and allylic protons, the fatty acid composition of a vegetable oil can be determined. Other related work is Miyake et al. (1998a). Docosahexaenoic acid and ω-3 fatty acids in fish oils can be determined by 1H NMR in good agreement with GC analyses (Igarashi et al., 2000). Other authors also reported on the determination of ω-3 fatty acids in fish oils (Aursand et al., 1993; Sacchi et al., 1993). Related, although less specific, parameters, such as the iodine value, which reflects the total amount of unsaturated fatty acids in an oil or fat, can also be determined by 1H NMR (Miyake et al., 1998b). The key to this determination is that the signal of the terminal methyl group in ω-3 fatty acids is shifted slightly downfield from that of fatty acids without this double bond. The triacylglycerols in whole vegetable seeds were quantitatively determined using a magic angle spinning technique (Wollenberg, 1991). Similar kinds of analyses can be carried out with 13CNMR. While regiospecific analysis (determination of the α or β positions of the fatty acids attached to the glycerol backbone of triacylglycerols) was carried out by 1H-NMR in presence of shift reagents (Frost et al., 1975), extensive 13C-NMR reports exist in this matter, although the two α fatty acid chains cannot be distinguished. The shifts of the carbonyl carbons are employed for this purpose. Semiquantitative analysis of mixtures is possible in this fashion. Some literature is Wollenberg (1990), Gunstone (1991b), van Calsteren et al. (1996), and Vlahov (1998). The levels of ∆5 or ∆6 unsaturated fatty acids (petroselinic and linoleic) (Gunstone 1991a) or γ-linolenic acid (Gunstone, 1990a) were determined with reasonable accuracy in a variety of less common oils by 13C-NMR as were polyene esters (Gunstone, 1990b) individual acids in fish oils (Gunstone, 1991c). Intact soybean seeds can be analyzed for their fatty acid composition by 13C-NMR (Yoshida et al., 1989). A semiquantitative determination of the composition of hydrogenated fats was achieved using the shift differences of allylic signals as discussed above (Gunstone, 1993b). Related quantitative aspects of 31P-NMR are discussed above in the section on 31P-NMR. Identification/verification of vegetable oils: While oils or fats may be deliberately mixed for specific reasons, the adulteration of high-value oils with oils of
Quantification and applications of NMR
Fatty Acid Profile and Related Issues. This method utilizes integration values in 1H-NMR applied to equations for determining the amounts of the unsaturated fatty acids. The results agree well with GC conducted as a control, although reportedly 13C NMR 466
Analysis
strong singlet peak of the resulting methyl esters is useful in quantification. Solid fat content (SFC): While the other methods discussed here usually rely on high-resolution, research-grade NMR instruments, a benchtop, low-resolution pulsed NMR instrument can be used for determining the SFC of an oil or fat. The method determines the amount of solid triacylglycerols in the oil or fat at different temperatures, with only the pulsed NMR signal of the liquid fat being measured (van Duynhoven et al., 1999). The lowtemperature signal is proportional to the total liquid at 60°C. The method is used for quality control purposes in hydrogenation, blending, and interesterification. The SFC is used to assess properties of food products, such as hardness and mouth feel. Crystallization mechanisms of fat blends can be studied by kinetic SFC measurements. The NMR-based SFC method is considered to be more accurate than the older dilatometric method giving the solid fat index (SFI) as shown by an interlaboratory collaborative study. The results of SFC cannot be directly compared to the dilatometric SFI. The dilatometric procedure is also more labor-intensive and cumbersome than SFC determination by NMR. Nonfatty (extraneous) materials: nonfatty or extraneous materials in fatty compounds can also be determined by 1H NMR. An example is the quantitation of other lipidic materials, such as sterols, in vegetable oils (Sacco et al., 2000). The signal of a methyl group at C-18 of sterols is reportedly especially useful for this purpose. Another example is for blends of biodiesel with petroleum-derived diesel fuel. 1H NMR may be applied to verifying the blend level of biodiesel with the petroleum-derived fuel (Knothe, 2001b), in which case again the strong peak of the methyl ester moiety is useful.
lesser value is a problem of economic and commercial significance when the adulterated oil is marketed as the pure, high-value oil. This constitutes a major problem for olive oil. Accordingly, numerous publications from major olive oil-producing Mediterranean countries are concerned with identifying lower-value oils, such as hazelnut oil, used for adulterating olive oil. The adulteration problem is complicated by the fact that the lower-value oils usually have fatty acid profiles similar to olive oil. Among the methods used for analyzing potentially adulterated olive oil is NMR, both 1H and 13 C. For example, NMR was utilized in a study applying multivariate statistical methods to certain peaks of olive oil diluted with hazelnut or sunflower oil (Fauhl et al., 2000). Besides analyzing olive oil for diluents, 1H NMR, together with analytical data, can be used in assessing the variety and geographical origin of the oil (Sacco et al., 2000). Even when not quantitating components, the peak differences in NMR can be used to distinguish vegetable oils by visual inspection of the spectra (Guillén and Ruiz, 2003). An example of the application of 13C NMR is the detection of soybean oil in olive oil with the DEPT procedure (Vlahov, 1997). The concentration dependence of the triacylglycerols in olive oil can also be used for its analysis (Mannina et al., 2000). Overviews of methods for assessing the quality and adulteration status of olive oil have been published (Sacchi et al., 1997; Mannina and Segre, 2002). Monitoring of oxidation: the oxidation of vegetable oils or their derivatives is an important quality problem and can lead to further deterioration of the oil. Especially, unsaturated fatty acids with bis-allylic methylene groups are susceptible to oxidation. In 1H-NMR-based oxidation studies, primary oxidation products, such as hydroperoxides, and secondary oxidation products, such as aldehydes, were detected (Guillén and Ruiz, 2001; Guillén and Ruiz, 2005). Oxidation of ethyl docosahexaenoate was also evaluated by 1H-NMR and compared with traditional methods (Falch et al., 2004). 1H NMR is especially useful for such studies since the samples do not require treatment, which could cause changes to the samples themselves. 2D techniques also utilizing 13C NMR have also been reported for studying the oxidation products in autoxidized linoleoyl/linolenoylglycerols (Silwood and Grootveld, 1999). Reaction monitoring (mixtures of vegetable oils with other fatty compounds): NMR can be used for reaction monitoring, which actually constitutes analyzing mixtures of vegetable oils with other fatty compounds. An example is the transesterification reaction of a vegetable oil to its corresponding methyl esters (Knothe, 2001a), a reaction that is steadily gaining significance due to the increasing production of biodiesel to which both 1HNMR (Knothe, 2001a) and 13C-NMR (Dimmig et al., 1999) have been applied. In the case of 1H-NMR, the
References (Additional information is also available at the website: www.lipidlibrary.co.uk.) Aursand, M. et al. (1993) Quantitative high-resolution 13C and 1H nuclear magnetic resonance of δ3 fatty acids from white muscle of Atlantic salmon (Salmo salar). J. Am. Oil Chem. Soc. 70, 971–981. Blaise, P. et al. (1997) Identification of cyclopentenyl fatty acids by 1H and 13C nuclear magnetic resonance. J. Am. Oil Chem. Soc., 74, 727–730. Carballeira, N.M. et al. (1999) Synthesis of racemic 9-methyl-10-hexadecenoic acid. Chem. Phys. Lipids, 97, 87–91. Diehl, B.W.K. (2001) High resolution NMR spectroscopy, Eur. J. Lipid Sci. Technol., 103, 830–834. Diehl, B.W.K. (2002) 31P-NMR in phospholipid analysis. Lipid Technol., 14, 62–65. Dimmig, T. et al. (1999) 13C-NMR spectroscopic determination of the conversion and reaction kinetics of transesterification of triglycerides to methyl esters. Chem. Tech. (Leipzig), 51, 326–329.
467
6.5 Nuclear Magnetic Resonance Spectroscopy
Falch, E. et al. (2004) Correlation between 1H and traditional methods for determining lipid oxidation of ethyl docosahexaenoate. J. Am. Oil Chem. Soc., 81, 1105–1110. Fauhl, C. et al. (2000) 1H NMR as a tool for the analysis of mixtures of virgin olive oil with oils of different botanical origin. Magn. Reson. Chem., 38, 436–443. Frost, D.J. and Gunstone, F.D. (1975) The PMR analysis of nonconjugated alkenoic and alkynoic acids and esters. Chem. Phys. Lipids, 15, 53–85. Frost, D.J. et al. (1975) PMR analysis of unsaturated triglycerides using shift reagents. Chem. Phys. Lipids, 14, 189–192. Fuchs, B. et al. (2005) The phosphatidylcholine/lysophosphatidylcholine ratio in human plasma is an indicator of the severity of rheumatoid arthritis: investigations by 31P NMR and MALDI-TOF MS. Clin. Biochem., 38, 925–933. Fürmeier, S. and Metzger, J.O. (2003) Fat-derived aziridines and their n-substituted derivatives: biologically active compounds based on renewable raw materials. Eur. J. Org. Chem., 649–659. Gosalbo, L. et al. (1993) Synthesis of deuterated cyclopropene esters structurally related to palmitic and myristic acids. Lipids, 28, 1125–1130. Guillén, M.D. and Ruiz, A. (2001) High resolution 1H nuclear magnetic resonance in the study of edible oils and fats. Trends Food Sci. Technol., 12, 328–338. Guillén, M.D. and Ruiz, A. (2003) Edible oils: discrimination by 1H nuclear magnetic resonance. J. Sci. Food Agric., 83, 338–346. Guillén, M.D. and Ruiz, A. (2005) Monitoring the oxidation of unsaturated oils and formation of oxygenated aldehydes by proton NMR. Eur. J. Lipid Sci. Technol. 107, 36–47. Gunstone, F.D. (1991a) The 13C-NMR spectra of six oils containing petroselinic acid and of aquilegia oil and meadowfoam oil which contain δ5 acids. Chem. Phys. Lipids, 58, 159–167. Gunstone, F.D. (1991b) 13C-NMR studies of mono-, di- and triacylglycerols leading to qualitative and semiquantitative information about mixtures of these glycerol esters. Chem. Phys. Lipids, 58, 219–224. Gunstone, F.D. (1991c) High resolution NMR studies of fish oils. Chem. Phys. Lipids, 59, 83–89. Gunstone, F.D., (1993a) High resolution 13C-NMR spectroscopy of lipids. In Advances in Lipid Methodology — Two, Ed., W.W. Christie, Oily Press, Dundee, U.K., pp. 1–68. Gunstone, F.D. (1993b) The composition of hydrogenated fats by high-resolution 13C nuclear magnetic resonance spectroscopy. J. Am. Oil Chem. Soc., 70, 965. Gunstone, F.D. and Ismail, I.A. (1967) Fatty acids, Part 15. Nuclear magnetic resonance spectra of the cis octadecenoic acids and of some acetylenic acids. Chem. Phys. Lipids, 1, 337–340. Gunstone, F.D. and Jacobsberg, F.R. (1972) The preparation and properties of the complete series of methyl epoxyoctadecanoates. Chem. Phys. Lipids, 9, 26–34. Gunstone, F.D. and Schuler, H.R. (1975) Fatty acids, Part 46. PMR spectra of several epoxyoctadecenoic, epoxyoctadecynoic, and diepoxyoctadecanoic esters. Chem. Phys. Lipids, 15, 189–197. Gunstone, F.D. et al. (1976) Fatty acids. Part 48. 13C nuclear magnetic resonance of acetylenic fatty acids. Chem. Phys. Lipids, 17, 1–13.
Gunstone, F.D. et al. (1977) Fatty acids. Part 50. 13C nuclear magnetic resonance studies of olefinic fatty acids and esters. Chem. Phys. Lipids, 18, 115–129. Gunstone, F.D. et al. Eds. (1994) The Lipid Handbook, 2nd ed., Chapman & Hall, London. Igarashi, T. et al. (2000) Nondestructive quantitative determination of docosahexaenoic acid and n-3 fatty acids in fish oils by high-resolution 1H nuclear magnetic resonance spectroscopy. J. Am. Oil Chem. Soc., 77, 737–748. Knothe, G. (2001a) Monitoring a progressing transesterification reaction by fiber-optic near-infrared spectroscopy with correlation to 1H nuclear magnetic resonance spectroscopy. J. Am. Oil Chem. Soc., 77, 489–493. Knothe, G. (2001b) Determining the blend level of mixtures of biodiesel with conventional diesel fuel by fiber-optic nearinfrared spectroscopy and 1H nuclear magnetic resonance spectroscopy. J. Am. Oil Chem. Soc., 78, 1025–1028. Knothe, G. (2002) Synthesis and characterization of long chain 1,2 dioxo compounds. Chem. Phys. Lipids, 115, 85–91. Knothe, G. and Kenar, J.A. (2004) Determination of the fatty acid profile by 1H-NMR spectroscopy. Eur. J. Lipid Sci. Technol., 106, 88–96. Knothe, G. and Nelsen, T.C. (1998) Evaluation of the 13C NMR signals of saturated carbons in some long-chain compounds. J. Chem. Soc., Perkin Trans., 2, 2019–2026. Lie Ken Jie, M.S.F. and Lam, C.C. (1995a) 1H-Nuclear magnetic resonance spectroscopic studies of saturated, acetylenic and ethylene triacylglycerols. Chem. Phys. Lipids, 77, 155–171. Lie Ken Jie, M.S.F. and Lam, C.C. (1995b) 13C-nuclear magnetic resonance spectroscopic studies of triacylglycerols of type AAA and mixed triacylglycerols containing saturated, acetylenic and ethylenic acyl groups. Chem. Phys. Lipids, 78, 1–13. Lie Ken Jie, M.S.F. et al. (1997) Synthesis and nuclear magnetic resonance properties of all geometric isomers of conjugated linoleic acids. Lipids, 32, 1041–1044. Lie Ken Jie, M.S.F. et al. (2003) Epoxidation of a conjugated linoleic acid isomer. Eur. J. Lipid Sci. Technol., 105, 391–396. London, E. and Feigenson, G.W. (1979) Phosphorus NMR analysis of phospholipids in detergents. J. Lipid Res., 20, 408–412. Mannina, L. et al. (2000) Concentration dependence of 13C NMR spectra of triglycerides: implications for the NMR analysis of olive oils. Magn. Reson. Chem., 38, 886–890. Mannina, L. and Segre, A. (2002) High resolution nuclear magnetic resonance: from chemical structure to food authenticity. Grasas y Aceites, 53, 22–33. Meneses, P. and Glonek, T. (1988) High resolution 31P NMR of extracted phospholipids. J. Lipid Res., 29, 679–689. Miyake, Y. et al. (1998a) Determination of unsaturated fatty acid composition by high-resolution nuclear magnetic resonance spectroscopy. J. Am. Oil Chem. Soc., 75, 1091–1094. Miyake, Y. et al. (1998b) Rapid determination of iodine value by 1H nuclear magnetic resonance spectroscopy. J. Am. Oil Chem. Soc., 75, 15–19. Porter, N.A. and Wujek, J.S. (1987) Allylic hydroperoxide rearrangement: ß-scission or concerted pathway? J. Org. Chem., 52, 5085–5089.
468
Analysis
Sacchi, R. et al. (1993) Proton nuclear magnetic resonance rapid and structure-specific determination of T-3 polyunsaturated fatty acids in fish lipids. J. Am. Oil Chem. Soc., 70, 225–228. Sacchi, R. et al. (1997) 1H and 13C NMR of virgin olive oil: an overview. Magn. Reson. Chem., 35, S133–S145. Sacco, A. et al. (2000) Characterization of Italian olive oils based on analytical and nuclear magnetic resonance determinations. J. Am. Oil Chem. Soc., 77, 619–625. Schiller, J. et al. (2002) Effects of thermal stressing on saturated vegetable oils and isolated triacylglycerols — product analysis by MALDI-TOF mass spectrometry, NMR and IR spectroscopy. Eur. J. Lipid Sci. Technol., 104, 496–505. Silwood, C.J.L. and Grootveld, M. (1999) Application of highresolution, two-dimensional 1H and 13C nuclear magnetic resonance techniques to the characterization of lipid oxidation products in autoxidized linoleoyl / linolenoylglycerols. Lipids, 34, 741–756. Spyros A. and Dais, P. (2000) Application of 31P NMR spectroscopy in food analysis. 1. Quantitative determination of the mono- and diglyceride composition of olive oils. J. Agric. Food Chem., 48, 802–805. Tulloch, A.P. (1977) Deuterium isotope effects and assignment of 13C chemical shifts in spectra of methyl octadecanoate and the sixteen isomeric oxooctadecanoates. Can. J. Chem., 55, 1135–1142.
Tulloch, A.P. (1978) Carbon-13 NMR spectra of all the isomeric methyl hydroxy- and acetoxyoctadecanoates. Determination of chemical shifts by deuterium isotope effects. Org. Magn. Res., 11, 109–115. van Calsteren, M.R. et al. (1996) 13C NMR of triglycerides. Bull. Magn. Reson., 18, 175–177. van Duynhoven, J. et al. (1999) Solid fat content determination by NMR. INFORM, 10, 479–484. Vlahov, G. (1997) Quantitative 13C NMR method using the DEPT pulse sequence for the detection of olive oil adulteration with soybean oil. Magn. Reson. Chem., 35, S8–S12. Vlahov, G. (1998) Regiospecific analysis of natural mixtures of triglycerides using quantitative 13C nuclear magnetic resonance of acyl chain carbonyl carbons. Magn. Reson. Chem., 36, 359–362. Wollenberg, K.F. (1990) Quantitative high resolution 13C nuclear magnetic resonance of the olefinic and carbonyl carbons of edible vegetable oils. J. Am. Oil Chem. Soc., 67, 487–494. Wollenberg, K. (1991) Quantitative triacylglycerol analysis of whole vegetable seeds by 1H and 13C magic angle sample spinning NMR spectroscopy. J. Am. Oil Chem. Soc., 68, 391–400. Yoshida, M. et al. (1989) Non-destructive analysis of the oil composition of soybean seeds by natural abundance carbon-13 nuclear magnetic resonance spectroscopy. Agric. Biol. Chem., 53, 1395–1400.
469
7 PHYSICAL PROPERTIES: STRUCTURAL AND PHYSICAL CHARACTERISTICS
I. Foubert, K. Dewettinck, D. Van de Walle, A. J. Dijkstra and P. J. Quinn
7.1
Fat crystallisation is also important in fractionation processes, when products with different melting and physical properties are made (Hartel, 1992; Walstra, 2003). Section 7.2 of this chapter will deal with crystallisation, polymorphism, and melting. It will discuss nucleation, crystal growth, crystal size, and morphology, polymorphism, solid fat content (SFC), microstructural development, macroscopical and mechanical properties, and melting behaviour. In the section on polymorphism, triglycerides, diglycerides, and monoglycerides are treated separately. The next section of this chapter (7.3) deals with phase behaviour. Phase diagrams of mixtures of triglycerides, real fats, and mixtures of triglycerides with mono- and diglycerides are discussed. Section 7.4 of this chapter deals with lipid/water interaction. After explaining the concept of a liquid crystalline phase, the interaction of water with monoglycerides, phospholipids, triglycerides, diglycerides, and soaps will be discussed.
Introduction
Food products in which fat crystallisation is important include chocolate and confectionery coatings, dairy products, such as butter and cream, and shortenings, margarine, and spreads. An understanding of fat crystallisation processes in these products plays a critical role in determining overall product quality (Bailey, 1950; Hartel, 1992; Walstra, 2003). Fat crystallisation largely determines the following important properties of a food: • The consistency of high-fat products like butter, margarine, and chocolate during storage and handling. This determines the shape retention of the product during storage and its plasticity during handling (e.g., when spreading on a slice of bread or mixing it into a dough). • Eating properties: This may concern fracture properties and also meltdown, sticky mouthfeel and coolness. • Physical stability: This concerns formation and sedimentation of crystals in oil, oiling out (oil separates under the influence of gravity), coalescence of aqueous droplets in butter and margarine, and partial coalescence in some oil-in-water emulsions, such as cream when fat crystals that slightly protrude from the fat globules trigger coalescence of two adjacent globules. • Visual appearance: Examples are the appearance of bloom on chocolate, gloss of chocolate and margarine, and turbidity in oils.
References Bailey, A.E. (1950), Melting and Solidification of Fats, Interscience, New York. Hartel, R.W. (1992), Solid-liquid equilibrium in foods, in Physical Chemistry of Foods, Schwartzberg, H.G. and Hartel, R.W., Eds., Marcel Dekker, Inc., New York, 47–81. Walstra, P. (2003), Physical Chemistry of Foods, Marcel Dekker, New York.
471
7.2
Crystallisation and melting
7.2
and not every droplet will contain an impurity that can give rise to heterogeneous nucleation (e.g., if a bulk fat contains one catalytic impurity per mm3 and is divided into emulsion droplets of 100 µm3, only about 1 in 107 droplets will contain an impurity). An example of the crystallisation of an oil-in-water emulsion is the physical ripening of dairy cream prior to the churning process. Crystallisation in one droplet cannot induce crystallisation in another droplet, so that a material in a finely dispersed state can be undercooled to a far lower temperature than a bulk material, and its nucleation rate will generally be much slower. The nucleation is in this case mostly primary and homogeneous (Walstra, 2003). Secondary nucleation occurs in the presence of crystals of the material being crystallised, and, therefore, can only occur after primary homogeneous or heterogeneous nucleation and subsequent crystal growth (Kloek, 1998).
Crystallisation and melting
Crystals can be formed from the vapour phase, the melt, or a solution. No special theories have been developed to treat each of these different phase changes, but general theories have been adapted to apply to specific cases (Garside, 1987). Crystallisation comprises several steps. First of all, a sufficiently large thermodynamic driving force must be provided. Once this has been attained, nucleation (Section 7.2.1) can occur, whereby crystals are generated as a result of bringing growth units together so that a crystal lattice can be formed. From then on, proper crystal growth (Section 7.2.2) can proceed. Although it is convenient to treat nucleation and crystal growth as consecutive events, it should be noted that nucleation does not stop when growth starts. So, in a crystallising fat, nucleation and growth occur side by side (Hartel, 1992; Timms, 2003). The combination of nucleation and crystal growth leads to crystals with a certain size distribution and morphology (Section 7.2.3). Polymorphism (Section 7.2.4) complicates the crystallisation of lipids. Ultimately a certain solid:liquid ratio (Section 7.2.5) is obtained. Various models have been developed to describe the rate of the crystallisation process. They will be compared and models describing multistep crystallisation and nonisothermal crystallisation will also be discussed (Section 7.2.6). When the crystals are large enough, they form aggregates. This aggregation leads to the formation of a continuous network of crystals, which may alter in several different ways, often involving sintering (i.e., growing together) of adjacent crystals (Walstra et al., 2001) (Section 7.2.7). All processes mentioned above determine the macroscopic and mechanical properties of a fat (Section 7.2.8). The melting behaviour, which is reverse of crystallisation, is discussed in Section 7.2.9.
7.2.1 7.2.1.1
7.2.1.2
Classical nucleation theory for homogeneous nucleation
The driving force for crystallisation is the difference in chemical potential ∆µ [J mol-1] (or partial molar Gibbs free energy) between the liquid phase (melt or solution) and the solid phase. The larger this difference in chemical potential, the larger the driving force for crystallisation. To obtain nucleation in a solution, it is necessary to supersaturate this solution by arriving at a concentration C [m-3 or equivalent] that is higher than the solubility, in other words the concentration at saturation Cs [m-3 or equivalent]. The difference in chemical potential then equals:
∆µ = Rg × TK × ln(C C s )
(7.1)
where Rg denotes the ideal gas constant [8.314 J mol-1 K-1] and TK the absolute temperature [K]. The term ln(C/Cs) is called the supersaturation, ln(σ) [-], while C/Cs is the supersaturation ratio σ [-]. For ideal solutions, the concentration at saturation Cs is given by the Hildebrand equation:
Nucleation Types of nucleation
lnC s =
Three types of nucleation can be distinguished: primary nucleation, which can be either homogeneous or heterogeneous, and secondary nucleation. If nucleation is not catalysed by fat crystals or foreign solid surfaces, it is called primary, homogeneous nucleation. In this case, supercooling up to 30 K should be applied before crystallisation can occur (Kloek, 1998). If foreign surfaces (e.g., dust, container walls, molecules of different compounds) are present and catalyse the nucleation process, this gives rise to primary, heterogeneous nucleation occurring at lower levels of supersaturation (1 to 3 K) than homogeneous nucleation (Garside, 1987). Most natural fats contain enough catalytic impurities for heterogeneous nucleation to take place. However, when the fat is divided into emulsion droplets, these impurities are also divided among the various droplets
∆H m ⎛ 1 1⎞ − Rg ⎜⎝ TKm Tk ⎟⎠
(7.2)
where TKm denotes the absolute melting temperature [K] and ∆Hm stands for the molar enthalpy variation in the system during the transition [J mol-1]. It should be noted that the values of TKm and ∆Hm are polymorph dependent and, therefore, the values for the specific polymorph concerned should be used. At low values, the supersaturation is often approximated by expanding the logarithmic term in Equation 7.1 by a Taylor series and using only the first term. This results in:
ln( σ ) ≈
C − Cs ∆C = = σr = σ − 1 Cs Cs
(7.3)
The term ∆C/CS is called the relative supersaturation σr [-]. 472
Physical Properties: Structural and Physical Characteristics
at a maximum so that the value of the critical radius r* [m] can be calculated by determining the first derivative of the function with respect to r and setting this to zero:
For crystallisation from the melt to occur, supercooling should be induced and the resulting difference in chemical potential can then be written as:
∆µ = ∆H m
TKm − TK TKm
The difference TKm – TK = ∆T is called supercooling [K]. In this case the relative supersaturation can be written as:
σr =
∆T TKm
(7.5)
4 π r 3∆µ 3Vm
∆G* =
(7.6)
J = AJ × e
(7.8)
−
∆G * kB TK
(7.9)
where AJ represents the global kinetic coefficient [s-1 m-3] and kB denotes the Boltzmann constant [1.380 10-23 J K-1]. AJ can also be written as:
∆GS
AJ = ∆G∗
16 π Vm2 γ 3 3 ( ∆µ )2
16π The factor 3 results from the spherical shape attributed to the nucleus and can, in general, be replaced by a dimensionless shape factor. The nucleation rate J [s-1 m-3] at which new nuclei are formed is a problem of kinetics and is determined by the rate at which embryos surmount the maximum in the free energy curve. Supposing the embryos follow a Boltzmann distribution as a function of their free energy, the nucleation rate can be written as:
where γ is the surface free energy per unit of surface area [J m-2], and Vm is the molar volume [m3 mol-1]. Figure 7.1 illustrates how ∆G depends on the embryo radius and shows that a critical radius r* [m] exists for which ∆G is ∆G (J)
(7.7)
Since growth of the embryo only leads to a decrease of ∆G once the embryo radius is higher than the critical radius r* [m], the embryo is stable above the critical radius and unstable below it. Inserting Equation 7.7 into Equation 7.6 yields the critical free energy of activation for nucleation ∆G* [J]:
Nucleation occurs when growth units meet, giving rise to a distribution of clusters, called embryos. The Gibbs free energy change ∆G [J] for the formation of such an embryo is determined by a surface term ∆Gs [J] due to surface tension that is positive, and a volume term ∆Gv [J]. This Gibbs free energy change ∆G is also called the free activation energy. For a spherical embryo with radius r [m], ∆G equals:
∆G = 4 π r 2 γ −
2 γ Vm ∆µ
r* =
(7.4)
N m × kB × TK hP
(7.10)
∆G
with Nm the number of molecules per m3 [m-3] and hP Planck’s constant [6.626 10-34 J s]. If nucleation occurs in the melt or a high viscosity solution, the difficulty encountered by a molecule in crossing the interface between the liquid and solid phases must be taken into account. Therefore, an additional critical free energy of activation for volume diffusion ∆G*vd [J] is added to Equation 7.9, which then becomes:
r (m) r∗
∆Gv
J = AJ × e
FIGURE 7.1 Variation of the Gibbs free energy change ∆G [J] for three-dimensional nucleation as a function of nucleus size r [m]. (From Boistelle, R. (1988), in Crystallization and Polymorphism of Fats and Fatty Acids, Garti, N. and Sato, K., Eds., Marcel Dekker, New York, 189–276. With permission.)
−
∆G * kB TK
×e
−
* ∆Gvd kB TK
(7.11)
Equation 7.11 is generally known as the Turnbull– Fisher equation. When the temperature decreases, the volume diffusion term may become rate determining,
473
7.2
Crystallisation and melting
Then the critical radius r* for which ∆Ghet is at its maximum can be calculated; it is the same as for homogeneous nucleation (cf. Equation 7.7):
causing the nucleation rate to drop off at these lower temperatures. 7.2.1.3
Heterogeneous nucleation
In heterogeneous nucleation, the process of molecular orientation is enhanced by the presence of a foreign surface, e.g., a dust particle or a microscopic structure in the vessel wall, which orients molecules preferentially so that a crystal lattice is more easily formed. To derive the critical free energy of activation for heterogeneous nucleation, ∆G*het [J], it is convenient to consider a cap-shaped nucleus making a contact angle with the foreign surface (Figure 7.2). The value of ω depends on the way the foreign surface is wetted by the nucleus. The surface-free energies involved in this process are γ0 [J m-2] between foreign surface and mother liquor, γ1 [J m-2] between nucleus and mother liquor, and γ2 [J m-2] between nucleus and foreign surface. When the nucleus is in contact with the foreign surface, two excess energies have been expended in the creation of the new surfaces, whereas one has been gained in losing a fraction of the area of the foreign surface. The Gibbs free energy change for heterogeneous nucleation ∆Ghet [J], therefore, equals their sum:
∆Ghet = − ∆Gv + A1γ 1 + A2 γ 2 − A2 γ 0
r* =
1 ⎛1 3 ⎞ ∆G *het = ∆G * × ⎜ − cos ω + cos3 ω⎟ ⎝2 4 ⎠ 4
7.2.1.4
the free energy for heterogeneous nucleation, ∆Ghet can be written as:
(7.14)
+ 2 πr 2 (1 − cos ω ) γ 1 − π r 2 (1 − cos 2 ω ) γ 1 cos ω in which r [m] is the radius of the underlying sphere. γ1 A1 γ2 A2
ω
Secondary nucleation
According to most studies, the main source of secondary nuclei is the crystal surface. Secondary nuclei form whenever tiny crystallites, embryos, are removed from the surface and exceed the critical size. Two main causes are invoked to explain the removal of embryos or nuclei from the surface. Except for partial dissolution, which gives rise to spontaneous removal, removal is due either to fluid shearing forces or to collisions between a crystal and its neighbours or with the walls, the agitator or other parts of the crystalliser (Boistelle, 1988). In some cases, it has also been assumed that the source of the secondary nuclei lies in the supersaturated solution itself (Boistelle, 1988). Walstra (2003) even states that true secondary nucleation means that nuclei are formed in the vicinity of a crystal of the same phase and not on its surface. Secondary nucleation has been studied mainly under conditions that are typical of industrial crystallisers and involve considerable agitation. However, Walstra (1998) has shown that secondary nucleation can also happen under quiescent conditions. He considered droplets of milk fat of typically a few µm in diameter at a temperature of around 20°C. No precooling to a lower temperature was applied. Under these conditions, earlier studies (Walstra and Beresteyn, 1975) showed that the number of catalytic impurities would be of the order of one per 100 µm3. As this is about the size of the droplets, there should be zero, one or occasionally two crystals in a droplet. However,
(7.13)
πr 3∆µ ( 2 − 3 cos ω + cos3 ω ) 3Vm
(7.16)
Accordingly, ∆G*het equals the product of ∆G* (the critical free energy of activation for homogeneous nucleation) and a term depending on the magnitude of the contact angle ω. For ω = 90° (the contact is limited to a geometrical point), ∆G*het equals ∆G* / 2 and for ω tending towards zero (perfect wetting of the substrate), ∆G*het also tends towards zero, which means that in this borderline case, no free energy of activation at all is required for nucleation.
By calculating the volume of the cap and the surfaces and taking into account that the different surface free energies depend on each other according to:
∆Ghet = −
(7.15)
However, since the spherical cap contains fewer molecules than the full sphere, the energy barrier is lower for heterogeneous than for homogeneous nucleation. Inserting Equation 7.15 in Equation 7.14 yields the critical free energy of activation for heterogeneous nucleation:
(7.12)
γ 0 = γ 2 + γ 1 cos ω
2 γ 1Vm ∆µ
γ0
FIGURE 7.2 Cap-shaped nucleus formed by heterogeneous nucleation on a foreign substrate. The arrows represent the different surface-free energies between foreign surface, mother phase, and nucleus. (From Boistelle, R. (1988), in Crystallization and Polymorphism of Fats and Fatty Acids, Garti, N. and Sato, K., Eds., Marcel Dekker, New York, 189–276. With permission.)
474
Physical Properties: Structural and Physical Characteristics
a typical droplet was measured to contain about 25% solid fat. Electron micrographs from earlier work (Mulder and Walstra, 1974), on the other hand, have shown that, under comparable conditions, crystals are typically 0.005 µm3, which would mean some 5000 crystals per droplet. There is, thus, an enormous discrepancy between the number of crystals present and the number that can originate from heterogeneous nucleation. As homogeneous nucleation is out of the question under these circumstances, the conclusion must be that the numerous small crystals formed originated from secondary nucleation. Walstra (1998) has also given a tentative explanation for this phenomenon. Clusters of more or less oriented molecules, which Larsson (1972) postulated to exist in liquid triglycerides, may diffuse away from the growing crystal and subsequently form a new nucleus. Two conditions should be fulfilled for this secondary nucleation to take place: the crystal surface must be rough and the crystal growth rate must be so slow that the clusters of more or less oriented molecules can diffuse away from the crystal face before they become incorporated in the crystal. There, however, is no generally accepted theory for secondary nucleation (Walstra, 2003). In the industrial dry fractionation process, secondary nucleation is considered to be responsible for the formation of additional small crystals that badly affect the filtration characteristics of the stearin cake and, thus, make the process less selective (Timms, 1997). It has been linked to the rate of agitation, but again, no mechanism has been put forward.
7.2.2
FIGURE 7.3 Schematic picture of different growth faces. Each small cube represents a growth unit. (From Aquilano, D. and Sgualdino, G. (2001), in Crystallization Processes in Fats and Lipid Systems, Garti, N. and Sato, K., Eds., Marcel Dekker, New York, 1–51. With permission.)
As a consequence, the growth rate of a K face is proportional to the relative supersaturation σr (Garside, 1987; Aquilano and Sgualdino, 2001). However, when growth occurs from the melt, the free energy of activation for self-diffusion of the molecules in the melt plays an essential role in the growth kinetics. Consequently, the growth rate is also inversely proportional to the viscosity. If the viscosity increases drastically with decreasing temperature, it may happen that at a given temperature, the growth rate passes through a maximum and begins to decrease with increased supercooling (Boistelle, 1988).
Crystal growth
7.2.2.2
Once nuclei are formed and exceed the critical size, they become crystallites, whose growth depends not only on external factors (supersaturation, solvent, temperature, impurities), but also on internal factors (structure, bonds, defects) (Boistelle, 1988). Consequently, the rate of crystal growth can vary by several orders of magnitude. Growth occurs by attachment of molecules to a crystal surface. On the other hand, molecules will also become detached. There is a continuous movement of molecules across the surface of the crystal in both directions. The net result of these two processes determines the growth rate (Walstra, 2003). The mechanism by which a crystal surface grows is determined by the nature of the interface between the crystal and the liquid (Garside, 1987). As illustrated in Figure 7.3, there are three types of faces: kinked (K), stepped (S), and flat (F) (Boistelle, 1988). 7.2.2.1
Growth of a stepped (S) face
From a kinetic viewpoint a stepped (S) face is similar to a K face. However, the number of kink sites per unit area is lower compared to a K face. Just as for a K face, the growth rate for an S face is proportional to the relative supersaturation σr, but due to the lower number of kink sites, the growth rate of an S face will be lower than that of a K face (Aquilano and Sgualdino, 2001). 7.2.2.3
Growth of a flat (F) face
Flat (F) faces grow layer after layer, either by a twodimensional nucleation mechanism or by a spiral growth mechanism (Boistelle, 1988). If the crystal surface is without any defect, growth takes place by two-dimensional nucleation. The growth units that adsorb on the surface must diffuse, make contact, and coalesce to form a stable two-dimensional nucleus (Boistelle, 1988). Once a surface nucleus is formed, the new crystal layer can be filled by attachment of growth units near the kink or by repeated surface nucleation and subsequent surface diffusion to existing surface nuclei (Kloek, 1998). Consequently, the growth rate of a perfect flat face is proportional to the two-dimensional nucleation rate and, hence, it is an exponential function of supersaturation. The growth rate is near zero below a certain
Growth of a kinked (K) face
A kinked (K) face looks like an infinite population of kink sites. Continuous growth will occur since there are no thermodynamic barriers to the growth process: Each growth unit reaching the surface in a supersaturated state will immediately be incorporated into the crystal. 475
7.2
Crystallisation and melting
further in Section 7.2.7 and Section 7.2.8, respectively. If the crystallisation is intended to isolate the crystalline fraction, as in the dry fractionation process, it is usually desirable to obtain large crystals, to ensure their efficient separation from the mother liquor (Walstra, 2003). Crystal morphology is controlled by the relative rates of growth of the different crystal faces, which in turn depend on internal and external factors. Crystals may grow with flat, well-defined faces or with rounded-off, roughened faces. An intermediate situation, when spherulitic crystals form, is common in fat crystallisation. The crystals then appear as many-faceted crystals growing out of an ill-defined nucleus. Like crystal size, crystal morphology can also play an important role in the functionality of a crystallised fat and during the dry fractionation of fats (Timms, 2003). Whether crystal faces become rough depends on the temperature. There is a critical temperature, the roughening temperature, above which crystal growth leads to rough crystal faces. Below this critical temperature, even at the same degree of supersaturation, flat, well-defined faces form. This effect is known as thermal roughening. Additionally, above a critical degree of supersaturation, a flat face can become rough. This effect is known as kinetic roughening. An excellent review of this complex topic and its theoretical treatment has been published by Bennema et al. (2001).
critical value of the relative supersaturation and increases dramatically above this value. This type of growth mechanism is rather rare (Boistelle, 1988; Aquilano and Sgualdino, 2001). The second mechanism involving layer growth is the spiral growth mechanism; this occurs much more frequently than the previous one. When a screw dislocation emerges on a face, it provides a step. When the growth units adsorb onto the face, they first diffuse towards and along the step. As soon as they encounter a kink, they are fixed more firmly to the surface. The step advances by rotating around the emergence point of the dislocation. After a complete rotation, one or several layers of growth units have been added to the crystal. There are several theoretical expressions for the growth rate and the general growth rate equation is very complex. In general, it can be said that at low supersaturation, the growth rate of the face is a quadratic function of supersaturation, whereas at high supersaturation it becomes proportional to the supersaturation (Boistelle, 1988).
7.2.3
Crystal size and morphology
Crystal size is controlled by the relative rates of nucleation and crystal growth. Since nucleation depends more strongly on supersaturation than crystal growth, the rate of nucleation mainly determines the size of the crystals. Since the nucleation rate increases roughly exponentially with increasing supersaturation, many nuclei are formed when the supersaturation is high, and this also means that small crystals will result. On the other hand, at low supersaturation only a few, large crystals are formed. Secondary nucleation can complicate this picture as it leads to a larger number of crystals and consequently a smaller average crystal size (Walstra, 2003; Timms, 2003). It should be noted that as nucleation and crystal growth proceed, the extent of supersaturation necessarily decreases causing the critical size for a stable crystal or nucleus to increase. Smaller crystals, which were stable at higher levels of supersaturation, now become unstable and redissolve. In theory, this process, which is called Ostwald ripening, would continue indefinitely until eventually only one large crystal would be left in the presence of a slightly supersaturated liquid. In practice, once crystals grow to about 10 µm, the thermodynamic driving force for Ostwald ripening has become negligible (Walstra, 2003; Timms, 2003). Crystal size may be an important quality parameter in foods. In many soft solid foods, it is often undesirable that crystals are felt in the mouth, a sensation described as sandiness when the crystals also exhibit a high melting point. To prevent this, crystals generally have to be smaller than about 10 µm. The macroscopic and mechanical properties of plastic fats also depend to a large extent on fat crystal size, among other factors, but this will be discussed
7.2.4 7.2.4.1
Polymorphism General principles
Polymorphism is defined as the existence of several crystalline phases with the same chemical composition that have a different structure, but yield identical liquid phases on melting. Polymorphism leads to the existence of multiple melting points. Two crystalline forms are enantiotropic when each has a definite range of stability. Either modification may be the stable one and transition can go in either direction, depending on the conditions. Two crystalline forms are monotropic if one is stable and the other metastable under all conditions. Transition will only take place in the direction of the more stable form. Natural fats are invariably monotropic (Nawar, 1996). Larsson (1966) proposed a classification of the basic polymorphs of triglycerides on the basis of both infrared spectroscopy and x-ray diffraction, which is now generally accepted. The polymorphs are classified into three crystallographic types: α (alpha), β′ (beta prime), and β (beta), according to their hydrocarbon subcell packing. A subcell is the smallest spatial unit of repetition along the chain axis (Nawar, 1996). The α-polymorph is associated with the hexagonal subcell packing in which the fatty acid chains are perpendicular to the methyl end group plane and are assumed to be oscillating with a high degree of molecular freedom. The β′-polymorph is associated with the orthorhombic subcell packing in which the fatty acid 476
Physical Properties: Structural and Physical Characteristics
in the equation for ∆G* (Equation 7.8) will vary considerably between the most unstable polymorph, which has a structure close to that of the liquid phase, and the most stable polymorph with a structure very much different from that of the liquid phase. This leads to a lower critical free energy of activation and, thus, a higher nucleation rate for the α-polymorph, despite the fact that the difference in chemical potential ∆µ is greater for the β-form (Rousset, 1997). According to Loisel (1996) the rate of crystal growth increases with the stability of the polymorph, while Kellens (1991) stated that the growth rate of the unstable polymorph is higher than that of the stable polymorph. However, this order reverses at higher driving forces.
chains are tilted with respect to the methyl end group plane and where adjacent zigzag fatty acid chains are in different planes. The β-polymorph is associated with the triclinic subcell packing where unlike the orthorhombic subcell packing, all zigzag fatty acid chains are in the same plane (Timms, 1984). Figure 7.4 shows schematic diagrams of these three basic polymorphs. Because each of the subcell packings is characterized by a unique set of x-ray diffraction (XRD) lines in the wide angle region between 3.5 and 5.5 Å (the short spacings), the three basic polymorphs can be unambiguously identified on the basis of wide angle x-ray diffraction (WAXD). The hexagonal subcell packing of the α-polymorph is easy to identify as it exhibits one strong diffraction line around 4.15 Å. The orthorhombic subcell packing of the β′-polymorph is characterized by two strong diffraction lines around 3.7 and 4.2 Å. The triclinic subcell packing of the β-polymorph gives a whole series of diffraction lines with one prominent line at 4.6 Å and two other, less intense lines around 3.6 and 3.8 Å. (Kellens, 1991). The polymorphs differ in stability, melting point, melting enthalpy, and density. The α-polymorph is the least stable and has the lowest melting point, melting enthalpy, and density. The β-polymorph is the most stable and has the highest melting point, melting enthalpy, and density. The β′-polymorph has intermediate properties (Walstra, 1987). Although differential scanning calorimetry (DSC) can accurately determine melting points and would be a useful technique to determine polymorphism (as each polymorph has its own melting point), this technique can only give an indication of the polymorphic form due to the presence of mixtures of polymorphs and the fact that the melting point of each polymorph depends on the chemical composition (Walstra and Beresteyn, 1975; Walstra, 2003). According to Equation 7.9, the polymorph with the lowest critical free activation energy for nucleation ∆G* shows the highest nucleation rate. The surface-free energy γ
α
β'
7.2.4.2
Polymorphism of triglycerides
Most triglycerides exhibit an α-form, although it is often very unstable. Some manifest both a β′- and a β-form, others only a stable β′- and no β-form or a stable β-form and no intermediate form. The existence or nonexistence of a polymorph depends heavily on the composition and the position of the fatty acids on the glycerol unit (Kellens, 1991). For mixed saturated/unsaturated triglycerides, the β′-polymorph is usually the most stable one if the triglyceride is asymmetrical, i.e., two saturated or two unsaturated acids occupy the 1,2- or 2,3-positions, the third position being occupied by an unsaturated or saturated acid, respectively. Examples of asymmetrical triglycerides are PPO and POO, with P standing for palmitic acid and O for oleic acid. The packing requirements for the β′-polymorph are less stringent than for the β-polymorph so that mixed fatty acid triglycerides of various types tend to be β′-stable (Timms, 1984). The polymorphism of triglycerides is complicated even further by the presence of other polymorphs and multiple submodifications of the basic polymorphs. Low-temperature XRD experiments have demonstrated the existence of the sub-α-polymorph, sometimes also called the γ-polymorph. The XRD pattern of this polymorph is similar to that of the β’-form, but its melting point is lower than that of the α-form (Hagemann, 1988; Kellens, 1991). For saturated monoacid triacylglycerols, it is generally accepted that at least two β’-polymorphs exist, which are denoted as β′2 and β′1 in order of increasing stability, and several authors have also revealed the presence of multiple β-forms (Kellens, 1991). In mixed-acid triglycerides, these submodifications are even more prevalent (Sato et al., 1999). SOS, with S standing for stearic acid, for example, has five polymorphs: α, γ, β′, β2, and β1. The two β-forms, which only show very subtle differences in their subcell structure, are also observed in POP, AOA, and BOB (A denotes arachidic acid and B behenic acid, saturated fatty acids with 20 and 22 carbon atoms, respectively). They correspond to the two β-forms, form V and VI, in cocoa butter. In the γ form, the hexagonal subcell structure of the α-form is also present in the oleoyl leaflets, but a specific type of subcell occurs in the stearoyl leaflet (Sato et al., 1999).
β
FIGURE 7.4 Schematic diagrams comparing the polymorphs α, β′ and β as exemplified by tristearin. (From Timms, R.E. (1984), Prog. Lipid Res., 23, 1–38.)
477
7.2
Crystallisation and melting
7.2.4.3
other polymorphs display their usual subcell chain packing. It is evident from the large number of polymorphic forms of rather similar energy in these mixed chain diglycerides, that the two chains have difficulty in deciding what is the most favourable composition. However, considering the difference in the number of polymorphs, it seems that the polyunsaturated linoleate chains pack marginally more effectively than monounsaturated chains with the saturated stearate chains.
Polymorphic behaviour in commercial fats
It is evident that the polymorphic behaviour of a fat is largely influenced by its fatty acid composition and the positional distribution of these fatty acids on the glycerol backbone, in other words by its triglyceride composition. In general, fats comprising relatively few and closely related triglyceride species tend to transform rapidly to stable β-forms. Conversely, heterogeneous fats tend to transform more slowly to stable forms since compound β′-crystals can host fewer different molecules than α-crystals and compound β-crystals hardly exist. Fats that tend to crystallise in β-forms include soya bean oil, peanut oil, low erucic acid rapeseed oil (canola), corn oil, olive oil, coconut oil, cocoa butter, and lard. On the other hand, cottonseed oil, palm oil, high erucic acid rapeseed oil (HEAR), milk fat, and tallow tend to produce β′-crystals that tend to persist for long periods (Nawar, 1996; Walstra, 2003). 7.2.4.4
7.2.4.5
Polymorphic behaviour of monoglycerides
For 1-monoglycerides it is generally agreed that four polymorphs exist: sub-α, α, β′, and β. Sub-α forms are a characteristic feature of 1-monoglycerides since they occur in all samples on which polymorphic studies have been performed. They form at low temperatures from the α-form via a solid-state transition. Some studies do not report a β’-form. A possible explanation for this is the difficulty of preparing this form (Hagemann, 1988; Krog and Sparso, 2004). But it is also true that unlike pure 1-monoglycerides, industrial distilled monoglycerides do not form any β′-polymorph (Krog, 2001). It seems that 2-monoglycerides are free from polymorphism, although Malkin (1954) has stated that the 2-monoglycerides probably separate from the melt in the α-form, but transform very rapidly into the β-form. However, it has to be stressed that the polymorphism of these compounds has not been studied frequently (Hagemann, 1988).
Polymorphic behaviour of diglycerides
Hagemann (1988) has reviewed the polymorphism of diglycerides. All 1,3-diglycerides ranging from CC (dicaprate), EE (dielaidate), OO (dioleate) to SM (M stands for myristic acid) possess stable β-polymorphs and no α-form. Even chain length 1,3-diglycerides from dilaurate to distearate can form two β–polymorphs, which Baur (1949) called a and b. Shannon et al. (1992) confirmed the existence of these two β-forms in PP and SS, but renamed them β2 and β1, respectively to be in accordance with the polymorph nomenclature rules of Larsson (1966). The two forms can be distinguished on the basis of subtle differences in short spacing values. A quite different polymorphism is displayed by 1,2diglycerides. Even monoacid 1,2-diglycerides have an α- and a stable β′-polymorph; this was erroneously called a β-form in earlier studies. The optically active and the optically inactive racemic 1,2-diglycerides exhibit very similar polymorphism but show contrasting thermal behaviour (Hagemann, 1988; Shannon et al., 1992). Di and Small (1993; 1995) studied the polymorphism of mixed chain diglycerides in which a saturated 18 carbon chain is attached to the 1-position and an unsaturated 18 carbon chain to the 2-position. This 1-stearoyl-2-oleoylsn-glycerol was shown to have eight polymorphs: γ2, γ1, α, β4, β3, β2, β1 and β′ in order of increasing melting point. In this case the β’-polymorph is thus the most stable. The γ2and γ 1 -polymorphs are formed reversibly from the α-polymorph on cooling. The β′-polymorph can only be formed by crystallisation from hexane at 4°C by slow evaporation. The β1-polymorph is only formed when a sample that is crystallised from a solvent is melted at 30°C for 1 h to remove the seeds of β′ and then cooled to below 17.4°C. The diglyceride 1-stearoyl-2-linoleoyl-sn-glycerol has four polymorphs: sub-α2, sub-α1, α, and β′ in order of increasing melting point. The polymorphs sub-α2 and sub-α1 have a pseudo-hexagonal chain packing, while the
7.2.4.6
Phase transitions
Most transitions are monotropic or irreversible and characterised by first-order kinetics. Only the sub-α to α transition is assumed to be reversible and of the secondorder type. Second-order transitions occur over a wide temperature range and do not involve heat exchange. Large changes in crystal structures do not occur in second-order transitions, which merely involve an increase in molecular movement (Hagemann, 1988; Kellens, 1991). However, Loisel et al. (1998) stated that, at least in cocoa butter, the sub-α to α-transition is also irreversible. They deduced this from the observation that cooling of the α-polymorph, obtained by transition of the sub-α polymorph on heating, did not restore the latter even after 1 day at low temperature. A metastable crystal can change into a stable one through the rearrangement of its structural unities until a complete transformation occurs (solid-state phase transition) or by melting and recrystallisation (melt-mediated phase transition). If a suitable solvent is involved, the metastable phase dissolves and a new stable phase is allowed to form from its supersaturated solution (solventmediated phase transition). The kind of transition may depend on the thermal history of the sample as is the case for 1,2-dipalmitoyl-sn-glycerol. For this diglyceride, Takahashi (1999) showed that when the diglyceride is incubated at around 3°C for more than 10 h, the α-phase 478
Physical Properties: Structural and Physical Characteristics
is converted directly into the β′-phase, but that without incubation at low temperatures, the α-phase is transformed into the β′-phase by way of a melted phase. If a number of different phase transitions from a less stable state to more and more stable states are possible, the closest more stable phase is usually formed and not the most stable one (Aquilano and Sgualdino, 2001). Phase transition kinetics can vary widely, ranging from almost instantaneous to extremely slow for some solid-state transitions (Aquilano and Sgualdino, 2001). 7.2.4.7
7.2.5
The solid fat content of a sample (SFC) is a measure of the percentage of solid, crystalline fat in a sample at a selected temperature. Often, the SFC is measured at selected points within a temperature range (Timms, 2003). A measure of the SFC can be determined by a variety of methods: dilatometry, pulsed nuclear magnetic resonance (p-NMR), or differential scanning calorimetry (DSC). The method used and differences in the way it is executed can seriously affect the final result. This should always be taken into account when comparing different results. Over the years several authors have compared the various methods, e.g., Walker and Bosin (1971), Lambelet (1983), Van Duynhoven et al. (1999), and Marangoni et al. (2000). Until the early 1970s, dilatometry was the standard method. It is based on the difference in specific volume of the liquid and the crystalline phase. It assumes all fats to have the same melting dilatation, which is incorrect since the melting dilatation of a fat varies according to the molecular weight of the triglycerides, the type of fatty acids and the polymorphic form of the fat. The result of a dilatometric measurement, called the Solid Fat Index (SFI) and expressed as [mm3/g] or [mm3/25 g] is fundamentally different from the solid fat content expressed as [%]. Moreover, a dilatometric determination takes a long time. In the early 1970s, Unilever research workers (Enden et al., 1978; 1982) pioneered the practical use of a pulsed NMR technique, also referred to as “time-domain” NMR (Todt et al., 2001) as a routine method for determination of the SFC that is expressed as a percentage between 0 and 100%. The principle of the pulsed NMR technique is as follows: After a short radio-frequency pulse, which rotates the magnetic field by 90°, the magnetisation signal in the detector decays over several hundred milliseconds. The decay of the signal from protons in the solid state is rapid, occurring over tens of microseconds, whereas the decay of the signal from protons in the liquid state is much slower, occurring over tens to hundreds of milliseconds. Accordingly, a distinction can be made between crystalline and liquid triglycerides. The SFC can be measured using the Direct Method, which involves the determination of the solid + liquid and of the liquid signals. Unfortunately, a “dead time,” during which no measurements can be made, occurs after the pulse, which makes it is impossible to measure the solid + liquid signal. To overcome this problem, an f-factor was introduced, which is often set at a value of 1.4 (f-factor for β′-margarine fats of Unilever). However, this f-factor depends on several factors, but most importantly on the polymorphic form of the fat. Because for many fat blends the actual f-factor is not known, it may be preferable to use the Indirect Method. In this method, the solid signal is ignored and the SFC is calculated by comparing the measured liquid signal to the liquid signal of a reference that is completely liquid at the temperature of measurement.
Arrangement in crystal space (cf. long spacings)
The x-ray scattering patterns at small angles (SAXS) of the various polymorphs show a series of lines that are related to the thickness of the layers formed by the sideby-side arrangement of the chains (long spacings). The layer thickness depends on the length of the molecule and, hence, on the number of carbon atoms in the fatty acid chains, and on the angle of tilt between the chain axis and plane of the methyl end groups. The triacylgly-cerols are arranged head to tail and form a chair-shaped structure with the fatty acid at the 2-position forming the back of the chair. Two packing modes are possible, resulting in pairs of two (2L packing) or three (3L packing) fatty acid chains (Figure 7.5) (Kellens, 1991). Pairs of three fatty acid chains are formed when the fatty acid chains are mixed, i.e., when there are large differences between the number of carbon atoms of the different chains and when saturated and unsaturated fatty acids form part of the same triacylglycerol (Sato and Kuroda, 1987).
1
3
2 1
3
L
2 2 L
2
3
δ
1
3
1
δ β–2
Solid Fat Content
β–3
FIGURE 7.5 Arrangement of the triacylglycerol molecules in the crystalline phase: double and triple chair arrangements of the form. L is the layer thickness and δ is the angle of tilt. (From Walstra, P. (1987), in Food Structure and Behaviour, Blanshard, V.M.V. and Lillford, P., Eds., Academic Press, London, 67–85.)
479
7.2
Crystallisation and melting
recent work concerning the modelling of multistep processes and the modelling of nonisothermal crystallisation.
When using calorimetry, the heat of melting per unit mass is measured. This value, however, varies significantly among triglycerides. When a constant average value is taken, this means that the resulting proportion of solid fat is significantly biased. The value of the latent heat also depends on the polymorph, which can further bias the result. An ultrasonic technique has also been proposed as an alternative way to measure the solid fat content (Povey, 1995). It is based on the observation that the velocity of ultrasound is greater in solid fat than in liquid fat. With this technique, the polymorphic form of the sample can bias the results, since polymorphic transitions lead to more densely packed crystals and, therefore, lower compressibility, resulting in a higher ultrasound velocity. Before its solid fat content can be determined, the fat must be exposed to a prescribed temperature profile: first it has to be melted completely to destroy all traces of crystals, then cooled to achieve virtually complete crystallisation, and finally it has to be held at the measuring temperature to come to equilibrium at that temperature. Sometimes, depending on the fat used, an extra step is introduced where the fat is held at a particular temperature, which is not the measuring temperature. This step is referred to as a tempering step. For confectionery fats, a tempering step of 40 hours at 26°C is mentioned in the standard methods to ensure that cocoa butter and similar fats like cocoa butter equivalents (CBEs) are converted to their β-polymorph before the SFC is measured.
7.2.6
7.2.6.1
The Avrami model
The Avrami model is the most widely used approach for the description of isothermal phase transformation kinetics. In the 1940s, various authors independently developed this kinetic formulation, which is sometimes also called the Johnson–Mehl–Avrami–Kolmogorov equation. The theory was initially developed for low molecular weight materials such as metals. Later it was extended to the crystallisation of polymers. Avrami (1939; 1940) stated that there is overwhelming evidence pointing to the conclusion that phases are nucleated by tiny germ nuclei that already exist in the liquid phase and whose effective number equals N 0 per unit nucleation region. The number of germ nuclei per unit region at time t decreases from N 0 in two ways: (1) some of them become active growth nuclei (N at time t) as a consequence of free energy fluctuations and with a probability of occurrence p per germ nucleus, and (2) some of them get swallowed by growing grains of the new phase. The number of growth nuclei may increase linearly with time (sporadic nucleation), or the large majority of the growth nuclei can be formed near the beginning of the transformation process (instantaneous nucleation). The variable V represents the volume of the crystalline phase per unit volume of space. Avrami also introduced a characteristic time scale, defined by pdt = dτ. This characteristic time scale is in fact rescaled time taking into account the value of p. Furthermore, Avrami made the assumption that growth ceases when one grain impinges upon another. The volume at rescaled time of any grain that began growth from a nucleus at rescaled time z is denoted as v(τ, z). The number of such grains is expressed by N(z). Thus, the total extended volume (the term “extended” refers to the volume the grains would have had, if growth had remained unimpeded) equals:
Modelling crystallisation kinetics
The crystallisation kinetics of fats (when, how fast, and to what extent fat components crystallise under certain conditions) is the basis for controlling operations in which (re)crystallisation is of concern. From the late 1970s onwards, and especially in the last few years, quite a number of articles have been published in which the isothermal single-step crystallisation of fats is mathematically modelled to enable the quantification of differences in crystallisation behaviour of different products and under different crystallisation circum-stances. A model is constructed on the basis of the experimental data sets, and providing parameters with a physical meaning. The most frequently used model to describe the isothermal one-step crystallisation kinetics of fats is the Avrami model. Some authors use a modified Avrami equation, also called the Avrami–Erofeev equation. Kloek (1998) and Vanhoutte (2002) used a reparameterised Gompertz equation to describe their crystallisation curves. Foubert et al. (2002) developed a new model, which is available as both an algebraic and a differential equation. In the following sections, further details of these four models used to describe the isothermal one-step crystallisation of fats will be given. This is followed by a paragraph comparing the different models with respect to curve fitting and their theoretical background. Finally, attention is given to some
τ
Vext = ∫ v( τ, z )N ( z )dz
(7.17)
0
Let r, the “radius,” be a one-dimensional measure of the size of a grain and let G be the direction average of the rate of growth of r. For a grain, which started to grow at time y, the radius r at time t is given by: t
r(t, y ) = ∫ G ( x )dx
(7.18)
y
or, if the rescaled time τ is introduced: τ
r( τ, z ) =
∫ p du z
480
G
(7.19)
Physical Properties: Structural and Physical Characteristics
Note that G differs from G(x) since G is expressed in the rescaled time. The grain volume then becomes:
particular case. To obtain the value for Vext, Equation 7.21 is integrated taking into account that N ( z ) = N0 e − z and x
⎡ G ⎤ v( τ, z ) = σr 3 = σ ⎢ ∫ du ⎥ ⎢⎣ z p ⎥⎦ τ
Eq ( − x ) =
3
(7.20)
1 ( x − z )q e − z dz q ! ∫0
= ( −1)
where σ is a shape factor, equal to 4π/3 for a sphere. Since the factors which govern the tendency of the growth nuclei to grow out of the germ nuclei are similar to those which govern further growth, Avrami assumed that p and G are approximately proportional throughout a considerable temperature and concentration range called the isokinetic range. Thus, if G/p is constant for a given substance in the isokinetic range, Equation 7.17 can be integrated according to:
q +1
⎡ −x ⎢e − 1 + x...( −1) ⎣
q +1
xq ⎤ ⎥ q! ⎦
(7.26)
_
Vext =
6 σG 3 N 0 p3
⎡ −τ τ 2 τ3 ⎤ 1 τ − + − + ⎥ = βE3 ( − τ ) e ⎢ 2 ! 3! ⎦ ⎣
(7.27)
where the following abbreviation has been introduced:
Vext = σ
G p3
3 τ
∫ ( τ − z ) N ( z )dz 3
(7.21)
_
0
β=
In any region, selected arbitrarily, the part of the volume still without crystallised matter is designated as the “nonoverlapped” volume. Then, on average, the ratio of the nonoverlapped volume v′ to the extended volume vext of a randomly selected region is equal to the density of untransformed matter 1-V at that time, so that:
v' = 1 −V vext
_
_ _ ⎫ ⎧ ⎡ ⎤ Vext = β ⎨E3 ( − τ ) − e − τ E3 ⎢ −( τ − τ ) ⎥ ⎬ ⎣ ⎦⎭ ⎩
(7.22)
(7.23)
V = 1− e
( − β τ 4 / 4 !)
= 1− e
_
( − σ G 3 N0 p t4 ) / 4
(7.30)
(7.24) Note that a not too large value for p and a very small value for t lead to similar values of τ and, therefore, support the same reasoning. However, Avrami did not take this case into account. On the other hand, when τ is very large, i.e., for p very large and t not too small, i.e., instantaneous nucleation, the exponential term e −τ and the terms up to the order of two in Equation 7.27 can be disregarded in comparison
Integrating and rearranging, this gives:
V = 1 − e −Vext
(7.29)
When N 0 is very large, i.e., the supply of germ nuclei is not exhausted until the end of crystallisation, two borderline cases can be considered. When τ is very small, i.e., when p is very small and t is not too large, i.e., in the case of sporadic nucleation, the first four terms of the series expansion of the exponential term e −τ in Equation 7.27 cancel out against the other terms between the square brackets. Hence, only the term of the fourth power in τ has to be taken into account, being the first term that does not cancel. By inserting the equation for Vext thus obtained in Equation 7.25 the following equation results:
since the nonoverlapped decrease of a grain is the same as the increment in transformed volume of that grain. For the unit volume this leads to:
dV = 1 −V dVext
(7.28)
This equation is valid up to τ = τ , the time corresponding to the exhaustion of the supply of germ nuclei. Beyond this, the upper limit of the integral should be replaced by _ τ and the result of integration may be expressed by:
The same reasoning may be applied to the nonoverlapping and extended portions of the increments of single grains in an element of time. The following equation is obtained for the average grain:
dv = 1 −V dvext
6 σG 3 N 0 p3
(7.25)
Thus, the entire problem of determining the kinetics of the crystallisation has been reduced to finding Vext in any
481
7.2
Crystallisation and melting
4. The growth morphology changes during the process. 5. Crystalline aggregates grow concurrently from both instantaneous and sporadic nuclei.
with the last term between square brackets; this results in the following equation for V:
V = 1 − e −β τ
3 / 3!
= 1 − e− σ G
3
_
N 0 t3
(7.31)
Evans (1945), seemingly without knowledge of the prior work of Avrami, arrived at the same model by solving the problem of expanding waves created by raindrops falling on a pond.
Again, τ can also be very large in the opposite case (t very large and p not too small), a case that Avrami did not take into account either. For intermediate values of p the way V depends on t will lie in between Equation 7.30 and Equation 7.31. In general, Equation 7.30 and Equation 7.31 can be written as:
V = 1 − e − kt
m
7.2.6.2
Apart from the original Avrami model as derived above, some authors also use a so-called modified Avrami model, also called the Avrami–Erofeev model. The modified equation is expressed by:
(7.32)
V = 1 − e( − k ' t )
which expression represents the equation that is generally known as the Avrami equation. For plate-like and linear growth, an analysis similar to the previous one leads to other values for k and m (Table 7.1). As can be seen from Table 7.1, the rate constant k depends on the nucleation (amount of germ nuclei N 0 for instantaneous nucleation and rate of nucleation p N 0 for sporadic nucleation) and on the growth rate. The exact relationship depends on the specific case. The Avrami exponent m depends on the type of nucleation (sporadic or instantaneous) and the growth morphology of the crystallising particles. The meaning of the value of m, however, is not straightforward, since values for m of 2 and 3 can have two different meanings. Theoretically, values for m should be an integer. However, analysis of experimental data frequently leads to Avrami exponents that are a noninteger. Several causes have been suggested (Long et al., 1995; Supaphol and Spruiell, 2001):
⎡ τ 2 τ3 ⎤ − ln(1 −V ) = const. ⎢e − τ − 1 + τ − + ⎥ 2 ! 3! ⎦ ⎣
⎡ p2t 2 p3t3 ⎤ − ln(1 −V ) = const. ⎢e − pt − 1 + pt − + ⎥ 2! 3! ⎦ ⎣
Sporadic Nucleation -n
K[s ] _
_
σ *G 3 * N 0* p 4
482
k [s-n]
m [-] _
2
σ ''* G * N 0
3
σ '* G 2 * N 0
4
σ * G 3 * N0
_
σ '* G 2 * N 0 * p 3 Spherical
Instantaneous Nucleation m [-]
σ ''* G * N 0 * p 2 Plate-like
(7.34)
When returning to the original time scale (τ= p*t), one obtains:
Summary of the values obtainable for k and m in the Avrami model
Growth Morphology Linear
(7.33)
m'
This equation differs from the original Avrami equation in that the rate constant k’ is also raised to the power m′ which is not the case in the original Avrami equation. This modified model originates from the work of Ng (1975) who described the development of the Erofeev model as well as from the work of Khanna and Taylor (1988) who modified the Avrami model to eliminate the dependence of k on m. Ng (1975) described the development of the Erofeev model in a work on thermal decomposition in the solid state. This development was based on the Avrami theory. It can be deduced from Equation 7.25 and Equation 7.26 that for three-dimensional growth:
1. The ratio of the density of the crystalline phase over the density of the liquid phase varies during the process. 2. The true nucleation rate varies during the process. 3. The growth rate changes during the process. TABLE 7.1
Modified Avrami model
_
_
1
2
3
(7.35)
Physical Properties: Structural and Physical Characteristics
extruded Nylon 6. Thus, if k is an overall rate constant, it should always be larger for the extruded Nylon 6 resin than for the virgin material. When calculating k by means of linear regression, it appeared that the value was higher for the extruded Nylon 6 at temperatures below 200°C, but lower at temperatures above 200°C. When calculating k′ from the modified Avrami model, the value was always higher for the extruded Nylon 6, as expected. The authors also cite other work where the original Avrami model has yielded rate constants that differ from the expected values. The authors further claim that despite the modification, the model retains its original connection to nucleation and crystal growth processes. The modification presented simply corrects the value of k by eliminating the influence of m. Khanna and Taylor (1988) conclude that attempts to obtain k values through the original Avrami model may lead to erroneous results, especially when comparing processes that have different values of m. Marangoni (1998) does not agree with this modified Avrami model and argues that Khanna and Taylor (1988) arbitrarily suggested a modification of the Avrami model without providing any theoretical justification. The only justification the authors provided was their opinion that k and m were correlated and that the modification would solve this problem. However, no proof of this was given in their paper.
This equation can be transformed into a simplified form in two borderline cases (see also the deduction of Equation 7.30 and Equation 7.31). When p x t is much smaller than 1, Equation 7.35 changes into:
⎛ p 4t 4 ⎞ − ln (1 −V ) = const ⎜ ⎝ 4 ! ⎟⎠
(7.36a)
or 1/ 4
[− ln (1 −V )] =
const pt = const '× t 24
(7.36b)
When p × t is much larger than 1, Equation 7.35 can be simplified into:
⎛ p3t3 ⎞ − ln (1 − V ) = const ⎜ ⎝ 3! ⎟⎠
(7.37a)
or 1/ 3
[− ln (1 −V )] =
const pt = const '× t 6
(7.37b)
These equations can be represented by the generalised Erofeev equation: 1/ m '
[− ln(1 − V )]
= k 't
7.2.6.3
Kloek (1998) and Vanhoutte (2002) fitted their crystallisation curves to a reparameterised Gompertz equation as derived by Zwietering et al. (1990). The latter authors compared several sigmoid functions for their ability to describe a bacterial growth curve. Most of these functions contain mathematical fitting parameters rather than parameters with a biological meaning, making it difficult to provide initial guesses for these parameters. Moreover, it is difficult to calculate 95% confidence intervals for biologically meaningful parameters if these are not derived directly from the equation but calculated from mathematical fitting parameters. Therefore, the growth models were rewritten to replace the mathematical fitting parameters with biologically meaningful parameters such as aG (the maximum value reached), µ (the maximum specific growth rate that is defined as the tangent in the inflection point), and λ (the lag time, which is defined as the x-axis intercept of that tangent) (Figure 7.6). This reparameterisation was performed by deriving an expression for the biologically meaningful parameters as a function of the mathematical fitting parameters of the basic function. The unmodified Gompertz equation is written as:
(7.38)
which is identical to Equation 7.33. Khanna and Taylor (1988) claimed that the value of k resulting from the original Avrami model may not be correct, since k is a function of m. According to the authors, this problem can be eliminated by using a modified equation, such as Equation 7.33. What these authors did was transform the Avrami constant k from a complex constant of an mth order process to a first order rate constant, despite the fact that crystallisation is not a first order process. It can be calculated that the k’ value of the modified Avrami model is the mth root of the k value of the original model (Marangoni, 1998):
k' = k
1/ m
The Gompertz model
(7.39)
Therefore, the modified Avrami model is simply a reparameterised Avrami model. Khanna and Taylor (1988) show that by modifying the Avrami model, more meaningful values for the reaction rate constant can be obtained. They compared, for example, the overall crystallisation rates of virgin Nylon 6 and extruded Nylon 6. By means of rate programmed DSC experiments, isothermal DSC experiments and optical microscopy, they showed that the crystallisation rate of virgin Nylon 6 is dramatically lower than the rate of
Y = A × exp ⎡⎣ − exp (B − D × t )⎤⎦
(7.40)
with Y being the logarithm of the relative population size.
483
7.2
Crystallisation and melting
7.2.6.4 Y
Unlike the former models, the model developed by Foubert et al. (2002) was originally expressed in the form of a differential equation. This type of equation has the advantage that (1) it is often easier to interpret the equation mechanistically, (2) it is easier to make minor changes to the equation on the basis of acquired knowledge, and (3) by incorporation of secondary models describing the temperature dependency of the parameters, the model can be used to describe nonisothermal crystallisation kinetics. An algebraic solution for isothermal conditions, however, offers the advantage that the estimation of the parameters is facilitated by readily available software packages capable of nonlinear regression of algebraic equations. An algebraic solution assuming isothermal conditions, therefore, was also developed. The differential equation of this model is expressed in terms of a variable h, which is the residual crystallisable fat:
a
µ λ
FIGURE 7.6
Time
Bacterial growth curve.
To obtain the inflection point (at t = ti) of the curve, the second derivative of the function with respect to t is set at zero. This leads to:
ti = B / D
(7.41)
An expression for the maximum specific growth rate can be derived by calculating the first derivative at this inflection point:
µ=
A− D with e equalling 2.718281 e
h=
( B − 1) D
(7.43)
µ×e λ +1 The parameter B can be replaced by A . The aG value equals the value of A since Y approaches A when t approaches infinity. Parameter A in the unmodified Gompertz equation can be substituted by aG, yielding the reparameterised Gompertz equation: ⎧ ⎡µ × e ⎤⎫ Y = aG × exp ⎨− exp ⎢ * (λ − t ) + 1⎥ ⎬ a G ⎣ ⎦ ⎪⎭ ⎩
aF − f aF
(7.45)
where f is the extent of crystallisation at time t, and aF is the maximum extent of crystallisation. In contrast to f, which increases sigmoidally over time, this variable h is related to the residual supersaturation (i.e., the driving force for crystallisation) and, thus, decreases sigmoidally over time. To obtain the model, the crystallisation process is represented as if it is a combination of a first-order forward reaction and a reverse reaction of order n with rate constants Ki for each of the reactions. The dynamics of h can then be written mathematically as:
(7.42)
The parameter D in the unmodified Gompertz equation, thus, can be replaced by µ ×e/A. To determine the lag time, the slope of the tangent through the inflection point is calculated and, subsequently, the intercept with the t-axis is calculated:
λ =
Model of Foubert (2002)
dh = K n × h n − K1 × h dt
(7.46)
K1 and Kn are the rate constants of the first order forward reaction and the nth order reverse reaction, respectively. To calculate the values of h as a function of time according to Equation 7.46, the initial value for h, h(0), needs to be elaborated according to:
(7.44)
Kloek (1998) and Vanhoutte (2002) used this reparameterised Gompertz equation after replacing Y by f, the extent of crystallisation at time t. At that point in time, A is the maximum fraction of solid fat. Kloek (1998) used this model because of several analogies between the crystallisation of fats and bacterial growth: production of bacteria is comparable with nucleation and growth of crystals, and consumption of nutrients is comparable with decrease of supersaturation.
h( 0 ) =
aF − f ( 0 ) aF
(7.47)
with f(0) representing the amount of crystals initially present, which can be related to the induction time of the crystallisation process. Extensive parameter estimation studies revealed that the relative difference between K1 and Kn amounts only to some 10-4 %. Furthermore, the quality of the 5-parameter
484
Physical Properties: Structural and Physical Characteristics
trend that could also be seen when the models were fitted to data pertaining to other cocoa butters at different temperatures and to data on milk fat and its fractions (Foubert et al., 2002). The Foubert model shows a better fit than both other models. The ability of the different models to describe isothermal fat crystallisation adequately has been tested statistically by Foubert et al. (2002). This study revealed that the Gompertz and Foubert models always perform better than the Avrami model and that the Foubert model performs better than the Gompertz model in the majority of cases. The theory behind the Avrami model was developed on the basis of some assumptions that may not always be valid in the case of fat crystallisation. In addition to the fact that this may lead to noninteger values for the Avrami exponent m, it may also raise questions about the applicability of the Avrami model. The modified Avrami model, advocated by some authors, has been criticized by others as having no theoretical foundation. In our opinion, the modified Avrami model simply is a reparameterisation of the original model, possibly leading to better parameter estimates. The theoretical basis for using the Gompertz model for fat crystallisation is rather weak. Bacterial growth can intuitively be compared with fat crystallisation, but this provides no real theoretical justification. The very good fits obtained with the Foubert model make it a useful tool to obtain a better quantitative description of crystallisation processes. Later work (Foubert et al., 2005) has already given some more insight into the meaning of the different parameters.
model was found to be not significantly better than that of a 4-parameter model for which K1 = Kn. It was decided to simplify the model according to:
dh = K × ( hn − h) dt
h( 0 ) =
aF − f ( 0 ) aF
(7.48)
in which aF is the maximum extent of crystallisation (expressed in percent [=solid fat potential] if measured by means of p-NMR, or expressed in J/g (latent heat) if measured by means of DSC), K is the rate constant (expressed in time unit-1), n is the order of the reverse reaction (dimensionless) and f(0) is the amount of crystals initially present (expressed in the same units as a). To simplify the calculation of the parameters, the differential equation (4-parameter model) was converted to its algebraic solution. Since the physical interpretation of the parameter “induction time” is more straightforward than that of the parameter h(0) (or the equivalent f(0)) and since the induction time can be more easily estimated from a crystallisation curve, it was decided to represent the equation as a function of tx instead of h(0). The parameter tx is defined as the time needed to obtain x% of crystallisation.
h = ⎡⎣1 + ((1 − x )1− n − 1) × e − 7.2.6.5
(1− n ) × K × ( t − tx )
⎤ ⎦
1 1− n
(7.49)
Comparison of models
Figure 7.7 compares the quality of fit of the Avrami, Gompertz, and Foubert models for a cocoa butter crystallisation followed by means of DSC. The modified Avrami model is not represented since it is not different from the original Avrami model with respect to curvefitting qualities. Figure 7.7 shows that the Gompertz model provides a better fit than the Avrami model, a
7.2.6.6
Models to describe mult-step crystallisation
Since fats are complex mixtures of triglycerides, their crystallisation can lead to the formation of many crystal types, either due to polymorphism or concomitant growth of several crystal types. This may lead to crystallisation curves in which two steps can be identified.
Released crystallization heat (J/g)
80 70 60 50 data Avrami Gompertz Foubert
40 30 20 10 0 0
FIGURE 7.7
0.2
0.4
0.6
0.8 1.2 1 Time (h)
Comparison of fit.
485
1.4
1.6
1.8
2
7.2
Crystallisation and melting
Released crystallization heat (J/g)
70 60 50 40 30 20 10 0
FIGURE 7.8
0
0.5
1
1.5 2 Time (h)
3
3.5
Two-step crystallisation curve.
Figure 7.8 shows an example of such a two-step process, which, of course, also makes the modelling more complex. To be able to fit a model to this kind of data Vanhoutte (2002) combined two Gompertz equations simply by numerically adding two algebraic Gompertz equations. Mazzanti et al. (2005) developed a quantitative model describing the growth of crystals and shear-related phase transitions in palm oil using a modified differential form of the Avrami model. Foubert et al. (2006) developed a model describing the isothermal two-step phase behaviour of cocoa butter as an extension to the original Foubert model. Both models have been developed on the basis of the known crystallisation mechanism and are combinations of several differential equations. 7.2.6.7
2.5
7.2.7
Microstructural development
The various stages of the microstructural development (aggregation, network formation, and sintering) are illustrated in Figure 7.9. Fat crystals attract each other by van der Waals’ forces. The only repulsive interaction is hard-core repulsion, which only operates over very short distances. Random aggregation of particles that encounter each other due to Brownian movement and then stick together, leads to formation of fractal aggregates. A specific property of such fractal aggregates is that their structure is self-similar, implying that they have, on average, the same structure when observed at different magnifications. The time
Models to describe nonisothermal crystallisation
A model able to describe nonisothermal crystallisation will be very interesting for the food industry, since most of their processes are of a nonisothermal kind. However, up to now most studies of nonisothermal fat crystallisation have been comparative in nature and have not applied a kinetic model. The difficulty in modelling nonisothermal crystallisation is that the rate of crystallisation depends upon the degree of crystallisation and the temperature, both of which vary. Smith et al. (2005) examined the determination of kinetic parameters from nonisothermal DSC crystallisation of a model fat, POP. They applied peak and isoconversional methods to determine activation energies and compared these techniques with a nonparametric method, which separates the temperature dependence and the degree of crystallisation dependence of the crystallisation rate. The Avrami model provides the best fit with the data, while the temperature dependence of the rate constant is best explained by a Vogel–Fulcher relationship, with the melting point of the crystallising species as the reference temperature.
1
2
3
FIGURE 7.9
486
Stages in microstructural development.
Physical Properties: Structural and Physical Characteristics
This section gives an overview of the macroscopic and mechanical properties of crystallised fats together with some information on how they are measured.
needed to form aggregates of a few particles is in the range of between 10 and 100 seconds. As soon as the volume fraction of particles in a fractal aggregate is about the same as the volume fraction of primary particles in the system, the aggregates start to impinge on each other. A continuous network or a gel is formed, giving the fat elastic properties. Values observed for the time it takes for a gel to be formed are between 2 and 5 minutes, i.e., at a very low fraction of solid fat (1%) (Kloek, 1998; Walstra et al., 2001). After formation of the primary network, the major part of the fat, thus, still has to crystallise. This additional crystallisation will lead to compaction of the aggregates that form the primary network. Whether the fractal nature of the network still exists at these higher solid:liquid ratios is a controversial point (Kloek, 1998; Narine and Marangoni, 1999). Another consequence of further crystallisation after aggregation is sintering, i.e., the formation of solid bridges between aggregated crystals and aggregates. Sintering of two crystals will occur if some triglyceride molecules are incorporated in the lattices of both crystals. This is more likely to occur at crystal surfaces that have defects due to lattice mismatches and may be related to the occurrence of compound crystals. Therefore, it is more likely that sintering occurs in fats that contain many different triglycerides (Kloek, 1998). Johansson (1994) showed that polymorphism is also important for the occurrence of sintering; sintering only occurs when the outer part of the crystals and the bridging molecules have the same polymorphic structure. The properties of a fat crystal network are affected by several formation process parameters, such as temperature, concentration, and agitation. A lower temperature means smaller crystals in the early stages, which would mean a somewhat faster aggregation and a clearly shorter gelation time. However, these conclusions are not universally valid because factors, such as polymorphism and secondary nucleation, disturb this pattern. Mild agitation may speed up aggregation of crystals, but at high shear rates, the resulting forces cause disaggregation. Prediction of these effects, however, is far from easy because the prevailing strain rates tend to vary enormously according to place and time (Walstra et al., 2001).
7.2.8
7.2.8.1
Small deformations
When discussing small-scale deformation behaviour (mostly measured by oscillatory experiments in a rheometer), the modulus, be it elastic or viscoelastic, is the most important parameter. It is a measure of the stiffness (not the strength) of the system. One can only speak of a true modulus if the strain is proportional to the stress, i.e. when measurements are performed in the so-called linear region. This region comprises only strains in the order of 10-4, which is a very low value. Outside this region, irreversible changes in the crystal network are induced (Kloek, 1998; Walstra et al., 2001). The rheological properties of fats are primarily determined by the fraction of crystallised fat. Another important factor is the interaction between the crystals. Initially, crystals are aggregated due to van der Waals’ attraction. Deformation then leads to an increasing distance between the crystals. However, if the crystals are sintered, the crystals have to bend when the network is deformed. In this case, the elastic modulus depends on the bending modulus of the crystals. Bending of crystals can also be important for nonsintered dispersions if the crystals are long and thin. The geometric arrangement of the crystals and the aggregates is also an important factor determining the elastic modulus. All these influences have been incorporated in various models that explain the elastic modulus of particle networks. 7.2.8.2
Large deformations
Although the theoretical background of small deformations has been studied quite extensively, the results are rarely of practical significance. Research on large-scale deformation behaviour, on the other hand, is of great practical importance as these experiments yield information about important quality characteristics, which are relevant to processing, handling, and eating. An example of such a characteristic is spreadability. A major criticism concerning these large-scale deformation tests, however, is that it is difficult to correlate the measured parameters with fundamental properties of the material (Shellhammer et al., 1997; Kloek, 1998). Care should also be taken when interpreting the results, since large-scale strain experiments can be applied in various deformation modes, which tend to give different results. Moreover, several factors may affect the results, such as deformation of the sample when transferring it in the measuring body, or inhomogeneous deformation (Walstra et al., 2001). Figure 7.10 gives an idealized stress-strain curve for a margarine-type fat with a fairly low SFC. Four regions can be distinguished. In zone 1 the behaviour is linearly elastic, which means that the strain is proportional to the
Mechanical and macroscopic properties
The mechanical and macroscopic properties of a fat are influenced by all the structure levels described above (molecular structure, primary crystals, crystal aggregates, three-dimensional network). Each step in this structural hierarchy is influenced by the processing conditions (Narine and Marangoni, 1999). The interaction between these structure levels has been the subject of recent research, e.g., Kloek (1998), Narine and Marangoni (1999), and Vanhoutte (Vanhoutte, 2002). 487
7.2
Crystallisation and melting
σ/kPa
Finally, plastic flow is attained in zone 4. The material flows, but it is very viscous. The high viscosity is primarily due to the presence of fairly large remnants of what was originally a continuous crystal network. Moreover, the irregular shape and spiky surface of these structural elements leads to their entanglement during flow. Possibly, van der Waals’ attraction between crystals or crystal aggregates may also contribute to these properties. The deformation rate can also be determined as a function of the stress applied. Often, this flow curve becomes virtually linear, and the Bingham yield stress σB can be obtained by extrapolation (Figure 7.11). It turns out that σB correlates well with the firmness (often called hardness) of the fats as determined by penetrometry, extrusion or wire-cutting tests. The firmness depends on the SFC, but the shape and steepness of the curve varies depending on the extent of sintering, the homogeneity of the network, and crystal size and shape. In practice, firmness tends to increase considerably when the temperature is decreased. This is because of an increase in solid fat content and an increase of sintering. Many fats increase in firmness during storage and the rate of increase tends to be faster at a higher storage temperature. The cause appears to be further sintering. Recrystallisation involving transition to the β-polymorph tends to lead to a decrease in firmness, often to a considerable extent. This is because β-crystals tend to be rather isometric and not very prone to sintering. The latter may be because hardly any compound crystal formation can occur in the β-polymorph. Working can decrease the firmness of a plastic fat and this is called work softening. The extent of work softening can vary widely: a relative decrease in firmness of between 10 and 75% can be observed, depending on the type of fat and the intensity of the working. Considerable setting, i.e., increase in firmness, may occur after working. Initially, setting will be caused by aggregation due to van der Waals’ attraction of the crystals and crystal network formation, but in a longer timescale, setting is presumed to be due to recrystallisation. The above considerations
100
2
2
1
2
3
4
50 1
0
0.002
0.004
0
0.2
0.4
0.6
Strain
FIGURE 7.10
Stress–strain curve.
stress. The deformation is fully reversible, which implies that no bonds between crystals are broken. For larger strains, the behaviour becomes nonlinear (zone 2) and it is increasingly viscoelastic. Although most of the deformation is still reversible, irreversibility increases with increasing strain. This means that an increasing number of bonds are broken. Presumably, van der Waals’ bonds are broken and re-form, whereas fairly weak sintered bonds are also broken, but do not re-form, at least not within a short time span. It should be realized that the inhomogeneity of the system leads to a wide variation in bond strength. Zone 3 is the region of stress overshoot, a phenomenon that typically leads to yielding of the material. Yielding is the transition from viscoelastic to viscous behaviour. It only occurs in materials composed of a solid network interspersed with a continuous liquid phase. In local planes, all of the bonds in the network are broken, but the “cracks” are immediately filled with liquid and the system remains continuous. In most systems exhibiting yielding, removal of the stress quickly leads to the partial recovery of the elastic properties. This means that significant interaction forces must remain between the solid structural elements. The stress needed to cause yielding of the fat is especially important since cutting, spreading, and shaping all involve yielding. σB/kPa 1
100
dε/dt/s–1
Working 0.5
50
0 50
FIGURE 7.11
σB
0 100
0
20
0
4
σ/kPa
Solid fat/%
Time/days
(a)
(b)
(c)
Bingham stress and firmness.
488
8
Physical Properties: Structural and Physical Characteristics
× × × × × ×× × × ×
80 200 60 160 40 120 20
× × × ×
0 80
×
–20 40 –40 –60 0
×
×
× ×
× ××
×
×
×× × ×
240
×
200
× ×
×
× ×
× ×
β-form β'1- form α-form α-forms β'-forms × β-forms
× ×
160
×
×
×
×
×
× × ×
120 ×
80
×
× ×
×
40 ×
×
α-forms β'-forms β-forms
0 6 8
10 10 12
FIGURE 7.12 a/b
22 14 18 26 14 16 in 18fatty20acid22 24 Carbons chain Fatty acid chain length
26
30 28
6
34 30
10
14
18
22
26
30
34
Carbons in fatty acid chain
Melting properties of fatty acids.
demonstrate that predicting the yield properties of a plastic fat is very difficult but that nevertheless some trends can be explained.
7.2.9
×
×
×
×
×
∆Hr (KJ/mol)
M.p. (°C) ∆H r (KJ/mol)
100 240
begin to soften, slip, and drop prior to reaching the final clear point, so the temperatures reported for these measurements are lower than the actual clear point of the fat (Kaylegian and Lindsay, 1995). The capillary tube melting point, sometimes referred to as the clear point (AOCS Official Method Cc 1-25, latest revision. 1997) corresponds to the melting point as ordinarily determined by an organic chemist. It is defined as the temperature at which the sample becomes completely clear and liquid when heated (at a rate of 0.5°C/min) in a capillary tube of 1 mm internal diameter. The filled tubes must be quickly chilled and held at 4 to 10°C overnight before the determination is made. This means that for some fats this method will give the melting point of an unstable polymorph. The softening point or open tube melting point (AOCS Official Method Cc 3-25, latest revision 1997, or rather the ISO Equivalent Cc 3b-92 revised in 2002) should be determined exactly as the capillary tube melting point, except that the lower end of the capillary tube is left unsealed. It is defined as the temperature at which the fat becomes sufficiently softened for the fat column to rise in the tube. This method is only satisfactory for relatively hard fats. The slip point (AOCS Official Method Cc 4-25, revised in 1989) determines the melting point of the sample in its finished state, unlike the preceding methods, in the course of which the material is melted and resolidified. Cylindrical brass rings (2 mm wall thickness, 9 to 11 mm inside diameter, 10 mm height) are filled by being forced into the sample and are then suspended in a bath with a saturated sodium chloride solution. The slip point is determined as the point at which the fat rises from the cylinder. The Wiley melting point (AOCS Official Method Cc 2-38, latest revision 1991) is quite commonly used in the U.S. as an indication of the temperature at which a fat becomes substantially liquid. Discs of the fat 3/8 inch in diameter and 1/8 inch thick are solidified and chilled in
Melting behaviour
The melting behaviour of saturated monoacid triglycerides depends largely on the fatty acid chain length. The longer the chains, the higher the melting point. Figure 7.12 shows the evolution of the melting point and the melting enthalpy of the different polymorphs of saturated monoacid triglycerides as a function of the fatty acid chain length. The melting points of the α-form rise smoothly as the chain length increases, while for the β-polymorph, the melting points rise in a zigzag pattern, whereby the triglycerides with even chain lengths exhibit higher melting points. The evolution of the β′-form melting point changes from an alternation between odd and even to a smooth increase as chain length increases. The melting enthalpies of monoacid saturated triglycerides (Figure 7.12b) can be described by parabolic equations with a maximum at a fatty acid chain length of 22 or more carbon atoms, depending on the polymorph. For unsaturated monoacid triglycerides, both the melting temperature and the melting enthalpy are lower than for the saturated monoacid triglycerides with the same fatty acid chain length. The melting temperature and enthalpy for cis-unsaturated monoacid triglycerides is lower than for the trans equivalent (Hagemann, 1988). Since natural fats always contain several different fatty acids and, thus, comprise an enormous number of different triglycerides, a natural fat has a melting range and not a sharp melting point. The melting point of a fat must be defined, for instance, as the temperature at which the fat becomes visually clear and free of crystals (clear point, final melting point). Other commonly employed measurements quantifying certain melting properties of a fat are dropping point, softening point, and slip point. A fat will 489
7.2
Crystallisation and melting
metal forms for 2 hours or more, after which the disc is suspended in an alcohol–water bath having the same density as the fat sample, and slowly heated. The Wiley melting point is defined as the temperature at which the fat disc becomes completely spherical. The dropping point (AOCS Official Method Cc 18-80, latest revision 2001) may be used as an alternate method. A sample cup with a drip hole is filled with the fat to be examined, crystallised in the freezer for 15 minutes and then heated at a constant rate. The dropping point is the temperature at which a drop of fat falls through the hole in the bottom of the cup. However, the melting point of a fat does not tell us everything we need to know. Cocoa butter and milk fat, for example, have about the same capillary tube melting point, but they have very different melting profiles. Cocoa butter consists largely of closely related, high melting triglycerides and, thus, exhibits a narrow melting range. Milk fat has a far wider compositional range. Consequently, its melting range spans about 80°C and its properties change more slowly as the temperature changes. These fats represent two extremes. Many animal carcass fats have a melting pattern more or less comparable to that of milk fat, while some vegetable fats, like coconut oil, behave more like cocoa butter (Walstra, 2003).
Enden, J.C. van den et al. (1978), A method for the determination of the solid phase content of fats using pulse Nuclear Magnetic Resonance, Fette Seifen Anstrichm., 80, 180–186. Enden, J.C. van den et al. (1982), Determination of the Solid Fat Content of hard confectionary butters, J. Am. Oil Chem. Soc., 59, 433–439. Evans, U.R. (1945), The law of expanding circles and spheres in relation to lateral growth of surface films and the grainsize of metals, Trans. Faraday Soc., 41, 365–374. Foubert, I. et al. (2002), Dynamic mathematical model of the crystallization kinetics of fats, Food Res. Int., 35, 945–956. Foubert, I. et al. (2005), Insight in model parameters by studying temperature influence on isothermal cocoa butter crystallisation, Eur. J. Lipid Sci. Technol., 107, 660–672. Foubert, I. et al. (2006), Modelling two-step isothermal fat crystallization, J. Food Eng., 75, 551–559. Garside, J. (1987), General principles of crystallization, in Food Structure and Behaviour, Blanshard, V.M.V. and Lillford, P., Eds., Academic Press, London, 35–49. Hagemann, J.W. (1988), Thermal behavior and polymorphism of acylglycerides, in Crystallization and Polymorphism of Fats and Fatty Acids, Garti, N. and Sato, K., Eds., Marcel Dekker, New York, 9–95. Hartel, R.W. (1992), Solid-liquid equilibrium in foods, in Physical Chemistry of Foods, Schwartzberg, H.G. and Hartel, R.W., Eds., Marcel Dekker, New York, 47–81. Johansson, D. (1994), Colloids in fats: the fat crystal as a functional particle, Ph.D. thesis, Lund University, Sweden. Kaylegian, K.E. and Lindsay, R.C. (1995), Handbook of Milkfat Fractionation Technology and Applications, AOCS Press, Champaign, IL. Kellens, M.J. (1991), Polymorphism of saturated monoacid triglycerides, Ph.D. thesis, K.U. Leuven, Belgium. Khanna, Y.P. and Taylor, T.J. (1988), Comments and recommendation on the use of the Avrami equation for physicochemical kinetics, Poly. Eng. Sci. 28, 1042–1045. Kloek, W. (1998), Mechanical properties of fats in their relation to their crystallisation, Ph.D. thesis, Landbouwuniversiteit Wageningen, The Netherlands. Krog, N. (2001), Crystallization properties and lyotropic phase behavior of food emulsifiers: relation to technical applications, in Crystallization Processes in Fats and Lipid Systems, Garti, N. and Sato, K., Eds., Marcel Dekker, New York, 505–526. Krog, N. and Sparso, F.V. (2004), Food emulsifiers: their chemical and physical properties, in Food Emulsions, Friberg, S.E., Larsson, K., and Sjöblom, J., Eds., Marcel Dekker, New York, 45–91. Lambelet, P. (1983), Comparison of NMR and DSC methods for determining solid fat content of fats, application to milk fat and its fractions, Food Sci. Technol. (Int.), 16, 90–95. Larsson, K. (1966), Classification of glyceride crystal forms, Acta Chem. Scand., 20, 2255–2260. Larsson, K. (1972), Molecular arrangement in glycerides, Fette Seifen Anstrichm., 74, 136–142. Loisel, C. (1996), Physico-chimie du chocolat: crystallisation du beurre de cacao et propriétés structurales, Ph.D. thesis, École Nationale Supérieure des Industies Agricoles et Alimentaires, Paris.
References Aquilano, D. and Sgualdino, G. (2001), Fundamental aspects of equilibrium and crystallization kinetics, in Crystallization Processes in Fats and Lipid Systems, Garti, N. and Sato, K., Eds., Marcel Dekker, Inc., New York, Basel, 1–51. Avrami, M. (1939), Kinetics of phase change. I. General theory, J. Chem. Phys., 7, 1103–1112. Avrami, M. (1940), Kinetics of phase change. II. Transformationtime relations for random distribution of nuclei, J. Chem. Phys., 8, 212–224. Baur, F.J. et al. (1949), The polymorphism of saturated 1,3diglycerides, J. Am. Chem. Soc., 71, 3363–3366. Bennema, P. et al. (2001), Morphological connected net-roughening transition theory: Application to β crystals of triacylglycerols, in Crystallization Processes in Fats and Lipid Systems, Garti, N. and Sato, K., Eds., Marcel Dekker, Inc., New York, 99–150. Boistelle, R. (1988), Fundamentals of nucleation and crystal growth, in Crystallization and Polymorphism of Fats and Fatty Acids, Garti, N. and Sato, K., Eds., Marcel Dekker, New York, 189–276. Di, L. and Small, D.M. (1993), Physical behavior of the mixed chain diacylglycerol, 1-stearoyl-2-oleoyl-sn-glycerol: difficulties in chain packing produce marked polymorphism, J. Lipid Res., 34, 1611–1623. Di, L. and Small, D.M. (1995), Physical behavior of the hydrophobic care of membranes: properties of 1-stearoyl-2linoleoyl-sn-glycerol, Biochemistry, 34, 16672–16677. Duynhoven, J. van et al. (1999), Solid fat determination by NMR, inform, 10, 479–484.
490
Physical Properties: Structural and Physical Characteristics
Loisel, C. (1998), Phase transitions and polymorphism of cocoa butter, J. Am. Oil Chem. Soc., 75, 425–439. Long, Y. et al. (1995), Kinetics of polymer crystallisation, Prog. Polymer Sci., 20, 651–701. Malkin, T. (1954), Progress in the Chemistry of Fats and other Lipids, Vol. II, Pergamon Press, Oxford, 1–50. Marangoni, A.G. (1998), On the use and misuse of the Avrami equation in characterization of the kinetics of fat crystallization, J. Am. Oil Chem. Soc., 75, 1465–1467. Marangoni, A.G. et al. (2000), Comment on the use of direct pulsed nuclear magnetic resonance solid fat content measurements in phase behavior studies of lipid mixtures, J. Am. Oil Chem. Soc., 77, 565–567. Mazzanti, G. et al. (2005), Modelling phase transitions during the crystallization of a multicomponent fat under shear, Phys. Rev. E, 71, 041607-1 – 041607-12. Mulder, H. and Walstra, P. (1974), The Milk Fat Globule, Emulsion Science as Applied to Milk Products and Comparable Foods, Pudoc, Wageningen and CAB, Farnham Royal. Narine, S.S. and Marangoni, A.G. (1999), Relating structure of fat crystal networks to mechanical properties: A review, Food Res. Int., 32, 227–248. Nawar, W.W. (1996), Lipids, in Food Chemistry, Fennema, O.R., Ed., Marcel Dekker, New York, 239–319. Ng, W.L. (1975), Thermal decomposition in the solid state, Aust. J. Chem., 28, 1169–1178. Povey, M.J.W. (1995), Ultrasound studies of shelf-life and crystallization, in New Physico-Chemical Techniques for the Characterization of Complex Food Systems, Dickenson, E., Ed., Chapman & Hall, London, 196–212. Rousset, Ph. (1997), Étude expérimentale et modélisation de la crystallisation de triacylglycérols et du beurre de cacao, Ph.D. thesis, École Polytechnique Fédérale de Lausanne, Switzerland. Sato, K. and Kuroda, T. (1987), Kinetics of melt crystallization and transformation of tripalmitin polymorphs, J. Am. Oil Chem. Soc., 64, 124–127. Sato, K. et al. (1999), Molecular interactions and kinetic properties of fats, Prog. Lipid Res., 38, 91–116. Shannon, R.J. et al. (1992), The polymorphism of diglycerides, J. Sci. Food Agric., 60, 405–417. Shellhammer, T.H. et al. (1997), Viscoelastic properties of edible lipids, J. Food Eng., 33, 305–320. Smith, P.R. (2005), Healthy high-quality fats from a food science perspective, Lipid Techn., 17, 177–181. Supaphol, P. and Spruiell, J.E. (2001), Isothermal melt- and coldcrystallisation kinetics and subsequent melting behavior in syndiotactic polypropylene: a differential scanning calorimetry study, Polymer, 42, 699–712. Takahashi, H. et al. (1999), Simultaneous x-ray diffraction and calorimetric study of metastable-tostable solid phase transformation of 1,2-dipalmitoyl-sn-glycerol, Biophys. Chem., 77, 173. Timms, R.E. (1984), Phase behaviour of fats and their mixtures, Prog. Lipid Res., 23, 1–38. Timms, R.E. (1997), Fractionation, in Lipid Technologies and Applications, Gunstone, F.D. and Padley, F.B., Eds., Marcel Dekker, New York, 199–222. Timms, R.E. (2003), Confectionary Fats Handbook. Properties, Production and Application, The Oily Press, Bridgewater, U.K.. Todt, H. et al. (2001), Quality control with time-domain NMR, Eur. J. Lipid Sci. Technol., 103, 835–840.
Vanhoutte, B. (2002), Milk fat crystallisation: fractionation and texturisation, Ph.D. thesis, Universiteit Gent, Ghent, Belgium. Walker, R.C. and Bosin, W.A. (1971), Comparison of SFI, DSC and NMR-methods for determining solid-liquid ratios in fats, J. Am. Oil Chem. Soc., 48, 50–53. Walstra, P. (1987), Fat crystallization, in Food Structure and Behaviour, Blanshard, V.M.V. and Lillford, P., Eds., Academic Press, London, 67–85. Walstra, P. (1998), Secondary nucleation in triglyceride crystallization, Progr. Colloid Polym. Sci., 108, 4–8. Walstra, P. (2003), Physical Chemistry of Foods, Marcel Dekker, New York. Walstra, P. and Beresteyn, E.C.H. van (1975), Crystallization of milk fat in the emulsified state, Neth. Milk Dairy J., 29, 35–65. Walstra, P. et al. (2001), Fat crystal networks, in Crystallization Processes in Fats and Lipid Systems, Garti, N. and Sato, K., Eds., Marcel Dekker, New York, 289–328. Zwietering, M.H. et al. (1990), Modelling of the bacterial growth curve, Appl. Env. Microbiol., 56, 1875–1881.
7.3
Phase behaviour
7.3.1
General principles
The physical properties (e.g., macroscopic and mechanical properties) of a fat are determined by its phase behaviour and this concept was first explained clearly by Mulder (1953). A natural fat is a mixture of many triglycerides and although each triglyceride has its own polymorphism and melting behaviour, natural fats cannot be considered in terms of their individual component triglycerides, but only in terms of their different phases (Timms, 2003). The phase behaviour of triglyceride mixtures has critical implications in fat blending and separation of component triglycerides from natural fats and oil resources as during fractionation processes. Therefore, the complex melting, crystallisation, and transformation behaviour of natural fats must be elucidated by examining the mixing behaviour of binary, ternary, or more multiple phases of specific triglyceride components (Sato, 2001b). 7.3.1.1
The concept of a phase
A phase is a state of matter (e.g., a triglyceride mixture) that is homogeneous and separated from another phase by a definite physical boundary. A phase can be defined in full by its composition, temperature, and pressure. In food products, the pressure can be ignored for most practical purposes. Natural fats always contain at least two phases: a liquid and a solid phase. There is usually only one liquid phase, but several solid phases may be present at the same time. Different solid phases in fats are usually only distinguishable under a microscope, although their presence may affect macroscopic properties (Timms, 2003). In the liquid state, the miscibility of triglycerides is almost “ideal”, i.e., no heat or volume changes occur on mixing and the ideal or Hildebrand solubility equation applies (Hannewijk et al., 1964; Knoester et al., 1972). 491
7.3 Phase behaviour
Measurable deviations from ideal behaviour occur only when the triglycerides differ appreciably in molecular weight (and, hence, also volume) (Timms, 1978; 2003). Wesdorp (1990) confirmed this ideal behaviour for triglycerides that do not deviate from the average carbon number of the fat mixture by more than ten carbon numbers. In practice, the triglycerides in real fats do mix in the solid state to form solid solutions (also called mixed crystals). A solid solution is an intimate mixture of two or more components in the solid state such that neither component can be easily distinguished (Timms, 2003). 7.3.1.2
Liquid (L)
Liquid (L) Eutectic
L+S
L + SB
L + SA
SB SA + SB
Solid solution (S)
A
The concept of a phase diagram
Since (at constant pressure) a phase is defined by its temperature and composition, a diagram with the temperature along one axis and the composition along the other is sufficient to define all the phases. Such a diagram is called a phase diagram and describes the phase behaviour of a mixture of triglycerides (usually at equilibrium) (Timms, 1984; 2003). One of the main purposes of drawing up phase diagrams is to obtain information about the solid phase behaviour, especially with respect to the miscibility of the components in the solid state. To this end, the mixing behaviour in the liquid state must be known. It is generally accepted that the liquid of a triglyceride mixed system may be treated as a close approximation to an ideal mixture. For such systems the equilibrium solid phase properties can be derived from the phase diagrams (Knoester and Dejong, 1975). A considerable number of phase diagrams of glyceride mixtures have been reported in literature. Rossell (1967) compiled a survey of the available data, and this is still very useful today. Although most practical fat systems comprise mixtures of at least 10 major triglycerides, most studies have investigated the phase behaviour of binary mixtures. For binary mixtures of triglycerides (A and B), five types of phase diagrams have been observed as shown in Figure 7.13 (Timms, 1984; 2003).
B
A
B
(a)
(b)
Liquid (L)
Liquid (L)
L + SB SB SA + M
A + SB A
B
A
M + SB
M
(c)
B
(d)
Liquid
L – SA
L + SB
SB SA
A
SA – SB
B
FIGURE 7.13 Phase diagrams of binary mixtures.
• In diagram (a) of Figure 7.13, the triglycerides A and B have similar properties (melting point, molecular volume, and polymorph) and mix to form a continuous solid solution. Noteworthy mixtures of this type are POSt/StOSt and StStSt/StStE (P stands for palmitic acid, O for oleic acid, St for stearic acid, and E for elaidic acid). • In diagram (b), triglycerides A and B are less similar and the solubility of one in the other is limited, leading to a mixture of solid solutions and a sharp dip and interruption to the liquidus line at the eutectic point. The eutectic system is the most common, noteworthy examples being PPP/StStSt, POSt/POP. Eutectic systems tend to occur when the components differ in molecular volume, shape, or polymorph, but not greatly in melting point. Diagrams (a) and (b) are the two types commonly found in many binary mixtures of triglycerides.
• In diagram (c) of Figure 7.13, the eutectic system tends to shift to a monotectic system, as differences in the melting points of the component triglycerides increase. For example the LLL/PPP (with L standing for linoleic acid) system (melting point difference = 20°C) shows a slight eutectic at 3% PPP, but the LLL/StStSt system (melting point difference = 27°C) shows either a monotectic or a eutectic at 1% StStSt. Similarly, the PPP/POP system (melting point difference = 28°C) shows a monotectic, whereas the PPP/StOSt system (melting point difference = 22°C) shows a slight eutectic at 5.5% PPP. In all such cases, the solid high melting component dissolves a substantial quantity, usually 20 to 30%, of the low melting component. • In diagram (d), A and B combine to form a special mixture called a molecular compound (M) that behaves like a new, pure triglyceride with unique 492
Physical Properties: Structural and Physical Characteristics
real fat, a multicomponent phase diagram with extra axes for each component triglyceride above two is necessary. An individual fat may be considered as a composition point in this multicomponent phase diagram. Small compositional changes in the diagram would reflect the natural variations in the properties of the fat; larger compositional changes would indicate a move to a completely different fat (Timms, 2003). It is not possible to depict such multicomponent phase diagrams, even if the data were available. Figure 7.14 shows the three types of phase diagrams that Timms (1984) deduced to be possible for binary mixtures of fats. Fp(AB), Fq(CD), Fs(CD), and Fr(CD) are to be considered as different types of fat formed by the mixture of just two triglycerides, A and B, or C and D. Essential differences between mixtures of fats and of pure component triglycerides are as follows:
properties differing from those of its component triglycerides. Hence, the diagram resembles two eutectic phase diagrams of the (b) type placed side by side. In the system StOSt/StStO, there is clear evidence for the formation of a compound. This system is then really two eutectic systems, StOSt/M and M/StStO, side by side. • Peritectic systems (e) have been observed to occur only in mixed saturated/unsaturated systems where at least one triglyceride has two unsaturated acids. Noteworthy examples are StOSt/StOO and POP/POO. Sato et al. performed major research on the phase behaviour and the thermodynamic and kinetic phase properties of binary mixtures of pure component triglycerides (Minato et al., 1996; Takeuchi et al., 2002; Takeuchi et al., 2003). The same group wrote extensive reviews on this matter (Sato et al., 1999; Sato and Ueno, 2001; Sato, 2001a). Lee (1978) presented a mathematical approach, which can simulate phase diagrams for binary mixtures of lipid molecules showing close agreement with experimental data and using a single parameter to describe the nonideal mixing in each phase. An index of lipid phase diagrams was composed by Koynova and Caffrey (2002).
7.3.2
• There is no unique temperature (the melting point) for a real fat at 0 or 100% composition, where liquidus and solidus or solindex lines meet. • There is usually no precise eutectic point for mixtures of real fats, although there is often a eutectic minimum in the liquidus curve and, therefore, in the melting point of Clear Point curves. • When the two components are not pure triglycerides, tie lines, i.e., lines drawn horizontallyin a phase diagram, do not define the ratio of solid to liquid phase present in between the solidus and liquidus lines.
Phase behaviour of real fats
Binary phase diagrams can display all the properties of a two-component mixture. To display the properties of a 50
Liquid (L)
50
Liquid (L)
45
45 L + S 1 + S2
L+S
40
S1
Temperature (°C)
35 Solid solution (S)
30
Fp(AB) 25
40 S2 35
S1 + S 2
30 Fq(CD)
Fs(CD) 25
Fp(AB)
(a)
(b)
20
20 Liquid (L)
15
15
10
10 L+S
5
5 0
0 0
10
20
Solid 30 solution 40 (S50 3)
S360 + Fr(CD) 70
80
90
100
Fully-hardened milk fat in CB/FHMF blend (%) Fp(AB) Fr(CD) (c)
FIGURE 7.14 Phase diagrams of fats. (Adapted from Timms, R.E. (2003), Confectionary Fats Handbook, Properties, Production and Application, The Oily Press, Bridgwater, U.K.)
493
Temperature (°C)
7.3 Phase behaviour
50
50
45
45
40
40
35
35
30
30
25
25
20
20
15
15
10
10
5
5
0
0 0
10
20
30
40
50
60
70
80
90
100
Fully-hardened milk fat in CB/FHMF blend (%)
FIGURE 7.15 Isosolid diagram for blends of cocoa butter (CB) and a fully hardened milk fat (FHMF). Isosolid lines are plotted at 5, 10, 20, 30, 40, 60, and 80% solids from the top downwards. (Adapted from Timms, R.E. (1979), Chem. Ind., 257–258.)
similarity between the components leads to a greater interaction or an increased tendency to form a solid solution. Knoester and De Jong (1975) determined deviations from ideality for several systems of mono- or diglycerides with triglycerides by measuring solubility temperatures and heats of mixing. The nonideal phase behaviour of glyceride systems reported in literature was predicted satisfactorily.
Moreover, phase diagrams essentially describe fat systems at equilibrium. They usually depict fats and their mixtures in their most stable polymorphs. In practice it takes time to reach equilibrium in the solid state, so that an individual phase diagram may be considered as a snapshot of the situation at a given time. Additionally, the phase behaviour of two fats may be different for different polymorphs (Kellens and Reynaers, 1992). Because the tie lines on a phase diagram of mixtures of two fats do not define the ratio of solid-to-liquid phases, the phase diagram alone is insufficient for practical purposes. Therefore, isosolid diagrams (Figure 7.15) should also be drawn up on the basis of solid fat contents (SFCs) to be determined by means of pulse NMR (Timms, 1979). Paulicka (1970; 1973) was the first to describe the construction and usefulness of phase diagrams of mixtures of cocoa butter and confectionery fats. A more upto-date typical step-by-step procedure for constructing phase and isosolid diagrams of fats has been outlined by (Timms, 2003).
7.3.3
References Hannewijk, J. et al. (1964), Dilatometry of fats, in Analysis and Characterisation of Oils, Fats and Fat Products, Vol. 1, Boekenoogen, H.A., Ed., Interscience, London, 119–182. Kellens, M.J. and Reynaers, H. (1992), Study of the polymorphism of saturated monoacid triglycerides II: Polymorphic behaviour of a 50/50 mixture ot tripalmitin and tristearin, Fat Sci. Technol., 94, 286–293. Knoester, M. et al. (1972), The solid-liquid equilibrium of binary mixtures of triglycerides with palmitic and stearic chains, Chem. Phys. Lipids, 9, 309–319. Knoester, M. and Dejong, D.J. (1975), Non-ideal phase behaviour of binary mixtures of triglycerides with mono-and diglycerides, Chem. Phys. Lipids, 14, 41–48. Koynova, R. and Caffrey, M. (2002), An index of lipid phase diagrams, Chem. Phys. Lipids, 115, 107–219. Lee, A.G. (1978), Calculation of phase diagrams for non-ideal mixtures of lipids, and a possible non-random distribution of lipids in lipid mixtures in the liquid crystalline phase, Biochim. Biophys. Acta, 507, 433–443. Lutton, E.S. and Jackson, F.L. (1966), Binary systems with monoglycerides, J. Am. Oil Chem. Soc., 43, 357–358.
Phase behaviour of mixtures of triglycerides and partial glycerides
Few studies of the phase behaviour of mixtures of triglycerides and partial glycerides have been reported in literature. Lutton and Jackson (1966) studied a number of binary systems, each involving a 1-monoglyceride and mono-, di-, and triglycerides. A general principle running through the behaviour observed is that a greater chemical
494
Physical Properties: Structural and Physical Characteristics
of the hydrocarbon chains. A few years later, the structure of the most common liquid crystalline phases was revealed by Luzzati et al. (1960). To obtain liquid crystalline phases, amphiphilic compounds should be present. These compounds exhibit both hydrophilic (polar) and hydrophobic (nonpolar) groups within the molecule (Larsson and Lundström, 1976). They orient on contact with polar solvent molecules, giving rise to polar and nonpolar regions that are separated by the polar end groups. The nonpolar hydrocarbon region may melt before the final melting point is reached. This occurs because the van der Waals’ forces among hydrocarbon chains are weaker than the hydrogen bonding between polar groups (Nawar, 1996). The term lyotropic mesomorphism is used to describe the formation of liquid crystalline systems by the penetration of a solvent in between the amphiphilic molecules. In contrast, liquid crystals formed without the aid of a solvent are said to be thermotropic. So far, lyotropic mesomorphism has been observed almost exclusively in lipid systems containing water. Lipids showing lyotropic mesomorphism frequently form thermotropic mesophases without any additions at high temperatures.
Minato, A. et al. (1996), Synchrotron radiation x-ray diffraction study on phase behavior of PPP-POP binary mixtures, J. Am. Oil Chem. Soc., 73, 1567–1572. Mulder, H. (1953), Melting and solidification of milk fat, Neth. Milk Dairy J., 7, 149–174. Paulicka, F.R. (1970), Phase behavior of fats in confectionary coatings, The Manufacturing Confectioner, 50, 73–74,76, 78. Paulicka, F.R. (1973), Phase behaviour of cocoa butter extenders, Chem. Ind., 17, 835–839. Rossell, J.B. (1967), Phase diagrams of triglyceride systems, Adv. Lipid Res., 5, 353–408. Sato, K. (2001a), Crystallization behaviour of fats and lipids — a review, Chem. Eng. Sci., 56, 2255–2265. Sato, K. (2001b), Uncovering the structures of βⴕ fat crystals: what do the molecules tell us? Lipid Technol., 13, 36–40. Sato, K. et al. (1999), Molecular interactions and kinetic properties of fats, Prog. Lipid Res., 38, 91–116. Sato, K. and Ueno, S. (2001), Molecular interactions and phase behavior of polymorphic fats, in Crystallization Processes in Fats and Lipid Systems, Garti, N. and Sato, K., Eds., Marcel Dekker, New York, 177–209. Takeuchi, M. et al. (2002), Crystallization kinetics of polymorphic forms of a molecular compound constructed by SOS (1,3-distearoyl-2-oleoyl-sn-glycerol) and SSO (1,2-distearoyl-3-oleoyl-sn-glycerol), Food Res. Int., 35, 919–926. Takeuchi, M. et al. (2003), Synchroton SAXS/WAXS study of polymorph-dependent phase behaviour of binary mixtures of saturated monoacid triacylglycerol, Cryst. Growth Des., 3, 373–374. Timms, R.E. (1978), The solubility of milk fat, fully hardened milk fat and milk fat hard fraction in liquid oils, Aust. J. Dairy Tech, 33, 130–135. Timms, R.E. (1979), Computer program to construct isosolid diagrams for fat blends, Chem. Ind., 257–258. Timms, R.E. (1984), Phase behaviour of fats and their mixtures, Prog. Lipid Res., 23, 1–38. Timms, R.E. (2003), Confectionary Fats Handbook. Properties, Production and Application, The Oily Press, Bridgwater, U.K. Wesdorp, L.H. (1990), Liquid-multiple solid phase equilibria in fats-theory and experiments, Ph.D. thesis, Technical University Delft, The Netherlands.
7.4
Lipid/water interactions
7.4.1
Introduction
7.4.2 7.4.2.1
General aspects Liquid crystalline phases
The liquid crystalline structure is determined primarily by the polar groups, while the nonpolar parts of the molecules form liquid-like associations (Söderberg and Ljusberg-Wahren, 1990). The water in such systems cannot be treated as a continuous free medium, but should be considered as part of the structural units. The most common mesomorphic structures are lamellar (termed “neat” in old literature), hexagonal (termed “middle” in old literature), and cubic phases. The lamellar liquid crystal structure (Lα) corresponds to that existing in biological bilayer membranes and is illustrated in Figure 7.16. It is made up of double layers of lipid molecules separated by water layers (Nawar, 1996). The polar groups are on the outside of the layers in contact with the water. Structures of this kind are usually less viscous and less transparent than the other mesomorphic structures. The capacity of a lamellar phase to retain water
Crystalline lipids are packed in a highly ordered, repeating pattern extending in all three spatial dimensions. In the liquid state, the molecules acquire freedom of movement and assume a state of disorder. Phases with intermediate properties between those of the crystalline and the liquid states are also known to occur. These so-called mesomorphic phases consist of liquid crystals (Nawar, 1996). A milestone in the elucidation of the liquid crystalline structures was the introduction of the liquid chain concept by Chapman (1958). By using IR spectroscopy he showed that a high temperature phase transition in anhydrous soap was caused by the melting
H2O
FIGURE 7.16 The lamellar phase consists of double layers of lipid molecules separated by water layers. (From Larsson, K. (1986), in The Lipid Handbook, 1st ed., Gunstone, F.D., Padley, F.B., Harwood, J.L., Eds., Chapman & Hall, London, 321–384.)
495
7.4 Lipid/water interactions
called viscous isotropic (V). A unique property is the presence of two media, one polar and one nonpolar, both continuous throughout the structure (Luzzati et al., 1968). The lattice is body-centred and the structure consists of networks of rods; in some cases, the rods are filled by the hydrocarbon chains (VI), in others by the polar moieties (VII). This rod system was considered to be the general structure for the cubic lipid-water phases (Figure 7.18). Cubic liquid crystals are stable in an excess of water. Lindblom et al. (1979) indicated that the cubic monoglyceride–water phase was closely related to the lamellar phase. The structure they proposed consisted of hexagonal lamellar bilayer units. This cubic structure is related to “Schwarz’s primitive cubic minimal surface” and consists of a three-dimensional continuous bilayer system separating two separate water channel systems (Figure 7.19).
depends on the nature of the lipid molecules. If the water content is increased above the swelling limit of the lamellar phase, a dispersion of spherical or cylindrical aggregates consisting of concentric, alternating layers of amphiphilic molecules and water gradually forms. This phase is called the lamellar dispersion (Krog and Larsson, 1968). The mechanism behind this change in structure is that the lamellar phase cannot exist in equilibrium with water, as there would be direct contact between water molecules in different states: bulk water and water that forms part of the lipid bilayers. By the formation of closed vesicles preserving the lamellar structure, such contacts are avoided (Larsson and Krog, 1973). In the hexagonal phase, the amphiphilic molecules form infinite cylinders arranged in a hexagonal array (Nawar, 1996) as shown in Figure 7.17a. The liquid hydrocarbon chains fill the interior of the cylinders, and the space between the cylinders is taken up by water. This type of liquid crystal is referred to as HI. In the reversed hexagonal structure (HII), water fills the interior of the cylinders and is surrounded by the polar groups of the amphiphile lipid (Figure 7.17b). The hydrocarbon chains make up the continuous phase between the cylinders. However, Clunie et al. (1965) concluded from calculations based on x-ray and density data that the average diameter of the cylinders is smaller than would be expected with this structure. They suggested that the hexagonal phase could be made up of linear chains of spherical micelles instead of cylinders. When the HI-phase is diluted with water, spherical micelles form. In contrast, dilution of the HII-phase with water is not possible. The phases described so far are two-dimensional mesophase structures. Three-dimensional phases of various types are also known and most of them have some form of cubic symmetry (Lis and Quinn, 1991). Because of its three-dimensional periodicity, such a phase is not a true liquid crystal, but all physical properties are closely related to those of the lamellar and hexagonal mesomorphic phases. The cubic phase is very viscous and could not be characterized by ordinary optical observations because it is isotropic. Hence, this phase is sometimes
(a)
7.4.2.2
Characterization methods
Polarisation microscopy is the most frequently used method permitting a definite distinction between lamellar and hexagonal phases (Morley and Tiddy, 1993). Both phases show their typical anisotropy when observed through crossed polarizers (D’Antona et al., 2000). The liquid crystalline phases can be identified unambiguously from their x-ray diffraction patterns (Luzzati et al., 1960). The arrangement of the hydrocarbon chains corresponds to that in a liquid hydrocarbon as shown by a weak and relatively diffuse x-ray diffraction band at wide angles corresponding to a Bragg spacing of about 4.5 Å (Kelker et al., 1980). At small angles, sharp diffraction bands are observed, which are characteristic of the mesomorphic phase. In the lamellar phase several orders are observed giving rise to calculated Bragg spacings in the ratio of 1:2:3:4. From these data, it is possible to determine the bilayer and water layer thickness and the cross section per polar head group (Larsson and Krog, 1973). A series of Bragg spacings in the ratio of 1: 3: 4: 7 is characteristic for the hexagonal phase, indicating a twodimensional hexagonal array of parallel cylindrical micelles. In the same way the diameter of the cylinders in
(b)
FIGURE 7.17 The hexagonal phase consists of cylinders of liquid hydrocarbon chains (a) or water (b) arranged in a hexagonal array. (From Larsson, K. (1986), in The Lipid Handbook, 1st ed., Gunstone, F.D., Padley, F.B., Harwood, J.L., Eds., Chapman & Hall, London, 321–384.)
FIGURE 7.18 The general cubic lipid-water system. (Adapted from Luzzati, V., et al., (1968), Nature, 220, 485–488.)
496
Physical Properties: Structural and Physical Characteristics
FIGURE 7.19 5465–5470.)
The cubic monoglyceride-water phase. (Adapted from Lindblom, G., et al., (1979), J. Am. Oil Chem. Soc., 101,
the HI- and HII-phase can be determined as well as the area of the cross section per polar head group and the distance between adjacent cylinders. As it has often been noted, the analysis of the x-ray diagrams of the cubic phase is generally hindered by a heavily spotted appearance, a consequence of the easy growth of fairly large crystals (Luzzati et al., 1960).
7.4.3 7.4.3.1
Monoglyceride/water interactions Pure monoglycerides
Although monoglycerides are virtually insoluble in water, they swell to form different liquid crystalline phases (Kelker et al., 1980). The mesomorphic behaviour of monoglyceride/water systems is represented by a phase diagram. In a binary system, it is possible to define where in a single diagram the different liquid crystalline phases exist by using composition and temperature as the parameters. The lipid organisation in these systems depends on the chemical structure of the compounds and on environmental conditions, such as temperature, water, and salt content (Nawar, 1996). This provides an opportunity to shift the transition temperature and drive the system towards any desired mesophase (Chupin et al., 2001). In the presence of water and above the Krafft temperature, at which the hydrocarbon chains of 1-monoglycerides “melt” and adopt a disordered state, water penetrates among the ordered polar groups. 1-Monoglyceride/water systems form up to three classes of liquid crystalline phases, namely the lamellar (Lα), reversed hexagonal (HII), and cubic (VII) phase. Hyde et al. (1984) have shown that the cubic phase can consist of two forms with different geometries. Two separate HII-phases have also been reported (Larsson, 1988). At low water concentrations an isotropic fluid phase (L2) is formed (Larsson, 1979). It was suggested that this structure consisted of water lamellae separated by lipid bilayers (Figure 7.20).
FIGURE 7.20 Proposed structure of L2-phase in aqueous monoglyceride systems. (Adapted from Larsson, K. (1979), J. Colloid Interface Sci., 72, 152–153.)
At very short chain length (C6), an ordinary micellar solution is formed in the water-rich region of the phase diagram. No other mesophase is formed; crystals + water transforms directly into an isotropic fluid above the Krafft point (Krog and Larsson, 1968). The lamellar phase and dispersion are the only mesomorphic phases of 1-monocaprylin, 1-monocaprin, and 1-monolaurin. The influence of chain length and unsaturation of the fatty acids on the phase behaviour of 1-monoglycerides is shown in Figure 7.21. In 1-monomyristin, the cubic phase is formed when the lamellar phase is heated (Lutton, 1965). Similarly, the lamellar dispersion phase breaks into a cubic phase and water (Pezron et al., 1990). In 1-monopalmitin and 1-monostearin, the lamellar region has shrunk considerably 497
7.4 Lipid/water interactions
Monolaurin 100
Monolinolein 100
Fluid Isotropic Fluid Isotropic
80 60
60
40
Viscous Isotropic
Neat
Monomyristin 100
40
Monoolein V.I .
Fluid Isotropic
80
Middle
100 Fluid Isotropic
Neat
80
60
60
Viscous Isotropic
40
40
Neat
Monopalmitin
140 Temperature °C
80
Middle
Neat
120
Monoelaidin
Fluid Isotropic
Visc o
100
60
60 Neat
Monostearin
120
80
Viscous Isotropic
Crystals + H2O
140
100
sotr opic
Neat
80
Fluid Isotropic
us I
40
Monoerucin
Fluid Isotropic
100
Viscous Isotropic
100
Fluid Isotropic
80
Neat
80
60
60
40
Neat
Monoarachidin
140 120
Middle
Fluid Isotropic
100
Monobehenin Fluid Isotropic
120
Middle
100
V.I.
Viscous Isotropic
80
140
Neat
Neat
80
60
60 0
FIGURE 7.21
10
20
30
40
100 0 % H2O
10
20
30
40
100
Aqueous 1-monoglyceride systems. (Adapted from Lutton, E.S. (1965), J. Am. Oil Chem. Soc., 42, 1068–1070.)
region. The strong cubic and hexagonal phase forming tendency shown with increasing chain length and with unsaturation is not the case for 1-monoerucin. Neither the cubic nor the hexagonal phase was observed. A maximum lamellar phase temperature of 45°C, and a liquid immiscibility line for melted monoerucin and water running between 15 and 20% water characterize this phase diagram. The polar head group of 2-monoglycerides requires a larger cross section area than that of the 1-isomer, which results in different phase properties (Larsson, 1986).
and the cubic phase has grown. In 1-monoarachidin, the cubic phase transforms into the hexagonal structure with closed water cylinders on further heating. The long hydrocarbon tails favour the formation of the HII phase. This is evidenced by the fact that the addition of nonpolar hydrocarbon-type matter to 1-monostearin can produce a hexagonal phase. The phase diagrams for systems of unsaturated 1-monoglycerides resemble those for systems of saturated compounds, but with the corresponding phase regions occurring at lower temperatures. Unlike the mesophases of saturated monoglycerides, the mesophases of unsaturated monoglycerides can exist at room temperature (Lutton, 1965). The diagram of trans-unsaturated 1-monoelaidin is much like that of 1-monostearin, but with the temperature scale shifted some 30°C downwards. 1-Monoolein shows a further drop in the lamellar region and an outcropping of the hexagonal phase near 100°C. 1-Monolinolein shows similar behaviour, but at slightly lower temperatures and with an expanded hexagonal
7.4.3.2
Distilled monoglycerides
Industrially produced distilled monoglycerides usually comprise a major component, although homologous members with longer and shorter chains are present (Krog and Larsson, 1968) as well as some diglycerides, free fatty acids, and glycerol (Brokaw and Lyman, 1958). Due to these differences in composition compared with pure monoglycerides, the phase diagram of distilled monoglycerides shows some specific features.
498
Physical Properties: Structural and Physical Characteristics
Temperature-programmed x-ray diffraction analysis has shown that when a mixture of distilled monoglycerides and water is heated a gel phase is formed a few degrees below the region of the lamellar region. In aqueous systems of pure 1-monoglycerides the transition of crystals + water into the lamellar phase always takes place directly, i.e., without formation of an intermediate gel phase (Krog and Borup, 1973). The structure of the gel phase is similar to that of the lamellar phase, with water layers alternating with lipid bilayers. The water penetrates into the crystalline lattice between the polar groups before the hydrocarbons are transformed into a liquid state. The exact temperature of gel formation depends on the chain length of the fatty acid and on the purity of the monoglyceride (Moonen and Bas, 2004). When water is added to the lamellar phase, it swells to a critical repeat distance. The swelling of the lamellar phase is limited by the long-range van der Waals’ forces between the lipid bilayers, which balance the osmotic pressure at an interlayer spacing of about 20Å (Larsson and Krog, 1973). Consequently, pure monoglycerides show limited swelling. At higher water concentration, a lamellar dispersion is formed. Distilled monoglycerides contain approximately 0.5 to 1.5% free fatty acids. When these fatty acids are neutral i s e d b y a d d i n g a b a s e, n e g at i v e l y c h a r g e d − groups ( −COO ) are formed on the surface of the lipid bilayer. Due to repulsion forces between these electric double layers, the swelling capacity of the lamellar system is greatly increased, in principle to infinite dilution. Neutralisation of the free fatty acids in the distilled monoglycerides, thus, expands the area of the lamellar phase into the higher water content region. The presence of salts in water decreases the swelling capacity of monoglycerides. The electric repulsion forces created by the ion-active surfactant in the bilayer are shielded, resulting in a decrease in the water layer thickness. This shielding effect supports the theory that the effect of ion-active surfactants in the bilayer is due to electric repulsion (Krog and Borup, 1973). In addition, the transition lamellar phase or dispersion to cubic or cubic + water phase is shifted to higher temperatures if the free fatty acids are neutralized in distilled monoglycerides. 7.4.3.3
Neat + Viscous Isotropic
100
Fluid Isotropic
90
Viscous Isotropic Viscous Isotropic + H2O
80 °C 70
Neat Dispersion
60 50 Gel or crystals + H2O
40
Gel + H2O or crystals + H2O
30 0
10
20
30
40
50
60
70
80
90
100
% H2O
FIGURE 7.22 The phase diagram of the monoglyceride/water system shows two lower boundaries of the lamellar phase. (Adapted from Krog, N. and Larsson, K. (1968), Chem. Phys. Lipids, 2, 129–143.)
lamellar phase, but the water is less ordered. This is caused by a reduced hydration of the gel phase (Morley and Tiddy, 1993). The occurrence of a single x-ray short spacing at 4,15Å shows that the lateral packing of the chains can be described as a hexagonal subcell. This indicates that the chains possess rotational or oscillational freedom (Krog and Larsson, 1968). The gel phase of monoglycerides is called the α-gel, as it is the hydrated form of the α-polymorph (Morley and Tiddy, 1993). The fatty acid chains are tilted about 54° towards the water layers in pure monoglycerides. In a neutralized, distilled monoglyceride–water system, the angle of tilt of the hydrocarbon chains towards the water layers is higher (Krog and Borup, 1973). The gel phase is stable just below the lamellar region and this stability can be related to a high molecular motion of the polar head groups (Cassin et al., 1998). This is indicated by a small region of α-gel stability just below the lamellar phase in the phase diagram on Figure 7.22. During extended holding at lower temperature, the water is expelled and the gel phase transforms into a microcrystalline suspension in water, called a coagel. The coagel phase consists of a three-dimensional network of plate-like crystals entrapping water domains. The molecular arrangement within the coagel phase is identical to that found in the bulk β-crystal of monoglycerides. 1-Monoglycerides are a racemic mixture of the stereoisomers D and L; the middle carbon atom of the glycerol moiety is optically active. In the lamellar phase the molecules undergo many gauche-trans transitions and have high lateral mobility. The two stereoisomers of monoglycerides are intimately mixed and the bilayer halves will be racemic. When the α-gel phase is formed, the molecular mobility is drastically reduced, but the racemic state of the bilayer is retained. The molecules can no longer undergo gauche-trans transitions, although they can still
The gel and coagel phase
If the lamellar phase is cooled, a metastable gel is formed. In the phase diagram as shown in Figure 7.22, this is expressed by two lower boundaries of the lamellar phase. The upper one corresponds to transformation of β-crystals + water to the lamellar phase on heating, and the lower one corresponds to the transformation of the lamellar phase into the metastable gel on cooling (Krog and Larsson, 1968). The structure of the gel phase is still lamellar, but the hydrocarbon chains are in the crystalline state. The hydrocarbon chain length is longer in the gel phase than for the 499
7.4 Lipid/water interactions
rotate around their (long) axis. Through chiral discrimination the D and L isomers rearrange within bilayers, after which the transition from α- to β-crystal can start (Sein et al., 2002). Figure 7.23 shows a diagram of the molecular processes during the phase transitions from lamellar phase to α-gel to coagel. The melting enthalpy of α-crystals is about twice the value obtained for the melting of α-crystals. The α-gel and coagel phase both convert into the lamellar phase when the Krafft temperature is reached. However, when cooled down, the lamellar phase converts only to the α-gel phase. The difference in melting enthalpy of the first heating (α-gel and/or coagel phase to lamellar phase) and the enthalpy of the second heating (α-gel phase to lamellar phase) can be employed to monitor the kinetics of the gel to coagel phase transition (Cassin et al., 1998). To do so, a parameter called the coagel index (CI) is defined according to:
CI =
∆H (first heat ) ∆H (second heat )
7.4.4
Like monoglycerides, phospholipids are substances that form highly ordered liquid crystal mesophases in water (Bueschelberger, 2004). Phospholipids, which have a large polar group and two hydrocarbon chains, have phase properties that are closely related to those of monoglycerides because of packing constraints (Pezron et al., 1990). This illustrates that lipids that have hydrophobic and hydrophilic regions of similar relative sizes (expressed for example by the HLB value) give the same type of water interaction (Larsson and Krog, 1973). As with monoglycerides, the ability of phospholipids to disperse in water occurs only above the Krafft temperature, which is nearly constant over a wide range of concentration. Most phospholipids form a lamellar phase that shows a high degree of swelling (Chupin et al., 2001). Their swelling pressure is higher than that of monoglycerides, demonstrating the decisive role of the polar head group (Pezron et al., 1990). Highly unsaturated phospholipids swell at room temperature while completely saturated phospholipids require higher temperatures to do so (Kelker et al., 1980). In the absence of ionic groups, the lamellar phase is transformed into a reversed hexagonal phase as the water content increases (Larsson and Lundström, 1976). As more water enters the layers, the area per head group increases, forcing the tails to spread out over larger areas. At some point, the tails must become too spread out for a flat layer to be stable. In the reversed hexagonal phase, the chains will have more space than in the lamellar structure. The lamellar and HII-phase can definitely be recognized in, for example, the phosphatidylethanolamine/water system. Lysolecithin is obtained from lecithin by removing one fatty acid moiety. This compound does not form a lamellar phase (Kelker et al., 1980). Because its polar group is large in comparison with the hydrocarbon fatty acid chain, this molecule preferentially forms a hexagonal phase (HI). In this phase, the heads take up more area than the tails (Matthew and Finn, 1988). In addition to lyotropic mesomorphism, phospholipids also exhibit thermotropic mesomorphism. Liquid crystalline phases are formed many tens of degrees below the published capillary melting points. Pure egg-lecithin at 40°C is in a form that is at least partially crystalline, and it transforms into a wax-like phase at about 80°C. It is in a cubic phase between 88 and 109°C. At higher temperatures up to the melting point at 231°C, it is in a lamellar phase (Kelker et al., 1980). These mesophases also exist in mixtures with water in the ranges indicated in the phase diagram in Figure 7.24.
(7.50)
A CI of 1 indicates that the sample is still in the gel state, a value of 2 means that the sample is in the coagel state. Depending on composition, processing and temperature this transition may take minutes or many months (Heertje et al., 1998). It is not possible to prepare completely stable α-gels from monoglycerides. Surfactants that are not polymorphic and which are stable in the α-crystalline state, form α-gels with water that are very stable. Some esters of monoglycerides, for example, show this behaviour (Heertje et al., 1998). These compounds are called α-tending because of the absence of a closely packed β-crystalline phase. Consequently, there is no driving force for phase transition. The concentration of the added co-emulsifiers is critical as it may prevent the formation of the lamellar phase. Mesomorphic phases with reversed hexagonal structures are formed instead.
TTk Lα
Coagel
Phospholipid/water interactions
T>Tk
α–gel
7.4.5
FIGURE 7.23 Schematic representation of the molecular processes during the phase transitions from lamellar phase to α-gel to coagel. (Adapted from Sein, A. et al. (2002), J. Colloid Interf. Sci., 249, 412–422.)
Triglycerides-diglycerides/water interactions
Due to their insolubility in water, tri- and diglycerides can form aqueous mesophases only in the presence of other components. Appreciable amounts can be solubilized in 500
Physical Properties: Structural and Physical Characteristics
1Phase (liquid)
250
point), giving rise to micelles, but the solubility is low at just a few degrees below this temperature (McBain and Lee, 1943). Reiss-Husson and Luzzati (1969) revealed the structure of potassium soaps in the micellar region of the phase diagram using x-ray diffraction. A common feature in soaps of saturated fatty acids is that spherical micelles exist at low concentrations, while at increased concentrations a transition into rod micelles occurs. Sodium oleate, however, was found to give rod-shaped micelles at all concentrations. The short-chain soaps are more soluble than the highchain soaps at low temperatures and low soap concentrations while at higher temperatures and higher soap concentrations the opposite is true (Vold et al., 1941). The existence of the liquid crystalline phases mentioned in Section 7.4.2.1 in certain areas of concentration is also a common feature of aqueous soap solutions. Soaps of different chain lengths show a lamellar phase at low water content and at higher water content there is a large region where the HI-phase exists. With increasing water demand to hydrate K+, Na+, and Li+ ions, the minimum water content required for the existence of the different phases increases in the isotropic and the mesomorphous phase (Ekwall and Mandell, 1968). As an example, the phase diagram of sodium myristate is given in Figure 7.25. In spite of a substantially higher water content, the HI-phase is much stiffer than the Lα-phase (Kelker et al., 1980). Between
2Phases (2 liquids)
Neat
T (°C)
200
2Phases (liquid + liquid crystal)
150 V.I
Neat (lamellar)
100
50
Crystal + Tc liquid crystal
0
10
20
30
40
50
100
% Water
FIGURE 7.24 Phase diagram of egg lecithin/water system. (Adapted from Kelker, H. and Hatz., and with a contribution by Schuman, C. (1980), in Handbook of Liquid Crystals, Verlag Chemie, Weinheim, 512–591.)
monoglyceride-water mesophases (Kelker et al., 1980). Little information is available about the influence of water on the crystallisation of triglycerides, although anhydrous milk fat (AMF) in particular, but also other commercial fats, may contain some water as a minor component. The role of water and phospholipids in vegetable oils has been investigated by Sambuc et al. (1980). When 4% lecithin and 16% water were incorporated, a delay in the onset of crystallisation was noticed for all the fats, while the final solid: liquid ratio was not affected. Savage and Dimick (1995) suggested that water forms micelles or inverse hexagonal mesophases with amphiphilic compounds, such as monoglycerides and phospholipids (see Section 7.4.3 and Section 7.4.4) and that such structures would then serve as templates for the crystallisation of the fat. Vanhoutte et al. (2002a) investigated the effect of low concentrations of water (up to 0.7%) and phospholipids (0.01-0.06%) on the isothermal crystallisation kinetics of anhydrous milk fat (AMF). The crystallisation was monitored by DSC and pNMR and described by means of the Gompertz model. Higher concentrations of water seemed to decrease the induction time, but no interaction effects between phospholipids and water were observed. In a second study (Vanhoutte et al., 2002b), even lower concentrations of phospholipids (up to 0.035%) were added while the concentration range of water remained the same. No significant effect could then be observed due to the water.
7.4.6
300
Temperature (°C)
250 Micellar solution 200 Lα 150 Hl 100
50
Crystals + H2O 20
40 60 H2O (%, w/w)
80
FIGURE 7.25 Phase diagram of sodium myristate. (From Larsson, K. (1986), in The Lipid Handbook, 1st ed., Gunstone, F.D., Padley, F.B., Harwood, J.L., Eds., Chapman & Hall, London, 321–384.)
Soap/water interactions
Soaps exhibit a good water-solubility above the melting temperature of the fatty acid chains (the so-called Krafft 501
7.4 Lipid/water interactions
the HI and the Lα-phase Luzzati et al. (1960) described two-phase mixtures also involving cubic phases.
F.D., Padley, F.B., Harwood, J.L., Eds., Chapman & Hall, London, 321–384. Larsson, K. (1988), 2 H-II types of phases in the same monoglyceride-water system, J. Colloid Interf. Sci., 122, 298. Larsson, K. and Krog, N. (1973), Structural properties of the lipid-water gel phase, Chem. Phys. Lipids, 10, 177–180. Larsson, K. and Lundström, I. (1976), Liquid crystalline phases in biological model systems, in Lyotropic Liquid Crystals and the Structure of Biomembranes: A Symposium Based on the Fifth International Liquid Crystal Conference Proceedings, Stockholm, Sweden, June 17-21, 1974, Friberg, S.E., Ed., American Chemical Society, Washington, D.C., 43–70. Lindblom, G. et al. (1979), Cubic phase of monoglyceride-water systems–arguments for a structure based upon lamellar bilayer units, J. Am. Oil Chem. Soc., 101, 5465–5470. Lis, L.J. and Quinn, P.J. (1991), The application of synchroton x-radiation for the study of phase transitions in lipid model membrane systems, J. Appl. Crystallogr., 24, 48–60. Lutton, E.S. (1965), Phase behavior of aqueous systems of monoglycerides, J. Am. Oil Chem. Soc., 42, 1068–1070. Luzzati, V. et al. (1960), La structure des colloïdes d'association. I. Les phases liquide-crystalline des systèmes amphiphileeau, Acta Crystallogr., 13, 660–667. Luzzati, V. et al. (1968), Structure of the cubic phases of lipidwater systems, Nature, 220, 485–488. McBain, J.W. and Lee, W.W. (1943), Vapor pressure data and phase diagrams of some concentrated soap/water systems above room temperature, Oil & Soap, 20, 17–22. Moonen, H. and Bas, H. (2004), Mono- and diglycerides, in Emulsifiers in Food Technology, Whitehurst, R.J., Ed., Blackwell Publishing Ltd., Oxford, 4058. Morley, W.G. and Tiddy, G.J.T. (1993), Phase behaviour of monoglyceride/water systems, J. Chem. Soc. Faraday Trans., 89, 2823–2831. Nawar, W.W. (1996), Lipids, in Food Chemistry, Fennema, O.R., Ed., Marcel Dekker, New York, 239–319. Pezron, I. et al. (1990), Repulsive pressure between monoglyceride bilayers in the lamellar gel states, J. Phys. Chem., 94, 8255–8261. Reiss-Husson, F. and Luzzati, V. (1969), The structure of the micellar solutions of some amphiphilic compounds in pure water as determined by Small-Angle-X-ray Scattering techniques, J. Phys. Chem., 68, 3504–3511. Sambuc, E. et al. (1980), Étude de la cristallisation des corps gras plastiques VI — Influence des glycérides partiels et des phosphatides en absence et présence d'eau. B. Cas des lécithines de soja, Rev. Franç. Corps Gras, 28, 13–19. Savage, C.M. and Dimick, P.S. (1995), Influence of phospholipids during crystallization of hard and soft cocoa butters, The Manufacturing Confectioner, 127–132. Sein, A. et al. (2002), Rheological characterization, crystallization, and gelation behavior of monoglyceride gels, J. Colloid Interf. Sci., 249, 412–422. Söderberg, I. and Ljusberg-Wahren, H. (1990), Phase properties and structure of a monoglyceride/sucrose/water system, Chem. Phys. Lipids, 55, 97–101. Vanhoutte, B. et al. (2002a), The effect of phospholipids and water on the isothermal crystallisation of milk fat, Eur. J. Lipid Sci. Technol., 104, 490–495.
Acknowledgment Writing a review for a book in between the regular scientific work of the laboratory can be quite hectic. Therefore, we would like to thank some people of the laboratory who helped us in accomplishing this. Nathalie De Clercq, Jeroen Vereecken, and Veerle De Graef did a good job in searching appropriate articles and Katleen Anthierens and Ümmü Ozdemire did almost all the administrative work.
References Brokaw, G.Y. and Lyman, W.C. (1958), The behavior of distilled monoglycerides in the presence of water, J. Am. Oil Chem. Soc., 35, 49–52. Bueschelberger, H.G. (2004), Lecithins, in Emulsifiers in Food Technology, Whitehurst, R.J., Ed., Blackwell Publishing Ltd., Oxford, 1–39. Cassin, G. et al. (1998), Investigation of the gel to coagel phase transition in monoglyceride-water systems, Langmuir, 14, 5757–5763. Chapman, D. (1958), An infrared spectroscopic examination of some anhydrous sodium soaps, J. Chem. Soc., 33, 784–788. Chupin, V. et al. (2001), Lipid organization and dynamics of the monostearoylglycerol-water system, Chem. Phys. Lipids, 109, 15–28. Clunie, J.S. et al. (1965), The structure of lyotropic mesomorphic phases, Proc. R. Soc, 285, 520–532. D’Antona, P.D. et al. (2000), Rheological and NMR characterization of monoglyceride-based formulations, J. Biomed. Mater. Res., 52, 40–52. Ekwall, P. and Mandell, L. (1968), Minimum water content of a number of reversed micellar and mesomorphous structures, Acta Chem. Scand., 22, 699–702. Heertje, I. et al. (1998), Liquid crystalline phases in the structuring of food products, Food Sci. Technol. (Int.), 31, 387–396. Hyde, S.T. et al. (1984), A cubic structure consisting of a lipid bilayer forming an infinite periodic minimum surface of the gyroid type in a glycerolmonooleate-water system, Z. Kristallogr., 168, 213–219. Kelker, H., Hatz, R., and with a contribution by Schuman, C. (1980), Lyotropic mesomorphism, in Handbook of Liquid Crystals, Verlag Chemie, Weinheim, 512–591. Krog, N. and Borup, A.P. (1973), Swelling behaviour of lamellar phases of saturated monoglycerides in aqueous systems, J. Sci. Food Agric., 24, 691–701. Krog, N. and Larsson, K. (1968), Phase behaviour and rheological properties of aqeous systems of industrial distilled monoglycerides, Chem. Phys. Lipids, 2, 129–143. Larsson, K. (1979), X-ray scattering study of the L2-phase in monoglyceride-water system, J. Colloid Interf. Sci., 72, 152–153. Larsson, K. (1986), Physical properties–Structural and physical characteristics, in The Lipid Handbook, 1st ed., Gunstone,
502
Physical Properties: Structural and Physical Characteristics
protein into its native configuration. On the other hand, there is evidence that the interaction of the proteins with membrane lipids is required to impose a bilayer conformation on the surrounding membrane lipids and is, therefore, an essential factor in preserving the structure and properties of the membrane itself. In general, water-soluble proteins interact only weakly if at all with nonionic surfactants, but intrinsic membrane proteins may be solubilized by such surfactants because they are able to interact with the hydrophobic domains that otherwise render them insoluble in aqueous media. Ionic surfactants, by contrast, interact strongly with all proteins and modify their functions and properties (Nielsen et al., 2005). One of the most studied of this type of interaction is between sodium dodecyl sulfate and proteins because of the importance in analysis of protein mixtures. The nonspecific cooperative binding of sodium dodecyl sulphate to soluble proteins results in unfolding of the polypeptide chain. After reduction of any disulfide bridges in proteins, sodium dodecyl sulfate, above the critical micellar concentration, interacts with the polypeptide in a stoichiometry of 1.4 g detergent per g protein. The interaction imposes a regular helical structure on the polypeptide chain, which becomes bent in the shape of a hairpin. The length of the resulting complex is a function of the length of the polypeptide chain and, because of the predictable conformation combined with a constant charge to mass ratio, differences in hydrodynamic and electrophoresis properties can be exploited in separation strategies of complex mixtures. The most notable system is the separation of proteins by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate (SDS-PAGE), which separates polypeptides on the basis of size. Another application of protein–surfactant complexes is in controlling colloidal stability in processed food products and in the stabilization of enzymes performing reactions under nonphysiological conditions. Where proteins interact with biological membranes at the aqueous-lipid interface the charges of the acidic membrane lipids provide a particular environment capable of interacting with basic amino acids of the protein. In the case of proteins that are interpolated into the hydrophobic interior of the membrane, the environment is more conducive to the location of amino acids with nonpolar side chains. Clearly, both proteins and membrane lipids have hydrophilic and hydrophobic groups, which interact to determine the structure and conformation of the complex.
Vanhoutte, B. et al. (2002b), Effect of phospholipids on isothermal crystallisation and fractionation of milk fat, Eur. J. Lipid Sci. Technol., 104, 738–744. Vold, R.D. et al. (1941), Phase rule study of the system sodium myristate/water, J. Am. Chem. Soc., 63, 1293–1296.
7.5
Interaction between lipids and proteins
7.5.1
Principles of lipid protein interactions
The basis for the interaction between lipids and proteins is related to their amphiphilic nature and is due to their influence on water structure, the so-called hydrophobic effect (Tanford, 1980). In general terms, four alternative types of phases can occur in lipid-protein-water systems. An aqueous lipid solution can coexist with a protein in the same solution; alternatively a solution of molecular lipoprotein complexes can be formed. It is also possible that the lipid forms a liquid-crystalline phase with water; such a phase can either solubilize a protein or coexist with a protein solution. The first two interaction alternatives in water solution have been thoroughly discussed (Tanford, 1980). When the lipid concentration is below the critical micelle concentration there is no interaction besides the eventual association of a few lipid molecules to the protein at certain high-affinity binding sites. At lipid concentrations above the critical micelle concentration there is a mass cooperative binding of numerous lipid molecules to each protein molecule and this may lead to unfolding of the protein. The interaction between proteins and surfactants has been characterized by a variety of biophysical methods, including light scattering, spectroscopic probe methods, surface tension measurements, equilibrium dialysis, surface-sensitive electrodes, and calorimetry. Interaction isotherms obtained from calorimetric titration to high surfactant concentration of proteins, such as bovine serum albumin, lysozyme, gelatin, and cellulase yield characteristic thermodynamic fingerprints, but the interpretations of the particular enthalpy changes so far defy detailed molecular interpretation. Thus, it is difficult to distinguish contributions to the changes in enthalpy due to surfactant binding, conformational changes in the protein and micellization. The way proteins fold depends largely on the local forces experienced by the different side chains and the need for the polymer chain to adopt a conformation with relatively high entropy. The forces at play are those due to the solvent environment as well as the proximity of other residues of the protein or residues of proteins with which it interacts. In the presence of lipids the hydrophobic residues of the protein may rearrange to produce a complex of greater entropy. With certain membrane proteins the “solvent” may include the lipid matrix of the membrane and interactions of this type may be required to fold the
7.5.2
Lipidprotein–water phase behaviour
Ternary complexes are formed between lipids in cubic phases with certain proteins and such complexes have been exploited in the crystallization of proteins for x-ray diffraction examination. A partial phase diagram of the ternary system of lysozyme-monoolein-water has been constructed (Ericsson et al., 1983) and is presented in 503
7.5
Interaction between lipids and proteins
of agents used to induce the transition are the alkyl glycosides in which hydrocarbon chains of hexyl, octyl, nonyl or decyl are combined with sugar residues of glucose or maltose.
Figure 7.26. The most remarkable feature of the phase diagram is that a relatively large proportion of the watersoluble protein can be incorporated to form a lipidprotein-water phase without any ionic interactions. It was also found that the protein occupied the aqueous phase of the complex in its native configuration. This discovery was to have considerable implications as is discussed below in the crystallization of both soluble and membrane proteins. The incorporation of protein results in an expansion of the cubic phase lattice formed by monoolein-water. The cubic domain of the phase diagram shown in Figure 7.26 contains all three fundamental cubic structures observed in lipid-water systems of the type found in monoolein-water, namely, the gyroid surface CG, the diamond surface CD, and the primitive surface CP in the order of increasing proteinwater ratio. Complex cubic phases are formed with ternary lipidwater systems like monoolein mixed in proportions of two parts protein solution or dispersion with three parts of lipid. When such mixtures are treated with precipitants such as non-ionic detergents or salts the protein begins to crystallize within hours of incubation at 20oC. The method can be used to grow crystals of soluble as well as membrane proteins and other organic and inorganic molecules. The precise process of crystallization from these tertiary lipid phases has been examined in some detail (Misquitta et al., 2004). Precipitants like Na+/K+ phosphate salts, for example, provoke a reduction in water activity, which favours protein–protein interactions. Three-dimensional structures are created when protein–protein contacts are established between successive layers. The key to successful crystallization of proteins is the action of the precipitant to destabilize the cubic phase of monoolein, which is the principle host in the lipid-protein complex, so that it tends to form a liquid-crystalline lamellar phase. A group
7.5.3
Some proteins are adapted to form stable water-soluble complexes with lipids and to function in the mobilization of complex lipids within living organisms. There are two main types of complex; complexes formed between monomeric proteins and lipids and, secondly, large lipoprotein complexes. One of the most ubiquitous monomeric proteins that bind lipids is albumin. Plasma albumin is a flexible protein that can adopt multiple conformations of approximately equal energy to accommodate the binding of ligands. One of the primary functions of albumin is to bind free fatty acids during mobilization of lipids in the body and transport them in the blood stream as a stable water-soluble complex. The protein has at least five fatty acid binding sites three of which are significantly higher affinity sites than the remainder. The mechanism of binding of the fatty acids and other lipophilic drugs to the protein has been investigated by NMR methods (Lucas et al., 2004). The average residence lifetime of a long chain fatty acid bound to a high-affinity site was found to be greater than 66 ms whereas short-chain fatty acids like octanoic acid have residence times of only a few ms. The lifetimes are relatively short because the fatty acids exchange readily between binding sites on the protein. Dissociation of the fatty acid from the protein, on the other hand, takes place on a time-scale of seconds. The dissociation of fatty acids from plasma albumin is the rate-limiting step in the delivery of fatty acids to target cells. Their dissociation from the complex at the site of entry into cells is assisted by the presence in the plasma membrane of proteins with a high affinity for fatty acids (McArthur et al., 1999). One such protein, membrane fatty acid-binding protein (FABPpm), binds tightly to free fatty acids and prevents destabilization of the membrane due to the presence of the free fatty acid in the structure. The entry of the free fatty acid into the outer leaflet of the membrane lipid bilayer matrix establishes a transmembrane gradient, which can be dissipated by flip-flop of the uncharged fatty acid from the outer to the inner membrane leaflet. This process can take place spontaneously or may be assisted by another fatty acid binding protein of the membrane, fatty acid translocase (FAT/CD 36). Another soluble fatty acid-binding protein present in the cytoplasm, FABPc, serves to remove free fatty acids present on the cytoplasmic leaflet and to transport them through the cytoplasm. There is also evidence that fatty acids also bind to caveolin-1. This protein is a component of caveolae that are said to function to deliver lipid to the different subcellular organelles. It remains unclear
Lysozyme
20
40
80 Unexamined region
60
60
80
H2O
40
Lysozyme solution Cu bi Cubic phase Cubic lamc + ella phase + lysozyme sol. r
20
40
60
20
80
Lipoproteins
Monoolein
FIGURE 7.26 Phase diagram illustrating the main regions of the monoolein-lysosome-water system. (Adapted from Ericsson, B. et al. (1983), Biochim. Biophys. Acta, 729, 23–27.)
504
Physical Properties: Structural and Physical Characteristics
Triacylglycerols and cholesterol are also synthesized in the liver and the excess of requirements is packaged into very low-density lipoproteins and exported to the peripheral tissues. The proteins stabilizing very low-density lipoproteins include apoB-100 and apoE. ApoB-100 is an extremely large protein comprised of more than 4500 amino acid residues and a molecular weight of 513 kD. ApoB-48 is formed from the first 48% of apoB-100 and arises from the posttranscriptional editing of apoB-100 mRNA in the intestine. The relationship between apoB-100 and B-48 has been the subject of considerable interest (Brodsky et al., 2004). The intermediate-density lipoproteins result from depletion of the triacylglycerol content of very low-density lipoproteins by the action of lipases associated with capillary surfaces and their consequent enrichment in cholesterol esters. These intermediate-density lipoproteins may be taken up by the liver and further processed or converted into low-density lipoproteins by hydrolysis of more triacylglycerol. Low-density lipoprotein is the major carrier of cholesterol and consists of a core of about 1500 cholesterol molecules esterified mainly to linoleate. The nonpolar lipid is stabilized by a monolayer of phospholipid and apoB-100. High-density lipoprotein is also involved in cholesterol transport, but the source of cholesterol is that scavenged from apoptosing and dying cells and membranes undergoing metabolic turnover. The cholesterol is esterified to a long-chain fatty acid in a reaction catalysed by an acyltransferase intrinsic to the high-density lipoprotein. The cholesterol esters are rapidly transferred to very low- or low-density lipoproteins by a specific transfer protein or targeted to the liver in their high-density lipoprotein vector. ApoA, apoC and apoE are referred to as exchangeable apolipoproteins and they are responsible for regulating the traffic of lipids into and out of cell by acting as cofactors for plasma enzymes and ligands for cell-surface receptors. The exchangeable apolipoproteins share the same genomic heritage and, therefore, possess structural similarities (Saito et al., 2004). One particular feature is a primary sequence arranged in αhelical motif in which the basic residues are located near the hydrophilic/hydrophobic interface and acidic residues are clustered at the center of the polar face. The helical segments are often interrupted by proline residues. This so-called Class A motif that is exemplified by apoA-1 lipoprotein is illustrated together
whether or not intracellular trafficking of fatty acids mediated by FABPc and that by the vesicular-caveolae mechanism act in concert with one another. Apart from transport to the cell, metabolic mobilization relies on activation of long-chain acyl-CoA. This process is mediated by a family of fatty acid-transport proteins referred to as FATP 1-5/6, which are known to possess acylCoA synthetase activity. There is evidence from the variable nature of the N-terminus that these fatty acid-binding proteins may be able to deliver fatty acids to particular membrane sites within the cell. The activation of fatty acids takes place at the highly conserved AMP-binding site located at the cytosolic domain of FATP from where they are primed for metabolism at the appropriate organelle. Serum lipoproteins are a group of proteins specifically adapted to form structures designed to transport lipids throughout the body. These lipiproteins circulate in the mammalian blood stream to distribute a cargo of lipids from their site of synthesis, usually the liver, to the peripheral tissues. There has been an extensive research effort to characterise lipoproteins because of their association with heart disease and atherosclerosis. The lipids, mainly triacylglycerols, cholesterol, and cholesterol esters, occupy a central core surrounded by a shell of polar lipids and proteins. The proteins act to stabilize the lipid droplet and provide recognition sites for targeting the complex to the appropriate site of delivery. The particles are usually spherical in shape and are classified according to their buoyant density. The characteristics of the different classes of human lipoproteins are presented in Table 7.2. As expected the density increases with increasing protein to lipid ratio. The protein components are synthesized in the liver and small intestine. While many of the 10 major apoproteins found in lipoproteins are common to more than one class of lipoprotein, their combination is distinct in each of them. Chylomicrons are the least dense of the lipoproteins and are responsible for packaging fats, cholesterol, and other lipids taken up from the diet in the blood stream and conducting them about the body. Because they consist almost entirely of triacylglycerols, they have a buoyant density of < 0.95 g/cm-3. The major protein is apolipoprotein B-48 (apoB-48), which has a molecular weight of 240 kD and forms an amphipathic shell around the spherical fat globule in which the interior surface of the protein is hydrophobic and the interior is hydrophilic. TABLE 7.2
The size and composition of mammalian lipoproteins
Lipoprotein Chylomicrons Very low density (VLDL) Intermediate density (IDL) Low density (LDL) High density
Density
Diameter (nm)
Triglyceride/Cholesterol
Apoprotein
< 0.95 0.95-1.006
1000 50
10 2.3
B-48,C,E B-100,C,E
1.006-1.019 1.019-1.063 1.063-1.210
30 20 10
1.0 0.2 0.03
B-100,E B-100 A
505
7.5
Interaction between lipids and proteins
The driving force for formation of complexes between the apolipoproteins and lipids appears to be a favourable change in enthalpy. A conformational transition from random coil to α-helix on binding of apoA-1 to lipid, for example, is associated with an enthalpy change in the order of -5kJ/α-helical segment. This represents a total enthalpy change of about -11 kJ.mole-1 and is additional to the enthalpy change accompanying the interaction of apoA-1 with lipid, which is about -40kJ. mole-1. The change in conformation of the protein, therefore, contributes significantly to the binding affinity between the protein and the lipid. Two conformers of high-density lipoprotein have been characterised, one discoidal in shape and the other spherical. The discoidal particles are comprised of a phospholipid bilayer disk with two molecules of apoA-1 encircling the edges where the acyl chains are exposed. The size is limited by the length of the apoA-1 molecules that are arranged in a belt of α-helices stacked one on the other in an antiparallel orientation. A similar discoidal particle has also been described for apoE, but, in this structure, four molecules of the protein are arranged at the periphery of the disk. The spherical form of highdensity lipoprotein varies in diameter and has more neutral lipid than the discoid form. The amphipathic helices of the apolipoproteins are believed to be interpolated between the phospholipid molecules rather than at the periphery. The extent of interaction of the protein with the lipid is greater as the proportion of protein in the particle decreases.
Nonpolar surface
Class A Positively charged residues Negatively-charged residues
Class Y
Class G
FIGURE 7.27 Helix wheel plots of the three classes of α-helical segments found in apoA-1, apoA-IV, and apoE apolipoproteins. (Adapted from Segrest et al. (1992), J. Lipid Res., 33, 141–166.)
with Class Y and Class G motifs in Figure 7.27 ApoE isoforms have a predominance of Class G helical motifs and the Class Y helical motif dominates the secondary structure of apoA-IV. The N-terminal amphipathic helices of apoE are bundled into four antiparallel strands forming an elongated globular structure with the hydrophobic faces oriented into the interior. The C-terminal adopts a coiled-coil helix structure that is more exposed to the aqueous phase. In the absence of lipid, apoA-1 adopts a two-domain configuration similar to apoE; an N-terminal helical bundle extending into the central part of the primary structure and a C-terminal helical domain that is less well organized. Both apoA-1 and apoE aggregated into oligomers in the absence of lipid via hydrophobic interactions between residues located at the C-termini of the respective molecules. The self-association in aqueous media via the C-terminal domains of apoA-1 and apoE is consistent with the role of the C-termini in the interaction of these proteins with lipids. When apoE binds to lipid it has been suggested that the initial binding takes place at the C-terminal, which then induces the 4-helicle bundle of the N-terminus to reorganize so that the hydrophilic faces of the helices open out and become available for binding to the lipid. Similar reorganizations are believed to occur when apoA-1 interacts with lipid. Thus, an initial binding takes place at the C-terminal domain, which is arranged in an elongated hairpin structure. Following this there is a conformational change involving residues 1-43, which serve to unmask a hydrophobic domain in residues 44-65 of the protein.
7.5.4
Enzyme-substrate interactions in lipolysis
A variety of enzymes are required to synthesize lipids and, in turn, hydrolyse them in the normal metabolic turnover of these structural components. Likewise, enzymes play an important role in the dietary uptake and assimilation of lipids in the body. The activity of many lipolytic enzymes share a common feature in that their hydrolytic activity against their normal substrates are greatly influenced by the presence of lipids that are not substrates of the reaction or known to be bound to the enzyme. For hydrolysis to take place, the water-soluble enzyme must be attracted to the substrate located at the lipid–water interface. The electrostatic charges carried by the reactants need to be favourable and the magnitude of the attractive force increases with the size of the opposite charges carried by the protein and the electro-static potential at the substrate interface. In addition, the enzyme must orient about the interface such that the substrate is presented in a favourable manner for the hydrolytic reaction to proceed. Finally, the products of the reaction must diffuse away from the reaction site to be replaced by a fresh substrate molecule. This process is often complicated by the fact that reaction products are often hydrophobic in character and remain concentrated 506
Physical Properties: Structural and Physical Characteristics
in the substrate thereby influencing the rate of hydrolysis simply by diluting the substrate and altering the manner in which the enzyme interacts with it at the substratewater interface. One of the characteristic features of hydrolysis of lipid substrates by their respective enzymes is the existence of a time delay or lag period before activation of the enzyme is observed. The lag period in activation of phospholipase C from Bacillus cereus has been investigated by assay of activity of the enzyme against large unilamellar vesicles composed of different substrate mixtures (Ruiz-Arguello et al., 1998). A lag period in activation of the enzyme was noted in substrate mixtures consisting of phosphatidylcholine (the normal substrate for the enzyme), phosphatidylethanolamine (hydrolysed to a lesser extent), sphingomyelin, and cholesterol (neither are substrates for the enzyme). The duration of the lag period was found to vary between 8 sec and more than 30 min depending on the proportion of the different lipids in the substrate mixture. Furthermore, the maximum rate of hydrolysis varied from 0.4 to more than 55 min–1, again depending on the particular substrate mixture. The presence of lipids in the substrate that tended to destabilize bilayers and favour hexagonal-II structure, such as phosphatidylethanolamine and cholesterol, were found to enhance activity and shorten the lag phase. Conversely, lipids that are known to stabilize bilayers, such as sphingomyelin, had the opposite effect on enzyme activity. This suggests that, while the substrate is presented in the form of a bilayer, the enzyme is sensitive to the local instability that assists orientation of the substrate about the active site of the enzyme. The interesting feature of all reaction mixtures, however, was that the proportion of the substrate hydrolysed during the lag period was invariably 0.10% of the total substrate present. The explanation for this observation is that the creation of local rafts of bilayer enriched in the diacylglycerol product of the reaction results in aggregation of the vesicular substrate, which, in turn, is responsible for the acceleration in the rate of hydrolysis. An indication of the effect of reaction products on the physical properties of the substrate has been obtained from studies of the effect of phosphatidic acid in monolayers of phosphatidylcholine subjected to hydrolysis by phospholipase D from Streptomyces chromofuscus (El Kirat et al., 2002). The enzyme is a member of a super family that includes endonucleases, helicases, lipid synthetases, and enzymes able to catalyze the formation or hydrolysis of phosphodiester bonds. In its reaction against a substrate of phosphatidylcholine, a phosphatidyl–enzyme intermediate is formed, which is subjected to a nucleophilic substitution by a molecule of water to release phosphatidic acid. The effect of phosphatidic acid in monolayers of substrate at the air–water interface on the surface elasticity modulus and the lag period before accelerated hydrolysis is observed is shown in Figure 7.28 Surface elasticity
1
Normalized Ks and lag time
Ks (at 30 mN.m–1) Lag time 0.75
0.5
0.25
0 0
10 20 Phosphatidic acid (mole %)
30
FIGURE 7.28 The value of surface elasticity modulus at 30 mN,m-1 and enzyme activity lag time determined as the time preceding a 5% decrease in monolayer area plotted against the proportion of phosphatidic acid mixed with the phosphatidylcholine substrate. (Adapted from El Kirat, K. et al., (2002), J. Biol. Chem, 277, 22131–21236.)
modulus, Ks, is derived from the surface pressure-area isotherm from the relationship: Ks = –A (∂π/∂A)
(7.51)
where A is the molecular area at the corresponding surface pressure, π. The greater the value of Ks for a monolayer the less it is subject to deformation. The correlation between Ks and lag time shown in the figure suggests that deformation of the substrate-water interface is an essential step in orientating the enzyme about the phospholipid substrate and that the product is instrumental in modulating this process. A number of studies have been reported that examine the effect of substrate presentation and molecular species preferences of secretory phospholipase A2. The enzyme hydrolyses the fatty acid esterified to the sn-2 position of the glycerol backbone of diacylglycerophosphatides. The activity of the different subclasses of this enzyme are known to be modulated by proteins and peptides, such as melittin and phospholipase A2-activating protein, that are believed to act by modifying the manner of presentation of the substrate (Koumanov et al., 2003). The effects of such proteins in activating the enzymes differs depending on whether the substrate is in the form of a dispersion of pure lipid or is present in a biological membrane. These differences highlight the role of regulatory peptides in the biological function of these phospholipases. In addition to regulation by proteins the presence of nonsubstrate lipids is also found to influence catalytic activity. Type-II secretory phospholipase A2 can be “activated” by displacement from an interface comprised of susceptible diacylglycerophospholipid substrate by binding to an interface of nonsusceptible phospholipids, such as 507
7.5
Interaction between lipids and proteins
emulsions such as margarine and ice cream have polar lipids and proteins at the fat–water interface. Baking performance of cereals depends upon lipid–protein interaction. The properties of starch products can be monitored by formation of an amylase–lipid complex as a direct consequence of hydrophobic interaction. Among pharmaceutical applications, the use of lipids and surfactants as formulation aids play an essential role in delivery and uptake of a variety of drugs into the body. Liposome vectors are commonly exploited for this purpose. An active component can be entrapped in liposomes or vesicles, giving a controlled-release system. By solubilizing proteins in the lipid bilayer that exposes a “label” on the surface, it may even be possible to achieve drug targeting. Clinical experiences so far, however, appear to be limited mainly because liposomal dispersions are not thermodynamically stable despite claims that phospholipid formulations of water-insoluble drugs achieve desirable versatility (Zhang et al., 2005).
sphingomyelin. The sphingomyelin may be presented in the form of a separate dispersion added to the assay mixture or as a phase-separated domain within the bilayer of substrate molecules. The affinity of the enzyme for the sphingomyelin interface may, in turn, be reduced by the association of the sphingomyelin with cholesterol to form a liquid-ordered phase. This effect is observed as an “activation” of the enzyme by cholesterol. The action of one phospholipase can trigger the activity of another phospholipase with different substrate specificity. Such actions may represent manifestations of the highly complex biochemical homeostatic mechanisms that are present in living cells responsible for maintaining the molecular species composition of cell membranes within relatively narrow limits. One example is the activation of secretory phospholipase A2 by ceramide, the product of sphingomyelinase on its substrate, sphingomyelin (Koumanov et al., 2002). The activation of secretory phospholipase A2 by ceramide is additional to “activation” by release of the enzyme bound to sphingomyelin domains. The activation is apparently due to the creation of phaseseparated domains of bilayer and hexagonal-II phase in the substrate dispersion. The presentation of diacylglycerophospholipid substrate at the interface between the two-phase structures would appear to be the most plausible explanation for the enhanced activity of the enzyme. This may be related to the observation that substrate molecular species with a polyunsaturated fatty acid, such as arachidonic acid, in the sn-2 position of the glycerol are the much-preferred substrates compared with molecular species with relatively saturated fatty acids acylated to the sn-2 position. The substrates for triacylglycerol lipases are usually presented in the form of an emulsion, such as dietary fat droplets or lipoproteins. These particles consist of substrate presented in a bulk phase, which is stabilized at the aqueous interface by a monolayer surface phase composed of phospholipid and other proteins. The enzyme catalysis takes place in the surface phase so both substrate and the enzyme must partition from their respective bulk phases into the interfacial region for reaction to occur. In many cases stabilization of lipid emulsions by phospholipids impedes access of the enzyme to its substrate and this is evidenced by a significant lag period that may be of duration of hours before lipase activity accelerates towards equilibrium. Such lag period can often be reduced or even eliminated by the introduction of reaction products, such as diacylglycerol or fatty acids; some lipases require cofactor proteins to augment their catalytic function (Brockman, 2000).
7.5.5
References Brockman, H.L. (2000), Kinetic behavior of the pancreatic lipase-colipase-lipid system, Biochimie, 82, 987–995. Brodsky, J.L. et al. (2004), Vesicular trafficking of hepatic apolipoprotein B100 and its maturation to very low-density lipoprotein particles; studies from cells and cell-free systems, Trends Cardiovacs. Med., 14, 127–132. El Kirat, K. et al. (2002), Role of calcium and membrane organization on phospholipase D localization and activity. Competition between a soluble and insoluble substrate, J. Biol. Chem, 277, 22131–21236. Ericsson, B. et al. (1983), A cubic protein-monoolein-water phase, Biochim. Biophys. Acta, 729, 23–27. Koumanov, K.S. et al. (2002), Ceramides increase the activity of the secretory phospholipase A2 and alter its fatty acid specificity, Biochem. J., 363, 45–51. Koumanov, K.S. et al. (2003), Bimodal regulatory effect of melittin and phospholipase A2-activating protein on human type II secretory phospholipase A2, Cell. Biol. Int., 27, 871–877. Lucas, L.H. et al. (2004), Epitope mapping and competitive binding of HSA drug site II ligands by NMR diffusion measurements, J. Am. Chem. Soc., 126, 14258–14266. McArthur, M.J. et al. (1999), Cellular uptake and intracellular trafficking of long chain fatty acids, J. Lipid Res., 40, 1371–1383. Misquitta, Y. et al. (2004), Rational design of lipid for membrane protein crystallization, J. Struct. Biol., 148, 169–175. Ruiz-Arguello, M.B. et al. (1998), Phospholipase C hydrolysis of phospholipids in bilayers of mixed lipid compositions, Biochemistry, 37, 11621–11628. Saito, H. et al. (2004), Contributions of domain structure and lipid interaction to the functionality of exchangeable human apolipoproteins, Prog. Lipid Res., 43, 350–380. Segrest, J.P. et al. (1992), The amphipathic helix in the exchangable apolipoproteins: a review of secondary structure and function, J. Lipid Res., 33, 141–166. Tanford, C. (1980), The Hydrophobic Effect, John Wiley & Sons, New York.
Technical applications
There are a number of important technical applications of lipid–protein interaction, mainly in the fields of food technology and the pharmaceutical industry. Food
508
Physical Properties: Structural and Physical Characteristics
a number of important constraints. There is, for example, a lateral heterogeneity in the relative mobility of lipids and proteins, which is due to the creation of domains of more ordered lipid arrangements dispersed amongst more disordered fluid domains. It is remarkable that membranes form such a stable unified structure considering the diversity of lipid molecular species and different types of proteins associated with each morphologically distinct membrane. It is also implicit that biochemical mechanisms must be responsible for creating and preserving this diversity, including that of the relative proportions of the different lipids and proteins. The nature of these processes, however, is largely unknown. Many membrane functions can be directly related to the physical properties of the lipid bilayer. The action of membranes as a barrier to the diffusion of water-soluble solutes, especially charged ions, is due to the creation of a hydrocarbon layer by the orientation of the hydrophobic moieties of the constituent lipids into the core of the structure. Variations in curvature and area per unit mass and motion of the constituent molecules can also be ascribed to the relatively fluid nature of the structure. The slow or prohibited translocation of lipids and proteins from one side of the membrane to the other creates functional asymmetries that are essential to the performance of lifesustaining processes. Thus, membrane proteins have a unique orientation with respect to the membrane bilayer, a property that is integral to the biosynthesis of cell membranes. This fact ensures that the functions performed by the proteins are vectorial. Since many membrane proteins are concerned with the selective translocation of solutes across the membrane, their mutual orientation means that transport takes place only in the required direction. Certain membrane lipids and proteins are glycosylated. The associated carbohydrates are particularly abundant on the outer surface of the plasma membrane where they form the glycocalyx. The function of this layer is to participate in recognition phenomena and as receptor sites for different hormones, toxins, and other ligands. In endothelial tissues the glycocalyx serves to shield the vascular wall from the shear stresses of blood flow, impede leakage of blood constituents across the endothelial lining, and prevent adhesion of leucocytes and platelets to the endothelium (Rehm et al., 2004). Membrane trafficking between subcellular compartments is a mechanism that operates to redistribute membrane components and their associated ligands within cells. The process is also central to the biogenesis and homeostasis of membranes. Endocytosis and exocytosis are processes that involve the uptake of plasma membrane into the cell and fusion of subcellular membrane vesicles with the plasma membrane, respectively. A model for endocysis to account for the participation of cytoplasmic proteins has been proposed recently (Hommelgaard et al., 2005). According to this model, small lipid domains
Zhang, J.A. et al. (2005), Development and characterization of a novel Cremophor EL free liposome-based paclitaxel (LEP-ETU) formulation, Eur. J. Pharm. Biopharm., 59, 177–187.
7.6
Biological membranes
7.6.1
Introduction
The structure of biological membranes has a history spanning more than a century and a chronology of developments of the topic can be sourced from the various models proposed for membrane structure published throughout this period (Gorter and Grendel, 1925; Danielli and Davson, 1935; Singer and Nicolson, 1972; Robertson, 1972; Green and Brucker, 1973). One of the most durable of the models is the so-called fluid mosaic model proposed by Singer and Nicolson in 1972. The model envisaged that membrane lipids form a bilayer configuration, which serves as a matrix for the association of the proteins. The principal arguments used to support this concept were thermodynamic; the experimental evidence was tenuous and based on assumptions that the arrangement of the lipid was indeed a bilayer. The most convincing evidence, which was not cited in the formulation of the model, is based on electron density calculations obtained from x-ray scattering intensity profiles of oriented films of hydrated egg lecithin, itself a lipid not found in a biological membrane (Levine and Wilkins, 1971). The electron densities were found to be consistent with an osmotically insensitive component assumed to be devoid of water and consisting of amphiphilic lipid molecules oriented with their hydrocarbon chains shielded from the aqueous component by polar groups aligned at the interface. It is now recognised that the ability of the membrane lipid molecules to associate in an aqueous environment in infinite lipid bilayers represents the basic mechanism underlying the formation of biological membranes. The property of such structures to separate aqueous compartments was also seen as a critical step in the evolution of living organisms. The significance of the ordered, but fluid, state of the lipid molecules, particularly for the interaction with membrane proteins, was revealed by studies of lipid–water systems preceding the formulation of the fluid mosaic model (Chapman et al., 1967). There is now a consensus that the lipid bilayer represents the matrix of biological membranes. In most cases the lipids are more or less fluid, but in some noteworthy examples they are crystalline. These include the purple membrane of Halobacterium halobium and related bacteria in which the lipids of the bilayer are arranged in a crystal structure in patches on the cell surface that support the light-driven proton pump, bacteriorhodopsin. In the overwhelming number of cases the lipid bilayer is thought to be predominantly in a fluid state such that the components are able to diffuse relative to one another subject to
509
7.6 Biological membranes
negatively charged lipids by screening the electrostatic charges on the lipid molecules with cations. The precise rearrangement of lipids to bring about fusion between adjacent membranes has been the subject of considerable interest. One model of the sequence of structures leading to fusion is the stalk hypothesis (Haque and Lentz, 2004). An initial intermediate structure (stalk) is said to form by the rearrangement of lipids in the outer leaflets of closely contacting bilayers in a manner that alters the normal topology of the bilayer. Evolution of the stalk into a trans-monolayer contact involves rearrangements in nonbilayer structures without a change in membrane topology. The free energy of such non-bilayer lipidic structures is thought to be dominated by bending of the monolayer and hydrophobic mismatch or “void” energies. The fusion process is facilitated by the strategic location of lipid molecular species that are intrinsically positively or negatively curved or long-chain lipids that favour a stable bilayer arrangement. Stable fusion pores can form only to the extent that the unfavorable free energy of these intermediates can be overcome by thermal energy. That molecular species of lipids present in cell membranes exhibit complex thermotropic and lyotropic mesomorphism has been well established from biophysical studies of lipid–water systems over many years. What is less clear is the manner in which these properties influence the creation and stability of local membrane lipid domains, the function of the different membrane proteins and the mechanisms operating to preserve membrane lipid composition. The overwhelming evidence supports the notion that the lipids of biological membranes are arranged in a bilayer configuration despite the fact that a considerable proportion of the molecular species of lipid present in a cell membrane may assume a nonbilayer arrangement if dispersed alone under physiological conditions. It is clear that the lipids of cell membranes play a crucial role in the function and properties of cell membranes. The properties of membrane lipids dispersed in aqueous systems have been examined in great detail by a range of biophysical techniques. The properties of lipid–water systems using such techniques have been extensively reviewed elsewhere in this volume and the results of such studies form a useful basis upon which their behaviour in biological membranes may be inferred. Some of the structural features of membrane proteins will be referred to below so that the factors that govern the interaction between the membrane proteins and lipids may be examined in an informed context.
consisting of ordered phase lipids and associated proteins associate laterally in the plasma membrane to form larger, endocytotic invaginations, which are stabilized by interaction with the cytoplasmic protein, caveolin. This interaction reduces the endocytic activity of these structures and, in addition, binding to the actin cytoskeleton via filamin makes the caveolae largely immobile. Nevertheless, a few caveolae may circumvent the restraints and become internalized. Short-range caveolar motility including membrane fission and fusion are also thought to take place so that clathrin-coated vesicles that bud off from the plasma membrane may fuse to become contiguous with the plasma membrane again. Internalization of caveolae can be specifically induced by complete reorganization of the actin cytoskeleton and removal of plasma membrane caveolae in a coordinated wave. In the budding and fusion of membrane vesicles from particular compartments, only membrane components that are destined to be translocated are included in the vesicle; resident membrane constituents that define the compartment are excluded from membrane forming the vesicle. The mechanism that operates to segregate membrane proteins involves short peptide sequences on the membrane proteins that represent recruitment signals. In addition to this segregation process, a recovery system is known to act to retrieve resident proteins that leak into the vesicular compartment and return them to their original location. Protein–protein interactions responsible for the targeting and docking of vesicles between appropriate membranes are mediated by the Rab family of GTPases. About 60 different Rab proteins, which are tethered to the membrane by covalently bound prenyl groups, have been found to be expressed in mammals and these serve to control the diversity of vesicle traffic in different cell types. They act to assemble soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) complexes on the vesicle (v-SNARE) and respective target (t-SNARE) membrane surface (Hong, 2005). The SNARE proteins are largely anchored to the membrane by a short C-terminal sequence of hydrophobic amino acids. Docking between complimentary v- and t-SNARES ensures interaction takes place between the appropriate membranes. There is evidence that some SNAREs are involved in the subsequent fusion between docked membranes (Chen et al., 2005). Fusion and fission are fundamental processes in membranes. The fusion between membranes can be regarded as a local phase transition at adjacent lipid bilayers where the action takes place. Whereas the choline phosphatides form the lamellar Lα structure over a wide range of pH, in the presence of various ions, other phospholipid classes in the membrane favour nonlamellar arrangements. Furthermore, certain polar lipids can form domains that favour the formation of nonlamellar structures at the growth temperature (Sot et al., 2005). Nonbilayer structures can also be induced in membranes containing certain
7.6.2
Lipid arrangement in biological membranes
The principal lines of evidence for the arrangement of lipids in biological membranes are based on studies of certain classes of lipid that form lamellar structures in aqueous dispersions and are supported by arguments 510
Physical Properties: Structural and Physical Characteristics
from chloroplast membranes compared to the intact membrane. Clearly there are bilayer as well as nonbilayer structures formed in the polar lipid dispersion. No such structures are observed in the intact membrane in which the membrane-associated particles are believed to represent the intrinsic proteins of the photosynthetic apparatus distributed in smooth regions said to be the central domain of a lipid bilayer. The implications of these observations is that either the arrangement of the polar lipid in the membrane is, in part, lamellar and coexisting with nonlamellar structures or that interaction with other membrane components, such as the protein, imposes a lamellar arrangement on those molecular species of lipid that would otherwise form a nonlamellar structure. It remains a possibility, however, that the particles themselves represent lipoprotein complexes, which may comprise lipids that are not components of the bilayer matrix. The methods used to determine the arrangement of lipids in biological membranes include diffraction and spectroscopic measurements as well as freeze-fracture electron microscopy and monomolecular film area determinations. Some of this evidence will be examined in the following sections.
100 nm (a)
7.6.2.1
Monomolecular film area measurements
The first demonstration that lipids form a monolayer at the air–water interface was reported by the polymath and statesman, Benjamin Franklin. In a letter to Doctor Brownrigg published in the Philosophical Transactions (1774), he stated:
(b)
FIGURE 7.29 Electron micrographs of freeze-fracture replicas prepared from (a) total polar lipids extracted from higher plant chloroplast membranes dispersed in chloroplast assay medium. (b) Intact chloroplast membranes from which the lipid extract was prepared. The scale bar and direction of shadowing is shown in (a).
This work largely lay fallow until taken up by Langmuir (1917) who received the 1932 Nobel Prize for chemistry for his work on surfactant films. This foundation of surface chemistry together with Overton’s earlier studies (1895) showing that hydrophobic molecules penetrate membranes more readily than hydrophilic solutes led to the experiments by Gorter and Grendel (1925) upon which the first model of membrane structure was formulated. Their experiment consisted of comparing the area occupied at an air–water interface by the lipids extracted from a known area of membrane; they examined erythrocyte membranes from six species, including man. The experimental methods used in these experiments, however, were found to be flawed; the use of acetone solvent resulted in incomplete extraction of the lipid, but this was compensated to some extent by inaccuracy in
along the lines of the predicted behaviour of amphiphiles when hydrated in aqueous systems. This approach, however, ignores the presence, often in very significant proportions, of lipids that do not form lamellar structures when dispersed in aqueous media at temperatures approximating to that of the growth temperature of the organism from which they are extracted. Indeed, it is commonly observed that dispersions of total polar lipid extracts of many membranes do not form pure lamellar phases, but exhibit a mixture of different phases consisting of both lamellar and nonlamellar structures. This is illustrated in Figure 7.29, which shows freezefracture electron microscopic images of replicas prepared from an aqueous dispersion of total polar lipids extracted
511
7.6 Biological membranes
insufficient for one covering the entire red cell surface in a bilayer. It was suggested that the discrepancy could be accounted for by nonlipid components if it was assumed that the lipid formed a bilayer.
measurement of membrane area. Furthermore, there is also uncertainty about the packing density of lipid molecules in the surface film at which the area should be measured; in fact, the measurements were made on films with a surface pressure of only 2mN.m–1. Under these conditions, it was found that the area occupied by the lipid at the air–water interface was about twice that of the area of membrane from which it was extracted. This led to the conclusion that the structure of cell membrane was composed of a lipid bilayer. It is now known that the density of phospholipids in a bilayer configuration is somewhat greater (equiva l e n t t o a m o n o m o l e c u l a r fi l m p r e s s u r e o f 30–35mN.m–1) than employed in the original experiments. Moreover, the Gorter and Grendel experiment has been repeated since by other workers to ensure complete lipid extraction and with accurate measurements of membrane area. The results invariably show that the area occupied by the lipids in a surface film is less than that required to form two layers over the entire membrane area. Table 7.3 shows the maximum proportion of membrane area that could be accounted for as lipid in bilayer arrangement. It can be seen that as the proportion of protein to lipid increases in a membrane the amount of lipid that could be accounted for as lipid bilayer decreases. It is important to recognise that no conclusions can be drawn about the arrangement of lipids in biological membranes using this approach. If, however, a bilayer configuration is assumed for the lipid then the remaining area could be accounted for by intrinsic proteins. This is the assumption used in calculating the surface area occupied by lipids in the human erythrocyte membrane (Engelman, 1969). Accordingly, the total volume of membrane occupied by the hydrocarbon chains of the phospholipids was calculated using an average chain length of 16.5 carbon atoms with 1.26 unsaturated bonds per chain. To this volume was added that occupied by the cholesterol, obtained from density measurements, assuming that the entire molecule resides in the hydrophobic region of the membrane. The value obtained was too large for a membrane consisting of a single layer of lipid molecules, but
7.6.2.2 Diffraction studies The principle methods of diffraction analysis rely on examination of assemblies of molecules that are arranged in a regular repeating lattice. This is necessary because scattering occurs only weakly and amplification can be achieved most conveniently with a large ensemble, e.g., crystals. As we have noted above, membranes are said to be fluid, which implies that the molecules are in a state of disorder. The notable exception is the purple membrane, referred to above, which exists in a rigid, twodimensional crystalline array. The purple membrane has proved to be amenable to detailed diffraction studies and the arrangement of the constituent bacteriorhodopsin and lipids has been determined to a resolution of 0.3 nm (Mitsuoka et al., 1999). In these studies, the two-dimensional order in the membrane has been examined using electron diffraction methods since the dose of x-rays required to detect a diffraction pattern from such a small assembly would cause radiation damage to the membrane. This problem is avoided in transmission electron microscopic methods, which have been modified to enable the detection of electron scattering intensities and their subsequent conversion into structural coordinates of the two-dimensional protein-lipid crystals. The intensity with which electrons are scattered varies according to the electrical potential within the structure, which, for practical purposes, is roughly proportional to electron density and, in turn, to atomic number. In this sense the electron diffraction method is analogous to x-ray diffraction, since the Fourier transform of the electron scattering yields the profile of atomic density through the structure. Apart from the usual problems of phase determination, the advantage of the electron scattering method is that very low doses of electrons can be employed (less than 1 electron per unit cell) and, in order to reconstruct the image, information is combined from a large number of unit cells. This avoids radiation damage to the sample, which would otherwise destroy the structure. The strategy for constructing a 3-dimensional image of a 2-dimensional array of molecules is achieved by analysing specimens tilted with respect to the angle of incidence of the electron beam. This method relies on the fact that the Fourier transform of a transmission electron micrograph of a 2dimensional crystal is a central section to a 3-dimensional Fourier transform of the crystal, that is, through a lattice of line in reciprocal space perpendicular to the plane of the crystal. Thus, the amplitudes and phases of Fourier terms along each lattice line can be obtained by combining the Fourier transforms from an appropriate number of different, tilted views of the 2-dimensional array. As in the x-ray diffraction analysis and inverse Fourier transform
TABLE 7.3 Theoretical area of membranes consisting of lipids in bilayer arrangement calculated from area occupied by total lipid extracts at the air/water interface Membrane Plasma membrane Erythrocyte Myelin Acholeplasma laidlawii Mitochondria Inner membrane Outer membrane Endoplasmic reticulum
Protein:Lipid (wt/wt)
Lipid Bilayer (%)
1.50 0.28 1.78
67 103 62
3.55 1.22 0.90
40 72 83
512
Physical Properties: Structural and Physical Characteristics
to produce a model showing that the two ligand-binding α subunits have a different extended conformation from the three other subunits in the closed channel. Furthermore, the acetylcholine-coordinating amino acid side-chains of the α subunits were found to be far apart in the closed channel, indicating that a localised rearrangement, involving closure of loops B and C around the bound acetylcholine molecule, takes place upon activation (Unwin, 2005). Conventional crystallographic methods cannot be used to investigate the structure of biological membranes because they are fluid structures with two rather than three-dimensional order. Likewise, the structure of membrane proteins when extracted from the lipids is not usually amenable to diffraction analysis because they are notoriously difficult to crystallize. Dispersions of membrane lipids, however, form well-recognised phases when dispersed in aqueous systems and these can be characterised from powder reflections. Methods have been developed to orient biological membranes into stacks using Langmuir–Blodgett methods or ultracentrifugation from which powder diffraction patterns can be recorded, but the stacking order is generally low and, hence, analyses tend to be of relatively low resolution. Selective deuterium substitution for protons can provide valuable positional information within the profile using neutron diffraction methods. Electron density profiles through the unit cell of membrane repeats have been constructed from x-ray diffraction data collected from biological membranes that are oriented naturally within the biological tissue, such as myelin and retinal rod outer segment disc membranes, or oriented by processes, such as centrifugation. Studies of a variety of membranes at resolutions of about 1.1 nm have shown the structures to be generally asymmetric with respect to the distribution of electron density. The sarcoplasmic reticulum membrane has been examined in detail by both x-ray and neutron diffraction methods. Electron density profiles for water, lipid, and protein have been constructed from this data that show that the major protein of the membrane, Ca2+ATPase, spans the membrane, and is in contact with water on both surfaces. A major portion of the protein protrudes substantially into the extra vesicular aqueous domain. The extent of this protrusion depends on the functional configuration of the enzyme (Blasie et al., 1990). By the use of deuterated lipids, it was possible to determine the arrangement of the lipid in the form of a bimolecular structure with the polar groups of the lipids residing at the aqueous interface on either surface of the structure and with the hydrocarbon chains extending into the interior. As with the protein, the membrane lipid and water profiles are asymmetric consistent with the complementary asymmetry in the protein mass distribution within the lipid bilayer domain of the sarcoplasmic reticulum membrane profile structure. A model of the structure of the sarcoplasmic reticulum membrane constructed from the diffraction and other data is illustrated in Figure 7.30.
produces a 3-dimensional map of the density distribution within the monomolecular lattice. The picture that emerges from these studies is that of a protein that consists of 7 rods, each of about 4 nm long and 1 nm apart. The rods span the lipophilic portion of the membrane, three approximately vertical to the plane of the membrane and four tilted by 10o to 20o to the vertical. Each rod is an α-helix of the polypeptide chain and, since the complete sequence of the protein is known, it has been possible to show that some of the chain connecting the seven rods is on one side of the membrane and some on the other. Most of the amino acid side chains of the rods are neutral. By contrast, those parts of the protein that are near or extend into the water on either side of the membrane contain 19 charged amino acid residues and the parts that protrude into the water on the outside of the membrane contain 6 charged groups. Within the rods there are 9 charged groups, 5 negative, 4 positive, which may be sufficiently close to neutralise each pair of charges in a predominantly nonaqueous environment. The length of the protein in contact with the lipid is approximately equal to the length 2 molecules of lipid. The individual bacteriorhodopsin molecules are aggregated within the structure into trimers, which produce a three-fold axis of symmetry within the membrane. The lipids of the membrane occupy the interstitial spaces between the protein molecules, including the central region between the protein trimers. The lipids of the purple membrane are typical archibacterial lipids containing phytanyl chains and a major sulphated triglycosyl archiol lipid dominating the composition with minor amounts archiol-based phospholipids reflecting the high salt concentrations of the environment in which these bacteria live. The lipids in the purple membrane, which consist largely of branched-chain phytanylether phospholipids and glycolipid sulphate, are arranged in a bilayer configuration. This is not the arrangement adopted by the lipids when the proteins are removed as revealed by freezefracture electron microscopic studies of the type illustrated in Figure 7.29a. Spin label studies have indicated that the hydrocarbon chains of the lipids in the membrane are relatively immobilized while the lipids in aqueous dispersions have hydrocarbon chains, which are comparatively mobile even towards the ether linkage of the phytanyl chains to the glycerol. All these studies indicate that there is a strong interaction between the lipid and the protein in the membrane of the halophilic bacteria, which leads to a considerable modification of the phase behaviour of the lipids observed in isolated dispersions. Electron diffraction methods have been used to investigate the structure of other membrane proteins including the nicotinic acetylcholine receptor, which is a member of the pentameric “Cys-loop” super family of transmittergated ion channels associated with neuronal acetylcholine receptors, GABAA receptors, 5-HT3 receptors, and glycine receptors. An atomic resolution 0.4 nm was reported 513
7.6 Biological membranes
most conspicuous features are the membrane-associated particles in the fracture planes running longitudinally in the membrane. These were originally identified as enzyme complexes, which were supposedly rearranged during conventional fixation and staining required for thin section electron microscopy. Several lines of evidence were used to establish that the fracture plane occurs through a central region of the membrane rather than along the membrane–water interface. One was the observation that when a layer of radioactively labelled fatty acid applied to a glass surface was sandwiched with another layer of unlabelled fatty acid and thermally quenched, all radioactivity remained on the original plate when the plates were prised apart. This experiment is somewhat removed from the situation in biological membranes and brings into question the nature of the forces that anchor the fatty acids to the substrate surface. More convincing evidence was obtained from etching studies. These showed that by allowing water to sublimate from the fracture surface before shadowing with heavy metal to form the replica different membrane surfaces could be discerned as the water table was lowered. Failure of etching to alter the appearance of the initial fracture surface indicated that this surface was devoid of water and represented a hydrophobic domain within the membrane. Those surfaces of membrane originally in contact with the aqueous medium were exposed by the etching process. Interpretation of features exposed in lateral fracture planes through biological membranes is still conjectural. Evidence that membrane-associated particles consist, at least in part, of intrinsic membrane proteins is supported by observations that removal of lipid by digestion with phospholipases is found to result in an increase in density of particles suggesting that the surrounding smooth areas of fracture face represent lipid. Nevertheless, there are claims that vertical movement of proteins in the membrane matrix is possible so that domains of intrinsic proteins that are normally present in the hydrophobic domain of the membrane do not appear as membrane-associated particles on the longitudinal fracture planes. Figure 7.31 illustrates this, showing a freeze-fracture electron micrograph of the plasma membrane of the bluegreen alga, Synecoccus. The organism had been grown at 38oC at which temperature the distribution of membraneassociated particles was fairly random. The organism from which the replica was made had been cooled to 15°C prior to thermal quenching and this has produced a phase separation of the membrane components. Large regions of smooth fracture plane can be seen that correspond to the creation of gel phase, as judged by calorimetric and x-ray diffraction evidence (Furtado et al., 1979). The conventional interpretation is that the high melting point lipids phase separate into a gel phase from which intrinsic membrane proteins are excluded. By phase separation it is usually inferred that this is a lateral phase
Out
4 nm
In 0
4
8 nm
FIGURE 7.30 A model of the sarcoplasmic reticulum membrane constructed from electron density distribution profiles derived from diffraction studies of oriented membranes. (Adapted from Herbette, L. et al. (1985a), Biochim. Biophys. Acta, 817, 103–122.)
The high brilliance of synchrotron radiation has been exploited to provide information on structural changes that take place in membrane proteins and lipids on timescales appropriate to membrane function. Examination of membranes from halobacterium placed on substrate surfaces have revealed changes in the structure of the bacteriorhodopsin associated with the light-driven pumping of protons that occur on time-scale of ms (Oka et al., 2005). Similarly, synchrotron x-ray diffraction intensities have been recorded through temperature and pressure jumps used to induce structural transitions between phases of hydrated lipid bilayers from which relaxation rates and mechanisms could be determined (Rappolt et al., 2003). 7.6.2.3 Microscopy With the advent of the electron microscope in the 1950s, the ultrastructure of the cell was visualized in considerable detail and the morphology of cells were defined largely by the various cell membranes. The appearance of cell mem-branes in stained thin cross section was uniform and consisted of two darkly staining bands sandwiching an electron lucent layer. The outer bands were said to represent the membrane proteins in an extended β-sheet configu-ration coating a lipid bilayer. Although this model was inconsistent with hydrophobic contact between the proteins and lipids it did recognize that membranes were asymmetric with different proteins located on opposite sides of the membrane. The introduction of freeze-fracture electron microscopy shortly afterwards soon highlighted discrepancies in the unit membrane hypothesis current at the time. This method in which fixation of tissues was achieved by rapid thermal quenching techniques avoided the problems of uncertainty of heavy metal stain deposition. The structure of cell membranes is revealed by examination of templates replicating the topography of surfaces created when the frozen specimen is fractured. The replica is formed by shadowing with a thin layer of platinum, which in turn is supported by a layer of electron translucent carbon. Figure 7.29b shows a replica produced when the fracture plane occurs along the plane of the membrane. The 514
Physical Properties: Structural and Physical Characteristics
Nevertheless, taken together with other evidence, the current assignments of smooth regions to the internal cleavage along the central plane of lipid bilayer and membrane-associated particles as representing intrinsic membrane protein complexes are consistent with the fluidmosaic model of membrane structure. A recent development for visualizing the surface topography of membranes is the atomic force microscope. The principle of the method relies on recording the vertical displacement of a sharp probe tip as it is drawn across the surface of the membrane. Electrostatic and van der Waals’ interactions between the probe and the membrane surface produce the forces deflecting the probe in the vertical (z) direction, which is then amplified by a laser optical system. Scans in the x-y direction give a three-dimensional contour map of the membrane surface with a resolution in the vertical direction approaching 0.1 nm and between 0.5 and 1 nm in the lateral dimensions. Atomic force microscopy has been used to characterise the phase separation of domains in supported lipid bilayer membranes. The resolution is sufficient to demonstrate that bacterial toxins, for example, are targeted specifically to liquid-ordered phase that is organized into rafts in the fluid bilayer membrane (Geisse et al., 2004). The method has also been applied to map the cytoplasmic and periplasmic surfaces of the photosynthetic membrane of the purple bacterium, Rhodobacter, to obtain a model for the arrangement of the light-harvesting and reaction centre complexes (Scheuring et al., 2005). The pigmentprotein complexes are inserted across the lipid bilayer in an arrangement that allows the two light-harvesting complexes to collect the incident light and funnel the energy to a pair of bacteriochlorophylls located in the reaction center. Electrons are then transferred to two ubiquinone receptors reducing ubiquinone to ubiquinol, which is subsequently oxidised by the cytochrome bc1 complex. The net result is a cyclic electron transfer coupled to the generation of a protonmotive force, which is utilized to synthesize ATP. The topography of the native membrane is characterised by two features, small rings about 5 nm in diameter, representing the light-harvesting complexes, in contact with larger S-shaped complexes approximately 10 × 20 nm in size with protruding central proteins which are dimers of the reaction centre complexes. The two surfaces of the membrane can be distinguished by the extent to which the light-harvesting and reaction centre complexes, respectively, protrude from the membrane surface.
Gel
100 nm
FIGURE 7.31 Freeze fracture of a Synocuccus cell membrane showing gel phase separation of membrane-associated particles. The density of particles around the periphery of the phaseseparated domain is not sufficient to account for all particles originally present in the membrane.
separation and this is, in part, true as can be seen by an increase in density of particles around the periphery of the smooth regions of the membrane fracture plane. The additional particles in this high-density region cannot, however, account for all the particles originally present in the membrane. This suggests that membrane proteins have been ejected from the hydrophobic interior of the membrane into the aqueous phase or the constituents of the particles have undergone a reorganization that enables them to accommodate into a lamellar gel phase. From this we may conclude that some intrinsic proteins are arranged in oligomeric complexes and it has been argued that nonbilayer forming membrane lipids not only promote protein complexes, but serve to seal the irregular protein–lipid interface to the passage of solutes (Quinn, 1983). It is clear that intrinsic membrane proteins can integrate into gel and liquid-ordered phases, but not in the same manner as in a fluid matrix. What is perhaps more difficult to explain is that in complementary fracture faces there are no holes or pits corresponding to the particles present in the opposite surface. Plastic deformation of the lipid between the fracture and shadowing operations resulting in obliteration of the pits has been suggested as a possible explanation, but contemporary methods of freeze-fracture make this unlikely. Another problem that is difficult to explain is that the diameter of the particles is often more than twice the thickness of a lipid bilayer (Tsvetkova et al., 1994) and the size of the protein complexes associated with membranes are consistent with these dimensions (Williams et al., 1982). The integration of intrinsic membrane proteins into the bilayer, therefore, is far from clear and we do not yet have an unambiguous interpretation of all the features seen in freeze-fracture replicas of biological membranes.
7.6.2.4 Magnetic resonance spectroscopy Magnetic resonance methods have been used to characterise the polymorphism of membrane lipids as well as provide information on the order, motion, and interaction between lipids in bilayer and other arrange-ments. One of the advantages of nuclear magnetic resonance spectroscopy is that the properties of nuclei with nonzero magnetic moments in natural abundance, such as 1H, 13C, 515
7.6 Biological membranes
and 31P, can be examined in biological samples. Alternatively, isotopic labelling or enrichment with suitable isotopes, such as 2H or 19F, that introduce minimal perturbation into the system can also be used to probe molecular motion and configuration in lipid assemblies. The average conformation of the lipid acyl chains and the polar head groups can be deduced from 2H-NMR methods. Furthermore, specific and nonspecific interactions between molecules in the structure can be characterised. An important parameter that can be obtained is the order parameter representing a measure of the angular distribution of molecules about a preferred molecular orientation. NMR methods exploiting quadrupolar splitting of deuterium nuclei have also been widely used to study lipid polymorphism and the structure of lipids in biological membranes. The versatility of the method relies on the fact that it is possible to substantially substitute 2H for 1H in all domains of the lipid without significantly perturbing the physical properties of the membrane. It is also possible to substitute deuterated water for 1H2O. When deuterated lipids are in an anisotropic arrangement the deuterium quadrupolar resonance is split to an extent that depends on the size of the lipid structures and decreases as motional averaging tends to isotropic. It is possible to derive an order parameter from the extent of splitting of the deuterium resonances for particular residues of lipids in different polymorphic phases. An example of the use of deuterium magnetic resonance methods to define the structure of lipids in a biological membrane is illustrated in Figure 7.32. This shows a 2HNMR spectra, recorded at 30oC, of a fatty acid auxotroph of Escherichia coli grown on 2H-oleic acid together with a spectrum of an aqueous dispersion of a total lipid extract of the cells recorded under the same conditions. A rather broad powder pattern is observed for the deuterons incorporated into the cell membranes with two peaks separated by a frequency, ∆υq of 6.7 kHz and a sharp isotropic signal, which can be assigned to a small amount of residual HOD in the sample. The value of ∆υq is related to the segmental order of the C-D bond and the large intrinsic line width is typical of biological membranes and reconstituted membrane systems. The 2H-NMR spectrum obtained from an aqueous dispersion of a total lipid extract of the cells is characterised by two sharp peaks, separated by 9.9 kHz, which is considerably greater than ∆υq of the lipids in the membranes. The spectrum is typical of lipids in a liquid-crystalline bilayer configuration. The general conclusion from studies of this type is that the lipids in biological membranes are arranged in a bilayer with the motion of the lipids constrained by their interaction with other membrane components. 31P-NMR signals from phospholipid assemblies can be used to characterise the arrangement of the constituent molecules and the particular phase that is formed. The information is obtained from the line shapes of proton decoupled spectra, which indicate chemical shielding
(a)
10
0
–10
kHz (b)
FIGURE 7.32 2H NMR spectra recorded at 30°C of (a) a suspension of a fatty acid auxotroph of Escherichia coli grown on [2H]-oleic acid, and (b) an aqueous dispersion of a total lipid extract from the organism used to record spectrum (a). (Data from Killian, A.J. et al. (1992) Biochim. Biophys. Acta, 1105, 253–262.)
anisotropy of the lipid phosphate moiety. Thus, different orientations of the phosphate segment yield characteristic resonances at different frequencies. In the case of lipids in a bilayer configuration, the 31P-NMR spectrum is typical of a shielding tensor that is axially symmetric around a director axis consisting of a low-field shoulder and highfield peak. Oriented bilayers give rise to individual narrow resonances with an angular-dependent chemical shift. The interpretation of these features is that molecules with their long axis perpendicular to the magnetic field contribute to the high-field peak and molecules with a parallel orientation correspond to resonances at a position in the shoulder region and that the director axis coincides with the bilayer normal about which the phosphate segment rotates. The rate of reorientation resulting from Brownian motion of lipid molecules in bilayers can also be determined from the line shape. When the rate of reorientation of the phosphate segment is fast, the line shape narrows until eventually, when motion is effectively isotropic on the timescale of the measurement (5.6 the surface of the protein interfaced with the lipid 518
Physical Properties: Structural and Physical Characteristics
includes 10 putative transmembrane domains. An ATP binding site and a phosphoenzyme formation site are located within the largest cytosolic loop, whereas a sequence implicated in the coupling to transport activity was identified in another hydrophilic loop. The sequence of mammalian ATPase II is homologous to a yeast ATPase encoded by the drs2 gene, of which the null mutant of a yeast strain lacks a specific phosphatidylserine internalization activity that is otherwise present in wild type yeast strains. Four isoforms of the ATPase II enzyme have been identified in bovine brain. The substrate specificities and selectivities were characterized and the results are summarised in Table 7.4. This shows relative specificities of phosphatidylcholine and phosphatidylethanolamine compared with phosphatidylserine for activation of the ATPase by different isoforms of recombinant ATPase II. It was found that there was no significant difference between the presence or absence of phosphatidylcholine suggesting that this phospholipid is not an activator of ATPase activity. The ratio of activation phosphatidlyserine/phosphatidylcholine shown in the table is a reliable indication of the relative specific activity of the different ATPase II isoforms. The activity is in the order α2 > α1 > β2> β1. A different order is observed in relative selectivity of phosphatidylserine compared with phosphatidylethanolamine in which the order is α2 > β2 > α1 > β1. The involvement of Drs2 protein in the transbilayer movement and distribution of phospholipids in the plasma membrane of the S. cerevisiae end4∆ mutant in which both growth and the internalization step of endocytosis are blocked at a restrictive temperature of >34°C has been investigated (Pohl et al., 2005). It is known that several proteins, such as ATP-binding cassette multidrug transporters Pdr5 and Ste6, accumulate in the plasma membrane of the yeast after incubation under nonpermissive conditions and it was thought that this may have consequences for the organization and dynamics of the plasma-membrane lipid phase. It is known that the Ste6 protein transports synthetic alkyl phospholipids, but not whether the protein-mediated transbilayer movement and the transbilayer asymmetry of lipids in the plasma membrane of the S. cerevisiae end4∆ mutant are altered. The transbilayer movement of fluorescently labelled analogues of choline and serine phosphatides has been measured in an S. cerevisiae end4∆drs2∆ strain, to determine whether
presented a positively charged domain that presumably represented sites of interaction with negatively charged lipids. Nevertheless, the dissociation constants of spectrin for binding to neutral and negatively charged phospholipid interfaces could not be distinguished suggesting a multipoint attachment of the whole protein to the interface with specific domains on the protein preferentially interacting with phosphatidylserine. Another example of preferential interaction of membrane-associated proteins with acidic phospholipids is dynamin. Dynamin acts at the cytoplasmic surface of the plasma membrane to effect receptor mediated endocytosis. The protein locates at the neck of deeply invaginated coated pits where it induces membrane constriction and the pinching off of endocytotic vesicles. The action is mediated by a conformational change in the protein driven by GTP hydrolysis. Studies of the binding of dynamin to different phospholipid monolayers has revealed that the protein binds very strongly to acidic phospholipids, particularly phosphatidylinositol and phosphatidic acid. The binding to these phospholipids is said to induce penetration of the protein into the hydrophobic domain of the lipid, causing destabilization of the membrane and fission (Burger et al., 2000). While examples such as these provide evidence of strong interactions of negatively charged membrane lipids with membrane proteins, the role of these interactions in maintaining asymmetric distribution of lipids across biological membranes is unclear. In any event such effects are likely to be of minor importance relative to actively mediated phospholipid translocation processes. Active translocation of phospholipids across the plasma membrane has been demonstrated both from the inner to the outer leaflet and from the outer to the inner leaflet. The translocation processes specifically transport phosphatidyserine and phosphatidylethanolamine from the outer to the cytoplasmic surface of the membrane, while choline phosphatides are transported from the cytoplasmic to the outer surface. The rate of translocation, in general, is greater for the amino phospholipids compared with the choline phospholipids. The aminophospholipid translocase is an ATPase IItype enzyme that requires Mg2+ and is activated by phosphatidylserine and to a lesser extent by phosphatidylethanolamine and is sensitive to the sulphydryl group reagent, N-ethylmaleimide. Because the enzyme is inhibited by vanadate ions, it is categorised as a P-type ATPase. The properties with respect to substrate specificity, biochemical, and physical properties of the enzyme and strategies for identification and isolation from various sources have recently been reviewed (Daleke, 2003). The protein has a molecular weight of about 116 kDa and has been isolated from the plasma membranes from various sources. Sequence data of ATPase II from human, mouse, and bovine tissues indicates that the protein has several P-type consensus sequences and a membrane topology that
TABLE 7.4 Relative phospholipid specificities for activation of different isoforms of recombinant ATPaseII Relative Specificity PS/PC PS/PE
α1
α2
β3
β4
53.0 2.7
68.0 10.9
25.0 1.3
32.5 3.5
PS, phosphatidylserine; PC, phosphatidylcholine; PE, phosphatidylethanolamine. Source: Data from Ding, J., et al. (2000), J. Biol. Chem., 275, 23378–23386.
519
7.6 Biological membranes
the use of membrane lipid analogues that incorporate para-magnetic, fluorescent, or other probes and it is important to recognise that it is the motion of the probe that is observed and not the native molecules. It is not unknown for probe molecules to exert a considerable perturbation on the motion and structure of the surrounding lipids or to behave in a manner distinctly different from the lipids they are designed to mimic. Another problem that also has to be addressed is the need to introduce the probe into the membrane without disturbing the structure. The most common method is to incubate cells or membrane suspensions in a medium into which a small aliquot of lipid probe has been injected from a concentrated solution in organic solvent, such as ethanol. Alternatively the probe may be codispersed with unlabelled phospholipid or mild detergent. The strategy of choice often depends on whether the probe increases or decreases the hydropathy of the lipid analogue compared to the native lipids, which comprise the membrane. Possibly the least perturbing method is the use of phospholipid transfer proteins to perform the task, but this method relies on close physical resemblance between the analog and the phospholipid for which the protein has affinity. Nevertheless, with these restrictions in mind the general picture that has emerged is that lipid molecules are relatively mobile within the membrane and constraints on this motion occur by interactions with the proteins and other membrane constituents.
internalization of analogues by transbilayer movement depended on a functional drs2 gene in endocytosis-deficient yeast cells. It has been shown that exposure of endogenous aminophospholipids to the exoplasmic leaflet of the mutant cells is altered with respect to wild-type cells. There is also a family of membrane P-glycoproteins involved in transmembrane phospholipid movement. The translocase responsible for nonspecific movement of phospholipid from the cytoplasmic leaflet to the cell surface has been shown to be identical to the multidrug resistance protein-1. The proteins appear to translocate particular phospholipids on different cell types (Sugawara et al., 2005). Sphingolipids, for example, are preferentially transported across the plasma membrane of LLC-PK1 kidney epithelial cells transfected with MRP1 cDNA, but phosphatidylcholine was the preferred phospholipid translocated by MRP3 in bile canicular cells where the protein predominates in the plasma membrane of fibroblasts of transgenic mice. The loss of membrane phospholipid asymmetry is known to be an important signalling mechanism as it results in the appearance of specific phospholipids on opposite sides of the membrane, a situation that triggers a variety of cellular responses (Zwaal et al., 2005). A family of membrane proteins referred to as phospholipid scramblases (PLSCR) have been implicated in dissipating phospholipid asymmetry in a process that depends on elevated cytoplasmic calcium concentration. Four genes have been identified that code for phospholipid scramblases in human and mouse and all have been conserved through evolution. The amino acid sequence deduced for one of the human phospholipid scramblases, HuPLSCR1, indicates that the scramblase is a Type-2 membrane protein of molecular weight 35 kDa (318 amino acid residues) with a transmembrane helical domain located towards the C-terminus (Wiedmer et al., 2000). The N-terminal cytoplasmic domain was found to possess a putative phosphorylation site and a calciumbinding segment. There is evidence that regulation of scramblase activity may be mediated by phosphorylation in the presence of PKCδ, the action of which results in the surface exposure of phosphatidylserine in apoptosing cells (Frasch et al., 2000). Regulation of scramblase activity and its localization in the nucleus may also occur via palmitolylation at cysteine thiol residues of the protein (Ben-Efraim et al., 2004).
7.6.4
7.6.4.1 Motion of lipids in membranes The two methods that have been exploited to greatest effect to examine lateral diffusion of lipids in the plane of biological membranes and from one leaflet of the membrane lipid bilayer to the other are electron paramagnetic resonance and fluorescence probe techniques. A range of probe molecules have been synthesized as analogs of membrane lipids. One example is the synthesis of Lucifer yellow derivatives of phospholipids and cholesterol, which have been introduced into human erythrocyte ghost membranes and living melanoma cells to measure lateral diffusion rates. The rates of diffusion were then compared with diffusion in phospholipid bilayer membranes. Lateral diffusion rates were obtained using a spot fluorescence photo-bleaching recovery method. The method involves bleaching of probe molecules in a defined surface area of membrane using an intense pulse of laser light and monitoring recovery of fluorescence as unbleached probe molecules from the surrounding area of membrane diffuse into the bleached domain. It was found that Lucifer yellow derivatives of cholesterol and phosphatidylethanolamine diffused rapidly with a diffusion coefficient >1 µm2 s–1 in lipid dispersions at temperatures greater than the gel to liquid-crystalline phase transition temperature, but the diffusion rate decreased dramatically for probes in gel
Membrane lipid dynamics
With the notable exception of lipids in the purple membrane of Halobacterium, which adopt more or less a crystal structure, the lipid matrix of biological membranes is said to be fluid. The term fluidity is not a precise quantitative parameter, but it implies that the molecules of the structure exhibit motion with respect to one another. The measurement of this motion is often performed by 520
Physical Properties: Structural and Physical Characteristics
phase lipid ( T > TC
T > TH (a)
(b)
T > TH
TC > T
Bilayer-forming lipid (d)
Non-bilayer-forming lipid (c)
FIGURE 7.38 Illustration of the lipid phase separation model for low temperature to cell membranes. The scheme shows the consequences of cooling from the growth temperature (a) to a temperature below the hexagonal-II to liquid-crystalline phase transition temperature of the nonbilayer forming lipids (b) and subsequently below the gel to liquid-crystalline phase transition of all the membrane lipids (c). The effect of reheating to the growth temperature is shown in (d). (Data from Quinn, P.J. (1985), Cryobiology, 22, 128–146.)
cultured at 28oC and 38oC that occur at 2o and 4oC, respectively. This observation clearly shows that it is the high melting point lipids that phase separate from the membrane proteins in the intact biological membrane. The question remains as to which lipids have higher melting points and which lipids have low melting points. This has important implications with regard to the ability of the membrane to restore a random distribution of components after thermally induced phase separations. It is well known, for example, that the phase separations of the type observed in Synococcus result in irreversible changes and loss in viability of the cells, which are unable to repair their leaky membranes. It is seen that the lipid composition of Synococcus is relatively simple, and appreciable proportions of the membrane lipids undergo phase transitions over a relatively narrow range of temperatures. With more complex mixtures, typical of many biological membranes of higher organisms, the transition endotherms observed on heating membranes previously cooled to low temperatures exhibit transitions that are invariably broad and extend over tens of degrees. Attempts have been made to resolve these broad transitions in erythrocyte membranes and lipid mixtures simulating the membrane lipid composition into components contributed to by each of the major lipid classes present in the membrane. The results of such experiments using human erythrocyte phospholipids have shown that the phase separation of the sphingolipids dominated the higher temperature phase behaviour of the outer leaflet of the membrane, and the phosphatidylethanolamines appear to phase separate initially in the inner leaflet upon cooling the membrane. 524
Physical Properties: Structural and Physical Characteristics
at higher temperatures causes a phase separation of nonbilayer lipids into stable aggregates of cylindrical inverted micelles. Interpretation of the effects of temperatures greater than 45oC is based on phase conditions that result in a release of the constraints imposed by interaction of the major nonbilayer-forming lipid of the membrane, monogalactosyldiacylglycerol, with other membrane components and its segregation into domains of nonbilayer lipid structure. Gross phase separations of this type require that the shift in thermal stability of the stacked membrane is relatively large because three-dimensional aggregates of lipid are not observed if the chloroplast membrane is destacked by manipulation of the ionic environment before heat treatment. It has been suggested that a shift in the phase of the nonbilayer lipid, monogalactosyldiacylglycerol, underlies these structural changes and that the functional role of this nonbilayer lipid may be to package the light-harvesting chlorophyll a/b-protein complexes together with the photosystem-II core protein complex into an efficient functional unit localised within the grana stack. All membranes contain molecular species of lipid that at physiological temperatures do not form bilayer structures. These include phosphatidylethanolamine, monoglycosyldiacylglycerol, and cardiolipin. While it is generally assumed that such lipids are constrained into a bilayer arrangement by interaction with other membrane constituents, their phase separation to form inverted micelles is an event that is thought to be associated with membrane fusion. Recent NMR evidence, however, has identified lipids undergoing isotropic motion in intact cell membranes that are said to be associated with raft domains (Ferretti et al., 2003). Motion of this type may be a feature of the boundary region separating fluid from liquidordered domains where packing faults and molecular mismatches are expected to occur.
Non-bilayer lipid structures 100 nm
FIGURE 7.39 An electron micrograph of a freeze-fracture replica prepared from bean chloroplasts heated for 5 min at 55°C before thermal quenching.
that the higher melting point lipids present in gel-phase domains will be rich in the hexagonal-II-forming lipids. This assumption is based on the fact that, with equivalent hydrocarbon substituents, lipids that tend to form hexagonal-II structures have liquid-crystalline to gel-phase transitions at temperatures that are considerably higher than those corresponding to bilayer-forming lipids. Where phase separations between lipid classes of this type are created, the changes are not likely to be reversed on reheating to temperatures above the hexagonal-II to lamellar-phase transition temperature. In practice, it would be improbable for there to be a single temperature where all the nonlamellar lipids would co-exist together with the bilayer-forming lipids in the fluid phase, allowing the original distribution of the lipids and the proteins to be restored. Damage to the membrane in these circumstances would be expected to result when the membrane is reheated to temperatures where the domains of the phase-separated hexagonal-II-rich lipids tend to form nonbilayer structures. The creation of nonbilayer structures, such as inverted lipid micelles, unless suppressed or dealt with by normal homeostatic mechanisms operating within the membrane, would serve to destroy the permeability barrier properties of the membrane. If this breakdown of membrane barrier is of sufficient duration to permit loss of essential components or irreversible alterations in the intracellular compartmentation, loss of cell viability will result. Phase separations of the constituents of biological membranes can also be driven by exposure of membranes to salts or pH that screen charges on the acidic phospholipids as well as to high temperatures. The structural changes in thylakoid membranes of higher plant chloroplasts subjected to thermal stress is illustrated in Figure 7.39. Chloroplasts maintain a normal morphology during a brief exposure (5 min) to temperatures up to 35oC. Incubation at temperatures of between 35o and 45oC causes complete destacking of the grana and incubation
7.6.4.4
Domain creation by lipid–lipid interactions
Great emphasis has more recently been placed on lipid phase separations that are brought about by the association between particular membrane lipids to create lateral domains within the bilayer. The phase separation in these instances is driven by the order created in the bilayer by the formation of complexes. Complexes have been described between phosphatidylcholine and diacylglycerol, phosphatidylcholine and α-tocopherol, phosphatidylethanolamine and glucosylceramide, and choline phosphatides and cholesterol. The ordered phase in each case depends subtly on the molecular species of lipid involved in the complex and the extent of order created within the phase-separated domains again depends on the molecules involved. Diacylglycerol accumulates transiently in cell membranes as a consequence of phospholipase C-type enzymes activated by a variety of hormones, growth factors, and neurotransmitters. The amount of diacylglycerol formed 525
7.6 Biological membranes
that result in repackaging of intrinsic proteins or membrane fusion. Diacylglycerol generation from proteolipid complexes containing polyphosphoinositol lipids by intranuclear phospholipase C is a recognised signalling pathway in the control of gene expression (Martelli et al., 2003). These complexes have been localized to interchromatin granule clusters by immunostaining with monoclonal antibodies in preparations that have been extracted with cold TritonX100. On this basis it is said that the complexes do not originate from the membrane bilayer lipid matrix of the nuclear envelope; however, the possibility that they represent detergent-resistant membrane domains of the envelope cannot be discounted on the present evidence. The consequence of the presence of ceramide in membranes is dominated by its tendency to self assemble into rigid bilayer structures especially in the presence of sphingomyelin and cholesterol. This has led to the suggestion that ceramides play an important role in membrane signaling processes, membrane fusion, and permeability of solutes through bilayers (Goni et al., 2005). Evidence that is consistent with these functions is that ceramides are recovered in significant amounts in detergent-resistant membrane fractions even from cells in an unstimulated state. Ceramide domains are more hydrophobic than the parent sphingolipid domains, promoting membrane fusion. In relatively minor proportions (3 to 6 mol%), ceramide synergises with cholesterol in stabilising ordered lipid domains. Ceramide alone is able to promote order in the lipid chains of phosphatidylcholine, even with monounsaturation in the sn-2 chain, and it partitions into liquid-ordered domains with an affinity markedly higher than that of other (fatty acid-matched) sphingolipids (Wang, 2003). The generation of ceramide in membranes takes place at different subcellular sites. De novo biosynthesis occurs at the endoplasmic reticulum, whereas at the plasma membrane ceramide is derived by hydrolysis of sphingomyelin and complex sphingoglycolipids. The hydrolytic enzymes responsible for production of ceramide at the plasma membrane are designated by their pH optima into acid and neutral sphingomyelinases. Cellular modulators of ceramide production include diacylglycerol and other protein kinase C activators and serine proteases. Degradation of ceramide by ceramidases is enhanced by a variety of agents including cytokines, cell differentiating agents, death receptor ligands, cancer chemotherapeutic agents, and ionizing radiation (Jaffrezou et al., 2002). It has been proposed that the cellular response to ceramide elevation or depletion depends directly on where the ceramide is located (Blitterswijk et al., 2003). According to this view ceramide in the Golgi influences biosynthetic pathways in formation of sphingolipids and their transport in vesicles to the plasma membrane where they promote domain formation. Clustering of receptors on the cell surface, promoted by rafts, induces endocytosis in
can reach 2 mol% in some physiological situations and the resulting change in membrane lipid composition represents a molecular signal transducing mechanism responding to the interaction of agonists with their respective cell surface receptors. Structural changes in the lipid bilayer have been detected in phospholipids containing diacylglycerol and these physical perturbations may be responsible for the molecular signal, such as activation of protein kinase C or to enhance membrane fusion processes. Diacylglycerols when codispersed with disaturated phosphatidylcholines in molar ratios of up to 30 mol% create lateral phase separated domains of a stoichiometric 1:1 complex of diacylglycerol and phosphatidylcholine within bilayers of pure phospholipid (Quinn et al., 1995). It may be expected that physiological proportions of diacylglycerol may result in lateral phase separations in small domains within membranes. The domains created by the demixing process tend to destabilize the phospholipid bilayer. The presence of much higher, and unphysiological, proportions of diacylglycerols (greater than 30 mol%) is known to induce nonbilayer phases such as hexagonal-II (HII) and cubic phases in disaturated phosphatidylcholines. The transient nature of lamellar phase complexes between diacylglycerol and membrane phospholipids is necessary in the process of switching off the signal generated by the formation of diacylglycerol. This is accomplished by diacylglycerol kinase, which converts the substrate into phosphatidic acid. Control is presumably exercised by the molecular species of diacylglycerol that are generated from the parent phospholipid. Studies on the effect of acyl chain length of diacylglycerol on activation of protein kinase C has shown that diacylglycerol with shorter hydrocarbon chains (C6 to C10) are more effective than molecular species containing longer acyl chains. The miscibility of the short-chain diacylglycerols with the phospholipid in liquid-crystalline bilayers could be the explanation for the high activity of the short-chain diacylglycerol. By contrast, longer chain diacylglycerols form complexes with the phospholipids, which exist in gel phase at temperatures higher than the fluid phase transition of the phospholipid and are unable to activate protein kinase C. Complex formation has been characterised in binary mixtures of both dipalmitoylphosphatidylcholine/ dipalmitoylglycerol and dimyristoylphosphatidylcholine/ dimyristoylglycerol. In terms of the effect of diacylglycerol on domain formation in membranes it is clear that this is highly dependent on the molecular species of diacylglycerol and the degree of unsaturation of the polar lipids comprising the membrane lipid bilayer matrix. Interaction and complex formation between saturated molecular species may generate liquid-ordered domains, whereas interaction between unsaturated molecular species tends to destabilize the bilayer in favour of the formation of inverted phases 526
Physical Properties: Structural and Physical Characteristics
surface, whereas phosphatidylcholine has only acceptors (esteric O). The principal hydrogen bond site appears to be with the -NH group of the sphingomyelin, which can be detected indirectly by effects, such as the accessibility of 3-OH group of cholesterol to cholesterol oxidase, the extractability of cholesterol from the membrane, or the effect of cholesterol on interfacial elasticity. The multiple hydrogen-bonding propensity of the sphingomyelin molecule results in the formation of relatively large hydrogen-bonded clusters (4 to 8 molecules) of sphingomyelin in the membrane (Mombelli et al., 2003). Cholesterol can only form a single hydrogen bond and so acts as a terminator of the hydrogen bonding networks. The gain to the membrane of the single hydrogen bond formed with cholesterol is at the expense of larger clusters of hydrogenbonded sphingolipids, and so makes a relatively minor contribution to membrane properties. The condensing effect of cholesterol on phospholipids to create a liquid-ordered phase can also be viewed from the simple action of the sterol in spacing out the phospholipid molecules at the bilayer interface. The polar groups of the phospholipids tend to repel each other via charge–charge interactions and this effect is reduced by the interposition of uncharged cholesterol molecules. Computer modeling of the interacting molecules shows that cholesterol makes closer contacts with the sn-1 rather than sn-2 fatty acid chain of glycerolipids, which explains why unsaturation is allowed in the sn-2 chain. Moreover, such studies show that the main interaction is not that of cholesterol with the lipid chains, but of the tightly packed lipid chains with each other (Rog and Pasenkiewicz-Gierula, 2001). Cholesterol acts as a hydrophobic spacer that allows the saturated lipids to pack tightly and thereby interact more with each other along their entire length. Presumably constraint upon the flexibility of the first C12 atoms of the fatty acid chains by the presence of cholesterol promotes van der Waals’ interactions between the chains. The small -OH head group of cholesterol is probably of key importance in allowing room for tighter packing of the bulky phosphocholine head group, which can orient on the bilayer surface to prevent exposure of the hydrophobic region of the bilayer to water. One of the material advances in understanding the role of liquid-ordered phases in cell biology has been the isolation of these phase-separated domains from living cells. This is achieved by exploiting the relatively low solubility of the liquid-ordered phase in weak nonionic detergents. The fraction of membrane that survives solubility can be separated by density gradient centrifugation from the remaining solubilized membrane components. The isolation of detergent-resistant membrane has been modeled using giant unilamellar vesicles containing a lipid composition in which liquid-ordered phase is created in a matrix of fluid lipid. An example is shown in Figure 7.40. Giant unilamellar vesicles formed from a mixture of phosphatidycholine, sphingomyelin, and cholesterol (mole ratio
response to transient ceramide formation. Finally, changes in membrane permeability associated with the action of ceramidases on mitochondrial and plasma membrane are said to mediate apoptotic events. Ceramides are known to be one of the main mediators of apoptosis in cells. Increasing the levels of ceramide accumulating during inhibition of ceramidases, which convert ceramide to sphingosine and free fatty acid, results in cell death. Likewise, inhibition of ceramide production by blocking the de novo synthesis pathway or inhibiting neutral sphingomyelinase activity slows down apoptosis in response to a variety of factors including chemotherapeutic agents, tumor necrosis factor-ß, angiotensin-II, and B-cell activation (El Bawab et al., 2002). Perhaps the most studied of the complexes is that between choline phospholipids and cholesterol. Cholesterol interacts with lipids containing long-chain (>C14) saturated fatty acids in a manner dominated by van der Waals interactions. In order for van der Waals’ forces to operate, the sterol and lipid must fit closely. This steric constraint requires of the sterol that it have the 3-OH head group, the sterol rings must be planar, the hydrocarbon tail must be of the appropriate length, and it should be the natural enantiomer (Miao et al., 2002). On the phospholipid side, the primary requirement is for long, saturated lipid chains; in particular, cis double bonds are poorly tolerated in that part of the fatty acid chain (approximately the first 12 carbons) that interacts with the planar sterol rings. Glycerolipids, made in the endoplasmic reticulum, often have 1 to 4 double bonds in this region of their sn-2 acyl chain, whereas sphingolipids (made in the Golgi) are either fully saturated or have a double bond at the C15 position where it is accommodated by the flexible hydrocarbon tail of cholesterol. However, this requirement for saturation applies strictly only to the fatty acid in the sn-1 position for a glycerolipid; a single double bond, even in a relatively superficial position abutting the sterol rings of cholesterol, is tolerated nearly as well as a fully saturated chain in this position. It is the presence of multiple double bonds in one chain, or unsaturation in both chains, that strongly prevents condensation. The range of membrane lipids that can condense with cholesterol and partition into a liquid-ordered (LO) phase is much greater than is commonly assumed. This includes, in particular, the glycerolipids of the inner leaflet of the plasma membrane, and of internal membranes within the cell, provided they meet the condition of saturation in the sn-1 chain and no more than one double bond in the sn-2 chain. The fatty acid attached at sn-1 to glycerolipids is normally fully saturated except in certain tissues, such as brain and spermatozoa, and sphingolipids have a linear hydrocarbon chain at this position as part of the sphingosine base. Interactions between cholesterol and phospholipids are also believed to take place through hydrogen bonds. Sphingomyelin contributes both hydrogen bond acceptors (esteric O) and donors (-OH and -NH) to the membrane 527
7.6 Biological membranes
(a)
(b)
produces a marked effect on membrane characteristics as its mol% in a monolayer or bilayer of saturated lipids rises above 20 mol% to a maximum of 50 mol% after which it forms a separate cholesterol-only phase. Normal homeostatic processes operate to ensure that the level of cholesterol in the plasma membrane is strictly regulated and can only be manipulated experimentally by removing cholesterol (typically with methyl ß-cyclodextrin or cholesterol oxidase). However, red blood cells have only one membrane, which is exposed continuously to a rich supply of cholesterol in the blood plasma. In membrane rafts isolated from both human and goat red blood cells (ruminant cells have twice as much sphingolipid as human red cells), there is a strict 1:1 stoichiometry between cholesterol and sphingomyelin, although in these cells the cholesterol level is higher in the solubilised (disordered) membrane than in the rafts (Koumanov et al., 2005). This suggests that lipid-ordered microdomains impose an absolute equimolar stoichiometry between cholesterol and the saturated lipid, and as readily exclude excess sterol as they include it when needed. Although the condensing effect of cholesterol on lipids becomes marked at >20 mol% cholesterol, other effects occur at concentrations too low (60%) of 2-ethylhexanol is to serve in the production of plasticizers, such as diethylhexyl phthalate (DEHP; dioctyl phthalate) and diethylhexyl adipate (Bahrmann et al., 2001). Other uses of 2-ethylhexanol include the production of 2-ethylhexyl acrylate, which in turn is used in coating materials, adhesives, and inks, 2-ethylhexylnitrate used as cetane improver additives for diesel fuel, and 2-ethylhexylphosphates as an additive for lubricating oils (Bahrmann et al. 2001). An alternative synthesis of the corresponding acid, 2-ethylhexanoic acid, is achieved by hydrogenating the δ-lactone (depicted in Figure 8.61) in presence of a watersoluble rhodium-phosphine catalyst (Behr and Brehme 2002). The δ−lactone is obtained from a reaction of butadiene with carbon dioxide.
disproportionation of aldehyde to alcohol and acid, which terminates the chain reaction initiated in the previous step. More details on the mechanism are given in the literature (Veibel and Nielsen, 1967; Burk et al., 1985a). The Guerbet reaction can be applied on a commercial scale, therefore, and due to their interesting properties, Guerbet compounds are of significant commercial interest. Other Guerbet Compounds. Guerbet alcohols themselves can be used to prepare other derivatives. A straightforward example is Guerbet acids obtained by oxidation of Guerbet alcohols, for example, using Pd, Pt, or Ru catalysts (Behr and Döring, 1992). Guerbet alcohols also can be esterified with a variety of carboxylic acids, such as fatty acids, or by Guerbet acids, the latter being termed di-Guerbet esters (Knothe and Carlson, 1998). Carbonates synthesized from Guerbet alcohols (Kenar et al., 2004, 2005) and esters of fatty and dicarboxylic acids with Guerbet alcohols (Knothe, 2001) have also been reported. The patent literature provides information on various derivatives of Guerbet compounds, including alkoxylated Guerbet alcohols and esters, Guerbet amines, betaines, branched amine oxides, carbonates, esters of meadowfoam oil and ricinoleic acid (castor oil), fluorinated citrate esters, lactams, polyoxyalkylene glycol esters, sorbitan esters, etc. A list of some patents in this area through 2001 is provided in the literature (O’Lenick, 2001). Alcohols other than the typical ones shown in Figure 8.60 can be obtained from a Guerbet reaction. Secondary alcohols have been condensed under Guerbet conditions (Burk et al., 1985b). Several cross-condensations of mixtures of alcohols in Guerbet reactions have also been reported. These include (1) reaction of methanol with 1-butanol or 1-pentanol in presence of a rhodium catalyst (Burk et al., 1985b); (2) synthesis of 1-propanol and 2-methyl-1-propanol from ethanol and methanol using magnesium oxide (Ueda et al., 1990); and (3) synthesis of iso-butanol from methanol and n-propanol using various metal-based catalysts (Carlini et al., 2002, 2003a/b, 2004, 2005) and a related system for the condensation of methanol and ethanol or methanol, ethanol, and n-propanol (Carlini et al., 2003c). Reactions of fatty alcohols of chain length C8-C14 and benzyl alcohol yielded 2-benzyl fatty alcohols (Krause and Syldatk, 1985). Other classes of compounds can be subjected to reactions under Guerbet conditions. The reaction of amines under Guerbet conditions gave imines as major product (Miller, 1960).
8.18.1
Properties and Applications
CH3
Their wide liquidity range documented by melting points (especially) and boiling points is one reason why Guerbet compounds have significant practical applications. For example, 1-octanol has a melting point of −16.7°C and a boiling point of 194.4°C. The Guerbet alcohol with the same number of carbon atoms, 2-ethylhexanol, has a melting point of −70°C and a boiling point of 184.6°C.
H2C O
FIGURE 8.61
588
O
Chemical Properties
Carlini, C. et al. Guerbet condensation of methanol with npropanol to isobutyl alcohol over heterogeneous bifunctional catalysts based on Mg-Al mixed oxides partially substituted by different metal components, J. Molec. Catal. A: Chem., 232, 13–20, 2005. Gast, L.E. et al. Reactions of unsaturated fatty alcohols. VI. Guerbet reaction of soybean and linseed alcohols, J. Am. Oil Chem. Soc., 35, 703–707, 1958. Guerbet, M., Action de l’alcool amylique de fermentation sur dérivé sodé, Comptes rendus, 128, 511–513, 1899. Kenar, J.A. et al. Synthesis and characterization of dialkyl carbonates prepared from mid-, long-chain, and Guerbet alcohols, J. Am. Oil Chem. Soc., 81, 285–291, 2004. Kenar, J.A. et al. Physical properties of oleochemical carbonates, J. Am. Oil Chem. Soc., 82, 201–205, 2005. Knothe, G. and Carlson, K.D., Synthesis, mass spectrometry, and nuclear magnetic resonance characterization of di-Guerbet esters, J. Am. Oil Chem. Soc., 75, 1861–1866, 1998. Knothe, G., Characterization of esters of fatty acids and dicarboxylic acids with Guerbet alcohols, J. Am. Oil Chem. Soc., 78, 537–540, 2001. Knothe, G., Synthesis, applications, and characterization of Guerbet compounds and their derivatives, Lipid Technol., 14, 101–104, 2002. Krause, H.-J. and Syldatk, A., Neue Tenside aus gemischten Guerbet-Alkoholen, Fette, Seifen, Anstrichm., 87, 386–390, 1985. Markownikoff, W. and Zuboff, P., Ueber die Condensation höherer Alkohole: Tricaprylalkohol, Chem. Ber., 34, 3246–3249, 1901. Miller, R.E., The Guerbet reaction. I. The reaction of amines under guerbet conditions, J. Org. Chem., 25, 2126–2128, 1960. O’Lenick, A.J., Guerbet chemistry, J. Surfact. Deterg., 4, 311–315, 2001. Sulzbacher, M., The Guerbet reaction of cetyl alcohol, J. Appl. Chem., 5, 637–641, 1955. Ueda, W. et al. A low-pressure Guerbet reaction over magnesium oxide catalyst, J. Chem. Soc., Chem. Commun., 1558–1559, 1990. Veibel, S. and Nielsen, J.I., On the mechanism of the Guerbet reaction, Tetrahedron 23, 1723–1733, 1967.
References Bahrmann, H. et al. 2-Ethylhexanol, in Ullmann’s Encyclopedia of Industrial Chemistry, 6th ed., Wiley-VCH, Weinheim, Germany; online version, 2001. Behr, A. and Brehme, V.A., Homogeneous and heterogeneous catalyzed three-step synthesis of 2-ethylheptanoic acid from carbon dioxide, butadiene and hydrogen, J. Molec. Catal. A, 187, 69–80, 2002. Behr, A. and Döring, N., Synthesis of branched fatty acids by catalytic oxidation of alcohols. (Herstellung verzweigter Fettsäuren durch katalytische Oxidation von Alkoholen.), Fat Sci. Technol., 94, 13–18, 1992. Burk, P.L. et al. The rhodium-promoted Guerbet reaction. Part I. Higher alcohols from lower alcohols, J. Mol. Catal., 33, 1–14, 1985a. Burk, P.L. et al. The rhodium-promoted Guerbet reaction. Part II. Secondary alcohols and methanol as substrates, J. Mol. Catal., 33, 15–21, 1985b. Carlini, C. et al. Selective synthesis of isobutanol by means of the Guerbet reaction. Part 1. Methanol/n-propanol condensation by using copper based catalytic systems, J. Mol. Catal. A: Chem., 184, 273–280, 2002. Carlini, C. et al. Selective synthesis of isobutanol by means of the Guerbet reaction. Part 3. Methanol/n-propanol condensation by using bifunctional catalytic systems based on nickel, rhodium, and ruthenium species with basic components, J. Molec. Catal. A: Chem., 206, 409–418, 2003a. Carlini, C. et al. Selective synthesis of isobutanol by means of the guerbet condensation of methanol with n-propanol in the presence of heterogeneous and homogeneous palladium-based catalytic systems, J. Molec. Catal. A: Chem., 204–205, 721–728, 2003b. Carlini, C. et al. Selective synthesis of isobutanol by means of the Guerbet reaction. Part 2. Reaction of methanol/ethanol and methanol/ethanol/n-propanol mixtures over copper based / MeONa catalytic systems, J. Molec. Catal. A: Chem., 200, 137–146, 2003c. Carlini, C. et al. Guerbet condensation of methanol with npropanol to isobutyl alcohol over heterogeneous copper chromite/Mg-Al mixed oxides catalysts, J. Molec. Catal. A: Chem., 220, 215–220, 2004.
589
9 NONFOOD USES OF OILS AND FATS
F.D. Gunstone, J. Alander, S.Z. Erhan, B.K. Sharma, T.A. McKeon and J.-T. Lin
9.1
levels of burdensome stocks, to raise prices by removal of excess oil, or to meet targets for the reduction of carbon dioxide production. In contrast to fossil fuels, oils and fats represent a renewable resource produced by agricultural systems from carbon dioxide and water with sunlight providing the necessary energy. But some numbers should be applied to this issue to maintain a sense of perspective. Annual production of mineral oil, at 3.5 billion tonnes, is almost 30 times that of oils and fats, and the greater part of these latter are essential for food purposes. For example, biodiesel can never replace the demand for conventional fossil fuel — it can only diminish it marginally (Dumelin, 2005). Other environmental reasons are based on the fact that oils and fats are biodegraded more quickly than petrochemical products and, therefore, disappear more easily from the environment when used or spilled. Finally, when fully degraded, these materials liberate carbon dioxide trapped only months earlier and, therefore, do not add to total carbon dioxide, one of the greenhouse gases responsible for global warming. This contrasts with petrochemical products, which are oxidised to carbon dioxide trapped millennia earlier. Environmental issues are more complex than is generally appreciated. It must not be forgotten that the growing, harvesting, and transporting of oilseeds and their products are achieved only at some environmental cost. Hirsinger (2001) has reviewed the relation between oleochemicals and the environment and Urata et al. (2001) have described the contribution of surfactants and lipids to “Green Chemistry.” Yanagawa (2001) has discussed sustainable growth of the Asian-Pacific surfactant and detergent industries and Williams (2005) has reported on European detergent rules.
Introduction
Annual production of commodity oils and fats is now (2004) around 130 million tonnes and the general consensus is that ~ 80% is used for human food, ~ 6% for animal feed, and ~ 14% for the oleochemical industry though these ratios may change through the increasing use of oils and fats for oleochemical purposes, especially biodiesel. This chapter is devoted to the major uses of that last 14% (18 million tonnes). Before mineral oil and gas were used to supply many of our needs, oils and fats were widely used as illuminants and lubricants. Illumination was provided by a wick burning in olive oil in Mediterranean countries in biblical times or in the early days of the railways using rapeseed oil (then a high-erucic variety known as colza oil). At one time, lighthouses used seal oil or whale oil in their lamps. In a book written by T. P. Hilditch in 1927, an entire chapter was devoted to “the use of fats in candles and illuminants.” There are suggestions that the axles of early chariots were greased with a mixture of animal fat and lime producing calcium soaps. More importantly, soap made from fat and alkali (wood ash) has been used for centuries. Other uses are based on oxidative drying (hardening) of oil films to form coatings and have been exploited by artists, decorators, and producers of linoleum. With the development of fossil fuels and the inventiveness of chemists in the petrochemical industry, oils and fats ceased to be used as illuminants or as lubricants. Today there is a limited return to the use of oils and fats for some of these purposes, mainly on environmental grounds. Oil and gas supplies are not known with certainty, but they are finite and will not last indefinitely. Oils and fats can sometimes be used in place of fossil fuels and are even being burnt in electrical power stations to reduce the 591
9.2
Basic oleochemicals
The use of oils and fats as oleochemicals depends either on the physical properties of fatty acids and esters or on chemical properties relating to the carboxyl group, to olefinic centres, or to the whole molecule (as in its combustion). Linseed and castor oil are classed as industrial oils. Other oils used for nonfood purposes include significant proportions of the lauric oils (coconut and palm kernel), palm oil and palm stearin, tallow, rapeseed oil (both the high-erucic and the low-erucic oil), and lesser amounts of soybean and other oils. This listing may change when larger volumes of biodiesel are prepared. USDA figures for 2003/04 show the proportion of the commodity food oils used for industrial purposes in EU-15 as rapeseed oil (39%), palm kernel oil (29%), coconut oil (17%), palm oil (10%), soybean oil (7%), sunflower oil (6%), with olive oil, groundnut oil, and cottonseed oil at 1% or less. In the U.S. about 3% of total usage of soybean oil is for industrial purposes. Another important oleochemical feedstock is tall oil (a name based on tallolja, the Swedish word for pine oil). Tall oil fatty acids are by-products of the wood pulp industry and result when pine wood chips are digested, under pressure, with an aqueous mixture of sodium hydroxide and sodium sulfide during which the acids are converted to their sodium salts. Tall oil is the cheapest source of fatty acids rich in oleic and linoleic acids. In volume terms, surface-active compounds dominate among oleochemicals — a proportion of 90% has been reported, but this share is falling with increasing use as a fuel or by production of biodiesel. Surfactant molecules are mainly saturated or monounsaturated and find different uses according to their chain length. Three categories are recognised: C12 and C14 compounds from the lauric oils competing with identical products from the petrochemical industry, C16 and C18 compounds mainly from tallow and palm stearin, and C20 and C22 compounds from fish oils, high-erucic rapeseed oil, and crambe oil. These compounds may be used as acids or salts (soaps), as esters of methanol, other alkanols, or glycerol, as long-chain alcohols, or as nitrogen-containing compounds. Other oleochemical uses exploit the high unsaturation of oils, such as linseed and soybean oil, while castor oil is a unique source of several important chemicals. Several aspects of this topic have been covered in the sixth edition of Bailey’s Industrial Oil and Fat Products with chapters on soaps (Burke, 2005), detergents and detergency (Lynn, 2005), glycerine (Schroeder, 2005), biodiesel (Reaney et al., 2005), lubricants, hydraulic fluids, and inks (Erhan, 2005), polymers and plastics (Narine and Kong, 2005), paints (Lin, 2005), leather and textiles (Kronich and Kamath, 2005), and pharmaceutical and cosmetic uses (Hernandez, 2005). Hauthal and Wagner (2004) have reviewed household cleaning, care, and maintenance products.
References Burke, M.R., Soaps, in Bailey’s Industrial Oil and Fat Products, 6th ed., Shahidi, F., Ed., Wiley Interscience, New York, 2005, Chap. 3. Dumelin, E.E., Biodiesel — a blessing in disguise? Eur. J. Lipid Sci. Technol., 107, 63–64, 2005. Erhan, S.Z., Vegetable oils as lubricants, hydraulic fluids, and inks, in Bailey’s Industrial Oil and Fat Products, 6th ed., Shahidi, F., Ed., Wiley Interscience, New York, 2005, Chap. 7. Hauthal, H.G. and Wagner, G. Eds., Household Cleansing, Care, and Maintenance Products, Ziolkowsky GmbH publishing, Augsberg, Germany, 2004. Translation of the original German volume published in 2003. Hernandez, E., Pharmaceutical and cosmetic use of lipids, in Bailey’s Industrial Oil and Fat Products, 6th ed., Shahidi, F., Ed., Wiley Interscience, New York, 2005, Chap. 12. Hilditch, T.P., The Industrial Chemistry of Fats and Waxes, Baillière, Tindall and Cox, London, 1927. Hirsinger, F., Oleochemicals and the environment, in Oleochemical Manufacture and Applications, Gunstone, F.D. and Hamilton, R.J., Eds., Sheffield Academic Press, Sheffield, U.K., 2001, Chap 10. Kronick, P. and Kamath, Y.K., Leather and textile uses of fats and oils, in Bailey’s Industrial Oil and Fat Products, 6th ed., Shahidi, F., Ed., Wiley Interscience, New York, 2005, Chap. 10. Lin, K.F., Paints, varnishes, and related products, in Bailey’s Industrial Oil and Fat Products, 6th ed., Shahidi, F., Ed., Wiley Interscience, New York, 2005, Chap. 9. Lynn, Jr., J.L., Detergents and detergency, in Bailey’s Industrial Oil and Fat Products, 6th ed., Shahidi, F., Ed., Wiley Interscience, New York, 2005, Chap. 4. Narine, S.S. and Kong Xiaohua, Vegetable oils in production of polymers and plastics, in Bailey’s Industrial Oil and Fat Products, 6th ed., Shahidi, F., Ed., Wiley Interscience, New York, 2005, Chap. 8. Reaney, M.J.T., Vegetable oils as biodiesel, in Bailey’s Industrial Oil and Fat Products, 6th ed., Shahidi, F., Ed., Wiley Interscience, New York, 2005, Chap. 6. Schroeder, K., Glycerine, in Bailey’s Industrial Oil and Fat Products, 6th ed., Shahidi, F., Ed., Wiley Interscience, New York, 2005, Chap. 5. Urata, K. and Takaishi, N., A perspective on the contribution of surfactants and lipids toward “Green Chemistry:” present status and future potential, J. Surfact. Deterg., 4, 191–200, 2001. Williams, M., Europe tightens detergent rules, Oils Fats Int., 21(2), 20–21, 2005. Yanagawa, Y., Perspectives on the sustainable growth of AsianPacific surfactant and detergent industries towards the new millennium, J. Oleo Sci., 50, 281–393, 2001.
9.2
Basic oleochemicals
The basic oleochemicals are fatty acids, methyl esters, alcohols, amines, and glycerol. Traditionally these have been produced mainly in North America, Western Europe, and Japan from local or imported oils and fats. But this is changing, and countries in South East Asia, particularly Malaysia, have become major producers of 592
Nonfood Uses of Oils and Fats
basic oleochemicals using their increasing indigenous supplies of raw material. This is shown in Table 9.1 containing projected figures up to the year 2010. Over the 10-year period, the production of oleochemicals is expected to rise by one-third from 5.76 to 7.75 million tonnes. Although this rise is apparent in all regions, market share will fall in North America and Western Europe, but rise in Asia. It is predicted that by 2010 Asian production will equal production in Western Europe and North America combined. This statement refers particularly to Malaysia and possibly underestimates the contribution likely to be made by China by 2010 (MPOB, 2005). The materials used in the oleochemical industry and the processes by which oleochemicals are produced are summarised in Table 9.2 and Table 9.3. It is interesting to note in Table 9.3 the wide range of products that contain oleochemicals.
TABLE 9.3 Materials, processes, and products of the oleochemical industry Raw materials
Unit operations to produce basic oleochemicals Basic oleochemicals Operations to produce downstream products from the basic oleochemicals Oleochemical derivatives
End-use markets
TABLE 9.1 Estimates for 2000, 2005, and 2010 of basic oleochemicals (million tonnes) by region and by commodity (percentage figures are given in parentheses) 2000 World total By region Western Europe North America Asia Other By commodity Fatty acids Methyl esters Alcohols Amines Glycerol
5.76 1.76 1.36 2.27 0.37 3.05 0.66 1.44 0.57 0.75
2005 6.69
(31%) (24%) (39%) (6%)
1.87 1.52 2.79 0.51
2010 7.75
(28%) (23%) (42%) (7%)
3.50 0.73 1.73 0.62 0.86
1.96 1.66 3.54 0.59
(25%) (21%) (46%) (8%)
Source: Adapted from Gunstone, F.D., The Chemistry of Oils and Fats, Blackwell Publishing, Oxford, 2004.
9.2.1
4.00 0.80 2.07 0.70 1.00
TABLE 9.2 Basic oleochemicals and downstream products produced from triacylglycerols
Fatty acids
Methyl esters Alcohols Amines Glycerol
Fatty acids
Fat hydrolysis gives fatty acids and glycerol. The former are used in large quantities to make soaps and also as intermediates to produce methyl esters, amides, amines, and other important nitrogen-containing compounds, acid chlorides, anhydrides, ketene dimers, and peroxy acids and esters (Table 9.2). As there is often confusion about the weight relationships between fats and fatty acids, it is worth noting that hydrolysis of glycerol trioleate (100 g) involves reaction with water (6.1 g) to produce oleic acid (95.7 g) and glycerol (10.4 g). The contribution of water is often overlooked and it is erroneously argued that because the reaction produces 10 g of glycerol there will only be 90 g of fatty acids. The theoretical yield of free acid is close to 96%. The conversion of oils and fats to soaps (saponification) is carried out by a long-established process involving treatment with aqueous alkali at around 100ºC. Glycerol is obtained as a valuable by-product. The sodium and potassium salts are conventional soaps. Salts with other metals are used to promote polymerisation of drying oils, as components of greases and lubricants, and are incorporated into animal feeds for ruminants. Fats can also be hydrolysed by water itself in a fatsplitting process to yield free acids. This is probably a
Note: For more recent data on glycerol see Tables 9.4 and Table 9.5 and MPOB (2005). Source: Adapted from Gunstone, F.D., The Chemistry of Oils and Fats, Blackwell Publishing, Oxford, 2004.
Basic Oleochemicals
Tall oil, lauric oils, palm oil, highand low-erucic rapeseed oil, soybean oil, sunflower oil Splitting, distillation, fractionation, hydrogenation, methylation, hydrophilisation Fatty acids, methyl esters, fatty alcohols, fatty amines, glycerol Amidation, dimerisation, epoxidation, ethoxylation, quaternisation, sulfation, sulfonation, saponification, transesterification Amides, dimer and trimer acids, epoxidised oils and esters, ethoxylates and propoxylates, sulfates, sulfonates, esters, soaps, salts Building auxiliaries, candles, cleaning agents, cosmetics, detergents, flotation agents, food emulsifiers, inks, insecticides, leather treatment, lubricants, paints, pesticides, pharmaceuticals, plastics, soaps, textiles, tyres
Downstream Derivatives Esters, metal salts (soaps), amides and amines, ketene dimers, anhydrides, acid chlorides, peroxy acids, and esters Acids, other esters, alcohols, α-sulfonates Ethylene oxide adducts, sulfates, Guerbet alcohols and acids Various cationic surfactants – Section 9.3.6 Polyglycerol, mono- and diacylglycerols and their acetates, tartates, lactates, etc.
Note: Some of the compounds are not discussed in this chapter, but are described in Chapter 7, which is devoted to physical properties. Source: Adapted from Gunstone, F.D., The Chemistry of Oils and Fats, Blackwell Publishing, Oxford, 2004.
593
9.2
Basic oleochemicals
Claims have been made for the use of calcium carbonate as an environmentally acceptable catalyst (Suppes et al., 2001). Just as hydrolysis can occur without a catalyst at an elevated temperature, methanolysis is also possible without a catalyst at temperatures up to 200ºC, but this process is not used on an industrial scale. Candida antarctica lipase has been used to promote the methylation of waste fatty acids. Conversions of around 95% are achieved at 30ºC in 24 hours and under appropriate conditions the lipase maintains it activity through 45 cycles. Enzymatic methanolysis has also been examined with a range of vegetable oils and waste edible oil using a fixed bed reactor. Glycerolysis is another alcoholysis process employed on a commercial scale to convert triacylglycerols to mixtures of monoacylglycerols and diacylglycerols by reaction with glycerol in the presence of a basic catalyst when the following equilibrium is established:
homogeneous reaction between fat and a small amount of water dissolved in the fat. The procedure is usually carried out in a continuous, high pressure (20 to 60 bar), uncatalysed, counter-current process at 250ºC, though lower temperatures are desirable for highly unsaturated oils. Under these vigorous conditions, both the fatty acids and the glycerol will be discoloured and may have to be distilled. Splitting capacity in Malaysia has risen from 0.68 million tonnes in 1996 to 1.77 million tonnes in 2004 (Lim, 2005). This is in addition to the production of 0.47 million tonnes of soap noodles. Further information is found in Malaysian Oil Palm Statistics (MPOB, 2005). Toilet soaps are made from tallow (mainly in North and South America) or from vegetable oil (mainly palm oil) elsewhere in the world. The traditional process involves saponification (alkaline hydrolysis of vegetable oils or animal fats), but soaps are made increasingly by neutralisation of distilled fatty acids. There is a growing trade in soap noodles (sodium salts of fatty acids) that are converted elsewhere to coloured, scented, wrapped bars of toilet soap. This product is now in competition with liquid soaps (Table 9.7). Hydrolysis promoted by lipases, such as those from Rhizomucor miehei and Candida rugosa, takes about 20 hours at 20ºC or 6 hours at 45ºC, but gives cleaner products with less waste than the fat splitting process. Despite this and the saving in energy costs, it is not yet widely used on an industrial scale.
9.2.2
triacylglycerol + glycerol ] monoacylglycerol + diacylglycerol The composition of the product mix depends on the amount of glycerol dissolved in the fat phase. It can also be modified by the use of appropriate solvents. Concentrates of monoacylglycerol (90 to 95%), produced by molecular distillation, are widely used as emulsifiers. A similar process has been described using enzymes as catalysts, but this has not been used on an industrial scale.
Fatty esters 9.2.3
Esters can be made by esterification of acids by reaction with alcohols or from existing esters, including triacylglycerols, by reaction with alcohols (a process known as alcoholysis), acids (acidolysis), or other esters (interesterification). Alcoholysis is more widely practised than acidolysis and includes the important reactions of esters (especially triacylglycerols) with methanol (methanolysis) or with glycerol (glycerolysis). On an industrial scale ester production of methyl esters is most commonly undertaken by methanolysis of triacylglycerols (natural oils and fats) in the presence of an acidic, alkaline, or enzymatic catalyst. Large-scale methanolysis is used to make methyl esters for use as biofuel, as solvent, or as an intermediate in the production of alcohols. They can also be hydrolysed to acids. Oils low in free acid can be converted directly to methyl esters with an alkaline catalyst. Glycerol is also produced in this reaction and is recovered as a second marketable product. In a continuous process for the conversion of vegetable oils to methyl esters, conversion is >98% using sodium hydroxide as the catalyst. Under optimum conditions the reaction requires 6 to 8 minutes and may even take place during passage through the reaction plant.
Fatty alcohols
Long-chain alcohols (RCH2OH) with structures similar to the better-known acids (RCOOH) occur naturally in the free state and, more commonly, as esters. These last include wax esters made from long-chain alcohols and long-chain acids and occur in jojoba and other vegetable and animal waxes. Many insect pheromones are fatty alcohol acetates. Long-chain alcohols are important oleochemicals produced on a commercial scale by hydrogenolysis of acids, methyl esters, or triacylglycerols. This is a catalytic process and, depending on the choice of catalyst, olefinic double bonds may also be reduced or be left unchanged. These processes are generally applied to natural mixtures and the products, therefore, are mixtures varying in chain length. C8 to C14 alcohols are produced from lauric oils (coconut and palm kernel), C16 and C18 compounds from tallow, lard, palm oil, or palm stearin, and C22 alcohols from erucic acid-rich oils. Individual alcohols can be obtained by fractional distillation of the mixed products. Dodecanol and similar alcohols are also produced by the petrochemical industry through oligomerisation of ethene (ethylene). Commercial production of long-chain alcohols is now around 2 million tonnes annually of which twothirds or more is fat-based. About 75% of these fatty 594
Nonfood Uses of Oils and Fats
phases. Laboratory scale reactions, conducted under supercritical conditions with propane, are 5 to 10 times quicker than the conventional reaction. A copper-based catalyst free of chromium at 150 bar and 240 to 250°C is employed (van den Hark et al., 1999).
alcohols are used as alcohol sulfates, alcohol ethoxylates, or alcohol ethoxylate sulfates (see Section 9.3.3 to Section 9.3.5). Lim (2005) reports that since 2001 new production capacity for C12 to C18 alcohols has been 0.78 million tones, while capacity of only 0.24 million tonnes has been closed. An increase in the capacity to produce alcohols of 135,000 tonnes in Malaysia and 60,000 tonnes in Indonesia was reported for 2005. Malaysian exports of fatty alcohols rose from 288,000 tonnes in 2003 to 328,000 tonnes in 2004 (MPOB, 2005). Although commodity oils and fats are the starting point for most of these processes, the glycerol esters themselves are not generally used directly since, among other reasons, the valuable glycerol would be lost. More usually, hydrogenolysis is carried out on acids, methyl esters, or on wax esters made in situ from acids and alcohols. A wide range of minor products may also be formed during this reaction, including esters, aldehydes, alkanes, ethers, and acetals. In the methyl ester route, acid-free esters are first made from the natural oils by methanolysis (with release and recovery of glycerol) and then subjected to hydrogenolysis using pure hydrogen (>99.9%) and a copper chromite catalyst, usually in a fixed bed reactor, at 250 to 300 bar and 210°C. The volatile mixture of hydrogen and methanol can be separated and each component recycled. The alcohol product is stripped of methanol and the long-chain alcohols are used as such or are fractionated by distillation into individual components. This procedure can be adapted to produce olefinic alcohols with a copper-zinc catalyst free of chromium. Nickel catalysts activated with chromium, iron, or preferably rhodium can also be used for reactions at 200 to 230°C and 200 bar. Arguments have been presented for the acid route using some new technologies. This involves: (1) conversion of oil to acids and fractionation of these, some of which may be sold as acids; (2) preparation of methyl esters from acids using a resin bed as catalyst; and (3) reduction of esters to alcohols using a fixed bed catalyst (40 bar, 200 to 250°C, chromium-free catalyst). The combined procedures can be used flexibly to produce acids, methyl esters, and alcohols as required. In the wax ester route, the starting materials are fatty acids (distilled or fractionated) and some pre-made, longchain alcohols. At an appropriate temperature the acids and alcohols react without a catalyst to produce wax esters. These and pure hydrogen are then passed to the fixed-bed hydrogenation reactor charged with catalyst where hydrogenolysis takes place. Thereafter, hydrogen is separated from the alcohols. Some of these are returned to make more wax ester and the balance is distilled. This procedure has the advantage that it is not necessary to use or to recover methanol. These commercial processes are limited by the rates of hydrogen transfer between the gas, liquid, and solid
9.2.4
Fatty amines
Fatty amines, produced at a level of around 500 kt per annum, are the starting point for several types of nitrogencontaining compounds used as surfactants. Acids are first converted to nitriles — probably via amides — by reaction with ammonia at 280 to 360°C in the presence of alumina, thoria, titanium oxide, zinc oxide, manganese acetate, bauxite, or cobalt salts as catalysts. RCOOH → RCONH2 → RCN → RCH2NH2 → [RCH2]2NH Hydrogenation of nitrile occurs with nickel or cobalt as catalysts. Double bonds may be reduced at the same time, but conditions are usually selected to minimise this. Some conversion of cis to trans isomers may also occur. The major product is usually the primary amine (RCH2NH2), but this may be accompanied by secondary ([RCH2]2NH) and tertiary amines ([RCH2]3N). The formation of secondary and tertiary amines can be promoted by adjusting the reaction conditions. Aldimine (RCH = NH), the first product in the conversion of nitriles to primary amines, reacts with more hydrogen to form primary amine, with preformed primary amine to form secondary amine, or with preformed secondary amine to give tertiary amine. RC ≡ N → RCH = NH → RCH2NH2 reaction with hydrogen RC ≡ N → RCH = NH → [RCH2]2NH reaction with primary amine RC ≡ N → RCH = NH → [RCH2]3N reaction with secondary amine Tertiary amines with two methyl groups (RCH2NMe2) are made from primary amines by reductive alkylation with formaldehyde (methanal), from N,N-dimethylalkyl amides by catalytic reduction, or from fatty alcohols by catalysed reaction with dimethylamine. RCH2NH2 + 2CH2O → RCH2NMe2 RCONMe2 → RCH2NMe2 RCH2OH + Me2NH → RCH2NMe2 Still other nitrogen-containing, surface-active compounds may be made from carboxylic acids, alcohols, and amines (see Section 9.3.7 and Table 9.9 and Table 9.10).
595
9.2
Basic oleochemicals
9.2.5
In 2003, the annual production of glycerol (930 kt) came from countries with significant oleochemical industries including the U.S., Europe, Japan, and Southeast Asia. Significant importers were the U.S. (37% of its glycerol requirement) and Japan (50% of its glycerol requirement), while Southeast Asia was the major exporting region. Four ASEAN (Association of Southeast Asian Nations) countries (Malaysia, Indonesia, Philippines, Thailand, and Singapore) alone exported around 164,000 tonnes of glycerol in 2003. Table 9.4 clearly shows that ASEAN countries are now important producers of glycerol and have become the dominant exporter of this commodity. Sources of glycerol by oleochemical products are detailed in Table 9.5. Between 1999 and 2008, glycerol production is expected to rise 38% (from 804 to 1110 kt). Changes in the supply levels from various oleochemical processes over this 10-year period are soaps (–58kt), fatty acids (+88 kt), biodiesel (+293 kt), fatty alcohols (+32 kt), and petrochemical glycerol (–50 kt). These figures demonstrate the growing importance of biodiesel production and the continuing demand for fatty acids and fatty alcohols as sources of glycerol. The change through increased biodiesel production is mainly in Europe, but is expected to become more apparent in North America and elsewhere. The market for fatty acids continues to increase and new plants are being established in China and other developing countries. These will add to local supplies of glycerol and affect import requirements. In 2003, 41% of the European glycerol supply came from biodiesel production and that share is expected to increase. Following changes in taxation rules in Germany for a 95:5 blend of diesel with biodiesel, usage of biodiesel is expected to increase rapidly leading eventually to the production of 140 kt of glycerol in Germany, in addition to part of the 65 kt produced in 2003 for the existing pure biodiesel market. The biodiesel market in the rest of Europe will also move ahead strongly based mainly on locally produced rapeseed oil. European biodiesel capacity in Europe in 2004 was 2.2 million tonnes. New projects for large biodiesel plants continue to be announced
Glycerol
Oils and fats are mainly triacylglycerols and are generally used for dietary consumption in this form. However, in the oleochemical industry, oils and fats are used mainly to manufacture acids, soaps, methyl esters, alcohols, or nitrogen-containing derivatives and the production of these compounds will almost always involve the liberation of glycerol (1,2,3-propanetriol) at a level of around 10% of the oil or fat. This is a useful and valuable by-product and its economic value is an important part of the profitability of the oleochemical industry. Glycerol is also a product of the petrochemical industry where it is made from propene via epichlorohydrin (1-chloro2,3-epoxypropane). The increasing supply of glycerol from the oleochemical industry, the high price of propene, and the demand for epichlorohydrin for other purposes have together made the petrochemical supply route less important. It is now about 5% or less of total supply compared with 25% 20 years ago (Gunstone and Heming, 2004). Glycerol is available in several grades varying in purity and the requirements of the industries to which it is sold. Refined material is at least 86.5% pure and generally greater than 99.5%. Its value lies in its physical properties: it is hygroscopic, colourless, odourless, viscous, sweet-tasting, low boiling, nontoxic, emollient, a good solvent, and water-soluble. It is also easily biodegradable. Its major uses include oral care products, food and food emulsifiers, tobacco products, polyurethanes, prescription drugs, over-the-counter medicines, and cosmetics. Attempts are being made to develop new uses by conversion to other valuable compounds, such as glycidol (2,3-epoxypropanol), glycerol carbonate, and polyglycerols and their esters (Barrault et al., 2005, and Stepan website). In some of its uses, glycerol (produced at an annual level of around 1 million tonnes) competes with other polyols, such as pentaerythitol and trimethylolpropane (together 0.4 million tonnes), sorbitol (1.1 million tonnes.), propylene glycol (1.5 million tonnes), and ethylene glycol (7.5 million tonnes). The figures in parentheses represent annual production levels in 2003. TABLE 9.4
Production, consumption, exports, and imports of glycerol (kilotonnes) in 2003 Production
World U.S. Europe China Japan ASEANb Rest of world
Consumption
Exports
Imports
930a
936
251
248
142 315 20 45 197 211
201 325 65 85 33 227
24 25 2 2 164 34
75 35 45 43 – 50
a
Imports From:
SE Asia, Europe, S. America SE Asia Malaysia and Indonesia SE Asia
Details of sources by oleochemical products are given in Table 9.5. Malaysia, Indonesia, Philippines, Thailand, and Singapore. Source: Gunstone, F.D. and Heming, M.P.D., Glycerol – an important product of the oleochemical industry, Lipid Technol., 16, 177–179, 2004.
b
596
Nonfood Uses of Oils and Fats
Typically emulsifiers have HLB values of 5 to 6, wetting agents of 7 to 9, and detergents of 13 to 15. Surfactants are produced both by the petrochemical and oleochemical industries, though only the latter will be considered in this section. This is most obvious in the production of long-chain alcohols. There are environmental and economic reasons why lipid-based molecules find favour in a time of high-priced oil and gas. Amphiphilic molecules can exist comfortably at an oil–water interface and reduce the surface tension at such interfaces. This property is fundamental in all living systems and in many foods and other manmade systems. Depending on the HLB, appropriate amphiphilic molecules influence a range of important surfactant properties, such as emulsification, deemulsification, wetting, foaming, defoaming, water-repelling, dispersing, solubilising, detergency, sanitising, lubricity, and emolliency. The simplest and oldest examples are soaps, such as sodium palmitate in which the palmitic acid chain is lipophilic and the carboxylate group is hydrophilic. The lipophilic alkyl chain varies mainly in chain length with C12 and C14 chains from the lauric oils; C16 and C18 chains from tallow, palm oil, or palm stearin; and the C20 and C22 chains from fish oils or high-erucic rapeseed oil. The alkyl chain may also have some unsaturation and may be branched. The polar head group shows greater variations through carboxylates, sulfates and sulfonates, phosphates, partial esters of polyhydric alcohols, such as glycerol and carbohydrates, polyoxyethlene derivatives of alcohols or amines, derivatives of amino acids, and many other nitrogen-containing molecules. There are four major groups of surfactants – anionic, nonionic, cationic, and zwitterionic (amphoteric) – with these terms describing the nature of the head group. Of these, anionics are used in greatest amount, but nonionics are increasing faster than any of the other groups. Some figures of usage in 2000 are given in Table 9.6. For automatic dishwashers two, three, or four surfactants may be combined in a single washing tablet. Many complex molecules have been designed and synthesised as superior surfactants. Typical of these is the amphoteric surfactant shown below with four functional groups (amino, carboxy, hydroxy, and ether groups) (Hidaka et al., 2003).
TABLE 9.5 Production of glycerol (kilotonnes) in 1999, 2003, 2004, and forecast for 2008 by the oleochemical product of which it is a by-product 1999
2003
2004
2008
Total
804
930
970
1110
Soaps Fatty acids Biodiesel Fatty alcohol Synthetic Other
198 322 57 108 75 44
180 350 160 110 80 50
170 365 210 120 50 55
140 410 350 140 25 45
Source: Gunstone, F.D. and Heming, M.P.D., Glycerol – an important product of the oleochemical industry, Lipid Technol., 16, 177–179, 2004.
although doubts have been expressed about the economic viability of some of these programmes. Nevertheless, it is quite possible that European biodiesel capacity will reach 3 million tonnes by the end of 2005, with the largest capacities being in Germany (1.1 million tonnes/year), Italy (0.6 million tonnes/year) and France (0.4 million tonnes/year) (see Section 9.6).
References Barrault, et al., Polyglycerols and their esters as an additional use of glycerol, Lipid Technol., 17, 131–135, 2005. Dumelin, E.E., Biodiesel — a blessing in disguise? Eur. J. Lipid Sci. Technol., 107, 63–64, 2005. Gunstone, F.D., The Chemistry of Oils and Fats, Blackwell Publishing, Oxford, 2004, Chaps. 8, 11. Gunstone, F.D. and Heming, M.P.D., Glycerol — an important product of the oleochemical industry, Lipid Technol., 16, 177–179, 2004. Lim, S., Oleochemicals — biodiesel and glycerine overshadow market, Oils Fats Int., 21 (1) 37–39, 2005. MPOB (Malaysian Palm Oil Board), 2004 Malaysian Oil Palm Statistics, 2005. Stepan (Stephan Company): www.stepan.com Suppes, G.J., et al., Calcium carbonate catalysed alcoholysis of fats and oils, J. Am. Oil Chem. Soc., 78, 139–145, 2001. Van den Hark, S., et al., Hydrogenation of fatty acid methyl esters to fatty alcohols at supercritical conditions, J. Am. Oil Chem. Soc., 76, 1363–1370, 1999.
9.3
Surfactants
9.3.1
Introduction
C12H25OCH2CH(OH)CH2N(CH2CH2OH) CH2CH2COONa
Surfactants are surface-active molecules as a consequence of their amphiphilic nature. This means that one part of a surfactant molecule (the alkyl chain) is lipophilic (hydrophobic) and the other part (the polar head group) is hydrophilic (lipophobic). The balance between these forces is an important property of a surfactant molecule and is defined as the hydrophilic lipophilic balance (HLB) by the equation:
A detergent is a formulation of many components (Hargreaves, 2003) in which the surfactant is responsible for washing and cleansing properties. Detergents enter widely into daily life and play an important part in keeping the human environment clean and wholesome. Their many uses include: cleaning agents for floors, surfaces, laundry, dishes, and personal care products. They are frequently used in pharmaceuticals and as lubricants, and are employed in industries devoted to food, agriculture, metal-working, textiles, and building. Appel (2000) has summarised the
HLB = 20 (molecular weight of the hydrophilic portion/ molecular weight of the whole molecule) 597
9.3
Surfactants
modern methods of detergent manufacture and Berna et al. (1998) have reviewed laundry products still used in bar form for manual washing. Rosen and Dahanayake (2000) have written a book on the industrial utilization of surfactants and Hargreaves (2003) has authored a simple but useful work on formulation containing a large number of typical recipes some of which are cited in Table 9.7. There has been a general consolidation of the surfactants industry both of suppliers and retailers (Anon., 2001). Many natural products have surfactant properties. Hill (2001) has reviewed the use of oils and fats as oleochemical raw materials. Dembitsky (2004, 2005a-e) has demonstrated the diversity of these compounds and has reported their chemical structures and biological activities. Many microorganisms also produce biosurfactants (Solaiman, 2005). Urata and Takaishi (1999, 2002) have discussed a number of synthetic routes to novel compounds capable of self-assembly. Analytical methods for the examination of surfactants have been reported by Thin Sue Tang (2001), Morelli and Szajer (2000, 2001), and Waldhoff and Spilker (2005). A useful website is The Surfactants Virtual Library (see reference list).
TABLE 9.6 Consumption of surfactants (million tonnes) in 2000 by category and usage in three countries/regions of the world (excluding 7 million tonnes of soaps)
Total By category Anionic Nonionic Cationic Amphoteric By usage Household products Cosmetics and toiletries Cleaning products Textiles Mining and petroleum Plastics and paints Agrochemicals Other
Western Europe
North America
Japan
2.61
3.34
1.19
1.32 0.97 0.25 0.07
1.95 1.04 0.27 0.09
0.49 0.56 0.10 0.04
1.41 0.12 0.21 0.16 0.30 0.16 0.08 0.15
1.32 0.19 0.28 0.31 0.37 0.60 0.12 0.15
0.32 0.13 0.10 0.12 0.11 0.06 0.04 0.31
Note: Other information is available in references: Anon., 2000 2002a,b, 2003. Source: Adapted from Gunstone, F.D., The Chemistry of Oils and Fats, Blackwell Publishing, Oxford, 2004.
TABLE 9.7
Selected typical formulations for a range of products indicating the surfactants present
Product
Formulation
% 8.0 6.0 4.0 7.0 2.0
Shampoo for dry hair
40% triethanolamine lauryl sulfate 27% ammonium lauryl sulfate 70% sodium lauryl ether sulfate Coco amido propyl betaine Coconut diethanolamide and citric acid, salt, perfume, colour, preservative, and water
Foam bath
27% sodium lauryl ether sulfate 30% lauryl betaine and salt, citric acid, perfume, colour, preservative, and water
60.0 5.0
Shower gel
30% sodium lauryl ether sulfate Coconut diethanolamide 30% alkylamido propyl betaine Cocoamine oxide and salt, lactic acid, perfume, colour, preservative, and water
40.0 2.0 5.0 2.0
Liquid soap
27% sodium lauryl ether sulfate Monoethanolamine lauryl sulfate 30% lauryl betaine and salt, citric acid, perfume, colour, preservative, and water
20.0 10.0 7.0
Toothpaste
Glycerol Sodium lauryl sulfate and sodium carbomethoxycellulose, sodium monofluorophosphate, dicalcium phosphate dihydrate, flavour, preservative, and water
25.0 1.5
Moisturising cream
Caprylic/capric triacylglycerols Octyl cocoate Cetyl esters Cetyl/stearyl alcohol Polysorbate 60 Sorbitan stearate Glycerol and hydrolysed vegetable protein, perfume, colour, preservative, and water
13.0 3.0 3.0 3.0 3.0 2.0 3.0
Dish washing liquids (20% active)
Sodium dodecylbenzene sulfonate Coconut diethanolamide 27% sodium lauryl ether sulfonate and salt, perfume, colour, preservative, and water
23.3 2.0 13.3
Carpet shampoo
28% sodium lauryl ether sulfonate Coconut diethanolamide Isopropanol and perfume, colour, preservative, and water
35.7 3.0 10.0
Screenwash
30% sodium lauroyl sarcosinate (1.0%) Isopropanol (25.0%), colour, and water
1.0 25.0
Source: Adapted from Hargreaves, T., Chemical Formulation, RSC, Cambridge, U.K., 2003.
598
Nonfood Uses of Oils and Fats
9.3.2
Tracy and Reierson, 2002), sulfosuccinates (maleic anhydride and sodium sulfite), ethoxy carboxylates (sodium chloroacetate), or carbonate ethoxylates (dimethyl carbonate) by reaction with the reagents indicated in parentheses (Table 9.8). These are active components in detergents used in personal care products and for washing clothes and hard surfaces (floors, walls, dishes). Surfactant, in the range of 5 to 20%, is accompanied by other materials, such as phosphate, zeolite, bleaching agent, optical brightener, fragrance, and water. The sarcosinates, taurates, and isethionates (Table 9.8) are long-chain amides of sarcosine and taurine, respectively, or are esters of isethionic acid. α-Sulfonate esters are made from methyl esters of saturated acids by reaction with sulfur trioxide. The product, after neutralisation, is a mixture of monoester salts (from RCH(SO3H)COOCH3) and di-sodium salts (from RCH(SO3H)COOH) usually in a ratio of 80:20. Further details are given in chapters written by Porter (1997) and by Roberts (2001).
Anionic surfactants from carboxylic acids
As the name indicates, anionic surfactants have a negatively charged species and a counterion that is usually a metal, but may be a type of ammonium group. The anionics are used in greater volume than any other class of surfactants. Soap belongs to this category and is still the surfactant used in largest amount. In many countries, hard (tablet) soap is being replaced by a liquid soap that is not a carboxylate salt (see Table 9.7). Soap has the disadvantage that it can only be used at pH8 and above and that it forms an insoluble scum with the calcium salts present in hard water. Alternative and superior surfactants, therefore, have been developed. These are generally sulfates or sulfonates in place of carboxylates. In all of these, the alkyl chain is the most expensive component in the surfactant. Carboxylates Fatty alcohol sulfates Fatty alcohol ether sulfates α-Sulfonated esters
RCOOH ROSO3H R(OCH2CH2)nOSO3H RCH(SO3H)COOCH3
Sodium and potassium salts of fatty acids (traditional soaps) are still made by saponification of appropriate fats and also increasingly by neutralisation of carboxylic acids resulting from splitting (see Section 9.2.1). Cohen and Trujillo (1998) have reported the synthesis, characterization by GC-MS (gas chromatography-mass spectrometry), and by infrared spectroscopy, surface tension, and specific conductivity of methyl ester sufonates. The production of methyl ester sulfonates in the U.S. has been reported (Watkins, 2001). Scheibel (2004) has discussed changes in anionic surfactant technology in the laundry detergent industry.
9.3.3
9.3.4
Nonionic surfactants
Nonionic surfactants contain the usual type of lipophilic chain from a petrochemical or oleochemical source. The head group is not charged, but is polar through the presence of an appropriate collection of hydroxy, amino, or ether groups. The last come from ethylene oxide or propylene oxide products (see Section 9.3.5) and the former from glycerol, polyglycerol, low molecular weight carbohydrates, or amino acids or other amines. Typical structures include polyethylene oxide derivative of the fatty alcohol ROH (alcohol ethoxylate AEO), such as R(OCH2CH2)nOH and bis-polyethylene oxide derivatives of the fatty amine RNH2, such as
Anionic surfactants from alcohols
HO(CH2CH2O)n NR(OCH2CH2)mOH
Anionic surfactants of various kinds can be made from fatty alcohols or fatty alcohol ethoxylates (see Section 9.2.3 and Section 9.3.4) mainly as sulfates through reaction with sulfur trioxide or chlorosulfonic acid and used as sodium, ammonium, or monoethanolamine (HOCH2CH2NH2) salts. The long-chain alcohols may also be used as phosphates (phosphorus pentoxide,
Ethylene oxide is itself a product of the petrochemical industry and a hazardous chemical with undesirable environmental properties. It is for this reason that there is growing interest in the acyl and alkyl derivatives of glucose and other carbohydrates where all the reactants are
TABLE 9.8 Anionic surfactants produced from fatty alcohols and their ethoxylates Name Alcohol sulfates Ethoxy sulfates Monoacylglycerol sulfates Ethoxy phosphates Sulfosuccinates Ethoxycarboxylates Carbonate ethoxylates Sarcosinates Taurates Isethionates a
Structurea ROSO3H AEOSO3H RCOOCH2CH(OH)CH2OSO3H AEOPO3H2 or (AEO)2PO2H AEOCOCH2CH(OSO3Na)COONa AEOCH2COONa ROCOO(CH2CH2O)nH RCON(CH3)CH2COOH RCONHCH2CH2SO3H RCOOCH2CH2SO3H
Typical Uses Shampoo, toothpaste Shampoo, bubble bath Electroplating Carpet cleaner, oil spill dispersants
Corrosion inhibitor Toilet bars
RO represents a fatty alcohol, RCOO an acyl group and AEO an alcohol ethoxylate R(OCH2CH2)nO.
599
9.3
Surfactants
dermatological properties (Cox et al., 1998; Hreczuch et al., 2001). The reaction is not confined to methyl esters and has been applied to compounds with two or more ester functions. Interesting products result from triacylglycerols including natural mixtures, such as the lauric oils (coconut and palmkernel) and tallow. Reaction can occur at all three ester functions to give products with the structure shown in Figure 9.1 Subsequent partial hydrolysis gives a mixture of ethoxylated triacylglycerols (unreacted material), fatty acid soaps (RCOONa), and products in which 1, 2, or 3 acyl groups have been removed. The composition of the mixture depends on the degree of hydrolysis. Even without further separation this mixture has good surfactant properties (Cox and Weerasooriya, 2000). Another development is the replacement of ethylene oxide, wholly or in part, by propylene oxide in the reaction both with alcohols and esters. The repeating unit (-CH2CH2O-) is then replaced by (-CHMeCH2O-) or there is a mixture of both units in the poly-ether chain. The resulting branched-methyl compounds have interesting modified surfactant properties, important among which is their greater ability to reduce foaming when compared with the ethylene oxide derivatives. This reaction requires a calcium aluminum complex as catalyst (Cox et al., 1998; Weerasooriya, 1999). Further information is given by Bognolo (1997) and Gunstone (2001).
natural products coming from renewable resources (Section 9.3.7). Bognolo (1997) has written a full account of nonionic surfactants.
9.3.5
Ethoxylation and propoxylation of alcohols and esters
A substantial portion of the medium and long-chain alcohols are used only after conversion to ethoxylates or propoxylates. Ethoxylation of long-chain alcohols with ethylene oxide occurs at 135°C under pressure in 30 minutes in a reaction usually catalysed by ~ 0.2% of NaOH or KOH. The product is hydrophilic by virtue of one hydroxyl group and many ether links and may be a mixture of compounds with up to 20 ethylene oxide (EO) units (Figure 9.1). Since important surfactant properties vary with the number and range of EO units, there is a drive to produce materials with different narrow ranges of EO units. For example, when dodecanol is reacted with ethylene oxide (50% wt ≡ 4.4 mol) the product is a mixture of compounds with up to 20 or more EO units and no individual compound greater than 10%. With proprietary catalysts, such as ZrSO4(OR)2 or similar aluminium compounds, the products contain only 0 to 10 EO units with those having 4 to 6 EO units each around 20% and those with 3 and 7 units ~ 10%. These narrow range ethoxylates have good stability and skin mildness in liquid dishwashing products (Di Serio et al., 1998). The ethoxylation of esters, rather than alcohols, is an interesting development in this field since products with improved surfactant properties can be obtained from esters that are less expensive starting materials than the alcohols. Using a methyl ester and an appropriate catalyst, such as a composite aluminum and calcium metal oxide at 180°C and 3 bar the product has a narrow range of molecular weights with mainly 5 to 10 ethylene oxide units (see Figure 9.1). While alcohol ethoxylates can assume a linear arrangement ester ethoxylates are necessarily bent (boomerang shape) because of the trigonal ester carbon atom. The ester products have outstanding ROH
R(OCH2CH2)nOH
RCOOMe
RCO(OCH2CH2)nOMe
CH2O(CH2CH2O)xCOR
9.3.6
Alkyl polyglycosides
The term alkyl polyglycoside is the name given to technical products made from starch (or other source of glucose) and a fatty alcohol. These are sugar ethers in contrast to the sugar esters used in olestra. The alcohol is usually a mixture of C8/10, C12/14, or C16/18 alcohols derived from appropriate fatty acid sources. All the substrates are renewable resources. The reaction between starch and alcohol is usually catalysed by acids, such as sulfuric or 4toluenesulfonic and is accompanied by extensive depolymerisation of the carbohydrate polymer so that the product is mainly, but not entirely, an alkyl glucoside. In the two-step process (see equation below) butanol is used first and in the second step the fatty alcohol mixture (ROH) of desired chain length. The product is a mixture of compounds with R groups differing in chain length and values of n lying between 1 and 5. The average value of n (degree of polymerisation) is usually 1.3 to 1.7. Products with a value of 1.3 will contain molecules with one (~60%), two (~20%), three (~10%), and four and five glucose units. Products made from alcohols having 8 to 14 carbon atoms are water-soluble and are used as surfactants, those with 16 and 18 carbon atoms are not water-soluble, but are used as emulsifying agents and in cosmetic preparations (Hill et al., 1997).
Ethylene oxide derivative based on an alcohol
Ethylene oxide derivative based on a methyl ester
Ethylene oxide derivative based on a triacylglycerol
CHO(CH2CH2O)yCOR CH2O(CH2CH2O)zCOR
FIGURE 9.1 Ethylene oxide derivatives of alcohols, methyl esters, and triacylglycerols.
600
Nonfood Uses of Oils and Fats
starch + BuOH → [glucose]nOBu
again with acrylonitrile to give a triamine RNH(CH2)3 NH(CH2)3NH2.
[glucose]nOBu + ROH → [glucose]nOR
ROH + CH2=CHCN → RO(CH2)2 CN → RO(CH2)3NH2 (ether amine)
Piispanen et al. (2004) have described the structures and structure/property relationships of surfactants derived from natural products and Dembitsky (2004, 2005a-d) has reported the structure and biological activity of a large number of natural glycoside surfactants. The acyl esters of sugars are not yet commercial surfactants, but there is considerable interest in the preparation of acylated derivatives of fructose (Jung et al., 1998), glucose, or other monosaccharides, and of sucrose. These can be prepared chemically or, more specifically, with enzymatic catalysts and studies of their synthesis, hydrolysis, biodegradation, and some surfactant properties have been reported (Allen and Tau, 1999; Baker et al., 2000a,c; Polat et al., 2001).
9.3.7
RNH2 + CH2=CHCN → RNH(CH2)2 CN → RNH(CH2)3NH2 (diamine) Ether amines and diamines react with ethylene oxide to give ethyloxylated products (Table 9.8 and Table 9.9) as in the following equations: RO(CH2)3NH2 → RO(CH2)3NH(CH2CH2O)nH (ethoxylated ether amine) RNH(CH2)3NH2 → RNH(CH2)3NH(CH2CH2O)nH (ethoxylated diamine) RNH2 → H(OCH2CH2)nN(R)(CH2CH2O)mH (ethoxylated amine)
Cationic surfactants
Cationic surfactants are nitrogen containing-compounds. They show high substantivity (i.e., strong adherence) to natural surfaces and find extensive use in fabric softening, hair conditioning, corrosion inhibition, mineral flotation, and as bactericides (Karsa, 2001). Levinson (1999) has reviewed rinse-added fabric softener technology. Alcohols and amines add to acrylonitrile and, after catalytic hydrogenation, furnish ether amines (often written as one word) and diamines as shown in the equations below (Table 9.9 and Table 9.10). The diamine can react
RCONH(CH2)2NH(CH2)2NHCOR (diamidoamine or diacylated triamine) Diamidoamines (or diacylated amines) formed from carboxylic acids and diethylene triamine (H2N(CH2)2 NH(CH2)2NH2) have the structure shown and readily cyclise to imidazolines (Figure 9.2). Further information is provided by James (1997) and by Franklin et al. (2001).
Cationic surfactants made from fatty amines or fatty alcohols
TABLE 9.9 Product Name
Ether amine Diamine Ethoxylated ether amine Ethoxylated diamine Ethoxylated amine Diamido amineb
Product Structure
Starting Material
a
RO(CH2)3NH2 RNH(CH2)3NH2 a RO(CH2)3NH(CH2CH2O)nH RNH(CH2)3NH(CH2CH2O)nH H(OCH2CH2)nN(R)(CH2CH2O)mH RCONH(CH2)2NH(CH2)2NHCOR
ROH RNH2 RO(CH2)3NH2 RNH(CH2)3NH2 RNH2 H2N(CH2)2NH(CH2)2NH2
Reactant CH2=CHCN CH2=CHCN Ethylene oxide Ethylene oxide Ethylene oxide RCOOH
a
Product after catalytic hydrogenation of an intermediate nitrile. This is the name given to a diacylated triamine. The products are readily cyclised to imidazolones (Figure 9.2). Note: R represents an alkyl chain.
b
TABLE 9.10
Cationic surfactants made from fatty amines
Products Amines Quaternary salts (quats) Amine oxides Amido amines Imidazolines Ester amines Ether amines
Structure
Reactants
RNH2, R2NH, R3N [R2NMe2]+ X[RNMe2]+ ORCONH(CH2)3NMe2 Figure 9.2 RCOOCH2CH2NMe2 RO(CH2)3NH2
Nitriles – hydrogenation Tertiary amines and RX Tertiary amines and H2O2 Polyamines and RCOOH Polyamines and RCOOH Ethanolamines and RCOOH Fatty alcohols and acrylonitrile
Source: Adapted from Gunstone, F.D., The Chemistry of Oils and Fats, Blackwell Publishing, Oxford, 2004.
601
9.3
Surfactants
N
N
O
R
R N
N
R O
CH2CH2NHCOR
FIGURE 9.3
FIGURE 9.2 Imidazolines from RCOOH and diamine (H2N(CH2)2NH2) or triamine (H2N(CH2)2NH(CH2)2NH2).
9.3.8
O
R
OH O
CH2OH
Acetals from RCHO and glycerol.
possibility of using these compounds for drug delivery. The weak bonds can be broken when required with the help of enzymes, by reactions occurring at sewage plants, or by chemical or physical processes involving acid, alkali, ozone, heat, or ultraviolet light. These compounds are generally acetals/ketals or ortho esters (Hellberg, 2003). Cyclic acetals/ketals result when aldehydes or ketones react with polyhydric alcohols, such as glycerol (Figure 9.3), pentaerythryitol, or glucose. The products are 1,3dioxolanes (5-membered hetero ring) or 1,3-dioxanes (6-membered hetero ring). Any unreacted hydroxyl groups can be further functionalised. These compounds are made under anhydrous acidic conditions and are readily hydrolysed under aqueous acidic conditions (Hellberg et al., 2000a). Ortho-esters are made from ethyl orthoformate [HC(OEt)3], alcohols, and monomethyl polyethylene glycol (HO(C2H4O)nMe) in the presence of aluminum chloride. The product is a mixture of many compounds having the structures shown below in which x, y, and z have values 0-3 and x + y + z = 3. The products formed in largest amount have values of x, y, and z of 1,1,1 or 0,2,1, or 0,1,2. Such product mixtures are used for temporary emulsions, hard surface cleaners, textile treatment processing, etc. They are hydrolysed under mild acid or alkaline conditions and have good biodegradability (Hellberg et al., 2000b).
Gemini surfactants and cleavable surfactants
Most surfactants contain one lipophilic chain and one hydrophobic head group. Gemini or dimeric surfactants contain two of each of these linked together by a short aliphatic group or through an aromatic ring. A book devoted to this subject has been edited by Zana and Xia (2003). One example formulated below shows two quaternary groups linked through a tetramethylene spacer (Rosen and Tracy, 1998). Similar compounds with an aromatic spacer have been prepared and assessed for protection of steel fabrics against 2M hydrochloric acid. (Negm and Mohamed, 2004). R(Me)2N+(CH2)4N+(Me)2 R R3N+CH2COOC6H4OCOCH2N+R3 Unlike conventional surfactants, gemini consist of two molecules of monomeric surfactant linked through a flexible spacer. Such molecules are capable of wide variations in terms of each of their three components – lipophilic chain, head group, and spacer — and frequently show remarkable properties. Gemini surfactants generally display unusual patterns of self-assembly and some intriguing physicochemical properties. For example, ionic Gemini surfactants have critical micelle concentrations two orders of magnitude lower than their monomeric alkylsulfonate analogues and are very efficient at reducing surface tension both on their own and in combination with conventional surfactants. Gemini surfactants have much lower Krafft points and higher solubility in water compared to their monomeric counterpoints. Valivety et al. (1998) have described the synthesis of amino acid based gemini surfactants. One example is the compound formulated below and based on myristic acid (2 mols), serine (2 mols) and1,10-dihydroxydecane (1 mol).
[OR]y ⏐ H ⎯ C ⎯ [O(C2H4O)nMe]z ⏐ [OEt]x Ono et al. (2004, 2005) have described the preparation of cleavable surfactants from methyl pyruvate and from diethyl tartrate as shown below. CH3COCOOMe → CH3CH(OR)2COONaR = C8H17, C10H21, or C12H23
C13H27COOCH2CH(NH2)COO(CH2)10OCOCH(NH2) – CH2OCOC13H27
[EtO2CCH(OH)]2 → [RO2CCH(OSO3Na)]2R = C8H17, C10H21, or C12H23
decanediyl-1,10-bis-(3-O-myristoyl-L-serine) Quaternary ammonium compounds, much used as rinse-aids in the past, have been largely replaced by ester quats. Typical structures of these two categories of compounds are shown:
Another group of surfactants have a weak bond built into the molecule. These are interesting because this feature leads to improved biodegradabilty and opens up the
602
Nonfood Uses of Oils and Fats
Dembitsky, V.M., Astonishing diversity of natural surfactants: 3. Carotenoid glycosides and isoprenoid glycolipids, Lipids, 40, 535–557, 2005b. Dembitsky, V.M., Astonishing diversity of natural surfactants: 4. Fatty acid amide glycosides, their analogues and derivatives, Lipids, 40, 641–660, 2005c. Dembitsky, V.M., Astonishing diversity of natural surfactants: 5. Biologically active glycosides of aromatic metabolites, Lipids, 40, 869–900, 2005d. Dembitsky, V.M., Astonishing diversity of natural surfactants: 6. Biologically active marine and terrestrial alkaloid glycosides, Lipids, 40, 1081–1105, 2005e. Di Serio, M., et al., Narrow-range ethoxylation of fatty alcohols promoted by a zirconium alkoxide sulfate catalyst, J. Surfact. Deterg., 1, 83–91, 1998. Franklin, R., et al., Cationic and amine-based surfactants, in Oleochemical Manufacture and Applications, Gunstone, F.D., Hamilton, R.J., Eds., Sheffield Academic Press, Sheffield, U.K., 2001, Chap 2. Gunstone, F.D., Chemical reactions of fatty acids with special reference to the carboxyl group, Eur. J. Lipid Sci. Technol., 103, 307–314, 2001. Gunstone, F.D., The Chemistry of Oils and Fats, Blackwell Publishing, Oxford, 2004, Chaps. 8, 11. Gunstone, F.D. and Heming, M.D., Glycerol – an important product of the oleochemical industry, Lipid Technol., 16, 177–179, 2004. Hargreaves, T., Chemical Formulation, RSC, Cambridge, U.K., 2003. Hellberg, P-E., Cleavable surfactants — giving an unstable advantage, Lipid Technol., 15, 101–105, 2003. Hellberg, P-E., Ortho ester-based cleavable cationic surfactants, J. Surfact. Deterg., 5, 217–227, 2002. Hellberg, P-E., et al., Cleavable surfactants, J. Surfact. Deterg., 3, 81–91, 2000a. Hellberg, P-E., et al., Nonionic cleavable ortho ester surfactants, J. Surfact. Deterg., 3, 369–379, 2000b. Hidaka, H., et al., Preparation and properties of new tetrafunctional amphoteric surfactants bearing amino, carboxyl, and hydroxyl groups and an ether bond, J. Surfact. Deterg., 6, 131136, 2003. Hill, K., et al., Alkylpolyglycosides: Technology, Properties, and Applications, VCH, Weinheim, Germany, 1997. Hill, K., Fats and oils as oleochemical raw materials, J. Oleo Sci., 59, 433–444, 2001. Hreczuch, W., et al., Direct ethoxylation of longer-chain aliphatic ester, J. Surfact. Deterg., 4, 167–173, 2001. James, A.D., Cationic surfactants, in Lipid Technologies and Applications, Gunstone, F.D., Padley, F.B., Eds., Marcel Dekker, New York, 1997, Chap 24. Jung, S., et al., Structure and surface-active property determinations of fructose mono-oleates, J. Surfact. Deterg., 1, 53–57, 1998. Karsa, D. R., Quaternary ammonium surfactants, Lipid Technol., 7, 81–86, 2001. Levinson, M.I., Rinse-added fabric softener technology at the close of the twentieth century, J. Surfact. Deterg., 2, 223–235, 1999. Morelli, J.J. and Szajer, G., Analysis of Surfactants: Part I, J. Surfact. Deterg., 3, 539–552, 2000. Morelli, J.J. and Szajer, G., Analysis of surfactants: Part II, J. Surfact. Deterg., 4, 75–83, 2001. Negm, N.A. and Mohamed, A.S., Surface and thermodynamic properties of diquaternary bola-form amphphiles containing an aromatic spacer, J. Surfact. Deterg., 7, 23–30. 2004.
quat: R2N+Me2 Xester quat: (RCOOCH2CH2)2N+Me2 XThe ester quats are stable to acids, but are easily hydrolysed by alkali to soap and the compound ((HOCH2CH2)2 N+Me2X-). They show better environmental characteristics than the quats themselves (Hellberg, 2002). Alkyl ethoxylates (R(OC2CH2)nOH) are viscous oils that are not always easy to handle, but they react with carbon dioxide to form carbonates (R(OCH 2 CH 2 ) n OCO2Na). These are solid and are easily incorporated into granular detergents. In an alkaline solution, the ethoxylates are quickly regenerated from the carbonates.
References Allen, D.K. and Tao, B.Y., Carbohydrate-alkyl ester derivatives as biosurfactants, J. Surfact. Deterg., 2, 383–390, 1999. Anon., Publications and reports, J. Surfact. Deterg., 3, 139–141, 2000. Anon., Consolidation in the surfactants industry, Inform, 12, 872–882, 2001. Anon., Publications and reports, J. Surfact. Deterg., 5, 426, 2002a. Anon., LAS decline, Inform, 13, 704–705, 2002b. Anon., Publications and reports, J. Surfact. Deterg., 6, 396, 2003. Appel, P.W., Modern methods of detergent manufacture, J. Surfact. Deterg., 3, 395–405, 2000. Baker, I.J.A., et al., Sugar fatty acid ester surfactants: structure and ultimate aerobic biodegradability, J. Surfact. Deterg., 3, 1–11, 2000a. Baker, I.J.A., et al., Sugar fatty acid ester surfactants: biodegradation pathways, J. Surfact. Deterg., 3, 13–27, 2000b. Baker, I.J.A., et al., Sugar fatty acid ester surfactants: basecatalysed hydrolysis, J. Surfact. Deterg., 3, 29–32, 2000c. Berna, J.L., et al., Laundry products in bar form, J. Surfact. Deterg., 1, 263–271, 1998. Bognolo, G., Nonionic surfactants, in Lipid Technologies and Applications, Gunstone, F.D., Padley, F.B., Eds., Marcel Dekker, New York, 1997, Chap 25. Cohen, L. and Trujillo, F., Synthesis, characterisation, and surface properties of sulfoxylated methyl esters, J. Surfact. Deterg., 1, 353341, 1998. Cox, M.F. and Weerasooriya, U., Impact of molecular structures on the performance of methyl ester ethoxylates, J. Surfact. Deterg.. 1, 11–22, 1998. Cox, M.F. and Weerasooriya, U., Enhanced propoxylation of alcohols and alcohol ethoxylates, J. Surfact. Deterg.. 2, 59–68, 1999. Cox, M.F. and Weerasooriya, U., Partially saponified triglyceride ethoxylates, J. Surfact. Deterg.. 3, 213, 2000. Cox, M.F. et al. Methyl ester propoxylates, J. Surfact. Deterg.. 1, 167–175, 1998. Dembitsky, V.M., Astonishing diversity of natural surfactants: 1. Glycosides of fatty acids and alcohols, Lipids, 39, 933–953, 2004. Dembitsky, V.M., Astonishing diversity of natural surfactants: 2. Polyester glycosidic ionophores and macrocyclic glycosides, Lipids, 40, 219–248, 2005a.
603
9.4
Cosmetics and personal care products
of lipids make them suitable for controlling product consistency and to favourably influence the skin feel of the preparation. Knowledge about polymorphism and crystallisation kinetics is essential to optimise formulations based on lipids. The second aspect of lipids in cosmetic formulations being covered is the ability of essential fatty acids, tocopherols, and phytosterols to improve skin health. Lipids are important building blocks of the cell membranes; they act as antioxidants and anti-inflammatory agents and can have a fundamental effect on collagen production and break down in the skin. All of this illustrates the versatility of lipids as formulating tools in skin care and cosmetics. Lipids are important multifunctional ingredients commonly used in cosmetic and personal care products. They may function as emollients, moisturisers, emulsifiers, solubilisers, dispersing agents, texturisers and skin-feel improvers. Some lipids show interesting bioactivity (antiinflammatory and antioxidative) and influence protein synthesis and degradation in the skin. They find uses in skin-care emulsions, ointments, and balms as well as colour cosmetics and personal care products, such as shower gels, shampoos, and hair conditioners. The definition of lipids in cosmetic applications is variable and in this review the emphasis will be on acylglycerols, tocopherols, and phytosterols. The use of lipids and lipid-derived materials as texturisers will first be described, followed by examples of their bioactivity and interactions with the skin.
Ono, D., et al., Synthesis and properties of soap types of doublechain cleavable surfactants derived from pyruvate, J. Oleo Sci., 53, 89–95, 2004. Ono, D., et al., Preparation and properties of bis(sodium sulfate) types of cleavable surfactants derived from diethyl tartrate, J. Oleo Sci., 54, 51–57, 2005. Piispanen, P.S., et al., Surface properties of surfactants derived from natural products. Syntheses and structure/property relationships – Part 1, solubility and emulsification, Part 2, foaming, dispersion and wetting, J. Surfact. Deterg., 7, 147–159, 161–167, 2004. Polat, T. and Linhardt, R.J., Syntheses and applications of sucrose-based esters, J. Surfact. Deterg., 4, 415, 2001. Porter, M.R., Anionic detergents, in Lipid Technologies and Applications, Gunstone, F.D., Padley, F.B., Eds., Marcel Dekker, New York, 1997, Chap 23. Roberts, D.W., Manufacture of anionic surfactants, in Oleochemical Manufacture and Applications, Gunstone, F.D., Hamilton, R.J., Eds., Sheffield Academic Press, Sheffield, U.K., 2001, Chap 3. Rosen, M.J. and Tracy, D.J., Gemini surfactants, J. Surfact. Deterg., 1, 547–554, 1998. Rosen, M.J. and Dahanayake, M., Industrial Utilisation of Surfactants, Principles and Practice, AOCS Press, Champagne, IL, 2000. Schiebel, J.J., The evolution of anionic surfactant technology to meet the requirements of the laundry detergent industry, J. Surfact. Deterg., 7, 319–328, 2004. Solaiman, D.K.Y., Applications of microbial biosurfactants, Inform, 126, 408–410, 2005. The Surfactants Virtual Library: www.surfactants.net Thin Sue Tang, Analysis of oleochemicals, in Oleochemical Manufacture and Applications, Gunstone, F.D., Hamilton, R.J., Eds., Sheffield Academic Press, Sheffield, U.K., 2001, Chap 8. Tracy, D.J. and Reierson, R.L., Commercial synthesis of monoalkyl phosphates, J. Surfact. Deterg., 5, 169–172, 2002. Urata, K. and Takaishi, N., Newer synthetic approaches to surfactants, lipids, and related compounds based on C-3 building blocks: Recent advances related to fatty chemistry, J. Surfact. Deterg., 2, 91–103, 1999. Urata, K. and Takaishi, N., Self-assembly of compounds based on a glycerin skeleton as a C-3 building block, J. Surfact. Deterg., 5, 287294, 2002. Valivety, R., et al., Application of enzymes to the synthesis of amino acid-based bola and gemini surfactants, J. Surfact. Deterg., 1, 177–185, 1998. Waldhoff, H. and Spilker, R., Eds., Handbook of Detergents, Part C: Analysis, vol. 123, Marcel Dekker, New York, 2005. Watkins, C., Methyl ester sulfonates, Inform, 12, 11531159. 2001. Weerasooriya, U., Ester alkoxylation technology, J. Surfact. Deterg., 2, 373–381, 1999. Zana, R. and Xia, J., Eds., Gemini Surfactants: Synthesis, Interfacial and Solution-Phase Behaviour and Applications, Marcel Dekker, New York, 2003.
9.4
Cosmetics and personal care products
9.4.1
Introduction
9.4.2
Texture control using solid and semisolid lipids
A cosmetic skin-care formulation is a complex mixture of ingredients with varying functionality. A majority of such products are stabilised emulsions of an emollient in water with added bioactive materials for delivering real or apparent benefits to the well-being of the skin. The formulation needs to deliver both water-soluble actives and oil-soluble ingredients to the stratum corneum and preferentially control the penetration of these substances to the epidermis and dermis. The formulation regularly has to have a shelf life of more than 24 months, putting a lot of requirements on the emulsifying and stabilising system. Finally, the formulation also needs to be aesthetically and sensorially acceptable in order to fulfill the demands of the consumer. The main component in the oil phase of skin-care emulsions is the emollient. It comprises one or more oils of differing composition and chemical structure. Many lipidbased and lipid-derived materials have been suggested for use as emollients, including esters of long-chain fatty acids with both short- and long-chain alcohols, natural and synthetic triacylglycerols, and naturally occurring hydrocarbons, such as squalene and squalane. The primary function of these emollients is to lubricate the skin, decrease the water permeability, and to act as a carrier of
This review covers two aspects of using lipids in cosmetic and skin-care formulations. The physicochemical properties 604
Nonfood Uses of Oils and Fats
In general terms, the more solids at a given temperature, the harder and more brittle the emulsion will appear. It is important to optimise the solid phase content for the different temperature ranges that the product will encounter. For example, the solid phase content at body or skin temperature (34 to 37°C) determines the sensory properties and the spreadability of the product. The solids content at 40 to 45°C will have an influence on the high temperature storage stability of the formulation. Finally, the solids content at 20 70 25°C will determine the texture and consistency of the product at room temperature. Characteristic ranges for skin-care creams are less than 50% solids at room temperature, less than 30% solids at body temperature, and 5 to 15% solids at 40°C (Figure 9.4). With these solids ranges, a product will have good stability during storage, be nicely spreadable on the skin, and have good high temperature resistance.
the actives in the formulation. A secondary function of increasing importance is to influence the formulation texture and the sensory aspects when applying the emulsion to the skin. Important considerations when selecting emollients are their effects on emulsion structure and how that effect is translated to sensory properties on application (Wiechers et al., 2004). Physicochemical properties, such as viscosity and polarity, are strongly related to the ability of an emollient to spread onto and penetrate into the skin. By combining emollients with different characteristics in terms of spreading ability and viscosity, optimal performance on different timescales can be obtained. The ability of an emollient to interact with surfactants in liquid crystalline phases also plays a significant role in the performance (Wiechers, 2003). The microstructure and the resulting texture of cosmetic emulsions can be strongly influenced using semisolid and solid crystallising materials in the oil phase. By a careful selection of semisolid fats and waxes, the consistency of a cosmetic cream or lotion can be fine tuned to obtain optimal sensory properties and good product stability. In this context, the crystallisation behaviour of the ingredients used is strongly influencing the final result. There are three physicochemical concepts that are useful in the design of an emollient for cosmetic creams: the solids content of the oil phase, the polymorphic behaviour of the solids, and, finally, the kinetics of crystallisation.
9.4.3
9.4.4
The polymorphic behaviour of the solids used in cosmetic formulations is generally not well known by the formulators. Most simple waxes (wax esters and hydrocarbonbased waxes) crystallise in orthorhombic crystals similar to the beta-prime polymorph of triacylglycerols (Small, 1984). Semisolid and solid triacylglycerol based ingredients display a more complex crystallisation behaviour including the alpha-, beta-prime and beta forms. Many hydrogenated vegetable fats, such as hydrogenated soybean oil and hydrogenated palm oil, as well as the lauric oils, coconut oil, and palm kernel oil are generally stable in the beta-prime form. Many of the exotic “butters” used in cosmetic applications — cocoa butter, shea butter,
Solid phase content
First of all, the solid phase content is directly correlated with the consistency and sensory properties of the product.
Solid phase content (%)
Polymorphism
80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0
Soft emollient Waxy emollient
5
10
15
20
25
30
35
40
45
50
Temperature (°C)
FIGURE 9.4 Solid phase content as a function of temperature for two typical emollient blends: (a) an emollient with soft, plastic consistency at room temperature and good stability at elevated temperatures, and (b) an emollient with hard, waxy texture at room temperature and almost complete melting at body temperature. Solid phase content in the mixtures was determined by low resolution pulsed NMR after tempering at 26°C for 40 hours.
605
9.4
Cosmetics and personal care products
shorea butter and so on — crystallise in the triple-chain packed beta form. These butters normally require a good control of the crystallisation conditions (“tempering”) when used in higher concentrations if problems with the storage stability are to be avoided. When mixing ingredients from different groups, the resulting complex polymorphic behaviour can strongly influence the stability and sensory properties of the formulation. The best results in both aspects are obtained if the polymorphic behaviour of the system is matched; all ingredients should be either in the beta prime or in the beta modification. For further discussion of polymorphism, see Section 7.2.4.
9.4.5
9.4.6
The combination of polymorphic behaviour and response on cooling will determine the morphology of the crystals produced. Depending on the shape and size distribution of the crystals, the consistency can be softer or harder. Needle-like crystals, characteristic of the beta-prime polymorphs of fats and wax esters, will give a higher degree of elasticity compared to the more regularly shaped crystals associated with the beta polymorphs. Shear during cooling will also influence the shape and the size distributions of the particles. Rapid cooling can sometimes lock the particles in an unstable polymorph that will slowly transform on storage, resulting in undesired consistency changes. Both polymorphism and the solid phase content are dependent on the liquid phase used in the formulation. The solubility of the fats and waxes used will be different in different emollients and unexpected effects on product consistency and stability can be seen when the emollient composition is changed. In emulsions, the emulsion droplet sizes can also influence both solid phase content as well as polymorphic forms. If the emulsion droplets are small enough, a super-cooling of the oil phase can sometimes be observed with unstable crystal forms such as alpha being more stable. In anhydrous formulations, such as balms and ointments, as well as colour cosmetics (lipsticks, pencils, mascaras and foundations), the compatibility between the solid phases becomes very important for product performance (Matsuda and Yamaguchi, 2001). Co-crystallising solids that cover a wide melting point range can give very stable formulations with excellent consistency, stability, and skin feel.
Crystallisation kinetics
The response of a formulation to processing, especially cooling conditions, is dependent on the inherent crystallisation behaviour of the mixture of solids used in the product. Cooling conditions available in cosmetic product manufacturing frequently are not designed for rapid cooling, a fact the formulator needs to consider when selecting ingredients. Control of cooling conditions is also relevant when considering scale-up effects and product stability during storage. The most obvious influence of cooling conditions is on product consistency. In most systems, rapid cooling results in a massive burst of crystal nuclei that will have a limited potential of growth, resulting in small crystals and a harder consistency (Figure 9.5). The surface of the product is smooth and glossy if the crystals are small. Slow cooling results in fewer crystal nuclei that will grow to large sizes with time. Such systems are usually softer in consistency and spread more easily. However, uncontrolled growth can also lead to a grainy product with obvious “bloom” on the surfaces.
Penetrometer hardness (g)
500
Combined effects
Fast crystallisation Slow crystallisation
400
300
200
100
0 5
10
15
20
25
30
Solid fat content @ 22 C (%)
FIGURE 9.5 Hardness of emollient mixtures crystallised at different conditions: (a) rapidly cooled from melt to 20°C, cooling rate 10°C/minute, and (b) slowly cooled from melt to 20°C, cooling rate 0.3°C/minute. Solid phase content in the mixtures was determined by low resolution pulsed NMR at 20°C after tempering for 2 hours.
606
Nonfood Uses of Oils and Fats
9.4.7
(4 to 17%), cholesterol sulfate (1.5 to 4.5%), and triacylgly-cerols (5 to 25%) (Engblom, 1996).
Bioactive lipids in skin-care applications
The skin is the largest organ of the human body, covering about 1.5 to 2 m2, presenting a total thickness of less than 2 mm in most locations and weighing about 4 kg in adults. It is the only organ completely exposed to the environment, making the major role of the skin that of a barrier against air, microorganisms, and environmental pollutants. The outermost layer, stratum corneum or the horny layer, consists of cells named corneocytes embedded in a lipid matrix. The lipids form an intercellular lamellar sheet between the cornified cells and constitute the primary barrier of the skin. The structure and function of the stratum corneum is extensively studied and several models for the interaction between the epidermal lipids and the corneocytes have been proposed, including the “brick-andmortar” of Michaels (Michaels, Chandrasekaran, and Shaw, 1975), the “domain mosaic model” of Forslind (Forslind, 1994), and more recently the “single gel phase” model of Norlén (Norlén, 2001). A schematic cross section through human skin is shown in Figure 9.6. The outermost part of the epidermis is normally considered as the main target for cosmetic and personal care products. For optimal function, the skin requires a selection of different lipids. There are two main types of skin lipids; on the skin surface, there is the sebum, generated in the sebaceous glands and dominated by triacylglycerols, wax esters, and squalene. The epidermal lipids, which are generated in the epidermal cells, are essential for maintaining the integrity of the skin barrier by preventing the penetration of impurities, chemicals, microorganisms, and water, while also protecting against undesirable water loss through the skin. The main components of the epidermal lipids are ceramides (15 to 41%), free fatty acids (7 to 23%), cholesterol (13 to 34%), cholesterol esters
9.4.8
Lipids in skin care
Lipids are important ingredients in all skin-care categories and of special importance for dry and sensitive skin, and for antiaging and protecting skin-care formulations. Apart from acting as emollients, many lipids also function as delivery agents for various bioactive materials in the formulations. Dry-skin conditions have become a widespread problem in many parts of the world mainly due to life-style changes, including altered dietary patterns, changes in workplace conditions, and a comparatively older population. Dry skin is characterised by a reduced content of water and an altered lipid composition in the stratum corneum. A defective skin barrier results in increased water evaporation and an increased sensitivity to the environment. Thus, an ideal skin-care formulation should contain ingredients that improve barrier function and repair as well as supplement the natural epidermal lipids (Loden and Maibach, 2005; Park, 2001).
9.4.9
Essential fatty acids
Essential fatty acids are those polyunsaturated fatty acids (PUFA) that are necessary for good health, but cannot be synthesized in the body. Dry and atopic skin shows a decrease in linoleic acid content, the important precursor of ceramides essential for the barrier function of the skin (Horrobin, 2000). Both topical application and dietary intake of essential fatty acids have been shown to restore dry skin conditions as well as having therapeutic effects on skin disorders, such as atopic dermatitis, psoriasis, and
Intercellular lipid barrier
Epidermis Thickness 0.05-1 mm Corneocytes form in stratum basale and migrate to the stratum corneum
Stratum corneum Stratum granulosum Stratum spinosum
Dermis Thickness 3-5 mm Hair follicles and growing hair Sebaceous glands produce sebum Sweat gland Fibroblasts producing structural proteins Blood vessels, nerves
FIGURE 9.6
Schematic representation of skin structure.
607
9.4
Cosmetics and personal care products
acid, are shown to offer significant antioxidant and membrane stabilising properties in human skin (Kitazawa et al., 1997; Shindo et al., 1994). In addition, α-tocopherol has been demonstrated in vitro to reduce the age-dependent increase of collagenase activity, potentially delaying the progression of skin aging (Ricciarelli et al., 1999). The antiinflammatory action of γ-tocopherol has been demonstrated in dietary studies (Jiang et al., 2001; Jiang et al., 2000;, Jiang and Ames, 2003). The combination of α- and γ-tocopherol found in many seed oils is interesting for the dual activity these oils can have when used as emollients in topical applications. These nontoxic and biologically active substances are present at varying quantities in vegetable oils. Typical compositions for some common oils are given in Chapter 2.
acne (Conti, 1996; Spector, 1999). Common vegetable oils with a high content of essential fatty acids, such as linoleic acid, that can be used as emollients in skin-care products, are products derived from rapeseed, sunflower, corn, soybean, arachis (groundnut, peanut), and cottonseed. Another polyunsaturated fatty acid – γ-linolenic acid – has also shown potential for treating dry skin conditions caused by atopic dermatitis. γ-Linolenic acid metabolism leads to antiinflammatory prostaglandins that can ameliorate the effects of inflammatory skin disorders (Horrobin, 2000; Ziboh, Miller, and Cho, 2000). Oils derived from evening primrose, borage, blackcurrant, and echium are known for their high content of γ-linolenic and stearidonic fatty acids (Section 2.3.1.5). Due to the high sensitivity of polyunsaturated fatty acids with respect to oxidation, the addition of an optimised antioxidant system during oil processing is recommended. The added antioxidant will protect the oil during storage, but also contribute to the protection of the formulation.
9.4.10
9.4.11
Sterols and triterpene alcohols
Sterols (desmethylsterols), triterpene alcohols (4,4-dimethylsterols) and their derivatives such as hormone and vitamin precursors impart various important biological functions within the body. Desmethylsterols are commonly found in vegetable oils in concentrations ranging from 0.1 to 1% (Section 2.2.19; Table 2.47). The predominant sterol in seed oils is normally beta-sitosterol, but other sterols can also be found in significant amounts in specific plants. Triterpene alcohols are normally present in much lower concentrations in seed oils. Uniquely, high concentrations of triterpene alcohols are found in shea butter, the fat extracted from the kernels of the shea tree (Vitellaria paradoxa or Butyrospermum parkii) (Peers, 1977), growing in the arid regions of subSaharan Africa.
Tocopherols
The human skin is rich in lipids, proteins and DNA, all of which are extremely sensitive to oxidation and a proper protection against oxidation is necessary for the health of the skin (Kohen, 1999; Rengarajan, 1999). Beside the obvious oxidative damage to the skin constituents, free radicals and reactive oxygen species can initiate inflammatory reactions and activate matrix metalloproteinases (Brenneisen, Sies, and Scharffetter-Kochanek, 2002). Free radicals are closely linked with aging and oxidative stress in the skin, being associated not only with decreased cell viability and DNA damage, but as a significant agent in the skin-aging process when the skin loses its elasticity and regenerative power. It, therefore, is essential to protect the skin against the effects of UV radiation and oxygen-derived free radicals and many formulation strategies to achieve this have been developed during the past decades. The lipophilic tocopherols are essential for the stabilisation of biological membranes, especially those containing large amounts of polyunsaturated fatty acids. α-Tocopherol is particularly effective in protecting against oxidative damage of cellular membranes and biomolecules, such as lipids, proteins, and nucleic acids (Kohen 1999, Nachbar and Korting, 1995; Thiele et al., 2001). The tocopherols act as antioxidants primarily by a free radical scavenging mechanism. It has been shown to be of obvious physiological advantage to deliver tocopherols topically, which can ameliorate the early phase of an oxidative stress response. For example, topical application of vitamin E is shown to reduce the appearance of fine facial lines and wrinkles as well as increasing stratum corneum hydration and enhancing its water-binding capacity (Gehring, Fluhr, and Gloor, 1998). Synergistic effects between tocopherols and other antioxidants, such as flavonoids and ascorbic
9.4.12
Antiinflammatory effects of sterols and tocopherols
Phytosterols from rapeseed oil have been shown to impart antiinflammatory and healing effects on surfactant damaged skin (Loden and Andersson, 1996). The benefits using the combination of tocopherols and phytosterols in an emollient were demonstrated in a clinical study using rapeseed oil fractions with different levels of sterols and tocopherols. It was shown that a pretreatment with rapeseed oil fractions reduced the transepidermal water loss (TEWL) and erythema in skin treated with sodium lauryl sulfate solutions. The reference materials were petrolatum, an inert occlusive emollient, borage oil, which is rich in γ-linolenic acid, and a hydrocortisone cream, a known antiinflammatory preparation. The fractionated rapeseed oil showed activity comparable to the hydrocortisone cream, while no effects were observed with petrolatum and borage oil. Phytosterols are also known for having a structural role by interacting with the lameIlar lipid layers, strengthening the lipid barrier, and improving dry skin conditions, squamation, and erythema (Chlebarov, 1989). 608
Nonfood Uses of Oils and Fats
9.4.13
lupeol and its palmitate and linoleate were inhibitors of trypsin activity, while no effect on porcine pancreatic elastase was observed. Metalloproteases (e.g., collagenase) and serine proteases (trypsin, chymotrypsin, porcine pancreatic elastase, human leukocyte elastase) are inhibited in vitro by various types of triterpenes, including lupeol and its esters. Rajic et al. (2000) showed that esterification increases the degree of inhibition of trypsin and chymotrypsin. Lupeol palmitate, lupeol linoleate, and alpha-amyrin linoleate were potent trypsin inhibitors, while free lupeol and alpha-amyrin were less efficient. Chymotrypsin was inhibited by lupeol, the other tested compounds being weaker inhibitors. These examples show that there is a potentially useful effect of the triterpene alcohols from shea butter to prevent aging effects on the skin by inhibiting the degrading activity of proteases. Triterpene derivatives are also known for stimulating collagenesis of skin fibroblasts (Laugel, 1998), indicating a second mechanism for delaying skin aging and changes due to lowered content of the structural proteins.
Bioactivity of triterpene alcohols
There have been many bioactivity studies conducted with various types of triterpene alcohols. The studies performed with these alcohols indicate that there are at least two areas where interesting bioactive effects can be expected. The antiinflammatory effects of phytosterols, including the triterpene alcohols, are well demonstrated for both the free alcohols and their esters. The second effect is associated with the synthesis and degradation of the structural proteins collagen and elastin. Many individual triterpene alcohols and their natural mixtures have been investigated for their antiinflammatory properties. Several studies show that the two amyrins as well as lupeol and butyrospermol are antiinflammatory in different types of inflammation models. For example, Akihisa et al. (1997) presented data on the antiinflammatory effect of a large number of triterpene alcohols found in Theaceae oils (Camellia and Sasanqua), including butyrospermol, lupeol, alpha- and beta-amyrin, as well as taraxasterol, psitaraxasterol, and 24-methylenedammarenol (all of them also found in high concentrations in shea butter). All of these triterpene alcohols (in the form of free alcohols) showed inhibitory activities in the same concentration range as the control substance indomethacin, when tested in an inflammation model in mice. The mechanism for the inflammatory action of lupeol and its esters was investigated by Fernandez et al. (2001). The antiinflammatory activity of lupeol was studied in models demonstrating effects on two different inflammatory pathways. Topically applied lupeol had a significant antiinflammatory effect in the TPA model (cyclooxygenase pathway), while the effect was less pronounced in the arachidonic acid-induced oedema (lipoxygenase pathway), only the highest concentration tested gave any significant effect. It was concluded that lupeol is an inhibitor of certain proinflammatory mediators, such as prostaglandin E2 (a cyclooxygenase metabolite) and cytokines, but not leukotrienes (lipoxygenase metabolites). Some of the triterpene alcohols found in shea butter are also inhibitors of protein degrading enzymes, proteases. Different types of proteases are active in the skin, degrading collagen and elastin, two of the major structural proteins contributing to the toughness and strength of the skin. The production of collagen and elastin decreases with increasing age, resulting in thinner and less elastic skin. The effects of this natural aging process can be alleviated by stimulating collagen and elastin synthesis or by inhibiting the activity of the collagenases and elastases. Proteases are also implicated in the breakdown of connective tissue in rheumatoid arthritis and the triterpene alcohols and their derivatives have been investigated as alternatives to conventional pharmaceutical products, such as hydrocortisone and indomethacin. Several studies have been conducted to evaluate the inhibitory effect of triterpene alcohols on different types of proteases. For example, Hodges et al. (2003) showed that
9.4.14
Products enriched with minor lipids
Normally the natural levels of triterpene alcohols, sterols, and tocopherols are too low for having extensive effect on topical application. Vegetable oils with naturally high contents of sterols include wheat germ oil, rapeseed oil, and soybean oil. Shea butter, avocado oil, rice bran oil, and olive oil also contribute to high levels of triterpene alcohols (4,4-dimethylsterols). The naturally occurring levels of these minor components can be increased by different processing methods including saponification followed by extraction and distillation (Clark, 1996). A nondestructive way of isolating these components is based on low-temperature solvent fractionation (Alander et al., 2005; Mellerup, Bach, and Enkelund, 2002). These concentrated fractions can be used as part of the emollient system in a cosmetic formulation to deliver the bioactive minor components to the skin.
References Akihisa, T. et al. 1997, Triterpene alcohols from camellia and sasanqua oils and their anti- inflammatory effects, Chem.Pharm.Bull.(Tokyo), 45, 2016–2023. Alander, J. et al. 2005, Fractionation process, EP 1084215, Karlshamns AB, (patent). Brenneisen, P. et al. 2002, Ultraviolet-B irradiation and matrix metalloproteinases: from induction via signaling to initial events, Ann. N.Y. Acad. Sci, 973, 31–43. Chlebarov, S., 1989, Notabene Medici, vol. 2 and 3. Clark, J.P. 1996, Tocopherols and sterols from soybeans, Lipid Technol., 8, 111–114. Conti, A., 1996, Seasonal influences on stratum corneum ceramide I fatty acids and the influence of topical essential fatty acids, Int. J. Cosmet. Sci., 18, 1–12. Engblom, J., 1996, On the phase behaviour of lipids with respect to skin barrier function, Department of Food Technology, Lund University, Lund, Sweden.
609
9.5
Lubricants
protein kinase C inhibition, Free Radic. Biol. Med., 27, 729–737. Shindo, Y. et al. 1994, Enzymic and non-enzymic antioxidants in epidermis and dermis of human skin, J. Invest. Dermatol., 102, 122–124. Small, D.M., 1984, Lateral chain packing in lipids and membranes, J. Lipid Res., 25, 1490–1500. Spector, A.A., 1999, Essentiality of fatty acids, Lipids, 34, S1–S3. Thiele, J.J. et al. 2001, The antioxidant network of the stratum corneum, Curr. Probl. Dermatol., 29, 26–42. Wiechers, J.W., 2003, The EEC concept: how emollients and emulsifiers work together to create more efficacious cosmetic products, SÖFW-J., 129, 2225. Wiechers, J.W. et al. 2004, Formulating for efficacy, Int. J. Cosm. Sci., 26, 173–182. Ziboh, V.A. et al. 2000, Metabolism of polyunsaturated fatty acids by skin epidermal enzymes: generation of antiinflammatory and antiproliferative metabolites, Am. J. Clin. Nutr., 71, 361S–366S.
Fernandez, M.A. et al. 2001, New insights into the mechanism of action of the anti-inflammatory triterpene lupeol, J. Pharm. Pharmacol., 53, 1533–1539. Forslind, B., 1994, A domain mosaic model of the skin barrier, Acta Derm. Venereol., 74, 1–6. Gehring, W. et al. 1998, Influence of vitamin E acetate on stratum corneum hydration, Arzneimittelforschung., 48, 772–775. Hodges, L.D. et al. 2003, Antiprotease effect of anti-inflammatory lupeol esters, Mol. Cell Biochem, 252, 97–101. Horrobin, D.F., 2000, Essential fatty acid metabolism and its modification in atopic eczema, Am. J. Clin. Nutr., 71, 1 Suppl, 367S–372S. Jiang, Q. and Ames, B.N., 2003, Gamma-tocopherol, but not alpha-tocopherol, decreases proinflammatory eicosanoids and inflammation damage in rats, FASEB J., 17, 816–822. Jiang, Q. et al. 2001, Gamma-tocopherol, the major form of vitamin E in the U.S. diet, deserves more attention, Am. J. Clin. Nutr., 74, 714–722. Jiang, Q. et al. 2000, Gamma-tocopherol and its major metabolite, in contrast to alpha-tocopherol, inhibit cyclooxygenase activity in macrophages and epithelial cells, Proc. Natl. Acad. Sci. U.S.A., 97, 11494–11499. Kitazawa, M. et al. 1997, Interactions between vitamin E homologues and ascorbate free radicals in murine skin homogenates irradiated with ultraviolet light, Photochem. Photobio., 65, 355–365. Kohen, R., 1999, Skin antioxidants: their role in aging and in oxidative stress—new approaches for their evaluation, Biomed. Pharmacother., 53, 181–192. Laugel, C., 1998, Incorporation of triterpenic derivatives within an o/w/o multiple emulsion: structure and release studies, Int. J. Cosmet. Sci., 20, 183–191. Loden, M. and Andersson, A.C., 1996, Effect of topically applied lipids on surfactant-irritated skin, Br. J. Dermatol., 134, 215–220. Loden, M. and Maibach, H.I., 2005, Dry Skin and Moisturizers: Chemistry and Function, 2nd ed., CRC Press, Boca Raton, FL. Matsuda, H. and Yamaguchi, M., 2001, Separation and crystallization of oleaginous constituents in cosmetics —sweating and blooming, in Crystallization Processes in Fats and Lipid Systems, Garti, N. and Sato, K., Eds, Marcel Dekker, New York, pp. 485–503. Mellerup, J. et al. 2002, A process for preparing vegetable oil fractions rich in non-tocolic, high-melting, unsaponifiable matter, WO0250221, Aarhus Oliefabrik A/S, (patent). Michaels, A.S. et al. 1975, Drug permeation through human skin; theory and in vitro experimental measurement, AIChE J., 21, 985–996. Nachbar, F. and Korting, H.C., 1995, The role of vitamin E in normal and damaged skin, J. Mol. Med., 73, 7–17. Norlen, L., 2001, Skin barrier structure and function: the single gel phase model, J. Invest. Dermatol., 117, 830836. Park, W.S., 2001, Improvement of skin barrier function using lipid mixtures, SÖFW-J., 127, 9. Peers, K.E., 1977, The non-glyceride saponifiables of shea butter, J. Sci. Food Agric., 28, 1000–1009. Rajic, A. et al. 2000, Inhibition of serine proteases by antiinflammatory triterpenoids, Planta Medica, 66, 206–210. Rengarajan, H., 1999, Skin delivery of Vitamin E, J. Cosmet. Sci., 50, 249–279. Ricciarelli, R. et al. 1999, Age-dependent increase of collagenase expression can be reduced by alpha-tocopherol via
9.5
Lubricants
9.5.1
Introduction
Lubrication is the use of a material to improve the smoothness of movement of one surface over another. The material used to achieve this is called a lubricant. These are usually liquids or semiliquids, but may be solids or gases or any combination of solids, liquids, and gases. Smoothness of movement is improved by reducing friction. However, this is not always the case, and there may be situations in which it is more important to maintain steady friction than to obtain the lowest possible friction. In addition to simply lubricating the metal parts that come in contact, lubricants are expected to reduce or control friction between metal parts to save energy, reduce wear or prevent weld of metal surfaces, clean metal surfaces of dirt or sludge to prevent scratching or scoring, clean metal surfaces of water and acids to prevent corrosion, and often to prevent overheating. Annual consumption of oil-based lubricants in the U.S. is close to 10 million metric tons and valued at more than $8 billion (USD). The U.S. usage accounts for 27% of the world lubricant consumption (37.5 million metric tons). With a share of almost a third of global lubricant consumption, Asia-Pacific remains the leading lubricant region, followed by Europe and North America. More than 70% of total lubricant volume is used in motor oils for automotive engines and approximately 10% in hydraulic fluids. Other application areas, mostly industrial lubricants, are less significant. Lubricants are usually divided into four basic classes. 9.5.1.1
Liquid lubricants or oils
Liquid lubricants cover mineral oils, fatty oils, synthetics, emulsions, or even process fluids. Mineral oils are most often used as the base stock in lubricant formulation. Synthetic oils (synthetic esters, phosphate esters, silicones, and fluorocarbons) are used for lubricants, which are 610
Nonfood Uses of Oils and Fats
lubricant will not flow at all. Similarly, the advantages and disadvantages of gas lubricants are like the extremes of oils, where the flow properties are almost too good.
expected to operate in extreme conditions, i.e., high performance aircraft, missiles, and in space. Vegetable oils are used in formulating lubricants intended for the food and pharmaceutical industries, but even in these applications their use is quite limited. The advantages and disadvantages of oils stem from their ability to flow easily. Thus, on the credit side, it is very easy to pour them from a container, to feed them into a bearing by dripping, splashing or pumping, and to drain them out of a machine when no longer fit for use. Other advantages are the cooling of a bearing by carrying away heat, and cleaning it by removing debris. 9.5.1.2
9.5.1.4
The gas used in gas bearings is generally air, but any gas can be used so long as it does not attack the bearings or decompose.
9.5.2
Lubricant base oils
Lubricants are made from a base oil (80 to 100%) and suitable additive package. The additives are used to enhance the most important properties for each specific application. Most base oils originate from petroleum, including many synthetic esters and poly-alpha-olefins. Less than 2% of the base oils are the product of oleochemical and related industries. The primary area of their application has been as hydraulic fluids. The various base oils used in lubricant formulations are described below.
Greases
Greases are more than very viscous lubricating oils. Technically they are oils, which contain a thickening agent to make them semisolid. Thus, grease consists of oil constrained by microscopic thickener fibres to produce a stable and colloidal structure or gel. Greases contain three basic active ingredients: a base oil, additives, and thickener. The base oil may be mineral, synthetic, or vegetable oil. For thickeners, metal soaps and clays are mostly used apart from some nonsoap thickeners, which are inorganic (silica and bentonite clays) or organic (polyurea) materials. Metal soaps are prepared by heating fats or oils in the presence of an alkali, e.g., NaOH. Fats and oils can be animal or vegetable origin, and are produced from cattle, fish, castor bean, coconut, cottonseed, etc. The reaction products are soap, glycerol, and water. Soaps are very important in the production of greases. The most commonly used soap-type greases are calcium, lithium, aluminum, and sodium. In most cases the oil plays the most important role in determining the grease performance, but in some instances the additives and the thickener can be critical. Very often additives, which are similar to those in lubricating oils, are used. The behavior of greases is very similar to that of oils, but the former are used where the advantages of easy flow are outweighed by the disadvantages. Greases do not easily leak out of a machine or container, do not migrate away, and will form an effective seal against contaminants. 9.5.1.3
Gases
9.5.2.1
Mineral/petroleum oils
Since mineral oils are the most widely used lubricating oils, they are often the standard with which other oils are compared. Mineral oils are generally oils obtained from petroleum, although they may come from similar sources, such as oil shales and tar stands. The mineral oils used for lubrication were originally distilled fractions with suitable viscosity for lubrication. But now they are obtained through various steps of refining and extraction before being blended with specialty chemicals, called additives, to enhance existing performance characteristics. Mineral oils are mainly hydrocarbons of three basic types: paraffins, naphthenes, and compounds containing aromatic systems. Finally, there is usually a small proportion (~2%), containing aromatic ring systems. In addition to these hydrocarbons, there may be small quantities of compounds containing other elements, such as oxygen, sulfur, phosphorus, or nitrogen. Mineral base oils used for lubricants are generally molecules with 20 to 50 carbon atoms. Mineral base oils continue to be economical and provide superior performance characteristics in various applications, but they present a potential hazard because they are not readily biodegradable and are environmentally toxic. During the past few decades, the level of public awareness of environmental issues has risen considerably and materials that do not meet accepted standards of biodegradability are disapproved of by environmentalists and government bodies. It is believed that federal directives in the U.S. will be strictly imposed in the next 2 to 5 years, eventually resulting in newer regulations on the development and application of environmentally friendly base stocks.
Solids or dry lubricants
The lubricants used in solid form may be bulky solids, paint-like coatings, or loose powders interposed between two surfaces in moving contact. Depending upon the nature of the two surfaces, a wide variety of solid materials can reduce friction and prevent seizure. For example, dust, sand, or gravel on the surface of a road can cause vehicles to skid because they decrease friction between tires and the road surface. The majority of solid lubricant applications are met by only three materials: graphite, molybdenum disulfide, and PTFE (polytetrafluoroethylene). The advantages and disadvantages of solid lubricants are rather like the extremes for greases, where the 611
9.5
Lubricants
9.5.2.2
different. Tocopherols have important antioxidant properties. They are a series of benzopyranols with 1, 2, or 3 methyl groups attached to the phenolic ring along with a C16 side chain on the pyran ring. Antioxidant activity as a result of tocopherol content is high in corn oil, soybean oil, walnut oil, and cottonseed oil (Rossell et al., 1991 and Section 8.1). Fully refined oils (refined, bleached, and deodorized) have free fatty acid contents of less than 0.1% (normally 0.01 to 0.05% generally expressed as oleic acid). The quality of crude oils largely depends on the content of free fatty acids and quality generally deteriorates as the acid content rises. Free fatty acids are produced by hydrolysis of oils catalyzed by acids or enzymes.
Synthetic fluids
Many of the alternatives to mineral oils are synthetic materials manufactured from various feed stocks by chemical processes. There are several types of synthetic oil, which differ from each other in performance and properties. Widely used synthetic oils include hydrocarbons, diesters, polyol esters, phosphate ester, silicones, polyglycols, polyphenyl ethers and perfluoroalkyl polyethers. Out of these fluids, low molecular weight polyalphaolefins (PAO 2, PAO 4, essentially 20:1 and 10:1 mixtures of hydrogenated dimers and trimers of αdecene), dialkyl adipates (isodecyl, isotridecyl), or polyol esters (mostly neopentyl glycol or trimethylol propane with fatty acids) are biodegradable synthetic base oils (Rudnick, 2002). The most important are PAOs, which are branched-chain paraffins, and resemble highly refined mineral oils in their structure, properties, and performance. Their inherent oxidation resistance is good, but their boundary lubrication is not as good as that of the highly refined mineral oils. Synthetic oils offer improved performance but at a price. Most of the esters are biodegradable and offer superior thermal and oxidative stability. Prices for these niche products are higher than vegetable oils and significantly higher than petroleum base stocks. Although specialized synthetic lubricants have been successfully replacing mineral oil in various applications for many years, general-purpose synthetic lubricants have only recently been introduced on a large scale. They are generally more expensive, but have better oxidation and thermal resistance than mineral oils. Low resistance to oxidative degradation and poor low temperature behavior of vegetable base oils have triggered the development and rise in demand for biodegradable synthetic base stocks. 9.5.2.3
9.5.2.3.1 Economics and availability Until recently, mineral oil had a significant cost advantage over vegetable oils and so petroleum has been the base oil of economic choice. Recent rises in oil prices along with the low vegetable oil prices has narrowed the price difference to close to $0.05/lb (0.11/kg), and there is now more interest in vegetable oils base stocks (Table 9.11). Though most lubricants used currently originate from petroleum base stocks, vegetable oils have seen a promising increase as biodegradable fluids over the last decade. Environmental concerns as well as economics and performance issues will drive the market share for these oils and government legislation may force this issue. Today, less than 2% of the base stocks are products of the oleochemical and related industries with the primary area of their application in hydraulic fluids, which have the highest need for biodegradable lubricants (Padavich et al., 1995). This is consumed at approximately 5MMT/ year in the U.S. market. Because soybean oil provides nearly 80% of the seed oils produced annually in the U.S. and is the cheapest vegetable oil in the U.S. market, its relatively low cost and dependable supply make it one of the more important sources of lubricant base oil in the U.S. Another commonly used vegetable oil in lubricant applications is rapeseed oil due to its relatively good oxidative stability
Natural oils
These include vegetable oils and animal fats. They are usually excellent boundary lubricants, but they are much less stable than mineral oils, and tend to break down to give sticky deposits. Vegetable oils are used in various industrial applications, such as emulsifiers, lubricants, plasticizers, surfactants, plastics, solvents, and resins. The natural oils are mainly triacylglycerols (98%), diacylglycerols (0.5%), free fatty acids (0.1%), sterols (0.3%), and tocopherols (0.1%). The fatty acids exist mostly as esters of glycerol. They have a carbon chain length of 12 to 24 carbon atoms. The predominant carbon chain length of the fatty acids from plants is 18 carbon atoms. They are the fully acylated derivative of glycerol. The resulting structure is abundantly present in vegetable oils, and resembles a tuning fork in shape. Vegetable oils have 6 oxygen atoms and around 60 carbon atoms per molecule, compared to an average of 30 carbon atoms in mineral base oils. The fatty acid constituents of triacylglycerol molecules may be all identical (e.g., triolein in olive oil and tripalmitin in palm oil), two different, or all
TABLE 9.11
Cost of various base oils in 2004
Base Oils
Cost ($/lba)
Cost ($/kg)
0.22–0.25 ~0.29 0.45–0.55
0.48–0.55 0.638 0.99–1.21
~1.00 ~1.25 ~0.90
~2.20 ~2.76 ~1.98
~0.25
~0.55
Vegetable oils Soybean oil Canola oil High-oleic sunflower oil Synthetic oils TMPb trioleate TMPb trioleate (high-oleic) PAO8c Mineral base oil (Group I and II) a
b c
612
Price will vary based on quantity, customer supplier relationships, and market conditions. TMP, trimethylolpropane. PAO, polyalphaolefin.
Nonfood Uses of Oils and Fats
therefore, are necessary to suppress or eliminate triacylglycerol crystallization and to improve oxidation stability. The inherent problems of poor low temperature performance and oxidation stability in vegetable oils can be partially improved by a variety of reactions at either the fatty acid carboxy groups or the hydrocarbon chain depending on end use applications. More than 90% of chemical modifications have been those occurring at the fatty acid carboxy groups, while less than 10% have involved reactions at fatty acid hydrocarbon chain. Without sacrificing favorable viscosity–temperature characteristics and lubricity, unsaturated vegetable oils can be converted into thermo-oxidatively stable products by saturation of carbon–carbon double bonds using alkylation, arylation, cyclization, hydrogenation, epoxidation, and other reactions. Chemical modifications at the carboxyl group of vegetable oils include transesterification, hydrolysis, etc. Reactions at double bond and carboxyl positions of vegetable oils are discussed by Erhan et al. (2005) and Hwang et al. (2002). With improvements in their low temperature performance and oxidation stability, they can be used in automotive and industrial lubricant applications with the additional advantage of being clean, biodegradable, nontoxic, and requiring lesser amounts of expensive additives (e.g., VI improvers are not required and lesser amounts of antiwear/antifriction additives are required).
compared to other vegetable oils, its reasonable cost, and its wide availability in Europe and North America (Whitby, 2004). Other vegetable oils used as lubricants include olive, sunflower, and castor oil. 9.5.2.3.2 Natural oil advantages Environmental concerns over the use of petroleum-based products in activities such as forestry, farming, mining, boating, and others, has led to increased interest in the use of environmentally friendly fluids. The beneficial aspects of vegetable oils as lubricants are mainly their biodegradability and nontoxicity, which are not exhibited by conventional mineral base oils (Randles et al., 1992; Battersby et al., 1989). Their volatility is low due to the high molecular weight triacylglycerol structure and they have a narrow range of viscosity change with temperature (high viscosity index, VI) and high flash point. Lower volatility results in decreased exhaust emission and high VI means the oil is a naturally multigrade oil. The high VI index of vegetable oils eliminates the need for the polymeric VI improvers used with mineral oils, resulting in high shear stability of vegetable oils. The ester linkages deliver inherent lubricity on metallic surfaces due to their adhesive property. Higher lubricity or lower friction results in more power and better fuel economy. Vegetable oils also have superior solubilizing power for contaminants and additive molecules compared to mineral base fluids. The ester structures provide improved solvency for polar deposits and sludge containing worn metals. Further, vegetable oils have higher shear stability.
9.5.2.5
Monounsaturated fatty acids are more thermally stable than polyunsaturated fats and, therefore, are highly desired components in vegetable oils. Ideally, vegetable oils having high stability and low pour points contain only monounsaturated fatty acids. Advanced plant breeding and genetic engineering has enabled the development of vegetable oils with higher concentration of oleic acid and lower linoleic and linolenic acids. The oleic content of high-oleic varieties of rapeseed and soybean oil is 75 to 85%, while that of high-oleic sunflower oil is 80 to 92% (Whitby, 2004). The oxidative stabilities of these high-oleic oils are three to six times greater than normal vegetable oils. Specialty canola oil products that Monsanto (Monsanto Company, St. Louis, MO) expects to market include oil containing medium-chain fatty acids for lubricants as well as for nutritional and high-energy food products (Schmidt et al., 2005).
9.5.2.3.3 Natural oil disadvantages Performance limitations of vegetable oil base stocks are poor oxidative stability due to bis-allylic hydrogen atoms in the fatty acyl chain, deposit-forming tendency, low temperature solidification, and low hydrolytic stability. Oxidation results in increased acidity, corrosion, and viscosity and volatility of the lubricant. The inherently narrow viscosity range limits their use in various viscosity grades, especially at lower viscosities. The polar nature of triacylglycerols contributes to air entrainment and problems of foaming. On the other hand, parameters like lubricity, antiwear protection, load carrying capacity, rust prevention, foaming, demulsibility, etc., are mostly additive-dependent. Antioxidant additives (Becker et al., 1996) have limited capability to improve on oxidative stability; therefore, other approaches are required to improve the above characteristics. The performance limitations of vegetable oil base stocks can be overcome by genetic modification, chemical modification, processing changes, and development in additive technology. 9.5.2.4
Genetically modified vegetable oils
9.5.3
Lubricant additives
Lubricant additives are chemicals, nearly always organic or organo-metallic, that are added to oils in quantities of a few weight percent to improve the lubricating capacity and durability of the oil. Specific purposes of various lubricant additives are discussed below. Wear and friction improvers are used to improve the wear and friction characteristics by adsorption and
Chemically modified vegetable oils
Low temperature testing shows that vegetable oils solidify at –20°C on long-term exposure. Poor oxidative stability of vegetable oils is due to bis-allylic hydrogen atoms in the fatty acyl chain. Chemical modifications, 613
9.5
Lubricants
oxidation. Under high shear rates, VI improvers can suffer permanent or temporary viscosity loss. Pour point depressants (PPDs) enhance lubricant characteristics by reducing the pour point through interfering with the crystallization mechanism. PPDs do not prevent wax crystallization in the oil, but they are absorbed on the wax crystals and, thus, reduce the amount of oil occluded on the crystal. Reducing the crystal volume permits lubricant flow. Typically used PPDs are maleic anhydride-styrene copolymers and polymethacrylates. These can be used in mineral as well as vegetable oils, though a higher percentage is generally required in the latter. Antifoam agents/foam inhibitors include silicones and miscellaneous organic copolymers. The most common package of additives used in oil formulations contains antiwear and extreme pressure lubrication additives, oxidation inhibitors, detergents, dispersants, viscosity improvers, pour point depressants, and foam inhibitors.
extreme pressure (EP) lubrication. This includes adsorption or boundary additives, antiwear additives, and extreme pressure additives. The adsorption or boundary additives in current use are mostly the fatty acids and the esters and amines of these fatty acids. Common examples of antiwear additives are zinc dialkyldithiophosphate, tricresyl phosphate, dilauryl phosphate, diethyl phosphate, dibutyl phosphate, tributyl phosphate, and triparacresyl phosphate. The most commonly used EP additives are dibenzyldisulfide, phosphosulfurized isobutene, and chlorinated paraffin, sulfurchlorinated sperm oil, sulfurized derivatives of fatty acids and sulfurized sperm oil, cetyl chloride, mercaptobenzothiazole, chlorinated wax, lead naphthenates, chlorinated paraffinic oils, and molybdenum disulfide. Antioxidants improve the oxidation resistance and, thus, prevent a gradual increase in the viscosity and acidity of oil. Widely used antioxidant additives are zinc dialkyldithiophosphate, metal deactivators, phenol derivatives, amines, and organic phosphates. Sulfur-based EP and antiwear additives are also quite effective as antioxidants. Sulfur and phosphorus in elemental form or incorporated into organic compounds are also effective as antioxidants and antiwear additives. To control the corrosion of metal parts, corrosion inhibitors and rust inhibitors are used. Corrosion inhibitors are used to protect the nonferrous surfaces of bearings, seals, etc., against corrosive attack by oxidation products and additives containing reactive elements, such as sulfur, phosphorus, iodine, or chlorine. The commonly used corrosion inhibitors are benzotriazole, substituted azoles, zinc diethyldithiophosphate, zinc diethyldithiocarbamate, and trialkyl phosphites. Rust inhibitors are used to protect ferrous components against corrosion. Widely used rust inhibitors are metal sulfonates (i.e., calcium, barium, etc.), amine succinates, or other polar organic acids. Contamination control additives restrict contamination by reaction products, wear particles, and other debris. The other possible contaminants are soot from inefficient fuel combustion, unburnt fuel, breakdown products of the base oil, corrosion products, dust from the atmosphere, organic debris from microbiological decomposition of the oil, etc. Without proper control of contamination, the oil will lose its lubricating capacity, become corrosive and become unsuitable for service. Additives, which prevent the development of all these detrimental effects, are known as “detergents” or “dispersants.” Mild dispersants are typically low molecular weight polymers of methylacrylate esters, long chain alcohols, or polar vinyl compounds. Over-based dispersants are calcium, barium, or zinc salts of sulfonic, phenol, or salicylic acids. Viscosity index (VI) improvers reduce excessive decrease of lubricant viscosity at high temperatures. They are usually high molecular weight polymers that are dissolved in the oil and change shape from spheroidal to linear as the temperature increases. Unfortunately these additives are easily degraded by excessive shear rates and
9.5.4
Physico-chemical and performance properties of bio-based lubricant base oils
The physical and chemical properties of vegetable oils are determined mainly by the fatty acid (FA) profile. Table 9.12 shows the FA composition of some vegetable oils that are being used as potential lubricant base oils for industrial applications and also chemical properties associated with C = C unsaturation. High unsaturation in the triacylglycerol molecule (and particularly high levels of 18:2 and 18:3) increases the rate of oxidation resulting in polymerization and an increase in viscosity (Brodnitz, 1968). On the other hand, high saturation increases the melting range of the oil (Hagemann et al., 1972). Therefore, suitable adjustment between low temperature properties and oxidative stability must be made when selecting a vegetable oil base stock for a particular industrial application. The fatty acids can be saturated or unsaturated resulting in a straight chain or bent chain configuration, respectively. With the increase of double bonds in the chain, ability to gain a close packed conformation is prevented and, therefore, the oil remains liquid. The higher the IV, the more unsaturated (the greater the number of double bonds) the oil and, therefore, the higher is the potential for the oil to polymerize through oxidation. Any attempt to increase the saturation content of the oil through hydrogenation will increase the melting range temperature of the oil as shown in Table 9.13 (Swern, 1970). An IV of less than 25 is required if the neat oil is to be used for long-term applications in unmodified diesel engines. Triacylglycerols in the range of IV 50 to 100 may result in decreased engine life and, in particular, will reduce the life of fuel pump and injector. Most vegetable oils are unsuitable for lubricant applications due to their high saturated or polyunsaturated 614
Nonfood Uses of Oils and Fats
Analytical data of vegetable oils used as lubricant base oils
TABLE 9.12
Vegetable Oil
16:0
Soybean oil High-oleic soybean oil Sunflower oil High-oleic sunflower oil Safflower oil High-oleic safflower oil High-linoleic safflower oil Rapeseedc Corn oil Cottonseed oil a
b c
11.1 6.2 6.1 3.5 6.4 4.6 6.7 3.0 10.6 18.0
Fatty Acid Compositiona (%) 18:0 18:1 18:2 4.8 3.0 5.3 4.4 2.5 2.2 2.6 1.0 2.0 2.0
53.6 3.7 66.4 10.4 73.2 13.2 75.2 14.0 59.8 38.0
18:3 6.3 1.7 – – – – – 10.0 0.9 1.0
131.0 85.9 124.8 80.8 135.2 83.6 121.2 99.1 119.9 109.1
Gas chromatography analysis (16:0 palmitic, 18:0 stearic, 18:1 oleic, 18:2 linoleic, 18:3 linolenic), AACC Method 58-18, 1993. Iodine value (mg I2/g), AOCS method, Cd 1-25, 1993. Rapeseed oil also contains other fatty acids, such as 1% of 20:0, 6% of 20:1, and 49% of 22:1.
(Annual Book of ASTM Standards, 2000). The viscosity of any fluid changes with temperature, increasing as temperature decreases, and decreasing as temperature rises. Viscosity may also change with alterations in shear stress or shear rate. To compare base oils with respect to viscosity variations with temperature, ASTM Method D2270 provides a means to calculate a VI. This is an arbitrary number used to characterize the variation of kinematic viscosity of a base oil with temperature. The calculation is based on kinematic viscosity measurements at 40 and 100°C. For oils of similar kinematic viscosity, the higher the viscosity index, the smaller the effect of temperature. The benefits of higher VI are:
TABLE 9.13 Melting point and iodine values of some vegetable oils Oil Coconut oil Palm kernel oil Palm oil Olive oil Castor oil Rapeseed oil Cotton seed oil Sunflower oil Soybean oil Linseed oil
Iodine Value 18 16–19 54 81 85 98 105 125 130 178
Tristearin (18:0) Triolein (18:1) Trilinolein (18:2) Trilinolenin (18:3)
0 86 173 261
Approx. Melting Point (°C) 25 24 35 –6 –18 –10 –1 –17 –16 –24 74 5 –11 –24
1. Higher viscosity at high temperature, which results in lower oil consumption and less wear. 2. Lower viscosity at low temperature, which for engine oil may result in better starting capability and lower fuel consumption during warm-up.
fatty acid content. The oxidation stability of polyunsaturated fatty acids can be improved significantly by converting them to saturated fats; this, however, will cause the low temperature behaviour of the material to deteriorate. Optimally, vegetable oils with high oxidation stability and low pour points are the high oleic varieties. The oxidation stability of such oils is three to six times greater than that of conventional vegetable oils. Such oils will provide both high thermo-oxidation stability and reasonably low temperature flow properties. Bio-based lubricants that are accepted as environmentally friendly must pass bench tests designed to evaluate the potential performance as lubricants in addition to biodegradability and toxicity. 9.5.4.1
24.2 83.6 21.4 80.3 17.9 77.5 14.6 16.0 26.7 41.0
IVb
Other viscosity measurements are used to check properties at low or high temperatures. The Cannon minirotary viscometer is used to measure low temperature properties (yield stress and apparent viscosity) as per ASTM standard method D4684-97. ASTM method D4624-93 is used to determine if the test oils have the appropriate high-temperature characteristics required for an engine. This test is aimed at the performance of lubricants in the bearing systems of an engine. The viscosities of vegetable oils, synthetic biodegradable lubricant base oils polyalphaolefin (PAO), trimethylolpropane (TMP) ester, and adipate, and mineral base oil (Erhan et al., 2000) are compared in Table 9.14. The mineral oil is a nonbiodegradable base oil mostly used for formulations of automotive lubricants. Except for natural antioxidants, these fluids do not have any additives. The vegetable oils have excellent VI. The linear fatty acids contribute to the high VI because the molecule is generally long.
Viscosity
The primary consideration for any base oil to be used in lubricant formulation is its viscosity. Viscosities of base oils are mostly reported as kinematic viscosities measured at 40 and 100°C, according to ASTM (American Society for Testing and Materials) standard method D445-95 615
9.5
Lubricants
process when the microcrystalline structures initially formed become macrocrystalline and rapidly change to a solid-like consistency. This results in a rapid viscosity increase leading to poor pumpability, lubrication, and rheological behavior. Wax crystallization at low temperature is also controlled by steric and geometrical constraints in these molecules. The fatty acid chains of triacylglycerol molecule have a “tuning fork” conformation and undergo molecular stacking during the cooling process. Another problem with vegetable oils is the manner in which they solidify. Mineral oil, which is a mixture of short- and longbranched chains, solidifies a little at a time under cold conditions, displaying a cloud point when the first solid appears in the liquid, and a pour point when the apparent viscosity of the solid/liquid mix is too high to allow immediate gravity flow. Vegetable oil molecules are mostly of the same size, so they solidify at very nearly the same temperature, giving vegetable oils a freezing point rather than a pour point. This can result in oil solidification after more than few days in cold winter weather. In vegetable oil triacylglycerols, the presence of double bonds in the fatty acyl chain influences the low temperature behaviour as shown in Table 9.13. The decreased melting temperatures of these compounds are a result of disorganization of the crystalline lattice by the presence of double bonds. It has been firmly established (de Jong et al., 1991; D’Souza et al., 1991) that presence of cis unsaturation, lower molecular weights, and diverse chemical structures of triacylglycerols favour lower temperatures of solidification. This demonstrates the contradiction of having both low temperature properties and the best possible oxidative stability in a given triacylglycerol molecule. PPDs are used to suppress formation of large crystals during solidification, although the mechanism of PPD action on triacylglycerol crystallization remains undisclosed (Bentz et al., 1969). PPDs, like polymethacrylate, allow inclusion of the PPD molecule’s branches into the growing crystal (Erhan, 2004). The effect of PPDs on pour points of vegetable oils (Table 9.15) shows that an amount of 0.4% by weight of PPD significantly reduces rates of solidification. Increased amount of PPD may slow down solidification, but further depression ceases quite rapidly. Low temperature and cold storage properties of vegetable oils, thus, do not respond appreciably to the PPD, as opposed to mineral oils (Asadauskas et al., 1999) and have
TABLE 9.14 Viscosities and pour points of vegetable oil, synthetic oil, and mineral base oils KVa at 40°C, cST
KV at 100°C, cST
VIb
Pour Points, °C
ASTM D445
ASTM D445
ASTM D2270
ASTM D97
31.5 31.6 39 33 31.9 40.3 38.3 36.9 255.5
7.6 7.7
227 226
7.7 9.1 8.4 8.3 19.5
223 217 203 212 87
–9 – –12 –18 –15 –18 –9 3 –24
139 191
Vegetable oils Soybean Sunflower High-oleic sunflower Canola Corn Rapeseed Olive Peanut Castor Synthetic oils Diisotridecyl adipate TMPc trioleate PAO2d PAO4 PAO6 PAO8
27 46.8 5.54 16.8 31 45.8
5.4 9.4 1.8 3.9 5.9 7.8
129 138 140
–51 –39 –65 –70 –68 –63
Mineral oil
65.6
8.4
97
–18
a b c d
Kinematic viscosity. Viscosity Index. Trimethylolpropane. Polyalphaolefin (kinematic viscosity 2 mm2/s at 100°C).
9.5.4.2
Low temperature properties
Pour point measurement is most commonly used to check the low temperature properties of the oils. Pour points of various base oils are shown in Table 9.14. Overall, the data shows that low temperature properties of vegetable oils are inferior to those of synthetic base oils or even mineral oil. In vegetable oils, castor oil demonstrates PP notably lower than those of soybean, high-oleic sunflower, and canola oils, suggesting that hydrogen bonding between the hydroxy groups of ricinoleic acid interferes with the crystal growth. It must be noted that some fluids still pour after quite significant durations at slightly lower temperatures than their determined PP. A good example is castor oil, which pours after more than 24 hours storage at –25°C, although its PP appeared as –24°C in triplicate runs. Increasing molecular weight of fatty acids and full saturation contribute to the increase in PP, whereas cis unsaturation favors the decrease. The relatively poor low temperature flow properties of vegetable oils arise from the appearance of waxy crystals that rapidly agglomerate resulting in the solidification of the oil. A vegetable oil is a complex mixture; therefore, the transition from liquid to solid state occurs over a wide temperature range involving several polymorphic forms (α, β′, β) (Hagemann et al., 1983; Hagemann, 1988). Wax appearance and crystallization is a slow continuous
TABLE 9.15 Effect of PPD (polyalkylmethacrylate copolymer of 8000 amu, canola oil carrier 1:1.) on pour points (°C) of vegetable oils Amount of PPD % (w/w) Vegetable Oil Soybean High-oleic sunflower Canola
616
0
0.4
1
2
–9 –12 –18
–18 –21 –30
–18 –24 –33
–18 –24 –33
Nonfood Uses of Oils and Fats
loss in optical clarity. Therefore, to meet the viscometric properties of vegetable oil-based lubricants for engine applications, PPDs, synthetic hydrocarbons, and synthetic fluids have to be used in various combinations to produce base fluids meeting SAE 30, 5W-30, and 10W-30 viscosity requirements as shown in Table 9.17. These requirements can be met using a proper selection and combination of various vegetable oils along with PPD and synthetic oil diluent. All of the blends shown in the table contain antioxidants and antiwear additives and are compared with commercially available 5W-30 oil. Similar lubricants can be prepared using combinations of soybean, castor, canola, high-oleic oils, esters, and synthetic hydrocarbons.
shown unsatisfactory performance when exposed to low temperatures for longer durations (Antila et al., 1966). Studies show that diluents have a significant role in lowering the pour point of the vegetable oils (Asadauskas et al., 1999). However, high dilution does not necessarily translate to proportionate depression of pour point and no synergism exists between diluents and PPD molecules. During the cooling process, the response to diluents and PPD molecules is dependent to some extent on the vegetable oil FA composition and its geometry. Pour point determinations of safflower, high-oleic safflower, and high-linoleic safflower in the presence of diluent and PPD are shown in Table 9.16. The addition of a synthetic ester as a diluent to safflower and high-linoleic safflower oils showed a larger decrease in the pour point compared to the high-oleic oil. Due to the presence of multiple unsaturation in safflower and high-linoleic safflower oil, the triacylglycerol molecules encounter significant steric-hindrance from the “zigzag” nature of the FA chain during the cooling process. The presence of diluent molecules in the system enhances this effect by lowering the viscosity and by interfering with the stacking process during cooling. The addition of PPDs further lower the pour point. High-oleic and high-linoleic oils appear to show a better response in the presence of additive molecules. In addition to exhibiting good low-temperature behavior, base oils should be stable over extended time at low temperature to qualify for any industrial and automotive applications. Although high-oleic oils exhibit good thermaloxidative behaviour and acceptable PPD response, they fail in an industry-specified, low-temperature extended storage stability test. Table 9.16 shows the cold storage stability data of selected vegetable oils in the presence of diluents and PPD (Erhan et al., 2002). Using the optimized diluent and PPD concentration, safflower and high-linoleic safflower oils showed acceptable fluidity well beyond 7 days with some TABLE 9.16
9.5.4.3
Oxidation stability
Oxidation is the single most important reaction of oils resulting in increased acidity, corrosion, viscosity, and volatility when used as lubricant base oils for engine oils. A number of tests are used to evaluate the oxidation stability of lubricants (Booser, 1997). The more common choices for synthetic and vegetable oil fluids are the TOST test (dry) ASTM D943, Rotary Bomb Oxidation Test (RBOT) ASTM D2272, Modified Thin Film Oxygen Uptake Test (TFOUT), Pressurized Differential Scanning Calorimetry (PDSC) ASTM method D6186-98, and Penn State thinfilm Microoxidation (TFMO) test (Cvitkovic et al., 1979). Several benchtop oxidation tests are available as screening tools for oxidative stability of vegetable oils. Evaluation of oxidation is complex and a fully acceptable protocol has yet to emerge. Estimation of peroxide value (PV) can be used as an index of oxidation if the peroxides formed are stable and do not decompose after formation, which in most cases is not true. The activation energy for the formation of peroxide is 146-272 kJ/mol (Labuza, 1971) and that of decomposition of lipid peroxide is 84184.5 kJ/mol, suggesting peroxides are less stable than
Response of diluents and PPD on pour point and cold storage stability of vegetable oils Pour Point in °Ca
Fluid
Oil + Diluentb
Diluent + PPDc
Oil + Diluentb
Diluent + PPDc
–21 –21 –21
–39 –39 –27
–48 –45 –36
7+ 7+ 1
7+ 7+ 1
High-linoleic safflower oil Safflower oil High-oleic safflower oil a b c
ASTM D97. Oil:diluent ratio of 65:35 (vol/vol). Pour point depressants (PPDs) concentration of 1%.
TABLE 9.17
SAE requirements and properties of typical environmentally friendly lubricants
Lubricant Oil SAE requirements Rapeseed + castor High-oleic sunflower oil + PPD Corn oil + synthetic ester + PPD Commercial lubricant a
Number of Days at –25°C
Neat Oil
SAE Viscosity Grade
KV at 100°C
Pumping Viscositya (cP)
30 wt 10W-30 5W-30 5W-30
9.8–12.5 cSt 9.72 10.10 10.34 10.42
60,000 (max) Not required 4651 26,287 25,664
Measured using Cannon minirotary viscometer (MRV) as per ASTM D4684-97 method.
617
9.5
Lubricants
lipids (Swern, 1970). In the active oxygen method (AOM) (AOCS Official Method Cd-12-57, 1983), test oil is heated to 100°C and the oxidation is followed by measuring the PV of heated sample at regular time intervals until PV = 100 meq/kg is reached, which gives the AOM endpoint. A large amount of sample, numerous analysis, and critical control of airflow is required. With samples that form unstable peroxides, a PV = 100 meq/kg may never be reached and such measurements have no meaning. In the AOM method, consumption of O2 may also be a measure for induction period. The Rancimat method (Oxidationsstabilität, 1994; Laubli et al., 1988) is based on the fact that the volatile acids formed during oxidation (Loury, 1972; De Man et al., 1987) can be used for automated endpoint detection. Gordon and Mursi (1994) have shown good correlation of Rancimat results at 100°C with oil stability as measured by peroxide development during storage at 20°C. In another study, Jebe et al. (1993) pointed out the advantages of the Rancimat method at higher temperature. In the Sylvester test (Wewala, 1997), the sample is heated to 100°C in a closed vessel and pressure decrease due to O2 consumption is monitored. The Oxidograph (Wewala, 1997) is an automated version of this method and the induction period is determined from the sudden decrease in the O2 pressure. Oxidative status of oil can also be obtained by integrating the light curve during a chemiluminescence reaction (Matthäus et al., 1994). The method is highly sensitive for the measurement of lipid oxidation. Matthäus et al. (1993) described a linear correlation (R2 = 0.99) between the iodimetric peroxide determination (DGF Einheitsmethoden; 1984) and the chemiluminescence method. Another official method to measure induction period is the oil stability index (OSI) (AOCS Method Cd-12b-92, 1993). OSI values generally correspond well with AOM values if PV is 100 meq/kg or greater (Laubli et al., 1986). The method is automated and much easier compared to AOM. However, lengthy experimental time, large errors associated with small changes in O2/air flow rate (Hill et al., 1995) and inability to differentiate between small changes in vegetable oil matrix are major disadvantages. PDSC is also popular for the determination of oxidative stabilities of vegetable oils (Shankwalkar et al., 1993; Kowalski, 1989; Kowalski, 1991). The TFMO method (Lee et al., 1993) is often the method of choice for studying vegetable oils because it is simple and reproducible. The test is especially effective when thermally induced volatility is low and insoluble deposit formation through polymerization is to be considered rather than rates of inhibitor depletion. In vegetable oils, unsaturation, due to C = C from oleic, linoleic, and linolenic acid moieties, provides active sites for various oxidation reactions. Saturated FAs have relatively high oxidation stability (Brodnitz, 1968), but this decreases with increasing unsaturation in the molecule. The rate of oxidation depends on the degree of unsaturation of a fatty acyl chain. In general, the rate of oxidation
TABLE 9.18
Oxidation rates of simple triacylglycerols
Triacylglycerols Tristearin (18:0) Triolein (18:1) Trilinolein 18:2) Trilinolenin (18:3)
Oxidation Low Moderate High Very high
Relative Rate 1 10 100 200
of linoleic (18:2) is 10 (or more) times greater than oleic (18:1), while linolenic (18:3) is twice as great as the linoleic fatty acyl chain (Table 9.18). Oxidation usually takes place through a radical initiated chain mechanism (Murray et al., 1982). Initiation RH ÆR• R• + O2 Æ RO2• Propagation RO2• + RH Æ RO2H + R• R• + O2 Æ RO2• Branching RO2H Æ RO• + •OH RO• + RH + O2 Æ ROH + RO2• •OH + RH + O Æ H O + RO • 2 2 2 Inhibition In H + RO2•Æ In• + RO2H Peroxide decomposition RO2H Æ RO• + •OH + inert products The free radicals generated during the initiation stage react with O2 to form peroxy free radicals and hydroperoxides (Privett et al., 1962). During this period, O2 is consumed in a zero-order process (Labuza et al., 1983), apparently leading to intermediates that are not well characterized, prior to the formation of peroxides (Privett et al., 1962). The latter undergoes further reaction to form alcohols, ketones, aldehydes, carboxylic acids (Shahidi, 1997), leading to rancidity and toxicity (Grosch, 1979). The formation of polar functionalities further accelerates the oil degradation process (Steinberg et al., 1989; Harman, 1982). These compounds have molecular weights that are similar to vegetable oils and, therefore, remain in solution. As the oxidation proceeds, the oxygenated compounds polymerize to form viscous material that, at a particular point, becomes oil insoluble leading to oil thickening and deposits. The extent of oxidation and formation of oxidation products are further complicated by the amount of unsaturation, structural differences in the various triacylglycerol molecules, and the presence of antioxidants. All these factors, together or individually, can change the specific compounds formed and the rates of their formation (Coates et al., 1986). In addition to unsaturation in the molecule, oxidative degradation and kinetics of oxidation is influenced by methylene chain length, bisallylic methylene groups, etc. The cumulative effect of various structural parameters in the triacylglycerol molecule makes oxidation a highly complex process and no simple kinetic model alone would hold good for such systems. Oxidative stabilities of various base oils using the Penn State TFMO test are compared in Table 9.19. In this test, a thin film of oil is oxidized in air high temperatures. The 618
Nonfood Uses of Oils and Fats
interaction of the sample with the reactant gas (oxygen). A film thickness of less than 1 mm is required to ensure proper oil–O2 interaction and eliminate any discrepancy in the result due to oxygen diffusion limitations (Kowalski, 1993; Adhvaryu et al., 1999). Oxygen gas (dry, 99% pure, obtained commercially) is pressurized in the module at a constant pressure of 3450 kPa and maintained throughout the length of the experiment. The sample is then heated at 10°C/min to 250°C. From the DSC thermogram, the onset temperature (OT) is determined. This represents the temperature when rapid increase in the rate of oxidation is observed in the system. This temperature is obtained from extrapolating the tangent drawn on the steepest slope of reaction exotherm. A high OT would suggest a high oxidative stability of the vegetable oil. The OT for various vegetable oils is shown in Table 9.20. The OT is influenced by the degree of poly-unsaturation present in the vegetable oils. It is generally observed that a high polyunsaturation (linoleic and linolenic acid content) decreases, while high oleic content in the FA chain increases the OT. The increase in saturated fatty acids improves the resistance to initial thermal breakdown. The activation energy requirement for such system is considerably high. This results in delaying the onset of initial oxidation process where bond scission takes place to form primary oxidation products. The percentage of oleic acid (see Table 9.12) in the different vegetable oils explains the observed trends in oxidation stability. However, the polyunsaturated and saturated fatty acid contents do not conclusively explain the relative variation of OT among the vegetable oils. The role of different structural parameters obtained using 1H and 13C NMR on the oxidation behavior of unmodified and genetically modified vegetable oils has been explained elsewhere (Adhvaryu et al., 2000). Improvements in oxidation stability are needed due to more stringent demands being placed on lubricant performance. Use of antioxidant (AO) additives along with higholeic vegetable oils improves the oxidation stability. The AO package has to be optimized for vegetable oils. Typically a mixture of AO is required. Table 9.21 shows the effects of the AO package (commercial LZ7652) optimized for vegetable oils when evaluated in the ASTM D2272 RBOT and ASTM D943 TOST methods (Lawate, 2002; Rudnick, 2002). Also in these studies, it was found that high-oleic soybean oil is oxidatively more stable than conventional soybean oil, but less than mineral oil. In TFMO
TABLE 9.19 Oxidative degradation tendencies of various base oils using TFMO (30 min at 150°C) Base Oils Soybean oil High-oleic sunflower oil Diisotridecyl adipate PAO4a Mineral oil a
Deposits (%)
Evaporation (%)
48 13 3 6 5
2 0 5 45 5
Polyalphaolefin with viscosity of ~4 cSt at 100°C.
losses due to evaporation and oxidation and the depositforming tendencies (oxypolymerization) of the test sample are determined in the test as shown in Table 9.19. A temperature of 150°C and times of 30 to 60 minutes chosen for testing were high enough to cause a quantifiable polymerization in unsaturation-free base stocks, yet not too severe to result in oxidative gelation of vegetable oils. Therefore, the side processes, such as oxidative cleavage and formation of solids, were not too substantial. It appears from the data that vegetable oils oxypolymerize considerably faster than unsaturation-free fluids. Although high-oleic sunflower oil containing only 5% of linoleic acid shows higher resistance to oxypolymerization than soybean oil, its oxidative stability is still far less than those of PAO or adipate. Oxypolymerization proceeds much faster and slows down only when side processes, especially formation of solids, become more pronounced. It has been established that methylene-interrupted polyunsaturation is the key factor causing low oxidative stability of vegetable oils (Gardner, 1989). At higher temperatures, such as 175 and 200°C, evaporation was substantially higher in oils having high oleic content. Conversely, increase in polyunsaturation resulted in low evaporation of vegetable oil. A similar trend was observed in safflower oil: higholeic, and high-linoleic. The evaporative loss was greatest in high-oleic oil at 175 and 200°C. As the polyunsaturated FA content increased (as in high-linoleic safflower oil), the percent evaporation decreased at 175°C and remained comparable to safflower oil at 200°C. The deposit-forming tendency is the inverse of the evaporation trend with least deposit in high-oleic oils and more deposit in more polyunsaturated high-linoleic oils. The presence of polyunsaturation in the FA is the primary reason of low oxidative stability, as divinyl methylene hydrogen atoms are highly susceptible to free radical attack leading to substitution with O2 molecule and consequent formation of polymeric oxy-polar compounds. These compounds are the precursors of oil insoluble deposits often encountered with high temperature oxidation of vegetable oils. PDSC is another popular approach for rapid measurement of the oxidative stability of vegetable oils (Kowalski, 1993; Shankwalkar et al., 1993). The procedure is fast, requires only a small quantity of sample, and is extremely reproducible. A small amount of sample is placed in a hermetically sealed aluminum pan with a pinhole lid for
TABLE 9.20 Oxidation stability of vegetable oils using pressurized differential scanning calorimetry Vegetable Oil Cottonseed Safflower High-oleic safflower High-linoleic safflower Sunflower High-oleic sunflower
619
OT (°C) 150 166 178 166 145 177
9.5
Lubricants
9.5.4.6
TABLE 9.21 Oxidation stability of vegetable oils and synthetic ester using the RBOT method RBOT (min) Base Oils
There are two types of biodegradability tests: primary and ultimate biodegradation (Battersby, 2000; Product Review, 1996). Primary biodegradation involves the disappearance of the parent organic compounds under specific test conditions and may or may not indicate that the substrate will biodegrade completely. The primary biodegradation test method is CEC L-33-A-93 from the Coordinating European Council (CEC). This method measures the degree of degradation by the disappearance of specific hydrocarbons bands using IR (infrared) spectroscopy. Since the method does not identify the type of products produced, conclusions regarding the extent of degradation are limited. As regulations are tightened in Europe, the method may not be acceptable for certification of the fluid as environmentally friendly. Total or ultimate biodegradation is measured by tests that result in the complete degradation to carbon dioxide and water by microbial action within 28 days. The two most widely used tests are ASTM standard method D5846 (Annual Book of ASTM Standards, 2000) and the Organization of Economic Cooperation and Development (OECD) test method OECD 301B. The OECD method is accepted worldwide and is the basis for the German chemical laws, dangerous substances legislation (UBA WGK Water hazard), and Eco-labeling (Blue Angel) (Laemmle, 2002). The choice of base oil used is critical for environmentally friendly (EF) lubricants since it dictates the biodegradability and ecotoxicity of the finished lubricant. The biodegradability of different base oils is shown in Table 9.22. The use of PAO and mineral oils in EF lubricants is thus limited due to low biodegradability. Among the base oils listed, vegetable oils possess the highest biodegradability and lowest ecotoxicity.
a
No Antioxidant
3.0 % Antioxidants
15 15 15
79 169 232
15
280
15
170
TOST, hb
Vegetable oil Canola High-oleic canola High-oleic sunflower High-oleic soybean
< 100 ~ 350 ~ 500
Synthetic oil TMP trioleate Mineral oil (200 N) a b
350
~ 2000
Min to 25 lb pressure loss, ASTM D2272. Time to TAN = 2, ASTM D943.
tests the deposit-forming tendencies and oxidative volatility of vegetable oils are significantly reduced by the addition of AO additives (Perez et al., 2002). The proper combinations of high-oleic vegetable oils and additives have oxidative stability at par with off-the-shelf 10W-30 commercial mineral oil-based lubricant. 9.5.4.4
Friction-wear properties
Tests to evaluate the friction and wear characteristics of lubricants are numerous and range from bench tests to engine and pump stand tests (Annual Book of ASTM Standards, 2000; Booser, 1997). Pump stand tests include vane- (Dennison T-50, Vickers V104C) and piston-type tests (Dennison P-46). The DIN 51354 FZG test is popular both in Europe and the U.S. There are four different ASTM bench tests (Annual Book of ASTM Standards, 2000) used to evaluate wear and load-carrying ability: ASTM standard method D2266 (Four-Ball Wear), D2783 (Four-Ball EP), D2782 (Timken test), and D3233 (Falex EP). Each lab uses its own modification of the four-ball test, such as a sequential test correlated with the pump stand tests (Perez et al., 1986). In addition, the fluids are subjected to brake, clutch, and friction disc tests (Booser, 1997). Friction and wear performance of the vegetable oil based lubricants using a commercial additive package are acceptable for industrial and automotive lubricants. 9.5.4.5
Biodegradability
9.5.4.7
Toxicity
Aquatic toxicity tests measure the extent to which a fluid will poison selected environmental species, such as algae (OECD 201), Daphnia magna (OECD 202-12), flathead TABLE 9.22
Biodegradability of various base oils % Biodegradability by
Base Oils
Hydrolytic Stability
CEC Method
OECD Method
Vegetable oils
A serious threat to the stability of vegetable oils and synthetic esters is the presence of water. The reaction can yield organic acids that further catalyze the reaction resulting in further degradation of the base fluid and corrosive wear of metal surfaces. Additives can be used to improve hydrolytic stability (Booser, 1994), but good maintenance practices are the best insurance for long life. The stability of the fluids can be evaluated using the “coke bottle” test, ASTM standard method D2619 (Annual Book of ASTM Standards, 2000).
Soybean Canola High-oleic sunflower
100 100 100
>70 >70 >70
90 90 60 >60 30
15
5
Synthetic oils TMP trioleate TMP trioleate (high-oleic) PAO8 Mineral oil (150N)
620
Nonfood Uses of Oils and Fats
alternative oilseeds in the U.S. Sunflower oil is considered a premium oil due to its light colour, mild flavour, low level of saturated fats and ability to withstand high cooking temperatures. Although the sunflower has the potential for many industrial uses, in the U.S., it is mostly used for food or feed purposes.
minnows Oncorhyuchus mykis (OECD 203-13), and bacteria (OECD 209).
9.5.5
Suitability of natural oils as lubricants
The following vegetable oils are most suitable for use in lubricant applications. More details about these oils can be found in Chapter 2. 9.5.5.1
9.5.5.4
There are two types of safflower varieties: one that produces oil that is high in monounsaturated fatty acids (oleic acid), and the other with high concentrations of polyunsaturated fatty acids (linoleic acid). The high-linoleic safflower oil contains nearly 75% linoleic acid, which is considerably higher than corn, soybean, cottonseed, peanut, or olive oils. Higholeic safflower variety may contain up to 80% oleic acid and is comparable to olive oil, and stable when heated. Regular safflower oil is considered as a drying or semidrying oil, with properties intermediate between soybean and linseed oils, and is currently used in manufacturing paints and other surface coatings. The oil is light in color and will not yellow with aging. This oil can also be used as lubricant base oil, but like most vegetable oils, is currently too expensive for this use. High-oleic safflower oil is rapidly gaining recognition as one of nature’s most valuable vegetable oils with extensive industrial applications.
Soybean oil
Soy oil, extracted through pressing or via solvent extraction, is used for a number of industrial applications. Bioengineered (high oleic and/or low linoleic) soybeans may provide highly desirable improvements for fuels and other industrial products. Current research priorities are focused on developing improved soy oil-based lubricants with improved oxidation stability and cold weather pour properties. Developments in the area of total loss lubricants, hydraulic fluids, and crankcase lubricants are showing promising results. Products, such as biodegradable grease and 2-cycle engine oils, based on soybean oil are currently available. These are environmentally friendly when used in outboard motors, lawnmowers and other small engines. Soybean oil used in crankcases must exhibit properties such as high lubricity, viscosity index, flash point and low evaporation loss. 9.5.5.2
9.5.5.5
Canola oil
Sunflower oil
Traditional sunflower oil consists of 68% linoleic acid and about 20% oleic acid. Two varieties of the oil are currently available: normal and high-oleic. The high-oleic variety has higher oxidation stability and is suitable for industrial applications. Both canola and sunflowers are popular
9.5.5.6
TABLE 9.23 Fatty acid distribution and physical properties of natural and high-oleic canola oil Fatty Acid Distribution/ Technical Properties Palmitic acid (16:0) Stearic acid (18:0) Oleic acid (18:1) Linoleic acid (18:2) Linolenic acid (18:3) Melting temperature (ºC) PDSC (minutes) at 130 °C
High-Oleic Canola Oil
Canola Oil
3.5 1.5 84.0 3.5 4.0 –5.5 71
3.5 1.5 61.5 19.5 10.5 –9.5 25
Meadowfoam oil
Meadowfoam oil contains three previously unknown longchain fatty acids and resembles high erucic acid rapeseed oil in some respects (Bosisio, 1989). It is unusually high in long-chain fatty acids (over 90% C20 to C22 fatty acids) with very high levels of monounsaturation and very low levels of polyunsaturation. These characteristics make meadowfoam oil very stable, even when heated or exposed to air and crude meadowfoam oil is more oxidatively stable than other regular vegetable oils. Meadowfoam oil has the added benefit of enhancing the properties of other oils when mixed with them. Less expensive oils can be mixed with meadowfoam oil without the loss of the qualities of either oil and it can increase the stability of the oils to which it is added. The oil can be used as a lubricant, apart from other applications like light coloured premium grade solid wax, a sulfur compound valuable to the rubber industry, or used as a detergent or plasticizer.
Canola oil is typically referred to in the industry as a penetrating oil and generally has a higher level of linolenic fatty acid than soybean oil. High α-tocopherol content (19 mg/100gm), higher levels of oleic acid and lower levels of polyunsaturated acids contribute to the oil stability as compared to soybean oil. With the rapid development of high-oleic variety oils, various technical properties of the oil can be significantly improved, thereby meeting industrial specifications. Table 9.23 presents the comparison of natural and high oleic canola oil. 9.5.5.3
Safflower oil
Lesquerella oil
Lesquerella oil contains a hydroxy fatty acid and so resembles castor oil, which is an important raw material used by industry for making lubricating greases, resins, waxes, nylons, plastics, corrosion inhibitors, coatings, and cosmetics (Smith Jr. et al., 1961, 1962). Saturated hydroxy fatty acids produced by hydrogenation could be useful in the production of greases in the form of their lithium soaps. Many of the properties may be enhanced over those of castor oil because of the increased chain length of this new crop oil.
621
9.5
Lubricants
9.5.5.7
9.5.6
Cuphea oil
Cuphea oil contains high levels of short-chain saturated fatty acids, C8, C10, C12, and C14, which are used in the production of solvents, detergents, and emulsifiers. Other uses of the seed oil include cosmetics and motor oil. 9.5.5.8
Lubricants provide a well-established and highly competitive market, but growing only at an average rate of less than 1% per year (Padavich et al., 1995). More than 70% of total lubricant volume is used as motor oils for automotive engines and approximately 10% as hydraulic fluids. Other small areas of usage include: cutting oils, two stroke engine oils, chainsaw bar oils, wire rope oils, bicycle chain oils, railroad oils, pump oils, outboard engine oils, drilling oils, and other niche markets. Nonfood uses of vegetable oils have grown very little during the past 40 years (with the exception of the burgeoning demand for biodiesel). Although some markets have expanded or new ones added, other markets have been lost to competitive petroleum products. Vegetable oils are currently being used in various industrial applications, such as emulsifiers, lubricants, plasticizers, surfactants, plastics, solvents, and resins. Research and development approaches take advantage of the natural properties of these oils for lubricant applications, namely amphiphilic character to deliver inherent lubricity, high molecular weight, low volatility, high viscosity index, good solubilizing power for additive molecules and being eco-friendly (Randles et al., 1992; Battersby et al., 1989) among others. Vegetable oils and other lipid derivatives have shown significant increase in use as biodegradable lubricants over the last decade, but still contributing less than 2% of all base oils used in the market. A major application area is industrial hydraulic fluids, which represents a 222 million gallon market in the U.S., with potential use in waterways, farms, and forests. Vegetable base oils are suitable as metal cutting oils and fluids, and avoid the hazardous mist formation from mineral oils during use. Canola-based motor oils have rapidly evolved into a competitive product as a potential substitute for mineral oil-based products. In terms of pricing, they are highly competitive with synthetic motor oils. They are also the most “environmentally friendly” of the motor oils available maintaining properties of nontoxicity and biodegradability. In terms of functionality, they have exceeded expectations by surpassing both conventional and synthetic oils in the tests conducted. There has been significant reduction in tailpipe gas emissions of nitrogen oxides, carbon monoxide, and hydrocarbons, therefore providing an easy and effective way to reduce air pollution. As crankcase oil, though vegetable oil-based lubricants have limited contact with the environment, active development work is in progress on base stocks (e.g., canola, corn, soybean oil) for use in air-cooled engines. Other significant niche market areas are cutting and drive chain oils, two stroke engine oils, chain saw bar oil, wire rope oil, marine oils and outboard engine lubricants, oil for water and underground pumps, rail flange lubricants, agricultural equipment lubricants, metal cutting oils, tractor oils, dedusting, and several others.
Jojoba oil
Perhaps the most commercially advanced of the new crops is jojoba. Jojoba “oil” is a liquid ester wax rather than the familiar triacylglycerols produced in well-known oil seeds like canola. The principal oil structures contain 40 and 42 carbon atoms. The major market for jojoba oil continues to be within the cosmetic industry. An estimated 2000 metric tonnes per annum is consumed by this industry in the form of jojoba oil, hydrogenated jojoba oil, jojoba esters, hydrolyzed jojoba oil, ethoxylated jojoba oil, and other value-added jojoba oil derivatives. In lubricant applications, jojoba oil provides a market for approximately 100 tonnes annually. In general the price of jojoba is too high for this market compared to other available lubricant oils. The molecular structure of this oil is such that it is stable even at high temperatures and pressures unlike most of other lubricants. 9.5.5.9
Tallow and yellow grease
Tallow is inedible grease derived from animal fat renderings. Products such as lubricants, soap, cosmetics, and plastics are made from tallow. Tallow is gradually being replaced by yellow grease (waste vegetable oils) from used cooking oil from fast food restaurants. It contains a mix of unsaturated and saturated fatty acids (generally in a ratio of about 2.8:1). Other application areas of yellow grease are the manufacture of soap, makeup, clothing, rubber, and detergents. 9.5.5.10
Applications
Medium-chain triglycerides
Medium-chain triglycerides (MCT) are medium-chain fatty acid esters of glycerol. Medium-chain fatty acids are fatty acids containing from 6 to 12 carbon atoms. These fatty acids are constituents of coconut and palm kernel oils and are also found in camphor tree drupes. Coconut and palm kernel oils are called lauric oils because of their high content of the lauric acid (C12). MCT used for nutritional and other commercial purposes are derived from lauric oils. They differ from other fats in that they have a slightly lower calorie content (Bach et al., 1996) and they are more rapidly absorbed and burned as energy, resembling carbohydrate more than fat (Bach et al., 1982). With mineral oil falling out of favour in use as a processing lubricant and as a mould release and polishing agent in hard candy production, MCTs may be an option. Many European countries are banning mineral oil usage in food applications, a trend that may spread to the U.S. A recent development is an increase in use of castor oil and palm oil in manufacture of biodegradable- and food industry-grade lubricants.
622
Nonfood Uses of Oils and Fats
challenges that must be overcome in order to improve their usefulness as a sustainable alternative to petroleum base oils. These challenges are of such a scope that it is unlikely that one laboratory, or even one country's scientific community, will easily overcome them alone. New applications of vegetable oil-based lubricants are constantly gaining predominance in areas where stiff regulations require expensive clean up and disposal, and environmental safety plays a major role, despite the higher costs involved. One such class of lubricants is total loss lubricants. The applications where lubricants are lost directly to the environment (railroad rails and switches, wire cables on cranes, the bars of chain saws, and other power equipment) are the most likely application areas to initiate the use of biodegradable lubricants. In these limited-life applications, the stability of the lubricant is not a factor, giving vegetable oil an advantage as the base oil. Widespread use of vegetable oils will depend on how well they perform with a wide range of factors, including temperature, pressure, metal surface, other functional fluids and existing technology. Chemically and genetically modified derivatives of vegetable oils have resulted in significant improvement in thermo-oxidative, low temperature stability, and lubrication properties, thereby increasing their use in a variety of industrial lubricant applications. Development of genetically modified vegetable oil is a lengthy process and it is currently not extensively available as an all-purpose less expensive material capable of delivering all performance qualities for nonfood uses. Nevertheless, it is evident that over the last decade there has been a significant progress in the research and development of lipid technology for innovative industrial uses. In spite of all these technological advancements, an appropriate strategy is needed to convince users of lubricants. This drive toward the use of renewable resources was given a major boost in the U.S. government with the Federal Executive Order 13010 (which set a goal that 25% of all government purchases be biobased), which has been encouraging companies to make use of renewable, bio-degradable base stocks, rather than petroleum base stocks, in many applications (Patin et al., 2002). Achieving the U.S. federal government’s goal of tripling use of bio-based products and bioenergy by 2010 could create $15 to 20 billion in new income for farmers and rural America and reduce fossil fuel emissions by an amount of up to 100 million metric tonnes of carbon (Soya and Oilseed Industry News, 2003). Another opportunity to accelerate the development of this market exists in the 2002 Farm Bill (Farm Security and Rural Investment Act, published January 11, Federal Register). Section 9002 includes language directing all federal agencies to give preference to bio-based products (lubricants is one of the categories specified in the guidelines), unless it is unreasonable to do so, based on price, availability, or performance.
A recent product is a soy-based transformer dielectric fluid. This product is used to insulate and cool electrical distribution products, such as transformers, and is safe for the environment and the public. The fluid is biodegradable based on testing done by the U.S. Environmental Protection Agency (USEPA). The fluid also enhances the performance and life of a utility's transformer assets. The fluid extends paper insulation life five to eight times, lowering life cycle costs. The increased insulation life also translates to extended and enhanced transformer life or the ability to carry higher loads during peak demand periods without leading to premature insulation failure. The enhanced performance allows utilities to manage their assets more profitably and forestall costly capital expenditures. The fluid also has excellent fire resistant qualities. The fluid offers an ignition fire point of 360°C and flash point of 330°C, more than twice that of petroleum-based mineral oils. The soy-based fluid has been shown to enhance the loading performance of new transformers by up to 14% or extend their insulation life five to eight times. It also has a similar positive impact on larger units already in service, such as those found in electric substations. Although the cost of the fluid is slightly higher than mineral oils, its soy-based properties contribute to long-term savings that mineral oil cannot deliver. Accelerated aging tests have shown that this fluid extends transformer life well beyond that of units with mineral oil.
9.5.7
Future prospects of bio-based lubricants
Bio-based lubricants are becoming increasingly important in Europe, particularly for total loss lubricants. In Western Europe, bio-lubricants were 3.1% of total lubricant consumption in 2002 (Whitby, 2004). Their interest in biodegradable lubricants started in the mid 1980s. While the strongest environmental pressures and the resulting acceptance of biodegradable lubricants began in Northern Europe, the same pressures now appear in other European Countries. The first country in Europe to introduce bio-lubricants was Sweden, in 1988, when hydraulic fluids based on rapeseed oil were introduced. The City of Gothenburg published the first “Clean Lubrication” list in 1995, comprising 14 products from 10 lubricant suppliers. The list has been further expanded every year since then. Swedish standard SS 15 54 34 was revised in 1997 to include the environmental criteria in the “Clean Lubrication” list. By 2000, around 25% of all hydraulic fluids sold in Sweden were environmentally friendly, of which 80% were synthetic esters (Whitby, 2004). Almost all forestry operations now use environmentally friendly lubricants, including greases, gear oils, and chain bar oils, because this is now mandated for all forestry operators in Sweden. Biobased lubricants have the potential to create new market opportunities for farmers while easing society’s reliance on petroleum. Vegetable oils, however, offer some 623
9.5
Lubricants
The supply side of the market along with environmental pressure groups, agricultural companies and research institutes are more enthusiastic than the demand side of the market, mainly as a result of higher costs of bio-based lubricants. Users of lubricants are reluctant to use a more expensive product unless there are compelling economic or regulatory pressures to do so. Future outlook for vegetable oil-based lubricants, therefore, will be dictated by consumer awareness and regulations that mandate use of bio-based lubricants. The most encouraging development is OEM interest in these types of fluids, which provides an optimistic future to biobased lubricants.
Bosisio, M., Meadowfoam: pretty flowers pretty possibilities oilseed soaps limnanthes alba nutlets, Agric. Res., USDA Res. Serv., 1989, 37(2), 10. Brodnitz, M.H., Autoxidation of saturated fatty acids. a review, J. Agric. Food Chem., 16, 994, 1968. Coates, J.P. and Setti, L.C., Infrared spectroscopic methods for the study of lubricant oxidation products, ASLE Trans., 29, 394, 1986. Cvitkovic, E. et al. A thin film test for measurement of the oxidation and evaporation of ester type l ubricants, ASLE Trans., 22, 395, 1979. D’Souza, V. et al. Polymorphic behavior of high-melting glycerides from hydrogenated canola oil, J. Am. Oil Chem. Soc., 68, 907, 1991. de Jong, S. et al. Crystal structures and melting points of unsaturated triacylglycerols in the β-phase, J. Am. Oil Chem. Soc., 68, 371, 1991. de Man, J.M. et al. Formation of short chain volatile organic acids in the automated aom method, J. Am. Oil Chem. Soc., 64, 993, 1987. DGF Einheitsmethoden zur Untersuchung von Fetten, Fettprodukten, Tensiden un verwandten Stoffen, Deutsche Gesellchaft für Fettwissenschaft e.V., Munster, Wissenschaftliche Verlagsgesellschaft GmbH Stuttgart, Method C-IV 6a, 1984. Erhan, S.Z. and Adhvaryu, A., Vegetable-Based Base Stocks, in Biobased Industrial Fluids and Lubricants, Erhan, S.Z. and Perez, J.M., Eds., AOCS Press, Champaign, IL, 2002, Chap.1. Erhan, S.Z. and Asadauskas, S., Lubricant basestocks from vegetable oils, Ind. Crops Prod., 11, 277, 2000. Erhan, S.Z. et al. Chemically Functionalized Vegetable Oils, in Synthetics, Mineral Oils, and Bio-Based Lubricants Chemistry and Technology, Rudnick, L.R., Ed., CRC Press, Boca Raton, FL, 2005, Chap.22. Erhan, S.Z., Vegetable oils as lubricants, hydraulic fluids, and inks, in Bailey's Industrial Oil and Fat Products, Volume 6, Industrial and Nonedible Products from Oils and Fats, 6th ed., Shahidi, F., Ed., John Wiley & Sons, New York, 2004, Chap.7. Gardner, H.W., Oxygen radical chemistry of polyunsaturated fatty acids, Free Rad. Biol. Med., 7, 65, 1989. Gordon, M.H. and Mursi, E., Comparison of oil stability based on the Metrohm Rancimat with storage at 20°C, J. Am. Oil Chem. Soc., 71, 649, 1994. Grosch, W., Moll, C. and Biermann, U., Occurrence and formation of bitter-tasting trihydroxy fatty acids in soybeans, J. Agric. Food Chem., 27, 239, 1979. Hagemann, J.W. and Rothfus, J.A., Computer modeling of theoretical structures of monoacid triglyceride alpha-forms in various subcell arrangements, J. Am. Oil Chem. Soc., 60, 1308, 1983. Hagemann, J.W., Tallent, W.H., and Kolb, K.E., Differential scanning calorimetry of single acid triglycerides: effect of chain length and unsaturation, J. Am. Chem. Soc., 49, 118, 1972. Hagemann, J.W., Thermal behavior and polymorphism of acylglycerols, in Crystallization and Polymorphism of Fats and Fatty Acids, Garti, N. and Sato, K., Eds., Marcel Dekker, New York, 1988, 9–95.
References Adhvaryu, A. et al., Application of quantitative NMR spectroscopy to oxidation kinetics of base oils using a pressurized differential scanning calorimetry technique, Energ. Fuels, 13, 493, 1999. Adhvaryu, A. et al., Oxidation kinetic studies of oils derived from unmodified and genetically modified vegetables using pressurized differential scanning calorimetry and nuclear magnetic resonance spectroscopy, Thermochem. Acta, 364, 87, 2000. American Oil Chemist’s Society Official Method Cd 12-57, Fat Stability, Active Oxygen Method, Am. Oil Chem. Soc., Champaign, IL, 1983. American Oil Chemist’s Society Official Method Cd12b-92, Oil Stability Index (OSI), Am. Oil Chem. Soc., Champaign, IL, 1993. Annual Book of ASTM Standards, Section Five: Petroleum Products, Lubricants, and Fossil Fuels, vol. 05.02, ASTM, West Conshohocken, PA, 2000. Antila, V., Fatty acid composition, solidification and melting of Finnish butter fat, Finn. J. Dairy Sci., 27, 1, 1966. Asadauskas, S. and Erhan, S.Z., Depression of pour points of vegetable oils by blending with diluents used for biodegradable lubricants, J. Am. Oil Chem. Soc., 76, 316, 1999. Bach, A.C. and Babayan, V.K., Medium-chain triglycerides: an update, Am. J. Clin. Nutr., 36, 950–62, 1982. Bach, A.C. et al., The usefulness of dietary medium-chain triglycerides in body weight control: Fact or fancy?, J. Lipid Res., 37, 708–26, 1996. Battersby, N.S., The biodegradability and microbial toxicity testing of lubricants — some recommendations, Chemosphere, 41, 1011, 2000. Battersby, N.S. et al. A correlation between the biodegradability of oil products in the CEC L-33-T-82 and modified Sturm tests, Chemosphere, 24, 1989, 1992. Becker, R. and Knorr, A., An evaluation of antioxidants for vegetable oils at elevated temperatures, Lubr. Sci., 8, 95, 1996. Bentz, A.P. and Breidenbach, B.G., Evaluation of the differential scanning calorimetric method for fat solids, J. Am. Chem. Soc., 46, 60, 1969. Booser, E.R., CRC Handbook of Lubrication and Tribology, vol. III, CRC Press, Boca Raton, FL, 1994, 282. Booser, E.R., CRC Tribology Data Handbook, CRC Press, Boca Raton, FL, 1997, 136.
624
Nonfood Uses of Oils and Fats
Harman, D., Free Radical in Biology, vol. 5, Academic Press, New York, 1982, 255. Hill, S.E. and Perkins, E.G., Determination of oxidation Stability of Soybean oil with the oxidative stability instrument: operation parameter effects, J. Am. Oil Chem. Soc., 72, 741, 1995. Hwang, H.S. and Erhan, S.Z., Lubricant base stocks from modified soybean oil, in Bio-Based Industrial Fluids and Lubricants, Erhan, S.Z. and Perez, J.M., Eds., AOCS Press, Champaign, IL, 2002, Chap. 2. Jebe, T.A. et al. Collaborative study of the oil stability index analysis, J. Am. Oil Chem. Soc., 70, 1055, 1993. Kowalski, B, Evaluation of activities of antioxidants in rapeseed oil matrix by pressure differential scanning calorimetry, Thermochim. Acta, 213, 135, 1993. Kowalski, B., Determination of oxidative stability of edible vegetable oils by pressure differential scanning calorimetry, Thermochim. Acta, 156, 347, 1989. Kowalski, B., Thermal-oxidative decomposition of edible oils and fats. DSC studies, Thermochim. Acta, 184, 49, 1991. Labuza, T.P. and Bergquist, S., Kinetics of oxidation of potato chips under constant temperature and sine wave temperature conditions, J. Food Sci., 48, 712, 1983. Labuza, T.P., et al., Metal-catalyzed oxidation in the presence of water in foods, J. Am. Oil Chem. Soc., 48, 527, 1971. Laemmle, P., Biodegradable Hydraulic Fluids and Utto-Lubricants, SAE 2002-01-1455, NCFP I02-20.3, 2002. Laubli, M.W. and Bruttel, P.A., Determination of the oxidative stability of fats and oils: comparison between the active oxygen method (AOCS Cd 12-57) and the Rancimat method, J. Am. Oil Chem. Soc., 63, 792795, 1988. Laubli, M.W. and Bruttel, P.A., Determination of the oxidative stability of fats and oils: comparison between the active oxygen method (AOCS Cd 12-57) and the rancimat method, J. Am. Oil Chem. Soc., 63, 792, 1986. Lawate, S., Environmentally friendly hydraulic fluids, in Biobased Industrial Fluids and Lubricants, Erhan, S.Z. and Perez, J.M., Eds., AOCS Press, Champaign, IL, 2002, chap.3. Lee, C.J. and Klaus, E.E., Evaluation of deposit forming tendency of mineral and synthetic base oils using the Penn state microoxidation test, Lubr. Eng., 49, 441, 1993. Loury, M., Possible mechanisms of Autoxidative Rancidity, Lipids, 7, 671, 1972. Matthäus, B. et al. Fast chemiluminescence method for detection of oxidized lipids, Fat Sci. Technol., 96, 95, 1994. Matthäus, B. et al. Bestimmung von Hydroperoxiden in Fetten und Ölen durch Chemilumineszenz, Lebensmittelchemie, 47, 85, 1993. Murray, D.W. et al., Effect of basestock composition on lubricant oxidation performance, Petrol. Rev., 36 (421), 36–40, 1982. Oxidationsstabilität von Ölen und Fetten-Rancimatmethode, Metrohm AG, Herisau, Application Bulletin Metrohm Nr. 204/1 d, 1994. Padavich, R.A. and Honary, L., A market research and analysis report on the vegetable oil based industrial lubricants, SAE Tech paper 952077, p. 13, 1995. Patin, L.J. and James, D.K., Biodegradable and fire-resistant hydraulic fluids: now you can have both, Lubric. World, Nov. 26, 2002. Perez, J.M. and Boehman, A.L., Environmentally friendly fuels and lubricants, in Biobased Industrial Fluids and
Lubricants, Erhan, S.Z. and Perez, J.M., Eds., AOCS Press, Champaign, IL, 2002, Chap.6. Perez, J.M. et al. Comparative evaluation of several hydraulic fluids in operational equipment, a full scale pump test stand and the four-ball wear tester–Part III. New and used hydraulic fluids, Lubr. Engr., 52, 416, 1986. Privett, O.S. and Blank, M.L., The initial stages of autoxidation, J. Am. Oil Chem. Soc., 39, 465, 1962. Product review on biodegradable fluids and lubricants, Ind. Lub. & Tribol., 48(2), 17, 1996. Randles, S.I. and Wright, M., Environmentally considerate ester lubricants for the automotive and engineering industries, J. Syn. Lubr. 9, 145, 1992. Rossell, J.B. and Pritchard, J.L.R., Analysis of Oil Seeds, Fats and Fatty Foods, Elsevier, London, 1991. Rudnick, L.R., A comparison of synthetic and vegetable oil esters for use in environmentally friendly fluids, in Bio-Based Industrial Fluids and Lubricants, Erhan, S.Z. and Perez, J.M., Eds., AOCS Press, Champaign, IL, 2002, Chap. 4. Schmidt, M.A. et al. Biotechnological enhancement of soybean oil for lubricant applications, in Synthetics, Mineral Oils, and Bio-Based Lubricants Chemistry and Technology, Rudnick, L.R., Ed., CRC Press, Boca Raton, FL, 2005, Chap. 23. Shahidi, F., Natural Antioxidants Chemistry, Health Effects and Application, AOCS Press, Champaign, IL, 1997, 1. Shankwalkar, S. and Placek, D., Oxidation kinetics of tricresyl phosphate (TCP) using differential scanning calorimetry (DSC), Lubr. Eng., 50, 261, 1994. Smith, C.R. et al., Lesquerolic acid, a new hydroxy acid from Lesquerella seed oil, J. Org. Chem., 26, 2903, 1961. Smith, C.R. et al., Densipolic acid: a unique hydroxydienoid acid from Lesquerella densipila seed oil, J. Org. Chem., 27, 3112, 1962. Soya & Oilseed Industry News, ‘BioEnergy’ conference aims to help power U.S. energy independence, Oct. 30, 2003. Steinberg, D. et al., Beyond cholesterol: modifications of lowdensity lipoprotein that increase its atherogenicity, New Engl. J. Med., 320, 915, 1989. Swern, D., Organic Peroxides, vol. 1, Wiley/Interscience, New York, 1970, 115. Wewala, A.R., Natural Antioxidants. Chemistry, Health Effects and Applications, AOCS Press, Champaign, IL, 1997, 331. Whitby, R.D., Market share of bio-lubricants in Europe and the USA, Lipid Technol., 16(6), 125–130, 2004.
9.6
Biofuels
9.6.1
What are biofuels and why are they attracting so much interest?
For the last century and more, man has exploited fossil fuels (coal, mineral oil, and gas) to provide energy for warmth (and cooling), transport, and to drive industrial machinery. The supplies of fossil fuels are large but finite and in recent decades and particularly in the last few years there has been a revived interest in alternative forms of energy among which are two based on plant matter — bioethanol and biodiesel (usually methyl esters derived from vegetable oils and animal fats). The reasons for this 625
9.6
Biofuels
North and South America. However, biodiesel is probably more expensive to produce than petrodiesel and can only be economic with government assistance. In Europe this is provided in the form of lower taxation compared with conventional fuel. In the U.S. where fuels are less heavily taxed assistance can be provided via The Clean Air Act Amendments of 1990, the Energy policy Act of 1992, or by tax rebate. Currently this amounts to $0.01 per 1% of biodiesel from soybean oil or $0.005 from other sources. This is equivalent to 20 and 10 cents, respectively, for each litre of a 20% blend. The doubt about relative costs expressed earlier relates to the difficulty in determining the true cost of production free of taxes and subsidies and in the problem of assessing in financial terms the environmental and social benefits that result from producing and using biodiesel and the cost of global warming made worse by doing nothing. For example, in Brazil production of biodiesel from locally grown crops is seen not only as a way of reducing the environmental and economic cost of transporting fuel by road over long distances, but as a way of providing rural employment and slowing the drift of unemployed people to already overcrowded cities. Governments lend their support to the production and use of biodiesel for differing reasons including reduction in pollution, in carbon dioxide emissions, in the use of nonrenewable resources, and to encourage the use of domestically produced material. Despite the popularity of biodiesel as a way of overcoming some environmental concerns, some serious questions have been raised (Dumelin, 2005). First it has to be recognised that biodiesel can only make a small contribution to the total requirement for energy. Until we develop new energy sources, such as hydrogen or atomic fusion, we will remain dependent on fossil fuels. The annual production of mineral oils is about 30 times larger than that of commodity oils and fats. Secondly, oils and fats are needed as an important source of energy-rich food and of essential nutrients. The demand for food increases with the growth in population and in income and there remains a substantial unmet need. Why then should we divert this irreplaceable food source? The Indian government has decreed that biodiesel cannot be made from any oil or fat that could be used for food and alternative supplies are being sought, such as the nonedible oil from Jatropha curcas seeds. There are claims already that the supply/demand system is being disturbed and this is reflected in pressure on prices. However, it is worth noting that Fry (2001) has forecast that this situation will change during the present century. Production levels will continue to increase through rising yields, population will level off about mid-century, and there is a limit to how much fat we can eat with the consequence that from around 2050 onwards there will be quite large supplies available for nonfood purposes including biofuels. A third concern is that biodiesel is only economic through manipulation of taxation. When the supply of biodiesel is small, this may not matter, but as it increases the downturn in tax revenues will have to be met in another way. Perhaps this is a way of transferring a part of the cost from
new emphasis have included the rising cost of mineral oil, concern about security of supply and about environmental issues, and pressures from oil and fat producers and from governments. Europe has led the way in the production of biodiesel. North and South America have favoured bioethanol, but here too biodiesel is becoming more important. In 2004, the U.S. produced 4 billion gallons (19 million tonnes) of bioethanol, but only 30 million gallons (0.13 million tonnes) of biodiesel. In 1973 and again in 1978 the price of crude petroleum rose very rapidly and unexpectedly. This had serious consequences for the economies of importing countries in Asia, Europe, and North America and led to a serious global economic downturn. The power of the mineral oilproducing countries, particularly in the Middle East, to control prices and supplies made the western countries concerned about security of supply and caused them to consider alternative sources of petroleum supply and alternative sources of energy. Those concerns remain and are deepened by current fluctuating prices that are frequently high and occasionally very high. Environmental issues are immediate and local leading to pollution and also long-term and global leading to global warming. In cities, particularly, our dependence on cars, vans, and lorries (trucks) to transport goods and persons leads to poor atmospheric conditions (smog) and a consequent interest in fuels that are less polluting. There is also a growing concern about global warming that many consider to be manmade and arising in large part from the production of greenhouse gases of which carbon dioxide resulting from burning of fossil fuels is of greatest concern. There is also long-term worry about the depletion of our valuable but nonrenewable petroleum supplies. Despite the increasing demand for oils and fats as food and for oleochemicals there are times of over-supply and pressures then come from producers of oils and fats to find new uses for these materials. There is also pressure from governments who, having set and accepted targets for the use of nontraditional forms of energy, are struggling to meet those targets. These alternative sources of energy are further seen as a way of avoiding the use of nuclear energy, considered by many as a less acceptable nonfossil fuel solution. One way to meet some of these problems, in small part, is to replace fossil fuels with biodiesel (usually methyl esters derived from oils and fats). These materials have the advantage that they can be used neat or in various blends with petrodiesel without engine modification and with the approval of the vehicle manufactures. They can also be distributed easily through the existing fuel supply chain. Development and limited use of biodiesel goes back over 20 years, but these activities have accelerated in the last few years and many countries, both developed and developing, have plans to produce and use biodiesel if they are not already doing so. Several EU countries are in the lead, but they are being followed by rapid growth in Asia and in 626
Nonfood Uses of Oils and Fats
for this purpose. It is desirable to avoid too high a content of saturated esters that may solidify at ambient temperature and high levels of polyunsaturated esters, especially those with more than two double bonds, which lead to undesirable oxidation and may cause problems during storage of the fuel or at the moment of use. Nevertheless a wide range of materials is available for use with convenience of supply being the most important. Rapeseed or canola oil (Europe), soybean oil (North and South America), palm oil (Malaysia), coconut oil (Philippines), and tallow (New Zealand) have been used or examined in the countries indicated. There is even a report of fish oil being used in Canada after removal of the valuable EPA and DHA. The fatty acid composition of some of these oils is shown in Table 9.24. About 80% of the production cost of biodiesel lies in the price of the starting oil when this is an appropriately refined vegetable oil. Attempts to reduce cost have led to the use of animal fats (tallow, chicken fat, and animal fats unfit for human consumption), vegetable oils already used for frying and cooking, free acids removed during refining, and alternative vegetable oils, such as castor, palm, jatropha (Jatropha curcas), and karanja (Pongamnia pinnata) oils from India and Africa, babassu oil, jojoba oil (a wax that would yield methyl esters and fatty alcohols on methanolysis), and fatty acids recovered from refining processes (Table 9.25). The quality of these less common starting materials needs to be controlled and specifications for each
the customer to the taxpayer (frequently, but not always the same person). Reaney et al. (2005) also take an optimistic view. They consider that biodiesel will become the largest market for oils and fats exceeding demands for food, feed, and the other nonfood uses discussed in this chapter. This will involve developments in agriculture (mainly higher yields and new biodiesel crops), better use of by-products (meal and glycerol), and improved production technology. Although 100% biodiesel can be used in vehicles and is so used in Germany and Austria, it is more often offered as a blend with regular diesel. In the U.S. biodiesel generally contains 20% of added methyl esters, but in Europe 5% blends are more common. In addition to road vehicles, these products are being used in agricultural machinery, in boats — both for business and for pleasure — and in trains. Methyl esters can also be used in central heating systems and it is of interest that rapeseed methyl esters are used to heat the Reichstag building in Berlin, newly renovated after the union of West and East Germany into a single state in 1990. The methyl esters also serve in one route to fatty alcohols (Section 9.2.3).
9.6.2
From what material is biodiesel made?
Good quality biodiesel must meet certain criteria (see below) and this may limit the range of oils and fats that can be used TABLE 9.24
Fatty acid composition of selected oils and fats
Palmitic Stearic Oleic Linoleic Linolenic
Rapeseed
Sunflower
Soybean
Palm
Tallow
4–5 1–2 55–63 20–31 9–10
3–6 1–6 14–43 44–69
2–11 2–6 22–31 49–53 2–10
32–46 4–6.3 37–53 6–12
25–37 14–29 26–50 1–3
Source: Adapted from Knothe, G. and Dunn, R.O., in Oleochemical Manufacture and Applications, Gunstone, F D., Hamilton, R.J., Eds., Sheffield Academic Press, Sheffield, U.K., 2001, Chap. 5.
TABLE 9.25
Fuel-related properties of selected oils and fats and of the methyl esters derived from them IV
CN
35 61 94–120 117–143 110–143 35–48
42 37.6 37.9 37.1
HG
Viscosity
CP
PP
FP
–3.9 –3.9 7.2
–31.7 –12.2 –15.0
–15.0
–33.0
246 254 274 201 52
–2 2 0 12
–9 –1 -4 9
Oils and Fats Palm Rapeseed Soybean Sunflower Tallow Petrodiesel
47
39709 39623 39575 40054 45343
37.0 32.6 37.1 51.1 2.7
(37.8) (37.8) (37.8) (40) (37.8)
Derived Methyl Esters Rapeseed Soybean Sunflower Tallow
54.4 46.2 46.6
40449 39800 39800 39949
6.7 4.1 4.2 4.1
(40) (40) (40) (40)
84 171 96
Note: Iodine value, cetane number, gross heat of combustion (kJ/kg), kinematic viscosity min2/s) at the temperature indicated (°C), cloud point (°C), pour point (°C), flash point (°C). Source: Adapted from Knothe, G. and Dunn, R.O., in Oleochemical Manufacture and Applications, Gunstone, F D., Hamilton, R.J., Eds., Sheffield Academic Press, Sheffield, U.K., 2001, Chap. 5.
627
9.6
Biofuels
purification), methanol (for reuse), and biodiesel. Plants vary in size from 10 to 200 kt/year. The reaction is an equilibrium process involving two phases and interesterification may be carried out in two or three stages. At the end of each stage glycerol is removed and fresh methanol and a catalyst is charged to the reactor. Reaction proceeds at about 60°C with an overall residence time of about 2 hours. Plants are now available that are flexible enough to handle a range of different feed stocks (Meyer and Koerbitz, 2004). Information is available on the websites of companies, such as Ballestra, Bayer, Crown Iron Works, De Smet, Lurgi Life Science, and Westphalia available through Google.
would be required. Since large-scale production plants are generally preferred, it is necessary to secure a source of oil or fat available continually and in appropriate quantity.
9.6.3
How is biodiesel manufactured?
Despite the desire to use cheap starting materials for biodiesel production, the final product must meet defined specifications and this may be achieved more easily with good quality oil. The cost of upgrading a poor product may take up the saving in a cheaper starting material. For biodiesel production, crude oils should be degummed (removal of phospholipids) and neutralised (removal of free acids), but bleaching and deodorisation is not generally required. In general, oils and fats (triacylglycerols) are converted to methyl esters by reaction with methanol in the presence of an acidic, basic, or enzymatic catalyst (Section 9.2.2). Reaction with a basic catalyst (NaOH, KOH, or preferably NaOMe) is the preferred route, but is only appropriate with oils with low levels of free acid. The other product of this reaction is glycerol (Section 9.2.5) and the economic value of this by-product is an important part of the economics of the whole process. Since many other oleochemical processes are dependent on the economic value of glycerol, there is concern about the over supply of this by-product with the consequence of this for prices (Section 9.2.5). It is important that the methanolysis reaction proceed as far as possible as partial glycerol esters as well as methanol, glycerol, and free acids are considered to be undesirable impurities in biodiesel. Walker (2004) has reported that one tonne of rapeseed can be expected to produce crude oil (0.41 tonnes) and meal (0.58 tonnes) and that the refined oil (0.40 tonnes) when reacted with methanol will furnish 0.04 tonnes of glycerol and 0.38 tonnes (432 litres) of methyl esters. With an oilseed yield of 3 t/ha (common in Europe, but not elsewhere) this means that 1300 litres of biodiesel can be obtained with the crop from one hectare of land. Obviously, the energy recovered when the esters are used as a biofuel must exceed that required to produce the fuel with calculations starting from the planted seed. Walker has cited energy output/input ratios for rapeseed methyl esters generally between 2.0 and 2.6 after allowing for the value of the meal and the glycerol. One of the most energydemanding steps in the process is the fertiliser requirement. Walker (2004) has also discussed the economics of biodiesel production from rapeseed and produced figures of 30 to 60 pence/litre without allowance for distribution costs, taxation, or profit. The range of cost is related to the size of the operating plant. Several manufactures have designed new or adapted existing plants for efficient biodiesel production. Inputs are methanol, oil (extracted and appropriately pretreated), and catalyst (NaOH, KOH, or NaOMe). The steps involved include interesterification, separation of the glycerol and ester phases, and recovery of glycerol (for further
9.6.4
Specifications for biofuels
Biodiesel must meet specifications (which may differ between countries and regions of the world) that have been accepted by most vehicle manufactures. The EU biodiesel specification (EN 14214) provides minimum and/or maximum values for ester content, density, viscosity, flash point, sulfur content, cetane number, water content, oxidation stability, acid value, iodine value, content of methanol, glycerol, monoacylglycerol, diacylglycerol, linolenate ester, phosphorus, and some other properties. These relate to rapeseed methyl esters and the iodine value with an upper limit of 120 would exclude biodiesel from sunflower oil or soybean oil. ASTM standard D6751 in the U.S. covers flash point, water and sediment, carbon residue, sulfated ash, viscosity, sulfur, cetane, cloud point, acid number, and free and total glycerol. Iodine value is not included. Cetane numbers are an important index of ignition quality. For biodiesel these are affected by chain length, unsaturation, and branching and generally have higher minimum values (47 in ASTM D6751 and 51 in EN 14214) than those specified for petrodiesel. Other important properties are viscosity, pour point, and cloud point, which, if not appropriate, can affect behaviour at low temperatures. Low temperature problems may be overcome by using isopropyl esters in place of methyl esters or by winterisation of methyl esters to remove higher melting components. The latter solution adds cost to the manufacturing process by virtue of the extra step involved and leaves a higher melting fraction for which an alternative use has to be developed. Because biodiesel contains unsaturated centres, it is liable to oxidation during storage and it is important to have adequate storage conditions and possibly to add antioxidant. Reduction of sulfur dioxide emissions from vehicles has been achieved by use of ultra-low sulfur fuels, but this is disadvantageous insofar as lubricity is also reduced. However, lubricity in ultra-low sulfur diesel fuels can be restored by the addition of some 1 to 2% of biodiesel. Knothe and Steidley (2005) have shown that improved lubricity following addition of a little 628
Nonfood Uses of Oils and Fats
to meet manmade targets to reduce carbon dioxide emissions. The production of biodiesel in India is confined to nonedible oils by government decree (Kale, 2005a,b). Gunstone (2006) has forecast a demand for 40 to 50 million tonnes of biodiesel in 2020.
biodiesel is a consequence mainly of the free acid and monoacylglycerols present in this material. These have been considered as undesirable contaminants and there is perhaps a case for considering minimum as well as maximum limits for these components. Compared to regular diesel, biodiesel is considered to produce lower emissions of SO2, CO, hydrocarbons, PAH (polycyclic aromatic hydrocarbons), particulates, and smoke. Depending on engine tuning, NOx emissions are usually higher with biodiesel. Of course, carbon dioxide is produced as in the combustion of all organic products, but it is argued that this was trapped from the atmosphere during the growing season and, therefore, does not add to global supplies of this gas. Combustion of petrodiesel on the other hand liberates carbon dioxide trapped in fossil fuels millennia previously and, therefore, adds to the global carbon dioxide supply.
9.6.5
References Anon., BioX plans palm oil use for green energy, Inform, 16, 748, 2005. Dumelin, E.E., Biodiesel — a blessing in disguise, Eur. J. Lipid Sci. Technol., 107, 63–64, 2005. Fry, J., The world’s oil and fat needs in the 21st century: lessons from the 20th century. Lecture presented in 2001 and cited by Gunstone, F.D., Ed., Vegetable Oils in Food Technology, Blackwell Publishing, Oxford, U.K., 2002, pp. 15–16. Gunstone, F.D., Chang Lecture delivered at the AOCS Exhibition in St. Louis, May 2006. Kale, V., Biodiesel gaining popularity in India, Inform, 16, 466–468, 2005a. Kale, V., Jatropha — India’s crop for biodiesel production, Inform, 16, 532–533, 2005b. Knothe, G. and Dunn, R.O., Biofuels derived from vegetable oils and fats, in Oleochemical Manufacture and Applications, Gunstone, F.D., Hamilton, R.J., Eds., Sheffield Academic Press, Sheffield, U.K., 2001, Chap. 5. Knothe, G., et al., Ed., The Biodiesel Handbook, AOCS Press, Champaign, IL, 2004. Knothe, G. and Steidley, K.R., Lubricity of components of biodiesel and petrodiesel. The origin of biodiesel lubricity, Energ. Fuels, 19, 11921200, 2005. Meyer, S. and Koerbitz, W., Worldwide review of biodiesel production and studies on biodiesel production plants in Europe, Malay. Oil Sci. Technol., 13, 11–16, 2004. Mittelbach, M. and Remschmidt, C., Biodiesel — The Comprehensive Handbook, (published by Mittelbach), Graz, Austria, 2004. Reaney, M.J.T., Vegetable oils as biodiesel, in Bailey’s Industrial Oil and Fat Products, 6th ed., Shahidi, F., Ed., Wiley Interscience, New York, 2005, Chap. 6. Walker, K., Non-food uses, in Rapeseed and Canola Oil, Gunstone, F.D., Ed., Blackwell Publishing, Oxford, U.K., 2004, Chap. 7.
Present and future demand for biodiesel
Targets for the replacement of petrodiesel by biodiesel have been set in the EU, though not all member counties have accepted these. It is considered that in 2005 the level of replacement should be 2.5% of total vehicle consumption and that this should rise at the rate of 0.75% a year to 5.75% of total consumption. Targets set by individual countries for 2005 were 2.0% or above for Austria, Czech Republic, France, Germany, Italy, Latvia, Lithuania, Slovakia, Spain, Sweden, and the U.K., but not all these were met. Some other countries with targets of 1% or below are working actively to increase production and there are regular reports of new plants being planned and coming on stream. These levels of 2.5 and 5.75% will require about 2.5 and 6 tonnes of biodiesel. Currently about 80% of European biodiesel comes from rapeseed oil, but other sources will have to be found and this proportion must fall. Production of rapeseed in EU-25 in 2004/05 was around 5.5 million tonnes, 1 million tonnes higher than in the preceding year. Interest in producing biodiesel is not confined to Europe and many other countries are beginning to produce this material. Where consumption targets have been set (and not met), a growing export–import trade in biodiesel methyl esters will probably develop. For example, Malaysia and Indonesia with their ample supplies of starting material should be a good source of biodiesel and coco-biodiesel is being developed in the Philippines for export to Japan. Malaysia is expected to meet about 10% of the world demand for biodiesel. It is expected that the U.S. will use 0.5 million tonnes of biodiesel in 2006 and that a similar amount will be produced and used in Brazil by 2010. These figures relate to biodiesel and do not include the burning of around 0.5 to 1.0 million tonnes of vegetable oils and animal fats to produce electricity (Anon., 2005). This use is driven in part by low prices for oils and fats (especially palm oil) and in part by pressure
9.7
Surface coatings and inks
It has long been known that when thin layers of unsaturated fatty oils are exposed to air they harden to solid impervious films. Artists used this knowledge to provide a protective transparent film for their paintings. With incorporation of appropriate pigments, it is possible to produce a coloured layer. Surface coatings may be applied to wood, paper, metal and plastic surfaces to protect against moisture, oxygen, sunlight, radiation and pollutants, such as sulfur dioxide. They may also serve to decorate or disguise. About half the annual production of paint is used for internal and external use in buildings 629
9.7
Surface coatings and inks
from a polybasic acid and a polyhydric alcohol with unsaturated fatty acids acylated to additional hydroxyl groups. The polybasic acid is generally phthalic acid (benzene-1,2dicarboxylic acid) used in the form of its anhydride. The polyhydric alcohol may be glycerol or pentaerythritol [C(CH2 OH)4]. Other possible reactants are listed in Table 9.27. The fatty acids are mainly unsaturated and may be supplied as free acids, triacylglycerols, or often as monoacylglycerols derived from appropriate vegetable oils like linseed oil, soybean oil, or dehydrated castor oil. Castor, tung, coconut, safflower, sunflower, and tall oil may also be used. Alkyds usually contain 30 to 70% of fatty acids. Because the drying process involves oxidative polymerization, it is desirable that the vegetable oil serving as a source of the fatty acids contain as little natural antioxidant as possible. The fatty acids may also be modified to improve their properties as in reaction with maleic anhydride to increase functionality or by prior conjugation of the polyunsaturated system with a proprietary catalyst to speed drying. Alkyds are divided into three classes depending on their content of unsaturated oil, which may be >60% (long oil), 40 to 60% (medium oil), and 10-hydroxy-decanoic acid and 2octanol, resulting from caustic fusion with 1 mole of NaOH at 180-200oC; > sebacic acid and 2-octanol, resulting from caustic fusion with 2 moles of NaOH at 250-275oC; >10-undecylenic acid and heptaldehyde, resulting from pyrolysis in superheated steam at 550oC; >12-hydroxy-octadec-9-ynoic acid resulting from bromination and dehydrobromination of ricinoleate; and >9,10-oxy-12-hydroxyoctadecanoic acid, resulting from epoxidation or hot air oxidation.
634
Nonfood Uses of Oils and Fats
production of undecylenate, heptaldehyde is also produced, and is used in flavors and aromas. It is also oxidized to heptanoic acid. Esters of heptanoate remain liquid at low temperatures, retain viscosity, and are not volatile, so they are useful in such demanding applications as jet engine lubrication (Caupin, 1997). Esters of heptanoate are also used as plasticizers. Methyl ricinoleate can be converted to a hydroxyacetylenic fatty acid (Figure 9.9) by bromination, then dehydrobromination by sonic irradiation in ethanolic KOH (Lie Ken Jie et al., 1996). This compound presents three reactive moieties for polymerization reactions and has potential for more advanced applications, for example, in building nanomaterials. Ricinoleate can be epoxidized by hot air oxidation or by oxidation with peroxyacetic acid (Figure 9.9). The product formed can be used as a plasticizer and in paints or other coatings (Naughton and Vignolo, 1992) as a low volatile organic carbon (VOC) component.
9.8.4
Geller, D.P. and Goodrum, J.W. Effects of specific fatty acid methyl esters on diesel fuel lubricity. Fuel, 83, 2351–2356, 2004. Goodrum, J.W. and Geller, D.P. Influence of fatty acid methyl esters from hydroxylated vegetable oils on diesel fuel lubricity. Biores. Technol., 96, 851855, 2005. Guhanathan, S. et al. Studies on castor oil-based polyurethane/ polyacrylonitrile interpenetrating polymer network for toughening of polyunsaturated polyester resin. J. Appl. Polymer Sci., 92, 817–829, 2004. Hofer, R. et al. Oleochemical polyols — a new raw material source for polyurethane coatings and floorings. J. Coatings Tech., 69, 65–72, 1997. Kuhn, H. et al. Mechanism of the odor-absorption effect of zinc ricinoleate. A molecular dynamics computer simulation. J. Surf. Deterg., 3, 335–343, 2000. Lie Ken Jie, M.S.F. et al. Ultrasound-assisted synthesis of santalbic acid and a study of triacyl-glycerol species in Santalum album (Linn) seed oil. Lipids, 31, 1083–1089, 1996. Lin, J.T. et al. Identification and quantification of the molecular species of castor oil by HPLC and ELSD. J. Liq. Chrom. Rel. Tech., 26, 773–780, 2003. Magerl, A. et al. Allergic contact dermatitis from zinc ricinoleate in a deodorant and glyceryl ricinoleate in a lipstick. Cont. Derma., 44, 119–121, 2001. McKeon, T.A. et al. Developing a safe source of castor oil. Inform 13, 381−385, 2002. McKeon, T.A. et al. Hydroxy fatty acids. In Nutraceutical and Specialty Lipids and Their Co-Products, F. Shahidi, Ed., CRC Press, Boca Raton, FL, 2006. Naughton, F.C. and Vignolo, R.L. The chemistry of castor oil and its derivatives and their applications. Int. Castor Oil Assoc., Westfield, NJ, 1992. Qu, J. et al. Synthesis of castor oil water-borne polyurethane-acrylate hybrid emulsions. J. Chem. Ind. Eng., 56, 168–173, 2005. Roethli, J.C. et al. Lesquerella as a Source of Hydroxy Fatty Acids for Industrial Products, USDA-CSRS Office of Agricultural Materials. Growing Industrial Materials Series (unnumbered), Washington, D.C., 1991. Rojas-Barros, P. et al. Inheritance of high oleic/low ricinoleic content in the seed oil of castor mutant OLE-1. Crop Sci., 45, 157–162, 2005. Shikanov, A. et al. Poly(sebacic acid co-ricinoleic) biodegradeable carrier for paclitaxel — effect of additives. J. Cont. Rel., 105, 52–67, 2005. Slivniak, R. and Domb, A.J. Macrolactones and polyesters from ricinoleic acid. Biomacromolecules, 6, 1679–1688, 2005a. Slivniak, R. and Domb, A.J. Lactic acid and ricinoleic acid based copolyesters. Macromolecules, 38, 5545–5553, 2005b. Wurdack, K.J. et al. Molecular phylogenetic analysis of uniovulate Euphorbiaceae (Euphorbiaceae sensu stricto) using plastid rbcL and trnL-F DNA sequences. Am. J. Bot., 92, 1397–1420, 2005. Yeganeh, H. and Mehdizadeh, M.R. Synthesis and properties of isocyanate curable millable polyurethane elastomers based on castor oil as a renewable resource polyol. Eur. Polymer J., 40, 1233–1238, 2003.
Summary
Castor oil is a unique natural product, used in many industries and many types of products. It is generally thought that the uses of castor oil in industry are limited by the availability of the oil. Certainly, the presence of a midchain hydroxyl group and the polyol nature of castor oil provide nearly endless possibilities for chemists. More widespread cultivation of castor is very likely to expand utilization of castor oil by industry.
References Altafim, R.A.C. et al. The effects of fillers on polyurethane resinbased electrical insulators. Mat. Res., 6, 187–191, 2003. Anon., USDA Agricultural Statistics, Government Printing Office, Washington, D.C., 2005. Auld, D.L. et al. Development of castor with reduced toxicity. J. New Seeds, 3, 61–69, 2001. Bodalo-Santoyo, A. et al. Enzymatic biosynthesis of ricinoleic acid estolides. Biochem. Eng. J., 26, 155–158, 2005. Caupin H.J. Products from castor oil: past, present, and future. In Gunstone, F.D., Padley, F.B., Eds., Lipid Technologies and Applications. Marcel Dekker, New York, pp. 787–795. Cermak, S.C. et al. Synthesis and physical properties of estolides from lesquerella and castor fatty acid esters. Ind. Crops Prod., 23, 54–64, 2006. Clay, A.F. et al. Hydroxydecanoic acid for greasy and acne-prone skin. COSSMA, 1(14), 12–14, 2000. Dierig, D. et al. Improvement in hydroxy fatty acid seed oil content and other traits from interspecific hybrids of three Lesquerella species: Lesquerella fendleri, L. pallida, and L. lindheimeri. Euphytica, 139, 199–206, 2004.
635
10 LIPID METABOLISM
J.L. Harwood
10.1
Fatty acids
10.1.1
Fatty acid thiolesters
By contrast, mammals, fungi, and plant cytosols contain large (200 to 250 kDa) multifunctional polypeptides. The structures, regulation, and enzymatic mechanisms of the multisubunit ACCases are reviewed by Cronan and Waldrop (2002), while Kim (1997) has described mammalian ACCases and their regulation. In the first part of the ACCase reaction, the biotin moiety of BCCP (biotin carboxyl carrier protein) is carboxylated in an (adenosine 5-triphosphate) ATP-mediated process. Recent evidence has implicated carboxyphosphate as an intermediate. The carboxyl moiety is then transferred to the acceptor acetyl-CoA, probably by the use of carbon dioxide as the electrophile. Study of the halfreactions of acetyl-CoA carboxylase has allowed proposals concerning the mechanism of biotin participation to be made (De Titta et al., 1980). The whole subject of the mechanism of biotin-containing enzymes has been excellently reviewed (Knowles, 1989). Acetyl-CoA carboxylase has been purified to homogeneity from various animal tissues including rat liver (Wakil et al., 1983), chicken liver (Wada and Tanabe, 1983), rat and rabbit mammary gland (Wakil et al., 1983), and goose uropygial gland (Rainwater and Kollattukudy, 1982). It is a multifunctional protein of mass 220 to 260 kDa (kilodaltons).1 By means of combined protein, chemical, and molecular cloning techniques, Takai et al. (1987) elucidated the primary structure around the biotin-binding site of
For most of the reactions in which fatty acids participate in cells, they have to be “activated” (Gurr et al., 2002). This usually involves the generation of a thiolester, such as acylCoA or acyl-ACP. Because of the importance of thiolesters, the synthesis of CoA (coenzyme A) and ACP is critical and, indeed, the CoA biosynthetic pathway has been suggested as a possible target for antibacterials (Leonardi et al., 2005). Activation of fatty acids to coenzyme A esters is discussed by Watkins (1997) and, for a general discussion of the production of acyl-thioesters in different organisms, refer to chapters in Vance and Vance (2002). Because of the poor aqueous solubility of fatty acids, even as their acyl-thioesters, and because the latter may have potentially harmful (detergent-like) effects on cells, there are binding proteins present in cells. These include fatty acid-binding proteins (FABP) as well as acyl-CoA binding proteins (ACBP) (Veerkamp and Maatman, 1995; Glatz and van der Vusse, 1996; Bernlohr et al., 1997).
10.1.2
De novo synthesis
De novo synthesis of fatty acids requires the concerted action of two multiprotein complexes or multifunctional proteins. These are acetyl-CoA (coenzyme A) carboxylase and fatty acid synthase. Acetyl-CoA carboxylase (ACCase) is a type I biotin-containing carboxylase. ACCases can have a multiprotein structure or exist as multifunctional proteins (Table 10.1). ACCase catalyses the first committed step of fatty acid and, hence, acyl lipid synthesis. Bacteria and most plant chloroplasts (Harwood, 1996) contain a multisubunit form that is readily dissociated into its component proteins.
1
637
The dalton (Da) is equal to 1/12 of the mass of an atom of 12C. Many biologists use the dalton as a unit of mass, although it is not a “recognised” unit. It is especially convenient for structures where the word “molecule” is incorrect, but is used more generally. One should write that the molecular mass (not weight) of a protein, for example, is 220,000 Da or 220 kDa.
10.1
Fatty acids
TABLE 10.1 Examples of different acetyl-CoA carboxylases Species
Protein structure
E. coli
Multiprotein complex
Yeast Dicotyledon plants
Multifunctional protein Multiprotein complex Multifunctional protein
Grasses (Poaceae)
Multifunctional proteins
Animals
Multifunctional proteins
Details Four proteins: biotin carboxylase, BCCP and carboxyltransferase (a heterodimer); transcription of all four acc genes is under growth rate control 190–230 kDa; activated but not polymerised by citrate In chloroplasts; similar properties to E. coli enzyme; probably three of the subunits coded by nucleus, one by chloroplast Presumed to be cytosolic; concentrated in epithelial cells in pea; molecular mass 200–240 kDa; functions as dimer Graminicide insensitive Two isoforms, both 200–240 kDa, which function as dimers Chloroplast isoform is graminicide sensitive, but cytosolic form (concentrated in epithelial cells) is insensitive Both are nuclear encoded Cytosolic; about 250 kDa, but functions as polymer of up to 107 kDa; aggregation increased by citrate; also regulated by phosphorylation in response to hormones
Source: From Gurr, M.I., Harwood, J.L. and Frayn, K.N. (2002). Lipid Biochemistry, 5th ed., Blackwell Scientific, Oxford, U.K. With permission.
and insulin. Total enzyme amounts also change during cell differentiation and development (reviewed by Volpe and Vagelos, 1976). Use of antibodies to rat liver acetylCoA carboxylase revealed unexpectedly that the mitochondrial outer membrane contained a major pool of rather poorly active enzyme. It was suggested that this protein represented a reservoir of acetyl-CoA carboxylase that could be released and activated under lipogenic conditions (Allred and Roman-Lopez, 1988). However, the role of ACCase-generated malonyl-CoA in the control of mitochondria β-oxidation (see below) should be noted. The control of mammalian acetyl-CoA carboxylases has been summarised well by Goodridge (1991). Animal acetyl-CoA carboxylases have been recently reviewed by Rangan and Smith (2002), who give further information about domain organisation. They also draw attention to the two major isoforms of ACCase, α and β, which are found in animals. The α-isoform is found in the cytosol of lipogenic tissues (e.g., adipose) where it is important for de novo synthesis of fatty acids. The βisoform is present in tissues, which generally have a low lipogenic activity and is associated with the outer mitochondrial membrane. The malonyl-CoA produced by ACCase-β functions primarily as a negative regulator of carnitine palmitoyltransferase I and, therefore, regulates the flux of fatty acids into mitochondria for β-oxidation. Until fairly recently the molecular nature of plant acetylCoA carboxylase was unclear. However, it now seems that the enzyme is present as a multifunctional protein of mass 220 to 240 kDa (see Harwood, 1996) in all plant cytosols. Reference to the purification of higher molecular mass forms of the enzyme from parsley and oilseed rape as well as earlier work will be found in Harwood (1988). A similar form has also been purified to near homogeneity from the diatom Cyclotella cryptica (Roessler, 1990). In most plants (the gramineae are an exception), the major ACCase isoform,
chicken liver acetyl-CoA carboxylase, and they (in 1988) described the complete amino acid sequence of the chicken liver enzyme by cloning and sequencing the DNA complementary to the appropriate messenger RNA. The chicken liver acetyl-CoA carboxylase was found to be composed of 2324 amino acid residues with a calculated molecular mass of 262,706 Da. The BCCP domain was in the middle region of the polypeptide. The amino terminal portion showed a primary structure homologous to that of carbamyl phosphate synthase, and was thought to be the site of the biotin carboxylase for which carboxyl phosphate is the postulated reaction intermediate (see Knowles, 1989). Although animal acetyl-CoA carboxylases have molecular masses of about 250 kDa, they function as polymers of molar masses 4 × 106 to 8 × 106 g/mol (Wakil et al., 1983). The regulation of mammalian acetyl-CoA carboxylase has been well studied owing to the key role of this enzyme in controlling overall fatty acid (and fat) synthesis. Control occurs in two ways: short-term control, which involves allosteric regulation and covalent enzyme modification, or long-term control, where the amounts of the carboxylase are changed (Gurr et al., 2002). Metabolite control is due to citrate (Moss and Lane, 1971) and long-chain acyl-CoAs. The latter cause depolymerisation of mammalian carboxylases. Palmitoyl-CoA, stearoyl-CoA, and arachidoyl-CoA are the most effective at inhibiting the carboxylase (Nikawa et al., 1979). Other effectors have been reported to regulate mammalian acetyl-CoA carboxylase (cf. Wakil et al., 1983), but their physiological importance is less certain. Mammalian acetyl-CoA carboxylase is also subject to regulation by phosphorylation/dephosphorylation (Kim, 1997). Phosphorylation with one mole of phosphate per mole of rat liver carboxylase subunit causes complete inactivation. The reactions have also been studied using preparations from the rat epididymal fat pad. Long-term regulation of the carboxylase is caused by diet, thyroxine, 638
Lipid Metabolism
CoA carboxylase has been summarised (see Volpe and Vagelos, 1976; Cronan and Waldrop, 2002). The DNA sequence of the gene encoding the biotin carboxylase subunit has been reported and encodes a protein of 449 residues. The sequence is strikingly similar to the amino terminal sequence of two biotin-dependent carboxylase proteins, yeast pyruvate carboxylase and the subunit of rat propionyl-CoA carboxylase (Li and Cronan, 1992). Neither citrate nor phosphorylation plays any role in regulating the bacterial enzyme. Instead, the guanosine nucleotides, guanosine 3′-diphosphate 5-di (and tri-) phosphate, were suggested for use. However, these observations have been questioned (see Cronan and Waldrop, 2002). On the other hand, it is clear that nutrient supply is important in controlling total ACCase activity. Fatty acid synthases can be categorised as Type I, Type II, and Type III. The distribution and some of the properties of the Type I and Type II enzymes are summarised in Table 10.2. The Type I synthases tend to occur in higher organisms, and all of the purified synthases of eukaryotes, with the exception of plants, are of this type. Two bacterial genera, Mycobacterium and Corynebacterium, also contain Type I synthases. These synthases are large molecular mass multifunctional proteins containing covalently bound acyl carrier protein (ACP) (see Wakil et al., 1983). On the other hand, Type II synthases, such as that from E. coli, consist of individual enzymes that can be isolated in an active form. The ACPs of Type II synthases also readily dissociate and can be purified and characterised (see Volpe and Vagelos, 1973). Type III synthases elongate preformed acyl chains and will be dealt with later. The partial reactions of fatty acid synthases are shown in detail in Table 10.3, where assay details for measuring activities in plants are given. A comparison of some features of the Type I and Type II fatty acid synthases is made in Table 10.4. The Type I fatty acid synthases from rat liver, adipose and lactating mammary glands, rabbit mammary glands, and uropygial glands have been purified and found to have many common features (Table 10.4). The two subunits appear to be identical and of molecular mass 200 to 250 kDa
however, is a multiprotein complex located in the chloroplast where it is used for de novo fatty acid synthesis. In gramineae (grasses), the chloroplast isoform is a multiprotein complex. In addition to acetyl-CoA carboxylase, cell-free extracts from several monocotyledonous (monocot.) and dicotyledonous (dicot.) plant species have been found to contain three additional biotin-containing enzymes — 3-methylcrotonyl-CoA carboxylase, propionyl-CoA carboxylase and pyruvate carboxylase (Wurtele and Nikolau, 1990) — the presence of which can complicate purification protocols making use of avidin affinity columns. However, propionylCoA carboxylase activity is also present in many purified plant acetyl-CoA carboxylases. Regulation of the plant acetyl-CoA carboxylase is not as well understood as it is for mammals. Tricarboxylic acids do not seem to function at all, and the enzyme in leaves may be activated, at least in part, through changes in the stroma medium as photosynthesis occurs (see Harwood, 1988). Nevertheless, recent measurement of the pool sizes of intermediates during lightactivated fatty acids synthesis in spinach leaves shows clearly that the activity of acetyl-CoA carboxylase is a key regulatory component (Post-Beittenmiller et al., 1991). Recent updates on plant acetyl-CoA carboxylase and its regulation are given by Schmid and Ohlrogge (2002) and by Harwood (2005). These refer to the different forms of ACCase found in graminaceae (grasses) compared to dicotyledons and the importance of ACCase in the regulation of lipid synthesis (see Section 10.8). In addition to the animal and some plant acetyl-CoA carboxylases, that from the yeast Saccharomyces cerevisiae also seems to be a multifunctional protein of about 250 kDa. It appears to be regulated by nutrient supply and coordinately with the fatty acid synthase genes (Trotter, 2001). In contrast to the above, the acetyl-CoA carboxylase from Escherichia coli contains four separate proteins: BCCP, biotin carboxylase and a heterodimer, carboxyl transferase (Cronan and Waldrop, 2002). In this regard, E. coli shares the same characteristics as the transcarboxylase from Propionibacterium shermanii (Samols et al., 1988). Early work on the structure of the component proteins of E. coli acetylTABLE 10.2
Distribution and major properties of Type I and Type II fatty acid synthases Distribution
Type I
Type II
Animals Birds Yeast Euglena gracilis (cytoplasm) Fungi Mycobacterium smegmatis Corynebacterium diphtheriae Plants Euglena gracilis (chloroplasts) Most bacteria Cyanobacteria
Major characteristics High-molecular-mass multifunctional proteins. Covalently bound ACPa. Release unesterified fatty acids (animals) or acyl-CoA (yeast).
Dissociable enzymes in a complex with ACP. Dissociable ACP. Release acyl thioesters or transfer final products directly to lipid.
a
ACP, acyl carrier protein. Source: From Harwood, J.L. and Russell, N.J. (1984) Lipids in Plants in Microbes, Allen and Unwin, Hemel Hempstead, U.K. With permission.
639
10.1
Fatty acids
TABLE 10.3
Partial reactions of fatty acid synthase (FAS) and assay principles for measuring their activity in plants
Enzyme
Method
Activities common to all FASs Acetyl-CoA:ACP transacylase [14C]Acetyl-CoA + ACP [14C]Acetyl-ACP + CoA
Precipitate acetyl-ACP Count
Malonyl-CoA:ACP transacylase [14C]Malonyl-ACP + CoA [14C]Malonyl-CoA + ACP
Precipitate malonyl-ACP Count
β-Ketoacyl-ACP synthasea Acetyl-ACP + Malonyl-ACP
Acetoacetyl-ACP + CO2 + ACP
Measure absorbance at 303 nm (acetoacetate formation)
or nAcyl-ACP + Malonyl-ACP
(n + 2)Ketoacyl-ACP + ACP + CO2
In presence of NADPHb and other FAS enzymes, measure absorbance at 340 nm With [14C]malonyl-CoA, measure counts in acyl chains Use of NaH[14C]CO2 permits a CO2 exchange assay
β-Ketoacyl-ACP reductase β-Ketoacyl-ACP + NAD(P)H
β-Hydroxyacyl-ACP + NAD(P)
Change in absorbance at 340 nm
β-Hydroxyacyl-ACP dehydrase β-Hydroxylacyl-ACP Enoyl-ACP + H2O Enoyl-ACP reductase Enoyl-ACP + NAD(P)H
Back-reaction followed with crotonyl-ACP and decrease in absorbance at 263 nm
Acyl-ACP + NAD(P)
Change in absorbance at 340 nm
Additional activities present in E. coli and higher plants Acetoacetyl-ACP synthase [14C]Acetyl-CoA + Malonyl-ACP
Reaction in presence of cerulenin Precipitate acetoacetyl-ACP and count
[14C]Acetoacetyl-ACP + CoA + CO2
β-Ketoacyl-ACP synthase II Palmitoyl-ACP + [2-14C]Malonyl-ACP ACP + ACP + CO2
[2-14C] β-ketooctadecanoyl-
Reaction products reduced and counts in acyl chain measured
β-Ketoacyl-ACP synthase I in plants. NAD(P)(H), nicotinamide adenine dinucleotide (phosphate), (reduced). Source: From Harwood, J.L. et al. (1990). a
b
TABLE 10.4
Types of fatty acid synthases in different organisms
Source
Subunit types
Subunit (mol. mass)
Native (mol. mass)
Major products
α
220–270 × 103
450–550 × 103
16:0 free acid
βα
200–270 × 103
400–550 × 103
4:0–16:0 free acids
Type I: Multicatalytic polypeptides Mammalian, avian liver Mammalian mammary gland Goose uropygial gland M. smegmatis S. cerevisiae Dinoflagellates
α α α, β α
2,4,6,8-tetramethyl10:0 16:0-, 24:0-CoA 16:0, 18:0-CoA
290,000 185,000, 180,000 180,000
2 × 106 2.3 × 106 4 × 105
Separate enzymes
–
–
16:0-, 18:0-ACP
Separate enzymes
–
–
Separate enzymes
–
–
12:0-, 14:0-, 16:0-, 18:0-ACP 16:0-, 18:1-ACP
Type II: Freely dissociable enzymes Higher plant chloroplasts E. gracilis chloroplast E. coli
Source: From Gurr, M.I., Harwood, J.L. and Frayn, K.N. (2002). Lipid Biochemistry, 5th ed., Blackwell Scientific, Oxford, U.K. With permission.
640
Lipid Metabolism
N
400
320
140
600
220
230
75
300
KS
MAT
DH
core
ER
KR
ACP
TE
C161 K326A
S581
H878
NBD-1673 S2151A NBD-1886
C
S2302A
FIGURE 10.1 Linear domain map of the animal FAS. The approximate number of residues in each domain is indicated above the map. Below are shown the locations of nucleophilic residues involved in the formation of covalent acyl-O-serine, acyl-S-cysteine, and acyl-S-phosphopantetheine intermediates, the dehydrase active-site histidine residue, and the beginning of the two glycine-rich motifs in the nucleotide-binding domains (NBD). The introduction of mutations at the critical sites indicated compromise activity of the specific domains. Residue numbering is for the rat FAS. Abbreviations: ACP, acyl carrier protein; DH, dehydrase; ER, enoyl reductase; KR, β-ketoacyl-ACP reductase; KS, β-ketoacyl-ACP synthase; MAT, malonyl acetyl transferase; TE, thioesterase. (From Smith, S., Witkowski, A. and Joshi, A.K. (2003) Prog. Lipid Res. 42, 289–317. With permission.)
The Type I fatty acid synthase of Mycobacterium smegmatis is unusual in several respects. First, the two reductases have different reduced pyridine nucleotide specificities: the β-ketoacyl-ACP reductase requires NADPH and the enoyl-ACP reductase requires NADH, whereas the reductases of other Type I synthases only use NADPH. The products of the M. smegmatis synthetase have a bimodal distribution with peaks at C16 and C24, and the overall rate of fatty acid synthesis is increased by two types of polymethylsaccharides found in the mycobacterial cell wall (Bloch, 1977). Mycobacteria also contain a Type II FAS and details of the action of both types of FAS will be found in Barry et al. (1988). The Type II synthase of E. coli was the first FAS studied. The individual proteins have all been isolated, purified, and characterised (see Volpe and Vagelos, 1973) and the genes coding for these proteins identified (White et al., 2005). The malonyl group of malonyl-CoA (produced by ACCase) is transferred to ACP by malonyl-CoA: ACP transacylase (FabD). All of the subsequent intermediates in E. coli FAS are attached to the terminyl sulfhydryl of ACP, which is one of the most abundant proteins in the bacterium. Acetoacyl-ACP is then formed by acetoacetylACP synthase (FabH), which, it should be noted, uses acetyl-CoA as the other substrate in contrast to the usual priming reaction for FAS (in Type I enzymes) (Figure 10.2). The remaining cycle of reactions continues via reduction (FabG), dehydration (FabA or FabZ), and a second reduction (FabI) to create a saturated fatty acid. Condensation then takes place with malonyl-ACP and is catalysed by FabB (β-ketoacyl-ACP synthase I) or, at later stages, with FabF (β-ketoacyl-ACP synthase II). There are two other isoforms of enoyl reductase formed in bacteria termed FabK and FabL. The FabK protein of Grampositive bacteria has an additional flavin cofactor (White et al., 2005). The E. coli FAS is also interesting in that it can produce monounsaturated (mainly cis-vaccenic) acids as well as saturated products. The ratio of the three principal products — palmitate, palmitoleate, and cis-vaccenate — is controlled by the activity of the enzymes shown in Figure 10.3. Regulation of the pathway is discussed well by Jackowski et al. (1991).
(cf. Wakil et al., 1983). In contrast, the Type I synthase of yeast is not only larger (2.4 × 106 Da) than those of animals (0.4 ×106 to 5 × 106 Da), but contains two nonidentical subunits of molecular masses 208 and 220 kDa. The native yeast synthase is, thus, an α6 β6 complex (cf. Schweizer et al., 1973). The 208 kDa α subunit of the yeast FAS, encoded by the FAS2 gene, is trifunctional and contains domains for β-ketoacyl-ACP synthase, β-ketoacyl-ACP reductase, and acyl carrier protein. The 220 kDa β subunit, encoded by the FAS1 gene, possesses acetyl-, malonyl-, and palmitoyl transferase, dehydratase, and enoyl-ACP reductase activities (Trotter, 2001). The partial reactions of the yeast Type I FAS were elucidated in a classic series of experiments by Lynen (1967). Since then we have much more information, including the mechanism of reaction termination by which the acyl chain is transferred from ACP on the α-chain of FAS to CoA (Schweizer, 1984). For a general summary, see Gurr et al. (2002). Animal FAS complexes consist of homodimers of native molecular masses of 450 to 550 kDa. The earliest attempts to generate a domain map of FAS utilised limited proteolysis. As the entire sequences of several animal FASs were deduced, a more detailed domain order emerged. The order of domains has now be confirmed by the use of mutants, compromised in specific activities and is shown in Figure 10.1 (Smith et al., 2003). Various models for animal FAS have been proposed with two identical polypeptide chains arranged antiparallel that together form two centres for palmitate synthesis at the subunit interface. By the use of modified FASs in which the activity of one of the functional domains was specifically compromised by mutations, details of the dimeric structure began to be revealed (Smith et al., 2003). In a recent update, Astrurias et al. (2005) carried out cryo-EM analysis of single FAS particles and showed that the images were of two coiled monomers in an overlapping arrangement. Only limited local rearrangements were needed for catalytic interaction among different functional domains. Monomer coiling was suggested to be useful for FAS. The above papers give important information of animal fatty acid synthase (FAS), which is summarised by Rangan and Smith (2002).
641
10.1
Fatty acids
O RCH2CHCH2C ACP
O CH3C CoA
OH Fab A Fab Z
Fab G
Acc –OOCCH C 2
CoA
O Fab D
Fab H
O RCH2CH2 = CHC ACP
O RCH2CCH2C ACP O Fab B Fab F
Fab I
O RCH2CH2CH2 C ACP
–OOCCH C ACP 2
O
FIGURE 10.2 Pathway of fatty acid synthesis in E.coli. The initial reaction is condensation between acetyl-CoA and malonyl-ACP, catalysed by FabH. After that a cycle of reduction, dehydration, and a second reduction generates a fatty acid. Later condensation utilises the gradually lengthening fatty acid and uses FabB except for the final stages when FabF is used (see text and Heath et al. (2001) for more details.) Abbreviations: Acc, acetyl-CoA carboxylase; Fab A, β-hydroxyacyl-ACP dehydratase isomerase; Fab B, βketoacyl-ACP synthase I; Fab D, malonyl-CoA: ACP transacylase; Fab F, β-ketoacyl-ACP synthase II; Fab G, β-ketoacyl-ACP reductase; Fab H, β-ketoacyl-ACP synthase III; Fab I, enoyl-ACP reductase; Fab Z, β-hydroxyacyl-ACP dehydratase. OH-10:0 β-hydroxydecanoyl-ACP 1 1 ∆2-10:1 trans-2-decenoyl-ACP
∆3-10:1 cis-3-decenoyl-ACP 2
3
∆9-16:1 palmitoleoyl-ACP
16:0 palmitoyl-ACP
4 ∆11-18:1 cis-vaccenoyl-ACP
FIGURE 10.3 Product diversification in fatty acid biosynthesis. Three main fatty acids are produced by the Escherichia coli fatty acid synthase system. The ratio of these fatty acids is controlled by the activity of three enzymes: (1) 3-hydroxydecanoyl-ACP dehydrase (encoded by the fabA gene) is a specific dehydrase that introduces the double bond into the acyl chain; (2) 3-ketoacyl-ACP synthase I (encoded by the fabB gene) catalyses an essential step in the unsaturated fatty acid elongation pathway; (3) both 3-ketoacyl-ACP synthases I and II can elongate saturated fatty acids; and (4) 3-ketoacyl-ACP synthetase II (encoded by the fabF gene) is responsible for the elongation of 9-16:1 to 11-18:1. (From Jackowski et al. (1991).)
then allows the condensation of palmitoyl-ACP with malonyl-ACP and, therefore, controls the production of stearate. These enzymes differ in their sensitivity to cerulenin and arsenite (see Harwood, 1988). A third condensing enzyme, KAS III, has recently been discovered and, like its counterpart in E. coli, is responsible for acetoacetyl-ACP formation (Jaworski et al., 1989; Walsh et al., 1990). Although it is believed generally that de novo synthesis of fatty acids in plants is concentrated in plastids (see Harwood, 1988), a report of acyl-ACP in plant mitochondria (Chuman and Brody, 1989), together with fatty acid synthesis mediated by ACP in Neurospora crassa mitochondria (Mikolajczyk and Brody, 1990), raise the possibility for some activity in this organelle. Much attention has been paid to plant ACP. Earlier work on its purification and properties has been well reviewed (Ohlrogge, 1987). ACP is synthesized in the cytosol and posttranslationally imported into plastids, where
The fatty acid synthase of Brevibacterium ammoniagenes has a high molecular mass (like Type I synthases), but produces both saturated and unsaturated acids (like E. coli type II synthetase). In most Gram-positive and some Gram-negative bacteria, branched-chain acids are made by Type II synthases that use short-chain branched acyl-CoA primers instead of acetyl-CoA and so produce iso or anteiso fatty acid products (Kaneda, 1977). The topic of lipid synthesis (and, especially, fatty acid synthesis) in relation to existing and new antibacterials has been reviewed by Heath et al. (2001, 2002) who give a summary of fatty acid synthesis in E. coli with reference to some other bacteria. Plants also contain a Type II FAS. The enzymology of the complex has been reviewed (Stumpf, 1987; Harwood, 1988, 2005). Like E. coli, there are several condensing enzymes. β-Ketoacyl-ACP synthase I (KAS I) is responsible for forming keto acids of up to 16 carbons. KAS II
642
Lipid Metabolism
Formation of the very long-chain saturated fatty acids involved in the surface coverings of plants is subject to inhibition by thiocarbamate herbicides (Harwood, 1990), probably as their sulfoxide metabolites (Abulnaja and Harwood, 1991; see Section 11.8). Plant fatty acid elongation was updated by Harwood (1996) and summarised in Schmid and Ohlrogge (2002). Elongation products in yeast have been extensively studied and three separate elongase genes ELO1, ELO2, and ELO3 identified in Saccharomyces cerevisiae. All three genes are for condensing enzymes in the elongation systems (Dittrich et al., 1998). More recently, progress has been made in identifying other components. The yeast systems elongate saturated and monounsaturated fatty acids with different chain-length specificities (Leonard et al., 2004). Although there are some animal elongases, which have substrate preferences for saturated or monounsaturated fatty acids, many of them are used to elongate polyunsaturated fatty acids (PUFAs). This is not surprising because most dietary PUFA are 18C molecules, whereas most biologically effective metabolites are 20 or 22C (see Sections 10.1.8 and 11.1). Very long-chain PUFAs are particularly important components of mammalian brain (especially 22:6n-3) and it is not surprising that some of the first studies identified different elongases in this tissue (Goldberg et al., 1973; Bourre et al., 1975). Details of the ELOV (very long-chain elongase) genes from humans, rats, and mice are given in Leonard et al., (2004). Like the plant enzymes, the mammalian elongases are microsomally located (endoplasmic reticulum) and use malonyl-CoA in four steps (like FAS) to produce products 2C longer (see Cook and McMaster, 2002; Leonard et al., 2004). Fatty acids can also be elongated in mammals by a mitochondrial system, which uses acetyl-CoA as the unit for C2 addition and NADH for reduction. In general, the mitochondrial system elongates fatty acids in the range C10 to C14, whereas the microsomal system uses C16 and longer acids. At first it was thought that the mitochondrial elongation system could operate by a reversal of β oxidation, but this is now not considered thermodynamically feasible. Indeed, the (flavin adenine dinucleotide) FAD-dependent acyl-CoA dehydrogenase of oxidation is substituted by a more thermodynamically favorable enzyme, enoyl-CoA reductase, which is rate-limiting for the overall process in mitochondria (Cook and McMaster, 2002). In contrast to animals and plants, bacteria do not usually contain significant amounts of acids longer than C18. A notable exception is the long-chain (up to C56) fatty acids found in mycobacteria. These are formed by an elongation system using acetyl-CoA and NADH, which may be a reversal of β oxidation (Harwood and Russell, 1984). Fatty acid elongation systems in lower eurkaryotes (the nematode Caenorhabditis elegans, fungi, microalgae, moss) are described by Leonard et al. (2004).
proteolytic processing to the mature protein takes place. The attachment of the prosthetic group needs holo-ACP synthetase, which is cytosolic (El Hussein et al., 1988). Further aspects of the biochemistry of plant ACP are covered by Ohlrogge et al. (1991) and by Slabas and Fawcett (1992). All of the proteins catalysing the partial reactions of plant FAS have been purified and, in many cases, genes coding for them have been identified and sequenced. Most of these proteins occur as isoforms (see Harwood, 1988). Updates on the genetics of fatty acid synthesis (Ohlrogge et al., 1991) and its molecular biology (Slabas and Fawcett, 1992) have been published. Fatty acid synthesis in plants has been summarised by Schmid and Ohlrogge (2002) and reviewed in some detail by Harwood (2005). Regulation of fatty acid synthesis in plants has been discussed by Harwood (1996), Ohlrogge and Jaworski (1997), and, more recently, by Harwood (2005). The regulation of fatty acid synthesis in animals is described by Rangan and Smith (2002). Euglena gracilis is an interesting organism because, when it grows heterotrophically (when it is “animallike”), it contains a Type I synthase. When it is grown photoautotrophically, it contains, in addition to its cytoplasmic Type I enzyme, a Type II synthase in its chloroplasts (Ernst-Fonberg and Bloch, 1971). See also Worsham et al. (1988).
10.1.3 Elongation systems As mentioned in Section 10.1.2, the elongation of fatty acids is generally carried out by Type III fatty acid synthases, which use malonyl-CoA as the source of C2 units for addition. In plants, the elongation of stearate to form very longchain saturated fatty acids (which are precursors of the various components of waxes, cutin, and suberin; see Kolattukudy, 1980, 1987) takes place via several chain length-specific systems (Walker and Harwood, 1986). Malonyl-ACP and NADPH are required (see Harwood, 1988, 1996) and a cycle of condensation, reduction, dehydration, and second reduction reactions using acyl-CoAs are established. Proteins catalysing these partial reactions have been demonstrated in leek (Lessire et al., 1989) and Lunaria annua (Fehling et al., 1992). In the latter case, monounsaturated acyl-CoAs are also substrates, and the elongation of such acids is important in brassicas, such as rapeseed, where erucate (∆13-22:1) is a major component of the seed oil (Section 2.2). The genetics of the rape elongation system are well described by Ohlrogge et al., (1991). Elongation in jojoba (which accumulates lipid as wax esters) uses a system with oleoyl-CoA and malonyl-CoA as substrates, and this plant has been studied in some detail (Pollard and Stumpf, 1980).
643
10.1
Fatty acids
10.1.4
cis double bond and results in the synthesis of a series of monoenoic fatty acids (Figure 10.4). Details of the enzymes and the position of the genes involved in their synthesis are summarised in Heath et al. (2002). The ratio of the three main products of E. coli FAS is controlled by the relative activity of three enzymes, as illustrated in Figure 10.3. Aerobic desaturation involves the stereospecific removal of two hydrogen atoms from an acyl chain. Along with reducing equivalents from NAD(P)H or other reductants, the hydrogen atoms are used to reduce molecular oxygen to water. The desaturases are membrane-bound multienzyme complexes, with the exception of the stearoyl-ACP desaturase found in chloroplasts. The basic details of aerobic desaturation appear to be the same in all organisms in that oxygen and a reduced cofactor are necessary, although the nature of the carriers varies in different systems (Figure 10.5) as well as their susceptibility to cyanide
Desaturases
Unsaturated fatty acids can be produced by anaerobic or aerobic pathways — the latter being the most usual mechanism. Aerobic desaturases have been studied in a large number of cases, although successful purifications have only been made a few times. A general summary is found in Gurr et al. (2002). The anaerobic method, which is used by many members of the Eubacteriales, including all the anaerobes as well as some aerobes and facultative aerobes, has already been summarised in Section 10.1.2. It relies on the activity of the β-hydroxy-decanoyl-ACP β, γ-dehydrase whose isomerase activity allows the synthesis of a cis-3-decenoyl-ACP from the usual trans-2-decenoyl-ACP. cis-3-Decenoyl-ACP cannot be reduced by the enoyl-ACP reductase, but can be condensed by the β-keto-acyl-ACP synthase. This reaction, therefore, allows chain lengthening, but preserves the n-7 Acetyl-ACP
β-Ketoacyl-ACP synthetase I (fabB) or II (fabF )
β-Hydroxydecanoyl-ACP β, γ-Dehydrase trans-2-Decenoyl-ACP
(fabA)
cis-3-Decenoyl-ACP β-Ketoacyl-ACP synthetase I (fabB)
β-Ketoacyl-ACP synthetase I (fabB) or II (fabF)
∆9-16:1 β-Ketoacyl-ACP synthetase II (fabF, Cvc−)
16:0 18:0
∆11-18:1
Saturated fatty acids
Unsaturated fatty acids
FIGURE 10.4 Anaerobic pathway of fatty acid biosynthesis in bacteria showing mutants of Escherichia coli. Here fab refers to deficient mutants described for E.coli. (From Harwood, J.L. and Russell, N.J. (1984) Lipids in Plants in Microbes, Allen and Unwin, Hemel Hempstead, U.K. With permission.) O (R1)CH Electron donor (reduced) Electron donor (oxidised)∗
Oxidised cytochrome Oxidoreductase protein Reduced cytochrome∗∗
Cyanide sensitive protein
CH(R2)C X 2H2O
O2 O (R1)CH2CH2(R2)C X∗∗∗
FIGURE 10.5 A generalized scheme for aerobic fatty acid desaturation: *e.g., NADH, NADPH, reduced ferredoxin; **e.g., cytochrome b5; ***e.g., acyl-ACP (stearoyl-ACP ∆9-desaturase in plants); acyl-CoA (stearoyl-CoA ∆9-desaturase in animals); oleoylphosphatidylcholine (∆12-desaturase in yeast or plants); linoleoyl-monogalactosyldiacylglycerol (∆15-desaturase in plant chloroplasts). (From Gurr, M.I., Harwood, J.L. and Frayn, K.N. (2002). Lipid Biochemistry, 5th ed., Blackwell Scientific, Oxford, U.K. With permission.)
644
Lipid Metabolism
noted that the “anaerobic” pathway utilizing fatty acid synthase is characteristic of many bacteria, particularly those of the orders Pseudomonadales and Eubacteriales. It is now established that several species of Gramnegative bacteria, mainly marine, can produce PUFAs (Tocher et al., 1998). Attention has been particularly focused on Shewanella spp. (Metz et al., 2001) where a polyketide synthase seems to be used (see also Valentine and Valentine, 2004). Most mammalian tissues can modify acyl chain composition by introducing more than one bond. Like ∆9-desaturation, further desaturation requires molecular oxygen and an associated electron transport system. Animal systems cannot generally insert double bonds beyond the ∆9 position. Consequently, double bonds are inserted at the ∆6, ∆5, and ∆4 positions. However, synthesis of linoleate from oleoyl-CoA has been demonstrated in some insects (de Ronobales et al., 1987). Protozoa, such as Tetrahymena (Umeka and Nozawa, 1984) and Acanthamoeba castellanii, also contain desaturases capable of producing linoleate. The latter system has been studied in some detail and shown to take place on phosphatidylcholine and probably is an n-6 (rather than a ∆12-) desaturase (Jones et al., 1993; Rutter et al., 2002). Interestingly, the Acanthamoeba n-6 desaturase is induced by low temperature (Avery et al., 1995) and, independently, by oxygen changes (Thomas et al., 1998). Polyunsaturated fatty acids in all organisms usually contain methylene-interrupted double bonds and conjugated systems are rare. However, so-called conjugases (that produce conjugated PUFAs) have been found in a number of plants and their genes have been cloned (e.g., Cahoon et al., 1999; Iwabuchi et al., 2003). The animal desaturases are also used to further modify polyunsaturated fatty acids from the diet (e.g., linoleate, α-linolenate), and the major pathways are depicted in Figure 10.6. Cook and McMaster (2002) have summarised the metabolism of acids of the n-6 family (linoleate, etc.) and the n-3 family (α-linolenate) and competition between the pathways. In contrast to the stearate desaturases discussed above, the enzymes forming linoleate and α-linolenate in lower and higher plants use complex lipids as substrates. The use of phosphatidylcholine in this way was proposed originally following experiments with the yeast Candida albicans (Pugh and Kates, 1975) and with the green alga Chlorella vulgaris (Nichols et al., 1967). However, some yeasts contain both oleoyl-CoA and oleoyl-phospholipid desaturases (Pugh and Kates, 1975). More recent work covering the ∆9, ∆12, and ∆6-desaturases from various yeasts or fungi is summarised by Tocher et al. (1998). In the fungus, Mortierella alpina, a number of desaturases including a ∆5 enzyme must be present because this organism can produce arachidonate. Other species can also produce 20:5n-3 (EPA) and, therefore, must have a ∆15 (n-3) desaturase as well (Tocher
inhibition. There is also some controversy as to whether the two hydrogen atoms are removed sequentially (with the involvement of a hydroxy intermediate) or in a concerted mechanism. Evidence with Corynebacterium diphtheria using tritiated substrates suggested a stepwise mechanism (Schroepfer and Bloch, 1965), whereas experiments with 2H-Iabelled substrates and Chlorella vulgaris indicated concerted hydrogen removal (Morris, 1970). The first double bond inserted into an acyl chain is usually ∆9. The stearoyl-CoA ∆9-desaturase was first purified from rat liver (Strittmatter et al., 1974). The complex consists of three major proteins: NADH-cytochrome b5 reductase, cytochrome b5, and a terminal desaturase (or cyanide-sensitive) protein. Usually the activity of the terminal desaturase limits the overall speed of desaturation. The NADH-cytochrome b5 oxidoreductase is a flavoprotein of mass 43 kDa, cytochrome b5 has a molecular mass of 16.7 kDa, and the terminal desaturase is 53 kDa (see Cook, 1991 ). The approximate location of the three components in the endoplasmic reticulum has been deduced from their amino acid sequence and the use of specific chemical reagents. A useful summary of fatty acid desaturation in animals is given in Cook and McMaster (2002). The ∆9-desaturases from plants and algae (and Euglena gracilis) use stearoyl-ACP (the final product of their FASs) as substrate. The enzyme was first purified substantially from developing safflower seeds by McKeon and Stumpf (1982). The 9-(stearoyl) desaturase has been cloned from castor seeds (Knutzon et al., 1991) and from safflower (Thompson et al., 1991). The cDNA from safflower includes a 33 amino acid transit peptide. Modulation of the steroyl-ACP desaturase level, using antisense technology in transgenic rapeseed, resulted in a marked decrease in oleate with a commensurate increase in stearate (Kridl et al., 1991). For summaries of the stearoyl-ACP desaturases from plants, refer to Harwood (1996). Mutations affecting the levels of oleic acid in crop plants are discussed in Schmid and Ohlrogge (2002). In contrast to animals and plants, bacteria are unique in producing ∆10-monoenoic fatty acids. Bacilli commonly possess ∆5- or ∆10-desaturases, and Bacillus licheniformis contains both enzymes when it grows at low temperatures (Fulco, 1974). Under these conditions, it synthesizes small amounts of ∆5,10-16:2. It will be noted that, although this acid is polyunsaturated, it does not possess the usual methylene-interrupted structure. In fact, a general distinction is often made between bacteria and other organisms in that the former are unable to synthesize methylene-interrupted polyunsaturated fatty acids. This is not always true. For example, the filamentous gliding bacteria (Flexibacter spp.) contain considerable amounts of (n-3)20:5, which in some species are the major fatty acids (Johns and Perry, 1977). Morever, marine bacteria may be able to produce methylene-interrupted PUFAs (see below). Other features of aerobic desaturation in bacteria are discussed by Schweizer (1989). However, it should be 645
10.1
Fatty acids
Diet
Linolenic acid
Endogenous synthesis Oleic acid
18:1(n - 9) ∆9
20:1(n - 9) ∆11
Linoleic acid
18:2(n - 6) ∆9, 12
∆6 18:2(n - 9) ∆6, 9
20:2(n - 6) ∆11, 14
∆5
20:2(n - 9) ∆8, 11
18:3(n - 3) ∆6 ∆9, 12, 15
18:4(n - 3) ∆6, 9, 12, 15
∆6 18:3(n - 6) ∆6, 9, 12
20:3(n - 6) ∆5 ∆8, 11, 14
20:3(n - 9) ∆5, 8, 11
20:4(n - 3) ∆5 ∆8, 11, 14, 17
20:5(n - 3) ∆5, 8, 11, 14, 17
20:4(n - 6) ∆5, 8, 11, 14 22:5(n - 3) ∆7, 10, 13, 16, 19
22:1(n - 9) ∆13
24:1(n - 9) ∆15
22:6(n - 3) ∆4, 7, 10, 13, 16, 19
Desaturation
β-oxidation 24:5(n - 3) ∆6 24:6(n - 3) ∆9, 12, 15, 18, 21 ∆6, 9, 12, 15, 18, 21
2-Carbon chain elongation
22:4(n - 6) ∆7, 10, 13, 16
22:5(n - 6) ∆4, 7, 10, 13, 16 β-oxidation
24:4(n - 6) ∆6 24:5(n - 6) ∆9, 12, 15, 18 ∆6, 9, 12, 15, 18
34:6(n - 3) ∆16, 19, 22, 25, 28, 31
34:4(n - 6) ∆19, 22, 25, 28
FIGURE 10.6 Major pathways for polyunsaturated fatty acid synthesis in animals. Note the alternating sequence of desaturation in the horizontal direction and chain elongation in the vertical direction in the formation of polyunsaturated fatty acids from dietary essential fatty acids. Type size for individual fatty acids reflects, in a general way, relative accumulation in tissues. (From Gurr, M.I,. Harwood, J.L. and Frayn, K.N. (2002). Lipid Biochemistry, 5th ed., Blackwell Scientific, Oxford, U.K. With permission.)
and by Browse and Somerville (1991). The use of complex lipids for substrates, the cooperation of extraplastidic with plastid pathways and the site of Arabidopsis mutations in the pathways are shown in Figure 10.7. A recent update of plant fatty acid desaturation is by Hildebrand et al. (2005). The fatty acid desaturation systems in cyanobacteria have been studied in detail by Murata and colleagues. This work was particularly in relation to chilling sensitivity. The research has been summarized in Murata and Wada (1995). The possible roles of signaling pathways in desaturase induction are reviewed by Mikami and Murata (2003).
et al.,1998). Genes isolated from Mortierella, as well as from various algae, have proven useful in the genetic manipulation of higher plants to produce very long-chain PUFAs (see Section 11.8). In plants, desaturation of oleate to linoleate uses phosphatidylcholine (in the endoplasmic reticulum) as the major substrate. The enzyme in safflower has been shown to use cytochrome b5 as an electron source (Kearns et al., 1991; Smith et al., 1992). However, oleate desaturation in chloroplasts uses ferredoxin (Schmidt and Heinz, 1991). Other complex lipid substrates probably are also used (see Harwood. 1988), especially as demonstrated for chloroplast lipids (see Jones and Harwood, 1980). Further desaturation of linoleate to α-linolenate can take place either using monogalactosyldiacylglycerol (Jones and Harwood, 1980) or with phosphatidylcholine. The relative importance of these two substrates depends on the type of tissue: leaves use monogalactosyldiacylglycerol mainly, but seeds (which have poorly developed plastids) use the nonchloroplastic phosphatidylcholine as the main substrate. These aspects are discussed by Harwood (1996)
10.1.5 Hydroxylation Hydroxy fatty acids are formed as intermediates in various metabolic sequences (e.g., fatty acid biosynthesis, β-oxidation) as a result of specific hydroxylation reactions, and following other activities, such as those of lipoxygenase (see Section 10.1.7). The hydroxyl group is usually introduced close to one end of the acyl chain 646
Lipid Metabolism
Endoplasmic reticulum PI, PG
CDP·DAG
PE fad 2
PA 18:1 18:1 (16:0)
16:0·CoA 18:1·CoA G3P LPA
PA 18:1 16:0 PG
MGD
act 1 18:1 16:0 18:1 16:0
DAG 18:1 18:1 (16:0)
PC 18:1 18:1 (16:0)
DAG 18:1 16:0 DGD
fad 3
PC 18:2 18:2 (16:0)
PC 18:3 18:3 (16:0)
DAG 18:2 18:2 (16:0) SL
18:1 16:0 18:1 16:0
MGD 18:2 18:2
DGD 18:2 18:2 (16:0)
SL 16:0 18:2
fad A 18:1·ACP
fad B
G3P
18:1 t16:1 18:1 16:1 (16:0) 16:0·ACP fad C 18:0·ACP
fad C
18:2 t16:1 18:2 16:2 (16:0)
18:2 16:0 18:2 16:0
18:3 t16:1 18:3 16:3 (16:0)
18:3 16:0 18:3 16:0
Fatty acid fad D synthase
fad D
fad D 18:3 18:3 18:3 18:3 (16:0)
16:0 18:3
PLASTID
FIGURE 10.7 An abbreviated diagram of the two-pathway scheme of glycerolipid biosynthesis in the 16:3 plant Arabidopsis. Widths of the lines show the relative fluxes through different reactions. (From Browse and Somerville, 1991.)
(e.g., α,β) and less commonly in the middle of the chain. A good example of the latter in plants is the formation of ricinoleic acid (the major acid of castor oil, Section 2.2.2 and Section 9.8) by hydroxylation of oleoyl-phosphatidylcholine substrate using NADH and oxygen as co-factors (Moreau and Stumpf, 1981; Smith et al., 1992). Few of the hydroxylases have been characterised, so it is difficult to compare their properties. However, they appear to be mixed-function oxidases and require molecular oxygen and a reduced pyridine nucleotide. Most of the enzymes appear to require cytochrome P450, although ferredoxin (or a related haem protein) may substitute in some bacterial or plant systems (see Gaillard, 1980). Cellfree hydroxylating systems have been studied extensively in Bacillus megaterium and Pseudomonas oleovorans and some enzymes have been purified (see Schweizer, 1989). α-Oxidation systems producing 2-hydroxy fatty acids have been demonstrated in yeasts, bacteria, plants (see Harwood and Russell, 1984), and animals, while ω-oxidation systems introduce a hydroxyl group to the methyl end of the acyl chain. These oxidations are described more fully in Section 10.1.6. The major hydroxy fatty acids in plants have an ω-OH and an in-chain OH group (e.g., 10,16-dihydroxypalmitic acid). Their synthesis seems to involve ω-hydroxylation with NADPH and O2 as cofactors, followed by in-chain hydroxylation using the same co-substrates. If the
precursor is oleic acid, then the double bond is converted to an epoxide, which is then hydrated to yield 9,10-hydroxy groups. These conversions involve CoA esters. In-chain plant hydroxylase is sensitive to inhibition by O-phenanthroline and by CO in a reaction that is reversed by 420 to 460 nm light (Kolattukudy, 1980, 1987). It should be noted that a few hydroxylations occur without a mixed-function oxidase enzyme. For example, the ergot fungus Claviceps purpurea forms 12-hydroxyoleic (ricinoleic) acid by hydration of linoleic acid (Harwood and Russell, 1984).
10.1.6
Oxidation of fatty acids
Oxidation of fatty acids can occur in a number of ways, depending on the position of oxidation (e.g., α-, β-, or ω-oxidation) and the nature of the substrate (e.g., lipoxygenase attack on polyenoic fatty acids). 10.1.6.1
α-Oxidation
The removal of a single carbon atom from the carbonyl end of a fatty acid is carried out by α-oxidation. This process is particularly active in plants, but is also found in mammals (notably brain tissue) and bacteria. The removal of a single carbon may be important when degradation of a fatty acid by β-oxidation (see below) is blocked by the presence of a methyl branch at position 3, such as in phytanic acid. Defects in α-oxidation of 647
10.1
Fatty acids
CH3 RCH2CH
CH3 CH2COOH
α-oxidation (4 steps)
Phytanic acid
RCH2CH
CH3 CO
SCoA
β-oxidation
Pristanoyl-CoA RCO
FIGURE 10.8
(acyl-CoA synthetase)
Pristanic acid
CH3 RCH2CH
COOH
RCOCHCO (Thiolase)
SCoA
SCoA + CH3CH2CO
SCoA
Phytanic acid metabolism.
(2H) RCH2COOH
O2
[RCH(OOH)COOH]
H2O RCH(OH)COOH
etc. RCOOH NADH
CO2 + H2O RCHO NAD+
FIGURE 10.9
The α-oxidation of fatty acids in plants. Adapted from Galliard (1980).
prostaglandin endoperoxide H synthase (see Section 10.1.8) (Graham and Eastmond, 2002). This has led to the suggestion that a major role for α-oxidation in plants is in defence responses to pathogens (Hamberg et al., 1999). The occurrence of bacterial fatty acid α-oxidation is not firmly established and, if present, represents a pathway of minor importance (Finnerty, 1989). It has been studied in few bacterial species. However, in E.coli, the D-2-hydroxy fatty acid is preferentially decarboxylated, unlike in plants and animals where it is the L form that is metabolised preferentially. Thus, the 2-hydroxy acids tend not to accumulate in bacteria (Lekakis, 1977).
phytanic acid give rise to Refsum’s disease and there are other disorders giving rise to defects in α-oxidation (Mukherji et al., 2003). Human metabolism of phytanic and pristanic acids is reviewed by Verhoevan and Jakobs (2001) (see Section 11.6) and is shown in simplified form in Figure 10.8. α-Oxidation is also important because α-hydroxy fatty acids are intermediates in the process and these acids are components of certain sphingolipids (Bowen et al., 1974). Thus, brain cerebroside fatty acids are highly enriched in cerebronic acid (α-OH 24:0). Furthermore, tissues that accumulate large amounts of fatty acids almost invariably also contain significant amounts of odd chain-length fatty acids (e.g., see Mcllwain, 1966). In plants the mechanism of α-oxidation in leaves and seeds is identical. The fatty acid substrate is nonesterified (cf. β-oxidation) and is usually C12 to C18. It is attacked by molecular oxygen to generate an unstable 2-hydroperoxy intermediate (Figure 10.9), which decomposes to an aldehyde with release of carbon dioxide. Under certain conditions (e.g.,, in the presence of an enzyme, such as glutathione peroxidase, which will reduce peroxides), then a D-2-hydroxy fatty acid may be produced, which cannot be metabolized easily. Under normal conditions, though, the aldehyde is oxidized in the presence of a source of reducing power (pyridine nucleotide or flavoprotein depending on the tissue) to give a fatty acid one carbon atom shorter than the original fatty acid. A significant breakthrough to our understanding of α-oxidation in plants has come from the discovery of a pathogeninducible oxygenase, which has significant homology to
10.1.6.2
b-Oxidation
The β-oxidation of fatty acids was the first metabolic process in which labeled compounds were used for its investigation. Knoop’s classic experiments at the turn of the century were later confirmed and extended by others to reveal the details of the process (for references, see Greville and Tubbs, 1968; Wakil, 1970; Kunau et al., 1995; Schulz, 2002). Key steps in the pathway are the activation of the fatty acid to a coenzyme A thioester, the α,β-dehydrogenation of the acyl-CoA, the hydration of the resultant double bond, oxidation of the β-hydroxyacylCoA, and thiolytic cleavage of the β-ketoacyl-CoA (Figure 10.10). The mechanism of fatty acid uptake by animal cells has not been fully elucidated. Plasma membrane carrier proteins responsible for saturable high-affinity uptake have been identified, but some workers argue that spontaneous and nonspecific diffusion of fatty acids across the 648
Lipid Metabolism
R CH2 CH2 COOH Cell membrane R CH2 CH2 COOH FABP?
CYTOSOL
R CH2 CH2 COOH ATP AS
CoASH
CoASH
AMP + PPi
Outer mitochondrial membrane
CPTI
R CH2 CH2 COSCoA Carnitine
R CH2 CH2 CO Carnitine Inner mitochondrial membrane
T Carnitine R CH2 CH2 CO Carnitine CoASH
R
CH
AD
CH COSCoA H2O
CPTII
R CH2 CH2 COSCoA R COSCoA CH3 COSCoA
EH
KT CoASH
R CH OH
CH2 COSCoA
HAD
NAD+
R C CH2 COSCoA O
MATRIX
NADH + H+
FIGURE 10.10 Pathway of mitochondrial fatty acid oxidation. Enzymes of the pathway are AS, acyl-CoA synthetase; CPT I, carnitine palmitoyltransferase I; T, carnitine:acylcarnitine translocase; CPT II, carnitine palmitoyltransferase II; AD, acyl-CoA dehydrogenase; EH, enoyl-CoA hydratase; HAD, L-3-hydroxyacyl-CoA dehydrogenase; KT, 3-ketoacyl-CoA thiolase. Other abbreviation: FABP, fatty acid binding protein. (From Schulz, 1991b.)
Three enzymes (differing in chain-length specificity) that catalyse α,β-dehydrogenation of acyl-CoAs have also been purified. The overlapping specificities of these enzymes allow efficient oxidation of fatty acids in the C4 to C20 range (Greville and Tubbs, 1968). The acyl-CoA dehydrogenases (EC 1.3.99.3) are flavoproteins containing FAD (flavin adenine dinucleotide) as a prosthetic group. They have molecular masses of 170 to 190 kDa and are composed of four identical subunits, each of which carries a noncovalently bound FAD. Nucleotide sequences for several rat liver and human acyl-CoA dehydrogenases have been determined, with a greater than 90% homology observed for the same enzyme from rat compared to human (see Schulz, 2002). Hydration of the α,β-trans double bond is catalysed by enoyl-CoA hydratase (EC 4.2.1.17; crotonase). This enzyme will also attack cis double bonds and 3-enoyl-CoAs. An L-OH fatty acid is produced (∆4, in the latter case, and ∆3, in β-oxidation) from the trans isomers. For more details, see Schulz (2002). The L-3-hydroxyacyl-CoA obtained in the above step is oxidized by NAD+ in the presence of 3-hydroxyacyl-CoA dehydrogenase (EC 1.1.1.35). The enzyme is specific for the L form and usually has little chain-length specificity (Wakil, 1970). The enzyme has been purified from pig
plasma membrane is also important (see Schulz, 2002). β-Oxidation is preceded by activation of fatty acids to their coenzyme A (CoA) thioesters. A group of acyl-CoA synthetases have been identified that differ in their chainlength specificities, but which catalyse the same type of reaction: R-COOH + ATP + CoASH
RCO-SCoA + AMP + PPi
The individual acyl-CoA synthetases are grouped as short-chain, medium-chain, or long-chain synthetases, depending on their substrate specificities. Short-chain synthetases, which act on acetate or propionate, are present in the mitochondria of most animal tissues, but, in addition, are found in the cytosol of lipogenic tissues, such as liver, intestine, adipose tissue, and mammary gland. Medium-chain synthetases are also located in the mitochondrial matrix. By contrast, long-chain acyl-CoA svnthetases in animals are membrane-bound and associated with the endoplasmic reticulum, mitochondrial outer membrane or peroxisomes (see Schulz, 2002). Acyl-CoAs that need to be taken up by mitochondria do so via a carnitine-dependent mechanism. Two carnitine palmitoyltransferases and a carnitine:acylcarnitine translocase are needed, as illustrated in Figure 10.10. 649
10.1
Fatty acids
SCoA 13 12
10 9
O Three cycles of β oxidation SCoA
7
6
4
3
O Enoyl-CoA isomerase O
3
SCoA 2 One cycle of β oxidation
7 6 O
SCoA
5 4
Acyl-CoA dehydrogenase O 3 SCoA
2
5 4
2, 4-Dienoyl-CoA reductase
4
SCoA 3 O Enoyl-CoA isomerase 2 3
SCoA O Four cycles of β oxidation SCoA O
FIGURE 10.11
The β-oxidation of linoleoyl-CoA. (From Schulz, 1991b.)
Oxidation of saturated straight-chain fatty acids proceeds smoothly using the above series of enzymes, but there are problems in the breakdown of other fatty acids. For example, odd-chain acids finally liberate propionyl-CoA, which is converted to succinyl-CoA for further metabolism. Unsaturated fatty acids must also be modified because their double bonds may be in the wrong position and because of incorrect configuration for normal catabolism. Epimerase and isomerase enzymes are present to cope with these problems (see Gurr et al., 2002 and Schulz, 2002; Figure 10.11). Apart from β-oxidation in mitochondria, peroxisomes are also important. The reactions and enzymes involved have been characterised by Hashimoto (1990), with the research being aided by the use of drugs, such as clofibrate or di(2-ethyl-hexyl)phthalate, which cause peroxisome proliferation and also induce the enzymes of peroxisomal β-oxidation (Lazarow and Moser, 1989). Although the intermediates of peroxisomal β-oxidation are the same as in mitochondria, there are differences in the enzymes concerned. The first reaction is catalysed by acyl-CoA oxidase, which generates hydrogen peroxide as a product.
heart and rat liver. It has a molecular mass of 65 kDa, is composed of two identical subunits, and its conformation at 2.8 Å resolution has been determined (Schulz, 1991). Some enzymes will utilize NADP+ more slowly and these dehydrogenases will also oxidize S-3-hydroxyacyl-Nacylthioethanolamine and S-3-hydroxyacyl-hydrolipoate. Thiolytic cleavage of the 3-ketoacyl-CoA is catalysed by acetyl-CoA acyltransferase (EC 2.3.1.16), which liberates acetyl-CoA and an acyl-CoA two carbons shorter than the original substrate (see Figure 10.10). The above enzyme has a broad chain-length specificity, but a second enzyme, acetyl-CoA acetyltransferase (EC 2.3.1.9), catalyses the same reaction with acetoacetyl-CoA as substrate. Kinetic and exchange studies indicate that these thiolase reactions occur in two stages (see Greville and Tubbs, 1968). Both types of thiolases are homotetramers of mass about 170 kDa. While the 3-ketoacyl-CoA thiolase with its broad chain-length specificity is essential for β-oxidation, the acetoacetyl-CoA thiolase probably functions mainly in ketone body metabolism (Schulz, 1991). All mitochondrial enzymes of β-oxidation are synthesised in the cytosol on free polysomes (see Ozasa et al., 1984). 650
Lipid Metabolism
In plants, it was observed a number of years ago that peroxisomes, such as the glyoxysomes of germinating oil seeds, contained enzymes necessary for fatty acid β-oxidation (see Galliard, 1980; Kindl, 1987). Early aspects of the work were discussed by Beevers (1980). A thorough review has covered aspects of β-oxidation in plants (Gerhardt, 1992). The author concluded that there was little evidence for any other location for β-oxidation apart from in peroxisomes. Nevertheless, the possibility that mitochondria could be involved for the oxidation of straight chain acids is still not resolved and these organelles seem to be utilized for β-oxidation of short branched-chain 2-oxo acids (Graham and Eastmond, 2002). The latter authors have made a thorough review of peroxisomal β-oxidation in plants and its regulation.
The latter is rapidly destroyed by peroxisomal catalase activity. The hydration and second dehydrogenation steps are catalysed by a trifunctional enzyme. The essential features are illustrated in Figure 10.12. Because the acylCoA oxidase is almost inactive towards octanoyl-CoA (or shorter chains), peroxisomes are incapable of complete oxidation of fatty acids. For the oxidation of unsaturated fatty acids, such as linoleate, a 2,4-dienyl-CoA reductase is present and a ∆3-cis, ∆2-trans-enoyl-CoA isomerase (as the third activity of the trifunctional enzyme) is also needed (see Schulz, 1991; Osmundsen et al., 1991). The main function of peroxisomal β-oxidation in animals seems to be in the chain shortening of very long-chain fatty acids, prostaglandins, dicarboxylic acids, and various xenobiotic compounds (see Schulz, 1991). AMP + PP
(HOOC) CH3(CH2CH2)nCH2CH2COOH ATP + CoASH
(HOOC) CH3(CH2CH2)nCH2CH2CO H2O
CoA
S
O2 Acyl-CoA oxidase
Catalase H2O2
(HOOC) CH3(CH2CH2)nCH CHCO H2 O
S
CoA
Bi(tri)functional enzyme
(HOOC) CH3(CH2CH2)nCHCH2CO
S
CoA
OH NAD+
RH2 R
Bi(tri)functional enzyme
NADH
(HOOC) CH3(CH2CH2)nCCH2CO
S
CoA
O CoA
Thiolase
(HOOC) CH3(CH2CH2)nCO Chain-shortened acyl-CoA
S
CoA + CH3CO
S
CoA
Acylcarnitine/FFA/acyl-CoA Acetylcarnitine/acetate/acetyl-CoA
FIGURE 10.12 Essential features of peroxisomal βoxidation. A schematic representation of the main metabolic features of peroxisomal oxidation. The enzymes of peroxisomal βoxidation are contained within the peroxisomal membrane (rectangular box). Possible end products of βoxidation are indicated at the bottom. The [HOOC] to the left of the substrate acyl-CoA ester indicates that monoCoA esters of dicarboxylic acids are also substrates for peroxisomal βoxidation. An incorrect stoichiometry of catalase-dependent decomposition of H2O2 is used for the sake of brevity. The arrows pointing down to Acylcarnitine/FFA/acyl-CoA, or to Acetylcarnitine/ acetate/acetyl-CoA are meant to indicate possible alternative forms of export of chain-shortened fatty acids out of the peroxisome (FFA = free fatty acid). The oxidant (R) of NADH active in vivo remains to be established. With isolated peroxisomes, pyruvate (and lactate dehydrogenase) can function in this capacity. (From Osmundsen et al., 1991.)
651
10.1
Fatty acids
O H3CCH2CSCoA
O
O CH2
CHCSCoA
HOCH2CH2CSCoA
O
O CoASCCH3 + CO2
FIGURE 10.13
HOCH2CH2COH
O
O
CoASCCH2COH
O
HC CH2COH
Propionate metabolism in plants.
Plants operate a different mechanism from animals when removing the propionate liberated from odd chainlength fatty acid oxidation. The plant system involves a hydratase and two hydrogenase enzymes, and the overall scheme is summarized in Figure 10.13. The pathway is superficially similar to modified β-oxidation in bacteria, but the two systems differ in the fate of the individual carbon atoms (Stumpf, 1970). Bacteria take up nonesterified fatty acids from the growth medium by a process that probably involves the formation of acyl-CoA. The fatty acid taken up, therefore, is in a form directly available for β-oxidation. In E. coli there has been some study of the genes coding for fatty acids degradation (fad). These genes, which code for the enzymes of uptake/activation and β-oxidation, are located in three sites on the chromosome and comprise a regulon (Klein et al., 1971). Several types of fad mutants are known and another class of mutants, fadR, are constitutive for β-oxidation and are probably repressor mutants. In addition to being a repressor for fad genes, the fadR gene product may also control isocitrate lyase and malate synthetase (Maloy et al., 1980), which are key enzymes of the glyoxylate shunt. There are some differences in the details of β-oxidation enzymes in E. coli compared to those from animals. For example, only one acyl-CoA synthetase is present and this can activate both medium-chain and long-chain fatty acids (Schulz, 2002). Even though E. coli does not synthesize polyunsaturated fatty acids, it can easily oxidize them by the reductasedependent pathway (above). Details of the enzymes involved in E. coli fatty acid β-oxidation and available mutants that are defective in individual steps are given by Finnerty (1989). The same author also discusses the features of peroxisomal β-oxidation in microbial eukaryotes, such as Candida tropicalis. Other sources of information on E. coli are Nunn (1986) and Black and Dirusso (1994) and for other bacteria (Kunau et al., 1995). The control of mitochondrial β-oxidation flux is reviewed by Eaton (2002). 10.1.6.3
O
long-chain acyl-CoA dehydrogenases; and deficiencies in the electron transferring flavoprotein or its dehydrogenase. Several disorders (e.g., Zellweger syndrome) associated with impairment of peroxisomal β-oxidation have also been described. See Schulz (2002) and Section 11.6 for more details. 10.1.6.4
Ketone bodies
In mammalian liver, the excessive amounts of acetyl-CoA liberated by β-oxidation of fatty acid are converted to various ketone bodies. This conversion takes place at high rates when an elevation in the glucagon/insulin ratio occurs, such as during fasting or in uncontrolled diabetes. Several enzymes are involved, as shown below. These enzymes are also found in extrahepatic tissues, such as heart, kidney and intestine. Acetoacetyl-CoA
2 Acetyl-CoA
Acetoacetyl-CoA + CoA thiolase HMG-CoA
Acetoacetyl-CoA + Acetyl-CoA
Hydroxymethylglutaryl-CoA + CoA synthase HMG-CoA
Hydroxymethylglutaryl-CoA
Acetoacetate + Acetyl-CoA lyase
D-3-Hydroxybutyrate
Acetoacetate + NADH + H+
3-Hydroxybutyrate + NAD+ dehydrogenase
The ketone bodies produced by various tissues can readily diffuse into the blood and be taken up by other extra hepatic tissues and converted back to acetyl-CoA for complete combustion in the tricarboxylic acid cycle. A thorough review of these reactions and of the regulation of ketogenesis (Figure 10.14) has been provided by McGarry and Foster (1980) and see also Gurr et al. (2002) for a summary. Ketone bodies provide important alternative fuels to body tissues when carbohydrate is in short supply or cannot be efficiently utilised. A particular example is the central nervous system, which cannot utilise plasma fatty acids for energy. Thus, in prolonged starvation, ketone bodies become more important than glucose as a fuel source. The possibility of the utilisation of ketone bodies obviates the harmful degradation of muscle protein for gluconeogenesis. In addition, acetoacetate and 3-hydroxybutyrate are thought to be important precursors for lipid synthesis in neonatal brain (Webber and Edmond, 1979).
Inherited diseases of b-oxidation
A number of diseases that compromise the functions of liver, muscle, and other organs have been ascribed to deficiencies of β-oxidation. These include myopathic carnitine deficiency; deficiencies of short-chain, medium-chain, or 652
Lipid Metabolism
VLDL Triacylglycerol
Carnitine
Fatty acyl-CoA
Fatty acid
Glucose Malonyl-CoA Glucagon Fatty acylcarnitine
Pyruvate
Acetyl-CoA
Glucagon Acetyl-CoA
Citrate
Ketone bodies
FIGURE 10.14 Interactions between fatty acid synthesis and oxidation in liver. In the fed state, malonyl-CoA levels are high. This allows rapid fatty acid synthesis and inhibits β-oxidation by lowering carnitine acyltransferase I activity. If triacylglycerol synthesis is impaired, then acyl-CoA will feed back to inhibit acetyl-CoA carboxylase. In the fed state, this does not normally happen, and triacylglycerols are incorporated into very low density lipoprotein (VLDL) for export to extrahepatic tissues. Glucagon excess in fasting leads to a suppression of glycosis, cessation of lipogenesis, and activation of β-oxidation and ketogenesis. (From McGarry, J.D. and Foster, D.W. (1980). Annu. Rev. Biochem. 49, 395–420. With permission.)
rubredoxins), a flavoprotein, and a final component required for hydroxylase activity called the ω-hydroxylase. The latter has a molecular mass of 42,000 Da with 1 atom of iron. It also contains a large amount of phospholipid, and activity is lost if this lipid is removed. Other studies have included electron paramagnetic resonance (EPR) experiments on the spectra shown by the nonhaem ironprotein (for reviews, see Coon et al., 1972; Gunsalus et al., 1975). In plants, the ω-hydroxylase system is responsible for synthesis of ω-hydroxy fatty acyl components of cutin and suberin (see Sections 1.1.2.10 and 1.2.12). The reaction has been studied in preparations from Vicia faba with NADPH and oxygen as the required cofactors (Kolattukudy, 1980). The true substrate for ω-hydroxylation of palmitate is the free acid, and the active subcellular preparation is the microsomal fraction. The reaction showed the properties of a classic mixed-function oxidase, being inhibited by O-phenanthroline, 8-hydroxyquinoline (metal-ion chelators), sodium azide, and thiol-directed reagents. The involvement of cytochrome P450 in the V. faba system is unproven. Although the hydroxylation is inhibited by carbon monoxide, this inhibition was not reversed by light at 420 to 460 nm. Thus, if a cytochrome P450 is involved in the system, it must have unusual properties when compared to other cytochrome-P 450 -containing enzymes (Kolattukudy, 1980). The recent description of three long-chain fatty acid oxidase genes from Candida has led to the identification of a gene family involved in ω-oxidation in yeast with homologues in bacteria and plants (Van Hanen et al., 2000). In Candida the omega carbon is oxidised successively by a cytochrome P450 alkane/fatty acid oxidase, a H2O2-generating alcohol oxidase, an aldehyde dehydrogenase producing
Although ketone bodies serve useful functions, excessive accumulation of such compounds in blood can cause clinical problems, such as ketoacidosis. This may be severe in diabetes or alcoholism. A comprehensive review of clinical aspects of ketone bodies is that of Soling and Seufert (1978). 10.1.6.5
x-Oxidation
Microsomal preparations from many tissues can oxidise fatty acids in the ω-position. An ω-hydroxy fatty acid is formed first, and this can be oxidised by NAD+ and cytoplasmic enzymes to yield a dicarboxylic acid, which can be further attacked by β-oxidation (Greville and Tubbs, 1968). The enzyme(s) responsible for the oxidation of the ω-hydroxy fatty acid have been studied much less than the ω-hydroxylation system. There may be an ω-hydroxy acid dehydrogenase, although liver alcohol dehydrogenase can convert such acids to semialdehyde derivatives. Further oxidation of the latter would be catalysed by an NADlimited aldehyde dehydrogenase (cf. Greville and Tubbs, 1968). ω-Oxidation is important for the further metabolism of fatty acids that are di-substituted in the 2- or 3-positions as well for the catabolism of various xenobiotics that have alkyl chains (see Lenk, 1972). The role of peroxisomes in the ω-oxidation of fatty acids and xenobiotics is discussed by Osmundsen et al. (1991). The ω-hydroxylase system (alkane 1-mono-oxygenase, EC 1.14.15.3) has nonhaem irons as, apparently, the only prosthetic groups directly involved in the hydroxylation reactions. The system from Pseudomonas oleovorans has been isolated by Coon and coworkers (Kusunose et al., 1964). The system was fractionated into three components: a nonhaem iron protein (similar to the 653
10.1
Fatty acids
iron-containing fatty acid dioxygenases, which are ubiquitous in plants and animals (Brash, 1999). LOXs catalyse the regio- and stereo-specific dioxygenations of PUFAs containing a 1Z, 4Z-(cis, cis-1,4) pentadiene system to produce conjugated hydroperoxydiene derivatives:
TABLE 10.5 Major products of ruminant biohydrogenation of C18 unsaturated fatty acids Substrate
Products
Oleic acid Linoleic acid
18:0 9c, 11t-18:2 11t-18:1 18:0 9c, 11t, 15c-18:3 11t, 15c- and 9c, 11t-18:2 11t-18:1 18:0
α-Linolenic acid
RCH = CHCH2CH = CHR´
↓ O2 RCH = CHCH = CHCH(OOH)R´ Thus, typical substrates for lipoxygenases would be the PUFAs linoleic, α- linolenic, arachidonic and eicosapentaenoic acids.
ω-alcohols, ω-aldehydes, and ω-fatty acids. Four genes with high homology to Candida ω-oxidation genes are present in the Arabidopsis genome (Graham and Eastmond, 2002). 10.1.6.6
10.1.7.1
Fatty acid catabolism in ruminants
Because 20C PUFAs are only minor fatty acids in plants, the plant LOXs are classified according to their positional specificity of linoleate oxygenation, either at the 9-(9-LOX) or the 13-carbon (13-LOX) (Feussner and Wasternack, 2002). Lipoxygenase activity is widespread in the plant kingdom, often in very high amounts. The enzymes are particularly important in the food plants, where they destroy the essential polyunsaturated fatty acids to produce derivatives with characteristic tastes and flavours. They are also used for bleaching natural pigments, such as wheat flour carotenoids or alfalfa chlorophyll. General reviews on lipoxygenases are those by Gaffney (1996) and Piazza (1996). Plant lipoxygenases are discussed by Galliard and Chan (1980) and by Siedow (1991), while the lipoxygenase pathway is reviewed by Grechkin (1998) and Feussner and Wasternack (2002). Other useful reviews cover the commercial importance of lipoxygenases (Eskin et al., 1977), their role in olive oil quality (Harwood and Aparicio, 2000) and in plant defence (Blée, 1988). The animal lipoxygenases are dealt with separately in section 10.1.8. Plant tissues containing high levels of lipoxygenase activity are shown in Table 10.6. Leguminous seeds generally contain lipoxygenase, but the absence of measured
A specialized situation for lipid breakdown is the rumen of such animals as sheep and ox. A large number of microorganisms, such as bacteria of the genera Ruminococcus, Bacteroides, and Butyrivibrio, as well as protozoa, play a part in the breakdown of leaf lipids (see Section 2.10). Thus, α- and β-galactosidases are present to cleave the galactose residues of galactosylglycerides. Active lipases hydrolyse the acyl residues, and the liberated fatty acids are often biohydrogenated to give various mixtures of conjugated fatty acids containing trans as well as cis double bonds (Table 10.5) (see Garton, 1977; Harwood and Russell, 1984; and Gurr et al., 2002). Fermentation results in the production of large quantities of acetic, propionic, and butyric acids — the proportions of which vary with the ruminants’ diet. These water-soluble, shortchain fatty acids are absorbed well, and much of their total quantity is metabolised to ketone bodies in the rumen wall. Other features of fatty acid metabolism in ruminants are discussed by Garton (1977).
10.1.7
Lipoxygenase
Lipoxygenases (LOXs) (linoleate: oxygen oxidoreductase (EC 1.13.10.12)) constitutes a large family of nonheme, TABLE 10.6
Lipoxygenases in plants
Plant tissues containing high levels of lipoxygenase activity
Plant
Family
Tissue
Lipoxygenase activity (µl O2 consumed in 10 min/gram fresh weight)
Yellow bean Potato Eggplant Soybean Artichoke Pea Cauliflower Avocado Tomato Lettuce
Leguminosae Solanaceae Solanaceae Leguminosae Compositae Leguminosae Cruciferae Lauraceae Solanaceae Compositae
Seed Tuber Fruit Seed Heart Seed Floret Fruit Fruit Leaf
6480 4560 4320 4150 3360 1769 1440 720 360 120
654
Lipid Metabolism
lipoxygenase activity) to wheat flour in order to bleach pigments for white bread production. Enzymes from different sources differ in their co-oxidation ability, e.g., soybean Type I enzyme has poor activity in this regard while soybean Type II enzyme has high co-oxidation activity. The reaction probably proceeds by a free radical process (Veldink et al., 1977) and requires the presence of a substrate (e.g., linoleic acid) as well as the cosubstrate. The extent of the co-oxidation may depend on the lifetime of the radical intermediates and the relative efficiency of the lipoxygenase-mediated radical reduction (Weber and Grosch, 1976). Whereas in animals, arachidonate is a major substrate for LOX attack, in plants α-linolenate is the major substrate. Breakdown of this acid is known variously as the α-linolenic acid cascade or lipoxygenase pathway (see Figure 10.15). The fatty acid hydroperoxides produced by LOXs can be converted to three main types of products (see Figure 10.15) with important functions (Gurr et al., 2002) (Fig. 10.16):
activity in certain plant tissues does not mean that the enzyme is absent, since there are often inhibitors present. The enzyme has been detected in particulate, cytosolic, and vacuolar fractions (Feussner and Wasternack, 2002). Several types of lipoxygenase have been described. These differ in various properties, and the “classic” soybean lipoxygenase (Theorell et al., 1947) may be atypical in several respects. Two main types are present in soybeans. The Theorell enzyme acts on free acids only and is sometimes called an “acid” lipoxygenase (Verhue and Francke, 1972). The same enzyme is referred to as an “alkaline” lipoxygenase by Grosch et al. (1977) because its optimal activity is found at pH 9. It is better referred to as a Type I enzyme. Similarly, a second lipoxygenase with an optimum pH of 6.5 has been referred to as an “ester” enzyme a “neutral” enzyme, or a “b” enzyme and it is better referred to as the Type II enzyme (Galliard and Chan, 1980). Isoenzymic forms of lipoxygenase have been purified to homogeneity from many different plants (see Eskin et al., 1977). Most of the enzymes have properties similar to the soybean Type II enzyme. Tissues that have been used for study include wheat, alfalfa, potato, barley, pea, and various other legumes. Most lipoxygenases appear to be single polypeptides of molecular masses in the region 70,000 to 100,000 Da. A single atom of nonhaem iron per molecule is found (Chan, 1973) and the pH optima are in the range of 5.5 to 7.0. Because lipoxygenase lacks cofactors other than nonhaem iron, the number of inhibitors is small. The trans-unsaturated fatty acids, acetylenic fatty acids, α-bromo fatty acids, and fatty alcohols will all inhibit, and antioxidants or oxygen scavengers are effective in certain cases (see Hamberg et al., 1974; Eskin et al., 1977). The remarkable increase in sequence information has allowed phylogenetic tree analysis of multigene families. For LOXs, the Type I and Type II enzymes, as well as those classified as the 9- or 13-LOXs, form individual groups in separate branches of the tree (Feussner and Wasternack, 2002). Soybean Type I lipoxygenase is particularly stable. Other lipoxygenases are less stable and activity is lost during purification. Heat treatment is often used in the food industry to cause inactivation so as to prevent offflavours. The purified lipoxygenases are generally unstable at 70°C, but higher temperatures may be necessary for inactivating the enzymes in foodstuffs. A number of assay methods have been used for measuring lipoxygenase. These techniques use radioisotopes or the oxygen electrode and colorimetric methods. These assays and necessary precautions are discussed by Galliard and Chan (1980). Lipoxygenases will also catalyse co-oxidation reactions. This is used both in assay methods for the enzyme and in commercial applications. An example of the latter is the addition of soybean or broad bean flours (both rich in
1. Co-oxidative reactions with peroxygenase give a mixture of epoxy and hydroxyl fatty acids (depending on the nature of the acceptor fatty acid for the monooxygenation) and the products have roles in cutin biogenesis and defence against pest attack (Blée, 1988). 2. Hydroperoxide lyase cleaves the hydroperoxide into an aldehyde and an oxo-unsaturated fatty acid. The products have a role in pest defence and appear to act as pollinator and herbivore attractants, especially for flowers and fruits. They are also important flavour and aroma components in foods, drinks, and perfumes (see e.g., Sanchez and Salas, 2000). 3. Allene oxide synthase gives rise to the precursor of jasmonic acid. The latter and associated compounds are commonly known as jasmonates although some workers prefer the term “jasmonins” to differentiate from jasmonic acid esters. The jasmonins have important effects on plant growth, development, and senescence. Hydroperoxide lyase activity under 2 (above), usually occurs with the 13-hydroperoxide derivative thus giving rise to 6C compounds, which are volatile. Combinations of the volatiles produced are effective attractants for both pollinators and herbivores and the easily detected smells of various fruits, cut grass, cucumber, etc. derive from them. Profiles of the volatiles produced can be easily analysed by headspace gas chromatography and even used for identification purposes (see Harwood and Aparicio, 2000). Physiological roles for lipoxygenase-derived products in plants are shown in Table 10.7 (Gurr et al., 2000). For more detail, refer to Seidow (1991), Blée (1998), Feussner and Wasternack (2002) and Rosahl and Feussner (2005).
655
10.1
Fatty acids
Other 9C compounds Nonadienal +9-oxononanoic acid Lyase 9-lipoxygenase 9-Hydroperoxylinolenic acid O2 COOH
α-Linolenic acid
Epoxy FA Peroxygenase
O2 13-lipoxygenase H O O
Hydroxy FA Hydroxy FA
Peroxygenase
COOH Allene oxide synthase 13-hydroperoxylinolenic acid
Epoxy FA Allene oxide COOH
Lyase O
Allene oxide cyclase
12-Oxo-dodeca3-cis 9-enoic acid hexenal
COOH
Traumatic acid
O 12-oxophytodienoic acid Reduction
Other 6C volatile compounds
β-Oxidation
COOH Jasmonic acid O Tuberonic Cucurbic Other jasmonates acid acid or jasmonins
Conjugates
FIGURE 10.15 The α-linolenic acid cascade and oxylipin formation in plants. (From Gurr, M.I., Harwood, J.L. and Frayn, K.N. (2002). Lipid Biochemistry, 5th ed., Blackwell Scientific, Oxford, U.K. With permission.)
10.1.7.2
lyase with 9-hydroperoxylinoleic acid. Other aldehydes or conjugated derivatives can also have significant activity (Uchida, 2003).
Lipid peroxidation
A constant problem with unsaturated lipids, particularly PUFAs, is the ease of their oxidation. This, in general, not only destroys the beneficial effects of the parent molecule, but often gives rise to products with harmful properties. For further discussion of this subject, see Section 11.1 and Gurr et al. (2002). Specific useful reviews are those of Morrow and Roberts (1997) on isoprostanes and by Itabe (1998) on oxidized phospholipids and atherosclerosis. Some of the products initiated by enzyme activity, such as in the lipoxygenase pathway, may also have important pathophysiological effects. For example, there is much interest in 4-hydroxy-2-nonenal as a product and mediator of oxidation stress. This compound originates from the reaction of hydroperoxide
10.1.8
Production and function of the eicosanoids
Eicosanoids are oxygenated 20-carbon fatty acids. Because the major precursor of these compounds in animals is arachidonic acid, the pathways leading to the eicosanoids are also often known as the “arachidonate cascade.” There are three reactions involved in the initial metabolism of arachidonic acid, which itself must be released from membrane lipids through the activity of phospholipases. The enzymes involved are cyclooxygenase, 5-, 12-, or 15-lipoxygenases and various cytochrome P450
656
Lipid Metabolism
R1
R2
Lipoxygenase
R1
R1
R2
OOH
R2
OOH
Hydroperoxides (fungicidal, induce proteinases)
Lyase pathway
Aldehydes Oxo-acids
Allene oxide synthase pathway
Cyclised products α and γ ketols
(Trigger fungicide production. (Jasmonic acid best known. Acts in signalling transduction Important as flavour and volatile components of foods during wounding, stress and pathogen attack. and beverages) Developmental roles)
Peroxygenase pathway
Epoxides, epoxy alcohols, dihydrodiols, triols (Important, cutin precursors, may have defence roles also)
FIGURE 10.16 Lipoxygenase products and physiological roles in plants. (From Gurr, M.I., Harwood, J.L. and Frayn, K.N. (2002). Lipid Biochemistry, 5th ed., Blackwell Scientific, Oxford, U.K. With permission.)
TABLE 10.7
Some physiological roles for lipoxygenase-derived products in plants
Phenomenon
Notes
Compounds involved
Plant resistance to pathogens
Activation of several defence systems in “hypersensitive response”. Lipoxygenase expression rapidly induced following infection 13-Lipoxygenase induced, translocates to lipid body and dioxygenates storage lipids, which are then hydrolysed and the oxygenated fatty acid catabolised Widespread phenomenon in plants is the drought-induced accumulation of 9hydroperoxy derivatives of membrane lipids; function unknown Lipoxygenase may be involved directly in photosystem inactivation and chlorophyll oxidation; jasmonate as senescencepromoting substance, by inducing production of certain proteins; has complementary effects with ABA (abscisic acid) and may influence ethylene production Differential effects of jasmonates on ethylene formation; several fruit-specific lipoxygenase genes identified Stimulate a number of associated phenomena, such as cell expansion, cytoskeleton structure and carbohydrate accumulation
Jasmonic acid; phytoalexins (e.g., hexenal); eicosanoids from pathogen arachidonate
Mobilisation of storage lipids
Drought stress
Senescence
Fruit ripening
Tuber induction
13-Hydroperoxygenated lipids
9-Hydroperoxygenated lipids
Jasmonins
Jasmonins
Tuberonic acid and other jasmonins
Source: From Gurr, M.I., Harwood, J.L. and Frayn, K.N. (2002). Lipid Biochemistry, 5th Blackwell Scientific, Oxford, U.K. With permission.
657
10.1
Fatty acids
Phospholipid
Stimulus
Precursor fatty acid (e.g., Arachidonate) Cyt P450 epoxygenase
Lipoxygenases Cyclooxygenase Leukotrienes + hydroxyfatty acids lipoxins hepoxilins
Cyclic endoperoxides Prostacyclin synthetase
Prostacyclin (PGl)
Reductase or isomerases Prostaglandins (PGD, PGE, PGF mainly)
Hydroxy fatty acids, fatty acid epoxides Thromboxane synthetase
Thromboxanes (TXs)
FIGURE 10.17 Overall pathway for conversion of essential fatty acids into eicosanoids. (From Gurr, M.I., Harwood, J.L. and Frayn, K.N. (2002). Lipid Biochemistry, 5th ed., Blackwell Scientific, Oxford, U.K. With permission.)
aspirin or ibuprofen, act against the cyclooxygenase binding site. Since PGHS-2 (rather than PGHS-1) is implicated in chronic inflammatory complaints, there has been much interest in developing specific inhibitors against this isoform (Gurr et al., 2002). However, unexpected sidereactions have led to withdrawal of the first (specific) PGHS-2 (COX-2) inhibitors (Drazen, 2005). Once PGH 2 has been formed by PGH synthase, further metabolism is cell-specific. For example, platelets form mainly TxA2, endothelial cells form PGI2 as their major prostanoid, while PGH2 is the major product in renal collecting tubule cells. A number of these enzymes have been purified and characterised (see Smith and Murphy, 2002). PGE2 is the best-studied prostanoid and the molecular biology and physiology of its biosynthetic pathway has been reviewed recently (Murakami and Kudo, 2004). Prostanoids (prostaglandins, thromboxane, prostacyclin) exert a wide variety of actions in the body, which are mediated by specific receptors on plasma membranes. These receptors are classified into five basic types, termed DP, EP, FP, IP, and TP, on the basis of sensitivity to the five primary prostanoids formed from arachidonate. Details of these receptors and their physiological role will be found in Sagimoto et al. (2000). Prostanoids are local hormones that act very near to the site of synthesis (Smith and Murphy, 2002) and which have a very short half-life (a few minutes) due to their rapid catabolism. For PGE2, the initial reaction is oxidation to the relatively inactive 15-keto derivative. Further catabolism involves reduction of the double bond between C-13 and C-14, ω-oxidation and β-oxidation. β-Oxidation of eicosanoids is summarised by Diczfalusy (1994). As an alternative to PGH attack, arachidonic acid can be metabolised through the action of one of three lipoxygenases. These enzymes catalyse reactions analogous to the well-known plant enzymes (Section 10.1.7). The
epoxygenases (Figure 10.17). For prostanoid structures, see Smith and Murphy (2002) and, for a general discussion, Gurr et al.(2002). Eicosanoid synthesis is initiated following the interaction of a stimulus with the cell’s plasma membrane. Interaction of the agonist with its receptor leads to the activation of one or more phospholipase. The mobilisation of arachidonate could be caused, in theory, by a number of lipid degradative enzymes. In practice, various combinations of cytosolic or secretory phospholipase A2 enzymes appear to be used, depending on the circumstance (Smith and Murphy, 2002). Activation of the phospholipase(s) requires or is accompanied by a significant rise in intracellular Ca2+. The release of arachidonate is selective (Dennis, 1987). Once arachidonic acid (or an equivalent PUFA, such as EPA, 20:5n-3) is released, it can be metabolised by prostaglandin endoperoxide (PGH) synthase, also known as cyclooxygenase (COX) or prostaglandin H synthase. This enzyme has two catalytic activities: a cyclooxygenase, which catalyses the formation of PGG2, and a peroxidase, which converts this intermediate to PGH2 (Figure 10.18). Subsequent metabolism of PGH 2 to one of the major prostanoids then takes place via a cell-specific pathway. Prostanoids with a “2” subscript are derived from arachidonic acid, those with a “1” subscript from 8,10,14eicosatrienoic acid and those with a “3” subscript from 5,8,10,14,17-eicosapentaenoic acid. The structure and catalytic properties of prostaglandin endoperoxide synthase (Smith and Marnett, 1991) and regulation of its expression (De Witt, 1991) have been reviewed. For a recent summary, see Kulmacz et al. (2003). Vertebrates, from humans to fish, have two main isoforms of prostaglandin H synthase, termed PGHS-1 and -2(COX-1, COX-2). These two isoforms are structurally very similar, but have very different physiological roles and are regulated very differently (Kulmacz et al., 2003). The nonstearoidal anti-inflammatory drugs, such as 658
Lipid Metabolism
Stimulus Cell surface Phospholipid Phospholipase COOH
Arachidonic acid Cyclooxygenase COOH
O
PGH synthase
O OOH PGG2 Peroxidase COOH
O O
HOOC
HO OH PGH2
COOH O
O
O
OH PGD2
HO
COOH
PGI2 OH
HO OH PGE2
COOH
HO
COOH
O O
HO OH PGF2α
FIGURE 10.18 1991a.)
OH TxA2
Structures and biosynthetic relationships among prostanoids derived from arachidonic acid. (From Smith et al.
exact chemical structure concerned and the tissue tested, the leukotrienes are potent bronchoconstrictors, arterioconstrictors, vasodilators, and chemotactic agents. 5HETE is a major product of 5-LOX activity in all cells. While 5-HETE has its own acute biological potency (Spector et al., 1988), it can also be dehydrogenated to 5oxo-ETE. The latter compound is a strong chemoattractant and its biochemistry and functions have been reviewed (Powell and Rokach, 2005). 12-LOX is another active enzyme, giving rise initially to 12-HPETE from arachidonate (see Figure 10.19). 12HPETE, together with its metabolite 12-HETE, has a variety of important effects on neurotransmission, white blood cells, airways, and other tissues. 12-HPETE can be converted to hepoxilins via hepoxilin synthase. Alternatively, 12-LOX can act as a lipoxin synthase in converting LTA4
lipoxygenases constitute a family of closely-related, nonhaem, iron-containing dioxygenases. The immediate products of their reaction with various eicosaenoic acids are hydroperoxy fatty acids. In the case of arachidonic acid, the products are peroxyeicosatetraenoic acids (HpETEs). Subsequent metabolism generates hydroxyeicosatetraenoic acids (HETEs) (Figure 10.19). The biochemistry and functional activity of HETEs have been reviewed by Spector et al. (1988). The activity of the 5-lipoxygenase has been studied more thoroughly than that of the other lipoxygenases, mainly because the leukotrienes are the end products of this metabolic pathway. The production of leukotrienes LTA4, LTB4, and LTC4 is shown in Figure 10.20. For summaries of the actions of leukotrienes, see Smith and Murphy (2002) and Gurr et al. (2002). Depending on the 659
10.1
Fatty acids
COOH
Arachidonic acid 5-Lipoxygenase
15-Lipoxygenase
12-Lipoxygenase COOH
COOH
OOH COOH
HOO HOO
5-HPETE
15-HPETE
12-HPETE OH
COOH
COOH
COOH HO 5-HETE
HO 15-HETE 12-HETE
FIGURE 10.19
Lipoxygenase pathways for the synthesis of the major HETE isomers. (From Spector, et al. 1988.) COOH
Arachidonic acid 5-Lipoxygenase OOH COOH 2e– Reduction
O2 4e–
5-HPETE Lipoxygenation
OH
OH
COOH
COOH Dehydration
H2O OH
5-HETE COOH O
5S, 15S-diHETE
Non-enzymic
H2O 5, 12-diHETE + 5,6-diHETE OH
Leukotriene A4
Hydration (enzymic)
Glutathione S-transferase
H2O
GSH OH COOH S
COOH
CH2 CHCONCH2COOH
NHCO(CH2)2CHCOOH Leukotriene C4 NH2
OH Leukotriene B4
FIGURE 10.20
Formation of leukotrienes from arachidonic acid. (From Gurr and Harwood 1991.)
660
Lipid Metabolism
than the EETs but varieties of physiological functions have been noted (Oliw, 1994). For a good general summary of the eicosanoids, refer to Smith and Murphy (2002).
into lipoxin A4. Hepoxilins are known to function in relation to the release of intracellular calcium and the opening of potassium channels, while lipoxins have roles as immunologic and hemodynamic regulators. Receptors have been identified for some of these products of 12-LOX metabolism (Yamamoto et al., 1997). 15-LOX produces 15-HPETE from arachidonate and 13-HPODE (13-hydroperoxy, 9Z,10E-octadecadienoic acid) from linoleate. These products can go on to produce 15-HETE and 13-HODE, respectively, which can then initiate various biological effects. Both 15-HETE and 13HODE are bound to cell membrane receptors and these or related metabolites have action on erythropoiesis, the cardiovascular system, skin, respiration, and the reproductive system (Kuhn, 1996). The third metabolic pathway for PUFAs to produce biologically active metabolites is via P450-mediated reactions (see Figure 10.17). Arachidonic and linoleic acids can be oxygenated by P450 in four main ways: epoxidation, hydroxylation of the ω-side chain, hydroxylation of allylic or bis-allylic carbons, and hydroxylation with double bond migration (Oliw, 1994). Epoxyeicosatrienoic acids (EETs) with epoxy groups in the 5/6, 8/9, 11/12, or 14/15 positions (Figure 10.21), which are produced from arachidonate, have important functions in vascular smooth muscle, endothelium, myocardium, and other tissues (Spector et al., 2004). EETs are rapidly metabolised including being β-oxidised. The hydroxy products have been less well studied
10.1.9
Other conversions
Fatty acids can be modified in a number of other ways. For example, cyclopropane fatty acids are formed by the addition of a methylene group from S-adenosylmethionine across the double bond of a monounsaturated fatty acid. The latter is esterified to a phospholipid, so that the actual substrate is a membrane lipid. The aldehyde residues in plasmalogens and the alcohol residues in alkyl ether lipids, such as those in Clostridium butyricum also act as acceptors for the methyl group — in this case, forming the corresponding cyclopropane aldehydes and alcohols (Goldfine and Panos, 1971). The control of cyclopropane fatty acid synthase in bacteria, such as E. coli, has been examined in some detail (see Harwood and Russell, 1984). Cyclopropane fatty acids in plants seem to be made by the same mechanism as for bacteria (Mangold and Spener, 1980). These authors have also reviewed work on the synthesis of cyclopentenyl fatty acids, such as chaulmoogric (13-(2-cyclopentenyl) tridecanoic) acid. Branched, cyclic, and unsaturated hydrocarbons in higher plants are formed from appropriate fatty acids by decarboxylation (see Kolattukudy, 1980, 1987). The process COOH
Arachidonic acid O
O COOH
5,6-EpETrE (5,6-EET)
HO OH
O 14,15-EpETrE (14,15-EET)
11,12-EpETrE (11,12-EET)
HO OH COOH
COOH
5,6-DiHETrE
O
8,9-EpETrE (8,9-EET)
COOH
COOH
COOH
COOH
HO OH 11,12-DiHETrE
8,9-DiHETrE COOH
COOH
HO OH 14,15-DiHETrE COOH
CH2OH OH 19-Hydroxyarachidonate
20-Hydroxyarachidonate
FIGURE 10.21
Structures of products of epoxygenase pathways of the arachidonate cascade. (From Smith et al. 1991a.)
661
10.1
Fatty acids
De novo synthesis
C16
Free fatty acids C16–C32 Aldehydes C22–C32
C18 + C2 units
C16–C34 fatty acyl chains
Alkan-1-ols C22–C32
CO2 Alkanes C21–C33
Alkan-1-ol esters C36–C58 Secondary alcohols C29–C31
Ketones C29–C31
FIGURE 10.22 Conversion of fatty acids to other wax components. (From von Wettstein-Knowles, P. (1979). In Advances in the Biochemistry and Physiology of Plant Lipids (Eds. L-A Appelqvist and C. Liljenberg) Elsevier, Amsterdam, pp. 126. With permission.)
has also been studied in cyanobacteria, insects, and other species (see Kolattukudy, 1976). However, the exact reaction mechanism is ill defined. Plants use long-chain and very long-chain fatty acids as sources of hydrocarbon and β-diketones. These and other conversions involved in the generation of plant wax constituents are reviewed by von Wettstein-Knowles (1995), while Nelson and Blomquist (1995) discuss the formation of insect waxes. The relationship of fatty acids to the other plant wax components is shown in Figure 10.22. A recent update on the formation of the plant cuticle’s components is that by Kunst et al. (2005).
Avery, S.V. et al. (1995) Temperature-independent changes in plasma membrane lipid order and the phagocytotic activity of the amoeba Acanthamoeba castellanii are closely related. Biochem. J. 312, 810–816. Barry III, C.E. et al. (1988) Mycolic acids: structure, biosynthesis and physiological functions. Prog. Lipid. Res. 37, 143–179. Beevers, H. (1980). The role of glyoxylate cycle. In The Biochemistry of Plants, vol. 4, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 117–130. Bernlohr, D.A. et al. (1997) Intracellular lipid-binding proteins and their genes. Annu. Rev. Nutri. 17, 277–303. Black, P.N. and DiRusso, C.C. (1994). Molecular and biochemical analyses of fatty acid transport, metabolism, and gene regulation in Escherichia coli. Biochim. Biophys. Acta. 1210, 123–145. Blée, E. (1988). Phytooxylipins and plant defense reactions. Prog. Lipid Res. 37, 33–72. Bloch, K. (1977). Control mechanisms for fatty acid synthesis in Mycobacterium smegmatis. Adv. Enzymol., 45, 1–84. Bourre, J.M. et al. (1975). Biosynthesis of lignoceric acid from behenyl-CoA in mouse brain microsomes. Comparison between normal and Quaking mutant. Biochem. Biophys. Res. Commun., 63, 1027–1034. Bowen, D.M. et al. (1974). The dynamic role of lipids in the nervous system. MTP Int. Rev. Sci., 4, 141–179. Brash, A.R. (1999) Lipoxygenases: occurrence, functions, catalysis and acquisition of substrate. J. Biol. Chem. 274, 23679–23682. Bressler, R. (1970). Physiological-chemical aspects of fatty acid oxidation. In Lipid Metabolism, Ed. S.J. Wakil, Academic Press, New York, pp. 49–75. Browse, J. and Somerville, C. (1991). Glycerolipid synthesis — biochemistry and regulation. Annu. Rev. Plant Physiol., 42, 467–506. Cahoon, E.B. et al. (1999). Biosynthetic origin of double bonds: production of fatty acid components of high-value drying oils in transgenic soybean embryos. Proc. Natl. Acad. Sci. U.S.A. 96, 12935–12940. Chan, H.W.S. (1973). Soya-bean lipoxygenase: an iron-containing dioxygenase. Biochim. Biophys. Acta, 327, 32–35.
References Abulnaja, K.O. and Harwood, J.L. (1991). Interaction of thiocarbamate herbicides with fatty acid synthesis in germin at i n g p e a s a n d t h e i r m i c ro s o m a l f r a c t i o n s Phytochemistry. 30, 2883–2887. Alberts, A. and Vagelos, P.R. (1972). Acyl-CoA carboxylase. In The Enzymes, vol. 6, Ed. P.D. Boyer, Academic Press, New York, pp. 37–82. Allred, J.B. and Roman-Lopez, C.A. (1988). Enzymatically inactive forms of acetyl-CoA carboxylase in rat liver mitochondria. Biochem. J., 251, 881–885. Amy, C.M. et al. (1989). Molecular cloning and sequencing of cDNAs encoding the entire rat fatty acid synthase. Proc. Natl. Acad. Sci. USA, 86, 3114–3118. Amy, C.M. et al. (1990). Molecular cloning of the mammalian fatty acid synthase gene and identification of the promoter region. Biochem. J., 271, 675–679. Amy, C.M. et al. (1992). Intron-exon organization of the gene for the multifunctional animal fatty acid synthase. Proc. Natl. Acad. Sci., 89, 1005–1008. Ansell, G.B. and Spanner, S. (1982). Phosphatidylserine, phosphatidylethanolamine and phosphatidylcholine. In Phospholipids, Eds. J.N. Hawthorne and G.B. Ansell, Elsevier, Amsterdam, pp. 1–49. Asturias, F.J. et al. (2005). Structural and molecular organisation of mammalian fatty acid synthase. Nature Struc. Mol. Biol. 12, 225–232.
662
Lipid Metabolism
Chuman, L. and Brody, S. (1989). Acyl carrier protein is present in the mitochondria of plants and eucaryotic micro-organisms. Eur. J. Biochem., 184, 643–649. Cook, H.W. (1991). Fatty acid desaturation and chain elongation in eukaryotes. In Biochemistry of Lipids, Lipoproteins and Membranes, 2nd ed., Eds. D.E. Vance and J. Vance, Elsevier, Amsterdam, pp. 141–169. Cook, H.W. and McMaster, C.R., (2002). Fatty acid desaturation and elongation in eukaryotes. In Biochemistry of Lipids, Lipoproteins and Membranes, Eds. D.E. Vance and J.E. Vance, 4th ed., Elsevier, Amsterdam, pp. 181–204. Coon, M.J. et al. (1972). Functional components and mechanism of action of the resolved liver microsomal enzyme system catalyzing fatty acid, hydrocarbon and drug hydroxylation. In Biological Hydroxylation Mechanisms, Eds. G.S. Boyd and M.S. Smellie, Academic Press, New York, pp. 45–54. Cripps, C. et al. (1990). The delta 12-desaturase from the house cricket, Acheta domesticus (Orthoptera: Gryllidae): characterization and form of the substrate. Arch. Biochem. Biophys., 278, 46–51. Cronan, J.E. and Waldrop, G.L. (2002) Multi-subunit acetylCoA carboxylases. Prog. Lipid Res. 41, 407–435. Dailey, H.A. and Strittmatter, P. (1979). Modification and identification of cytochrome b5 carboxyl groups involved in protein-protein interaction with cytochrome b5 reductase. J. Biol. Chem. 254, 5388–5396. Dailey, H.A. and Strittmatter, P. (1980). Characterization of the interaction of amphipathic cytochrome b5 with stearyl coenzyme A desaturase and NADPH:cytochrome P450 reductase. J. Biol. Chem. 255, 5184–5189. de Renobales, M. et al. (1987). Biosynthesis of linoleic-acid in insects. Trends Biochem. Sci. 12, 364–366. De Titta, G.T. et al. (1980). Carboxybiotin translocation mechanisms suggested by diffraction studies of biotin and its vitamers. Proc. Natl Acad. Sci. USA, 77, 333–337. De Witt, D.L. (1991). Prostaglandin endoperoxide synthase: regulation of enzyme expression. Biochim. Biophys. Acta, 1083, 121–134. Dennis, E.A. (1987). Regulation of eicosanoid production — role of phospholipases and inhibitors Biotechnology, 5, 1294–1300. Diczfalusy, U. (1994). β-Oxidation of eicosanoids. Prog. Lipid Res. 33, 403–428. Dittrich, F. et al. (1988). Fatty acid elongation in yeast — biochemical characteristics of the enzyme system and isolation of elongation-deficient mutants. Eur. J. Biochem. 252, 477–485. Drazen, J.M. (2005). A lesson in unexpected problems. New Eng. J. Med. 352, 1031–1032. Eaton, S. (2002). Control of mitochondrial β-oxidation flux. Prog. Lipid Res. 41, 197–239. Egmond, M.R. et al. (1977). On the mechanism of action of soybean lipoxygenase-1. A stopped-flow kinetic study of the formation and conversion of yellow and purple enzyme species. Eur. J. Biochem., 76, 469–479. El Hussein, S.A. et al. (1988). Plant holo-(acyl carrier protein) synthase. Biochem. J., 252, 39–45. Ernst-Fonberg, M.L. and Bloch, K. (1971). A chloroplastassociated fatty acid synthetase system in Euglena. Arch. Biochem. Biophys., 143, 392–400.
Eskin, N.A. et al. (1977). Biochemistry of lipoxygenase in relation to food quality. Crit. Rev. Food Sci. Nutr. 9, 1–40. Esselman, W.J. and Clagett, C.O. (1974). Products of linoleic hydroperoxide-decomposing enzyme of alfalfa seed. J. Lipid Res., 15, 173–178. Fehling, E. et al. (1992). Solubilization and partial purification of constituents of acyl-CoA elongase from Lunaria annua. Biochim. Biophys. Acta, 1126, 88–94. Feussner, I. and Wasternack, C. (2002). The lipoxygenase pathway. Annu Rev. Plant Biol. 53, 275–297. Finnerty, W.R. (1989). Microbial lipid metabolism. In Microbial Lipids, vol. 2, Eds. C. Ratledge and S.G. Wilkinson, Academic Press, London, pp. 525–558. Fulco, A.J. (1974). Metabolic alterations of fatty acids. Annu. Rev. Biochem., 43, 215–241. Gaffney, B.J. (1996). Lipoxygenase: structural principles and spectroscopy. Annu. Rev. Biochem, 25, 431–459. Galliard, T. (1974). Unusual fatty acids in plants. Recent Adv. Phytochem., 8, 209–241. Galliard, T. (1980). Degradation of acyl lipids: hydrolytic and oxidative enzymes. In The Biochemistry of Plants, vol. 4, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 85–116. Gaillard, T. and Chan, H.W.S. (1980). Lipoxygenases. In The Biochemistry of Plants, vol. 4, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 131–161. Galliard, T. and Matthew, J.W. (1975). Enzymic reactions of fatty acid hydroperoxides in extracts of potato tuber. I. Comparison 9D- and 13L-hydroperoxy-octadecadienoic acids as substrates for the formation of a divinyl ether derivative. Biochim. Biophys. Acta. 389, 1–9. Galliard, T. et al. (1974). The enzymic and non-enzymic degradation of colneleic acid, an unsaturated fatty acid ether intermediate in the lipoxygenase pathway of linoleic acid oxidation in potato (Solanum tuberosum) tubers. Biochem. J. 138, 23–31. Gardner, H.W. (1975). Decomposition of linoleic acid hydroperoxides. Enzymic reactions compared with nonenzymic. Agric. Food Chem. 23, 129–136. Gardner, H.W. (1991). Recent investigations into the lipoxygenase pathway of plants. Biochim. Biophys, Acta. 1084, 221–239. Garton, G.A. (1977). Fatty acid metabolism in ruminants. In Biochemistry of Lipids II MTP Int. Rev. Sci., 14, 337–370. Gerhardt, B. (1992). Fatty acid degradation in plants. Prog. Lipid Res. 31, 417–446. Glatz, J.F.C. and van der Vusse, G.J. (1996) Cellular fatty acidbinding proteins: their function and physiological significance. Prog. Lipid Res. 35, 243–282. Goldberg, I. et al. (1973). Fatty acyl-coenzyme A elongation in brain of normal and quaking mice. Science, 182, 497–499. Goldfine, H. and Panos, C. (1971). Phospholipids of Clostridium butyricum. IV. Analysis of the positional isomers of monounsaturated and cyclopropane fatty acids and alk-1'-enyl ethers by capillary column chromatography. J. Lipid Res. 12, 214–220. Goodridge, A.G. (1991). Fatty acid synthesis in eukaryotes. In Biochemistry of Lipids, Lipoproteins and Membranes,
663
10.1
Fatty acids
2nd ed., Eds. D.E. Vance and J. Vance, Elsevier, Amsterdam, pp. 111–139. Graham, I.A. and Eastmond, P.J. (2002) Pathways of straight and branched chain fatty acid catabolism in higher plants. Prog. Lipid. Res. 41, 156–181. Grechkin, A (1998). Recent developments in biochemistry of the plant lipoxygenase pathway. Prog. Lipid Res. 37, 317–352. Greville, G.B. and Tubbs, P.K. (1968). The. catabolism of long chain fatty acids in mammalian tissues. Essays Biochem. 4, 155–212. Grosch, W. et al. (1977). Co-oxidation of beta-carotene and canthaxanthine by purified lipoxygenases from soya beans. Z. Lebensm. Unters. Forsch. 163, 77–81. Gunsalus, I.C. et al. (1975). Oxygenase-catalyzed biological hydroxylations. Annu. Rev. Biochem. 44, 377–407. Gurr, M.I. (1974). Biosynthesis of unsaturated fatty acids. In Biochemistry of Lipids, Ed. T.W. Goodwin, Butterworths, London, pp. 181–236. Gurr, M.I. and Harwood, J.L. (1991) Lipid Biochemistry, 4th ed., Chapman & Hall, London. Gurr, M.I. et al. (2002) Lipid Biochemistry, 5th ed., Blackwell Science, Oxford. Hamberg, M. et al. (1974). Oxygenases in fatty acid and steroid metabolism. In Molecular Mechanisms of Oxygen Activation, Ed. O. Hayaishi, Academic Press, New York, pp. 29–85. Hamberg, M. et al. (1999). Alpha-oxidation of fatty acids in higher plants. Identification of a pathogen-inducible oxygenase (piox) as an alpha-dioxygenase and biosynthesis of 2-hydroperoxylinolenic acid. J. Biol. Chem. 274, 24503–24513. Harwood, J.L. (1988). Fatty acid metabolism. Annu. Rev. Plant Physiol. 39, 101–138. Harwood, J.L. (1990). Lipid synthesis. In Target Sites for Herbicide Action, Ed. R.C. Kirkwood, Plenum, New York, pp. 57–94. Harwood, J.L. (1996). Recent advances in the biosynthesis of plant fatty acids. Biochim. Biophys. Acta. 1301, 7–56. Harwood, J.L. (2005). Fatty acid biosynthesis. In Plant Lipids: Biology, Utilisation and Manipulation, Ed. D.J. Murphy, Blackwell Publishing, Oxford, pp. 27–66. Harwood, J.L. and Aparicio. R. Eds. (2000) Handbook of Olive Oil. Aspen Publishing, Gaithersburg, MD. Harwood, J.L. and Russell, N.J. (1984) Lipids in Plants and Microbes, Allen and Unwin, London. Harwood, J.L. et al. (1990). Enzymes of fatty acid synthesis. In Methods in Plant Biochemistry, vol. 3 (Ed. P.J. Lea), Academic Press, London, pp. 193–217. Hashimoto, T. (1990). Purification properties and biosynthesis of peroxisomal beta oxidation enzymes. In Fatty Acid Oxidation: Clinical, Biochemical and Molecular Aspects, Eds. K. Tanaka and P.M. Costes, Alan R. Liss, New York, pp. 137–152. Heath, R.J. et al. (2001). Lipid biosynthesis as a target for antibacterial agents. Prog. Lipid Res. 40, 467–497. Heath, R.J. et al. (2002). Fatty acid and phospholipid metabolism in prokaryotes. In Biochemistry of Lipids, Lipoproteins and Membranes, Eds. D.E. Vance and J.E. Vance 4th ed., Elsevier, Amsterdam, pp. 55–92. Hildebrand, D.F. et al. (2005). Fatty acid manipulation. In Plant Lipids: Biology, Utilisation and Manipulation (Ed. D.J. Murphy) Blackwell Publishing, Oxford, pp. 67–102.
Holman, R.T. et al. (1969). Substrate specificity of soybean lipoxidase. J. Biol. Chem. 244, 1149–1151. Itabe, H. (1998). Oxidised phospholipids as a new landmark in atherosclerosis. Prog. Lipid Res. 37, 181–207. Iwabuchi, M. et al. (2003). Delta 12-oleate desaturase-related enzymes associated with formation of conjugated transdelta II, cis-delta 13 double bonds. J. Biol. Chem. 278, 4603–4610. Jackowski, S. et al. (1987). Acetoacetyl-acyl carrier protein synthase, a potential regulator of fatty acid biosynthesis in bacteria. J. Biol. Chem. 262, 7927–7931. Jackowski, S. et al. (1991). Lipid metabolism in prokaryotes. In Biochemistry of Lipids, Lipoproteins and Membranes, 2nd ed., Eds. D.E. Vance and J. Vance, Elsevier, Amsterdam, pp. 43–85. Jaworski, J.G. (1987). Biosynthesis of monoenoic and polyenoic fatty acids. In The Biochemistry of Plants, vol. 9, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 159–174. Jaworski, J.G. et al. (1989). A cerulenin insensitive short chain 3ketoacyl-acyl carrier protein synthase in Spinacia oleracea leaves. Plant Physiol. 90, 4144. Johns, R.B. and Perry, G.J. (1977). Lipids of marine bacterium Flexibacter-polymorphus. Arch. Microbiol. 104, 267–271. Jones, A.V.M. and Harwood, J.L. (1980). Desaturation of linoleic acid from exogenous lipids by isolated chloroplasts. Biochem. J. 190, 851–854. Jones, A.L. et al. (1992). Induction of delta 12-desaturase activity during temperature adaptation in Acanthamoeba castellanii. Biochem. Soc. Trans. 20, 170S. Jones, A.L. et al. (1993). Temperature induced membrane–lipid adaptation in Acanthamoeba castellanii. Biochem. J. 290, 273–278. Kaneda, T. (1977). Fatty acids of the genus Bacillus: an example of branched-chain preference. Bact. Rev. 41, 391–418. Kearns, E.V. et al. (1991). The role of cytochrome b5 in delta 12 desaturation of oleic acid by microsomes of safflower (Carthamus tinctorius L.). Arch. Biochem. Biophys. 284, 431–436. Kim, K-H (1997). Regulation of mammalian acetyl-CoA carboxylase. Annu. Rev. Nutri. 17, 77–99. Kindl, H. (1987). β-oxidation of fatty acids by specific organelles. In The Biochemistry of Plants, vol. 9, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 31–52. Klein, K. et al. (1971). Fatty acid degradation in Escherichia coli. An inducible system for the uptake of fatty acids and further characterization of old mutants. Eur. J. Biochem. 19, 442–450. Knowles, J.R. (1989). The mechanism of biotin-dependent enzymes. Annu. Rev. Biochem. 58, 195–221. Knutzon, D.S. et al. (1991). Nucleotide sequence of a complementary DNA clone encoding stearoyl-acyl carrier protein desaturase from castor bean. Plant Physiol. 96, 344–345. Kolattukudy, P.E. Ed. (1976) Chemistry and Biochemistry of Natural Waxes, Elsevier, New York. Kolattukudy, P.E. (1980). Cutin, suberin, and waxes. In The Biochemistry of Plants, vol. 4, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 571–645. Kolattukudy, P.E. (1987). Lipid-derived defensive polymers and waxes and their role in plant-microbe interaction. In The Biochemistry of Plants, vol. 9, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 291–314.
664
Lipid Metabolism
Krakower, G.R. and Kim, K.H. (1981). Purification and properties of acetyl-CoA carboxylase phosphatase. J. Biol. Chem. 256, 2408–2413. Kridl, J.C., Knutzon, D.S., Johnson, W.B., Thompson, G.A., Radke, S.E., Turner, J.C. and Knauf, V.C. (1991) Int. Soc. for Plant Mol. Biol., 3rd Int. Congr., Tucson, AZ, October, Poster 723. Kuhn, H. (1996). Biosynthesis, metabolization and biological importance of the primary 15-lipoxygenase metabolites 15-hydro(pero)xy-5Z, 8Z, 10Z, 13E-eicosatetraenoic acid and 13-hydro(epro)xy-9Z, 10E-octadecadienoic acid. Prog. Lipid Res. 35, 203–226. Kulmacz, R.J. et al. (2003). Comparison of the properties of prostaglandin H synthase-1 and -2. Prog. Lipid Res. 42, 377–404. Kunau, W.H. et al. (1995) β-Oxidation of fatty acids in mitochondria, peroxisomes and bacteria: a century of continued progress. Prog. Lipid. Res. 34, 267–342. Kunst, L. et al. (2005). The plant cuticle: formation and structure of epidermal surfaces. In Plant Lipids: Biology, Utilisation and Manipulation, Ed. D.J. Murphy, Blackwell Publishing, Oxford, pp. 270–302. Kusunose, M. et al. (1964). Enzymatic omega-oxidation of fatty acids. II. Substrate specificity and other properties of the enzyme system. J. Biol. Chem. 239, 2135–2139. Lazarow, P.B. and Moser, H.W. (1989). Disorders of peroxisome biogenesis. In The Metabolic Basis of Inherited Disease, 6th ed., Eds. C.R. Scriver, A.L. Beaudet, W.S. Sly and D. Valle, McGraw-Hill, New York, pp. 1479–1509. Lekakis, N.J. (1977). α-Oxidation of fatty-acids in Escherichiacoli. FEMS Microbiol. Lett. 1, 289–292. Lenk, W. (1972). Biochemical acyl hydroxylations. Prog. Drug Res.16, 229–308. Lent, B. and Kim, K.H. (1980). Purification and properties of a kinase which phosphorylates and inactivates acetyl-CoA carboxylase. J. Biol. Chem. 257, 1897–1901. Leonard, A.E. et al. (2004) Elongation of long-chain fatty acids. Prog. Lipid Res. 43, 36–54. Leonardi, R. et al. (2005) Coenzyme A: back in action. Prog. Lipid Res. 44, 125–153. Lessire, R. et al. (1989). Involvement of a beta-ketoacyl-CoA intermediate in acyl-CoA elongation by an acyl-CoA elongase purified from leek epidermal cells. Biochim. Biophys. Acta. 1006, 35–40. Li, S.-J. and Cronan, J.E. (1992). The gene encoding the biotin carboxylase subunit of Escherichia coli acetyl-CoA carboxylase. J. Biol. Chem. 267, 855–863. Libertini, L.J. and Smith, S. (1978). Purification and properties of a thioesterase from lactating rat mammary gland which modifies the product specificity of fatty acid synthetase. J. Biol. Chem. 253, 1393–1401. Lynen, F.O. (1967). The role of biotin-dependent carboxylations in biosynthetic reactions. Biochem. J., 102, 381–400. Maloy, S.R. et al. (1980). Elevated levels of glyoxylate shunt enzymes in Escherichia coli strains constitutive for fatty acid degradation. J. Bacteriol. 143, 720–725. Mangold, H.K. and Spener, F. (1980). Biosynthesis of cyclic fatty acids. In The Biochemistry of Plants, vol. 4, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 647–663. Matsuda, S. et al. (1993). Decreased messenger RNA of arachidonate 12-lipoxygenase in platelets of patients with myeloproliferative disorders. Biochim. Biophys. Acta. 1080, 243–249.
McGarry, J.D. and Foster, D.W. (1980). Regulation of hepatic fatty acid oxidation and ketone body production. Annu. Rev. Biochem. 49, 395–420. McKeon, T.A. and Stumpf, P.K. (1982). Purification and characterization of the stearoyl-acyl carrier protein desaturase and the acyl-acyl carrier protein thioesterase from maturing seeds of safflower. J. Biol. Chem. 257, 12141–12147. Mcllwain, H. (1966) Biochemistry and the Central Nervous System, Churchill, London. Metz, J.G. et al. (2001). Production of polyunsaturated fatty acids by polyketide synthases in both prokaryotes and eukaryotes. Science 293, 290–293. Mikami, K. and Murata, N. (2003). Membrane fluidity and the perception of environmental signals in cyanobacteria and plants. Prog. Lipid Res. 42, 527–543. Mikolajczyk, S. and Brody, S. (1990). De novo fatty acid synthesis mediated by acyl-carrier protein in Neurospora crassa mitochondria. Eur. J. Biochem. 187, 431–437. Moreau, R.A. and Stumpf, P.K. (1981). Recent studies of the enzymic synthesis of ricinoleic acid by developing castor beans. Plant Physiol. 67, 672–676. Morris, L.J. (1970). Mechanisms and stereochemistry in fatty acid metabolism. Biochem. J. 108, 681–693. Morrow, J. D. and Roberts, L.J. (1997). The isoprostanes: unique bioactive products of lipid peroxidation. Prog. Lipid Res. 36, 1–21. Moss, J. and Lane, M.D. (1971). The biotin-dependent enzymes. Adv. Enzymol. 35, 321–442. Mukherji, M. et al. (2003). The chemical biology of branchedchain lipid metabolism. Prog. Lipid. Res. 42, 359–376. Murakami, M. and Kudo, I. (2004). Recent advances in the molecular biology and physiology of the prostaglandin E2biosynthetic pathway. Prog. Lipid Res. 43, 1–35. Murata, N. and Wada, H. (1995). Acyl lipid desaturases and their importance in the tolerance and acclimatization to cold of cyanobacteria. Biochem. J. 308, 1–8. Naggert. J. et al. (1988). Molecular cloning and sequencing of a cDNA encoding the thioesterase domain of the rat fatty acid synthetase. J. Biol. Chem. 263, 1146–1150. Naggert, J. et al. (1991). Expression in Escherichia coli, purification and characterization of two mammalian thioesterases involved in fatty acid synthesis. Biochem. J. 273, 787–790. Nelson, D.R. and Blomquist, G. (1995) Insect waxes. In Waxes: Chemistry, Molecular Biology and Functions, Ed. R.J. Hamilton, The Oily Press, Dundee, U.K., pp. 1–90. Nichols, B.W. et al. (1967). Interrelationships between fatty acid biosynthesis and acyl-lipid synthesis in Chlorella vulgaris. Biochem., J. 104, 486–496. Nikawa, J. et al. (1979). Inhibitory effects of long-chain acyl coenzyme A analogues on rat liver acetyl coenzyme A carboxylase. FEBS Lett., 102, 223–226. Nunn, W.D. (1986). A molecular view of fatty acid catabolism in Escherichia coli. Microbiol. Rev. 50, 179–192. Ohlrogge, J.B. (1987). Biochemistry of plant acyl carrier proteins. In The Biochemistry of Plants, vol. 9, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 137–157. Ohlrogge, J. B. and Jaworski, J.G. (1997) Regulation of fatty acid synthesis. Annu. Rev. Plant Physiol. Plant Mol. Biol. 48, 109–136.
665
10.1
Fatty acids
Ohlrogge, J. et al. (1991). The genetics of plant lipids. Biochim. Biophys. Acta., 1082, 1–26. Oliw, E.H. (1994). Oxygenation of polyunsaturated fatty acids by cytochrome P450 monoxygenases. Prog. Lipid Res. 33, 329–354. Osmundsen, H. et al. (1991). Metabolic aspects of peroxisomal beta-oxidation. Biochim. Biophys. Acta, 1085, 141–158. Ozasa, H. et al. (1984). Biosynthesis of enzymes of rat-liver mitochondrial beta-oxidation. Eur. J. Biochem., 144, 453–458. Piazza, G. Ed. (1996). Lipoxygenase and Lipoxygenase Pathway Enzymes. Am. Oil Chem. Soc., Champaign, IL. Polakis, S.E. et al. (1974). Acetyl- coenzyme A carboxylase system of Escherichia coli. Studies on the mechanisms of the biotin carboxylase- and carboxyltransferase-catalyzed reactions. J. Biol. Chem., 249, 6657–6667. Pollard, M.R. and Stumpf, P.K. (1980). Biosynthesis of C20 and C22 fatty acids by developing seeds of Limnanthes alba. Plant Physiol. 66, 649–655. Pollard, M.R. et al. (1991). A specific acyl-ACP thioesterase implicated in medium-chain fatty acid production in immature cotyledons of Umbellularia californica. Arch. Biochem. Biophys. 284, 306–312. Post-Beittenmiller, D. et al. (1991). In vivo pools of free and acylated acyl carrier proteins in spinach. Evidence for sites of regulation of fatty acid biosynthesis. J. Biol. Chem., 266, 1858–1865. Poulose, A.J. et al. (1981). Primary structure of a chymotryptic peptide containing the “active serine” of the thioesterase domain of fatty acid synthase. Biochem. Biophys. Res. Commun. 103, 377–382. Powell, W.S. and Rokach, J. (2005). Biochemistry, biology and chemistry of the 5-lipoxygenase product 5-oxo-ETE. Prog. Lipid Res. 44, 154–183. Pugh, E.L. and Kates, M. (1975). Characterization of a membrane-bound phospholipid desaturase system of Candida lipolytica. Biochim. Biophys. Acta, 380, 442–453. Rainwater, D.L. and Kolattukudy, P.E. (1982). Purification and characterization of acyl-CoA carboxylase from the goose uropygial gland, which produces multimethylbranched acids and evidence for its identity with avian acetyl-CoA carboxylase. Arch. Biochem. Biophys. 213, 372–383. Rangan, V.S. and Smith, S. (2002). Fatty acid synthesis in eukaryotes. In Biochemistry of Lipids, Lipoproteins and Membranes, Eds. D.E. Vance and J.E. Vance, 4th ed., Elsevier, Amsterdam, pp. 151–179. Roessler, P.G. (1990). Purification and characterization of acetylCoA carboxylase from the diatom Cyclotella cryptica. Plant Physiol. 92, 73–78. Rosahl, S. and Feussner, I. (2005). Oxylipins. In Plant Lipids: Biology, Utilisation and Manipulation, Ed. D.J. Murphy, Blackwell Publishing, Oxford, pp. 329–354. Rutter, A.J. et al. (2002) Oxygen induction of a novel fatty acid n-6 desaturase in the soil protozoon, Acanthamoeba castellanii. Biochem. J. 368, 57–67. Samols, D. et al. (1988). Evolutionary conservation among biotin enzymes. J. Biol. Chem. 263, 6461–6464. Sanchez, J. and Salas, J.J. (2000). Biogenesis of olive oil aroma. In Handbook of Olive Oil, Eds. J.L. Harwood and R. Aparicio, Aspen Publishing, Gaithersburg, MD, pp. 79–99.
Schewe, T. and Kuhn, H. (1991). Do 15-lipoxygenases have a common biological role? Trends Biochem. Sci. 16, 369–373. Schmid, K.M. and Ohlrogge, J.B. (2002). Lipid metabolism in plants. In Biochemistry of Lipids, Lipoproteins and Membranes, Eds. D.E. Vance and J.E. Vance, 4th ed., Elsevier, Amsterdam, pp. 93–126. Schmidt, H. and Heinz, E. (1990). Involvement of ferredoxin in desaturation of lipid-bound oleate in chloroplasts. Plant Physiol. 94, 241–220. Schroepfer, G.J. and Bloch, K. (1965). The stereospecific conversion of stearic acid to oleic acid. J. Biol. Chem. 240, 54–63. Schulz, H. (1991a). Oxidation of fatty acids. In Biochemistry of Lipids, Lipoproteins and Membranes, 2nd ed., Eds. D.E. Vance and J. Vance, Elsevier, Amsterdam, pp. 87–110. Schulz, H. (1991b). Beta oxidation of fatty acids. Biochim. Biophys. Acta. 1081, 109–120. Schweizer, E. (1984). Genetics of fatty acid biosynthesis in yeast. In Fatty Acid Metabolism and its Regulation, Ed. S.Numa, Elsevier, Amsterdam, pp. 59–83. Schweizer, E. (1989). Biosynthesis of fatty acids and related compounds. In Microbial Lipids, vol. 2, Eds. C. Ratledge and S.G. Wilkinson, Academic Press, London, pp. 3–51. Schweizer, E. et al. (1973). Pantetheine-free mutants of the yeast fatty-acid-synthetase complex. Eur. J. Biochem. 39, 353–362. Sekiya, J. and Hatanaka, A. (1977). Cis-3-hexenal and normalhexanal formation from linolenic and linoleic acids in alfalfa cells cultured in vitro. Plant Sci. Lett. 10, 165–169. Siedow, J.N. (1991). Plant lipoxygenase — structure and function Annu. Rev. Plant Physiol. 42, 145–188. Sigal, E. (1991). The molecular biology of mammalian arachidonic acid metabolism. Am. J. Physiol. 260, L13–28. Slabas, A.R. and Fawcett, T. (1992). The biochemistry and molecular biology of plant lipid biosynthesis. Plant Mol. Biol. 19, 169–191. Smith, M.A. et al. (1990). Electron-transport components of the 1-acyl-2-oleoyl-sn-glycero-3-phosphocholine delta 12desaturase (delta 12-desaturase) in microsomal preparations from developing safflower (Carthamus tinctorius L.) cotyledons. Biochem. J. 272, 23–29. Smith, M.A. et al. (1992). Evidence for cytochrome b5 as an electron donor in ricinoleic acid biosynthesis in microsomal preparations from developing castor bean (Ricinus communis L.). Biochem. J. 287, 141–144. Smith, S. et al. (2003) Structural and functional organization of the animal fatty acid synthase. Prog. Lipid Res. 42, 289–317. Smith, W.L. (1989). The eicosanoids and their biochemical mechanisms of action. Biochem. J. 259, 315–324. Smith, W.L. and Marnett, L.J. (1991). Prostaglandin endoperoxide synthase: structure and catalysis. Biochim. Biophys. Acta. 1083, 1–17. Smith, W.L. and Murphy, R.C. (2002). The eicosanoids: cyclooxygenase, lipoxygenase and epoxygenase pathways. In Biochemistry of Lipids, Lipoproteins and Membranes, Eds. D.E. Vance and J.E. Vance, 4th ed., Elsevier, Amsterdam, pp. 341–371. Smith, W.L. et al. (1991a). The eicosanoids: cyclooxygenase, lipoxygenase, and epoxygenase pathways. In Biochemistry
666
Lipid Metabolism
of Lipids, Lipoproteins and Membranes, 2nd ed., Eds. D.E. Vance and J. Vance, Elsevier, Amsterdam, pp. 297–325. Smith, W.L. et al. (1991b). Prostaglandin and thromboxane biosynthesis. Pharm. Ther. 49, 153–179. Soling, H.D. and Seufert, C.D. Eds. (1978). Biochemical and Clinical Aspects of Ketone body Metabolism, Georg Thiem, Stuttgart. Spaapen, L.J.M., et al. (1977). On the properties of a pea lipoxygenase. Biochim. Biophys. Acta, 488, 517–520. Spector, A.A. et al. (2004). Epoxyeicosatrienoic acids (EETs): metabolism and biochemical function. Prog. Lipid Res. 43, 55–90. Spector, A.A. et al. (1988). Hydroxyeicosatetraenoic acids (HETEs). Prog. Lipid Res. 27, 271–323. Stern, J.R. (1961). Crotonase. In The Enzymes, 2nd ed,, vol. 5, Eds. P.D. Boyer, H. Lardy and K. Myrback, Academic Press, New York, pp. 511–529. Streckert, G. and Stan, H.J. (1975). Conversion of linoleic acid hydroperoxide by soybean lipoxygenase in the presence of guaiacol: identification of the reaction products. Lipids, 10, 847–854. Strittmatter, P. et al. (1974). Purification and properties of rat liver microsomal stearyl coenzyme A desaturase. Proc. Natl Acad. Sci. USA, 71, 4565–4569. Strittmatter, P. et al. (1988). Bacterial synthesis of active rat stearyl-CoA desaturase lacking the 26-residue amino-terminal amino acid sequence. J. Biol. Chem. 263, 2532–2535. Stumpf, P.K. (1970). Fatty acid metabolism in plants. In Lipid Metabolism, Ed. S.J. Wakil, Academic Press, New York, pp. 79–106. Stumpf, P.K. (1987). The biosynthesis of saturated fatty acids. In The Biochemistry of Plants, vol. 9, Eds. P.K. Stumpf and E. E. Conn, Academic Press, New York, pp. 121–136. Sugimoto, Y. et al. (2000). Distribution and function of prostanoid receptors: studies from knockout mice. Prog. Lipid Res. 39, 289–392. Sumper, M. and Riepertinger, C. (1972). Structural relationship of biotin-containing enzymes. Acetyl-CoA carboxylase and pyruvate carboxylase from yeast. Eur. J. Biochem., 29, 237–248. Takai, T. et al. (1987). Primary structure of the biotin-binding site of chicken liver acetyl-CoA carboxylase. FEBS Lett. 212, 98–102. Takai, T. et al. (1988). Primary structure of chicken liver acetylCoA carboxylase deduced from cDNA sequence. J. Biol. Chem. 263, 2651–2657. Theorell, H. et al. (1947). Crystalline lipoxidase. Acta Chem. Scand., 1, 571–576. Thiede, M.A. and Strittmatter, P. (1985). The induction and characterization of rat liver stearyl-CoA desaturase mRNA. J. Biol. Chem. 260, 14459–14463. Thiede, M.A. et al. (1986). Construction and sequence of cDNA for rat liver stearyl coenzyme A desaturase. J. Biol. Chem. 261, 13230–13235. Thomas, K. et al. (1998). Oxygen induces fatty acid (n-6) desaturation independently of temperature. FEBS Lett. 425, 171–174. Thompson, G.A. et al. (1991). Primary structures of the precursor and mature forms of stearoyl-acyl carrier protein desaturase from safflower embryos and requirement of ferredoxin for enzyme activity. Proc. Natl Acad. Sci. USA, 88, 2578–2582.
Tocher, D.R. et al. (1998). Recent advances in the biochemistry and molecular biology of fatty acyl desaturases. Prog. Lipid Res. 37, 73–107. Trotter, P.J. (2001). The genetics of fatty acid metabolism in Saccharomyces cerevisiae. Annu. Rev. Nutr. 21, 97–109. Uchida, K. (2003). 4-Hydroxy-2-nonenal: a product and mediator of oxidative stress. Prog. Lipid Res. 42, 318–343. Umeka, S. and Nozawa, Y. (1984). Effects of transient anaerobic stress on fatty acid desaturation and electron-transport system. Biochim. Biophys. Acta. 792, 25–32. Valentine, R.C. and Valentine, D.L. (2004). Omega-3 fatty acids in cellular membranes: a unified concept. Prog. Lipid Res. 43, 383–402. Vance, D.E. and Vance, J.E. Eds. (2002). Biochemistry of Lipids, Lipoproteins and Membranes 4th ed., Elsevier, Amsterdam. Vanhanen, S. et al. (2000). A consensus sequence for long-chain fatty acid alcohol oxidases from Candida identifies a family of genes involved in lipid omega-oxidation in yeast with homologues in plants and bacteria. J. Biol. Chem. 275, 4445–4452. Veerkamp, J.H. and Maatman, R.G.H.J. (1995) Cytoplasmic fatty acid-binding proteins: their structure and genes. Prog. Lipid Res. 34, 17–52. Veldink, G.A. et al. (1977). Plant lipoxygenases. Prog. Chem. Fats Other Lipids. 15, 131–66. Verhoeven, N.M. and Jakobs, C. (2001) Human metabolism of phytanic acid and pristanic acid. Prog. Lipid. Res. 40, 453–466. Verhue, W. H. and Francke, A. (1972). Heterogeneity of soybean lipoxygenase. Biochim. Biophys. Acta. 284, 4353. Voelker, T.A. et al. (1992). Fatty acid biosynthesis redirected to medium chains in transgenic oilseed plants. Science, 257, 72–74. Volpe, J.J. and Vagelos, P.R. (1973). Saturated fatty acid biosynthesis and its regulation. Annu. Rev. Biochem. 42, 21–60. Volpe, J.J. and Vagelos, P.R. (1976). Mechanisms and regulation of biosynthesis of saturated fatty acids. Physiol. Rev. 56, 339–417. von Wettstein-Knowles, P. (1979). Genetics and biosynthesis of plant epicuticular waxes. In Advances in the Biochemistry and Physiology of Plant Lipids, Eds. L-A Appelqvist and C. Liljenberg, Elsevier, Amsterdam, pp. 1–26. von Wettstein-Knowles, P. (1982). Biosynthesis of epicuticular lipids as analysed with the aid of gene mutations in barley. In Biochemistry and Metabolism of Plant Lipids, Eds. J.F.G.M. Wintermans and P.J.C. Kuiper, Elsevier, Amsterdam, pp. 69–78. von Wettstein-Knowles, P. (1995). Biosynthesis and genetics of waxes. In Waxes: Chemistry, Molecular Biology and Functions, Ed. R.J. Hamilton, The Oily Press, Dundee, U.K., pp. 91–130. Wada, H. and Murata, N. (1989). Synechocystis PCC6803 mutants defective in desaturation of fatty acids. Plant Cell Physiol. 30, 971–978. Wada, H. et al. (1990). Enhancement of chilling tolerance of a cyanobacterium by genetic manipulation of fatty acid desaturation. Nature, 347, 200–203. Wada, K. and Tanabe, T. (1983). Dephosphorylation and activation of chicken liver acetyl-coenzyme-A carboxylase. Eur. J. Biochem. 135, 17–23.
667
10.2
Glycerophospholipids
acid, is a key intermediate in glycerolipid metabolism as illustrated in Figure 10.23. In addition to providing the backbone of the glycerophospholipids, phosphatidate (via diacylglycerol) also acts as a precursor of triacylglycerols (Section 10.3) and the glycosylglycerides (Section 10.4). The biosynthesis of phosphatidate in animals begins with the activation of fatty acids to their acyl-CoAs by one of several chain-length-dependent acyl-CoA ligases (synthetases) (Bloch and Vance, 1977; Groot et al., 1976; Brindley, 1991). The most important locations of the longchain acyl-CoA ligases are the endoplasmic reticulum and mitochondria (Brindley, 1991). Some ligases show preferences towards saturated or unsaturated fatty acids. Of course, some of the enzymes play a role more in degradation (by providing acyl-CoAs for β-oxidation) than in biosynthesis. The peroxisomal acyl-CoA synthetases come into this category (Section 10.1.6). A review of the biochemistry of acyl-CoAs is that by Waku (1992). The main acceptor for acyl-CoA in most tissues is snglycerol 3-phosphate, derived from glycolysis. Glycerol 3phosphate acyltransferase is divided equally between mitochondrial and endoplasmic reticulum in mammalian liver, but in other tissues the endoplasmic reticulum is the main site. The latter enzyme is on the cytoplasmic face. The mitochondrial and endoplasmic reticulum enzymes can be distinguished by their relative sensitivities to heat, proteolytic enzymes, and SH-reagents (Brindley, 1991). The substrate selectivities of different enzymes are discussed by Waku (1992). In general, a variety of saturated and unsaturated acyl-CoAs are used. The second acyltransferase (lysophosphatidate acyl transferase) has a strong preference for unsaturated fatty acids in animals. In most subcellular fractions (and most animal tissues), the second acyltransferase has much higher activity than the first acyltransferase, so that lysophosphatidate does not accumulate. The substrate specificities of the two enzymes are such that the glycerolipids of animals show a preferential location of saturated fatty acids at the sn-1 position and unsaturated fatty acids at the sn-2 position. Some properties of the lysophosphatidate acyltransferase are covered in Brindley (1991). See also Coleman et al. (2002) for details of the above two acyltransferases. The acylation of dihydroxyacetone phosphate and the subsequent reduction of acyldihydroxyacetone phosphate to monoacylphosphatidic acid provide an alternative route for phosphatidic acid synthesis (Brindley, 1991). Acyltransferases that utilize dihydroxyacetone phosphate have been detected in several tissues (Bell and Coleman, 1980) and purified from guinea pig liver peroxisomes (Webber and Hajra, 1992). Another possible alternative pathway for phosphatidic acid synthesis (involving diacylglycerol kinase) has been demonstrated in rat liver (see Bell and Coleman, 1980). However, this enzyme is probably not very important quantitatively for phosphatidic acid formation. See also
Wakil, S.J. (1970). Fatty acid metabolism. In Lipid Metabolism, Ed. S.J. Wakil, Academic Press, New York, pp. 1–48. Wakil, S.J. et al. (1983). Fatty acid synthesis and its regulation. Annu. Rev. Biochem. 52, 537–579. Walker, K.A. and Harwood, J.L. (1986). Evidence for separate elongation enzymes for very-long-chain-fatty-acid synthesis in potato (Solanum tuberosum). Biochem. J. 237, 41–46. Walsh, M.C. et al. (1990). The short chain condensing enzyme has a widespread occurrence in the fatty acid synthetases from higher-plants. Phytochemistry. 29, 3797–3799. Wardale, D.A. et al. (1978). Localization of fatty-acid hydroperoxide cleavage activity in membranes of cucumber fruit. Phytochemistry. 17, 205–212. Watkins, P.A. (1997) Fatty acid activation. Prog. Lipid Res. 36, 55–83. Webber, R.J. and Edmond, J. (1979). The in vivo utilization of acetoacetate, D-(-)-3-hydroxybutyrate, and glucose for lipid synthesis in brain in the 18-day-old rat. Evidence for an acetyl-CoA bypass for sterol synthesis. J. Biol. Chem. 254, 3912–3920. Weber, F. and Grosch, W. (1976). Co-oxidation of a carotenoid by the enzyme lipoxygenase: influence on the formation of linoleic acid hydroperoxides. Z. Lebensm, Unters. Forsch. 161, 223–230. White, S.W. et al. (2005). The structural biology of Type II fatty acid biosynthesis. Ann. Rev. Biochem. 74, 791–831. Witkowski, A. et al. (1987). Molecular cloning and sequencing of a cDNA encoding the acyl carrier protein and its flanking domains in the mammalian fatty acid synthetase. Eur. J. Biochem. 165, 601–606. Witkowski, A. et al. (1991). Structural organization of the multifunctional animal fatty acid synthase. Eur. J. Biochem. 198, 571–579. Worsham, L.M.S. et al. (1988). Chemical cross-linking and its effect on fatty acid synthetase activity in intact chloroplasts from Euglena gracilis. Biochim. Biophys. Acta. 963, 423–438. Wurtele, E.S. and Nickolau, B.J. (1990). Plants contain multiple biotin enzymes: discovery of 3-methylcrotonyl-CoA carboxylase, propionyl-CoA carboxylase and pyruvate carboxylase in the plant kingdom. Arch. Biochem. Biophys. 278, 179–186. Yamamoto, S. (1989). Mammalian lipoxygenases: molecular and catalytic properties. Prostaglandins, Leukotrienes Essent. Fatty Acids 35, 219–229. Yamamoto, S. (1991). “Enzymatic” lipid peroxidation: reactions of mammalian lipoxygenases. Free Radical Biol. Med. 10, 149–159. Yamamoto, S. (1992). Mammalian lipoxygenases: molecular structures and functions. Biochim. Biophys. Acta. 1128, 117–131. Yamamoto, S. et al. (1997). Arachidonate 12-lipoxygenases. Prog. Lipid Res, 36, 23–41. Zimmerman, D.L. and Coudron, C.A. (1979). Identification of traumatin, a wound hormone, as 12-oxo-trans-10-dodecanoid acid. Plant Physiol. 63, 536–541.
10.2
Glycerophospholipids
10.2.1
Biosynthesis
The first reactions in glycerophospholipid synthesis can be regarded as the stepwise acylation of glycerol 3-phosphate. The product of these acylations, phosphatidic 668
Lipid Metabolism
that affect triacylglycerol synthesis and decreases with some drugs that inhibit nonpolar lipid formation (Bell and Coleman, 1980). As shown in Figure 10.23, phosphatidylcholine (the major animal phosphoglyceride) and phosphatidylethanolamine are both synthesized by a (cytidine 5'-diphosphate) CDP-base pathway. This is their major route of formation (Bell and Coleman, 1980; Ansell and Spanner, 1982). Three enzyme steps are required. First, the base is phosphorylated by a kinase enzyme. Choline kinase (EC 1.7.1.32) and ethanolamine kinase (EC 2.7.1.82) are soluble enzymes and have been purified from several tissues. The activities may reside in the same protein (Ulane, et al., 1977), although separate enzymes have been purified from rat liver (Brophy and Vance, 1976). The purified choline kinase is a dimer of mass 42 kDa (Vance, 1991). The second enzyme of the pathway is the cytidylyltransferase (EC 2.7.7.15), and this enzyme is partly soluble and partly associated with the endoplasmic reticulum (see Bell
Brindley and Sturton (1982) for a review of phosphatidic acid formation. Once phosphatidate has been formed, it can be converted to diacylglycerol through the action of phosphatidate phosphohydrolase. This is a key enzyme in glycerolipid metabolism, controlling, on the one hand, the supply of carbon for the major membrane phospholipids, phosphatidylcholine and phosphatidylethanolamine, and the major storage lipid, triacylglycerol, and, on the other, controlling the relative production of anionic phosphoglycerides by a competing cytidylyltransferase (see Figure 10.23 and below). Phosphatidate phosphohydrolase has been well reviewed (Brindley, 1988; see also Brindley, 1991; Coleman et al., 2000). Phosphatidate phosphohydrolase in animal tissues seems to be subject to control by translocation between the endoplasmic reticulum and cytosolic compartments (Brindley, 1991). In rat liver it may catalyse the rate-limiting step in glycerolipid synthesis because it has the lowest in vitro activity of all the enzymes in the pathway, varies directly with dietary or hormonal regimes
S-Adenosylmethionine
OH P-Etn P-Choline (PtdCho) (PtdEtn) (DAG) CMP-Phosphorylcholine CMP-Phosphorylethanolamine N-Acyl sphingosine Serine Pi PPi (ceramide) PPi ∗ CTP CTP Phosphorylcholine ATP Sphingomyelin
Ethanolamine
Phosphorylethanolamine
ADP P (PtdOH)
Choline CTP Acyl-CoA PPi
ADP ATP Ethanolamine ATP
Inositol
Acyl-CoA Glycerol 3-phosphate
P-Ser (PtdSer)
Ptdlns P-CMP
PtdilnsP
CMP PtdilnsP2 P OH
P Pi CDP-DAG OH OH P (PtdGly)
CMP
P OH P (DiPtdGly)
FIGURE 10.23 Phospholipid metabolism in animals. The relative thickness of lines shows the approximate carbon flow down each pathway. The asterisk (*) shows that other phospholipids can act as acceptors for serine in base exchange.
669
10.2
Glycerophospholipids
more physiologically relevant (Bell and Coleman, 1980), since most of the phosphatidylglycerol (except in lungs; Section 11.7) is used for diphosphatidylglycerol formation. The phosphatidylglycerol phosphate synthetase has been localised in the inner mitochondrial membrane, but the phosphatase (see Figure 10.23) is soluble. Together with CDP-diacylglycerol, phosphatidylglycerol is a cosubstrate for diphosphatidylglycerol synthesis in animals (see Figure 10.23). This contrasts with the bacterial pathway (see Figure 10.28). Diphosphatidylglycerol (cardiolipin) synthesis in different organisms has been discussed in detail by Schlame et al. (2000) who also discuss the lipid’s function in mitochondria and relevance to a number of human diseases. A third polyglycerophospholipid, bis(monoacylglycerol) phosphate, comprises less than 1% of the phospholipids of most animal tissues except for alveolar (lung) macrophages, where it represents 14 to 18% of the phospholipids. Its pathway of formation was proposed following experiments with macrophage-like cells. Phosphatidylglycerol was hydrolysed by phospholipase A2 and the lyso product acylated using another phospholipid. This yields bis(monoacylglycero)phosphate (BMP). After reorientation of the glycerol backbone, a second acylation (again using a phospholipid as donor) forms sn-1: sn-1'BMP. Spontaneous rearrangement causes the acyl residues to move to the sn-3 positions (Amidon et al., 1995). Phosphatidylinositol is also formed from CDP-diacylglycerol, though in this case only one enzyme step is involved (see Figure 10.23). Further phosphorylation to phosphatidylinositol 4-phosphate and phosphatidylinositol 4,5-bisphosphate are catalysed by kinase reactions. The first kinase appears to be present in isoforms and these are integral membrane proteins. One enzyme with a low Km for ATP (20-70 µM) and a molecular mass of 55 kDa has been purified to homogeneity (Downes and MacPhee, 1990). Phosphatidylinositol 4-phosphate kinase can be isolated from cytosolic and membrane sources. It appears to be a peripheral membrane protein (Vance, 1991). The above two kinases are important in the “phosphatidylinositol cycle” (Figure 10.24), which is connected with the important function of phosphatidylinositol 4,5-bisphosphate as a precursor of second messages that activate many regulatory processes in animal cells (see Section 10.2.2 and Section 10.6.3; also McPhee, 2002). Over recent years, a large number of inositol phospholipids have been identified, particularly those with phosphates at the 3-position and which have important signaling functions (Section 10.6) and Rameh and Cantley, 1997). Interconversions of inositol phospholipids through the action of various kinases and phosphatases are shown in Figure 10.25. See Freeman et al. (1998) for a review on phosphoinositide kinases and Vanhaesebroeck et al. (2001) for one on the synthesis and function of the 3-phosphorylated inositol lipids. Phospholipids cannot only be remodeled by base exchange reactions (Taki et al., 1978), but also by reacylation
and Coleman, 1980). The soluble protein in animals can be aggregated under various conditions and its activity raised by association with membranes (see Vance, 1991). The enzyme seems to be the rate-limiting step in the CDPcholine pathway in animals (Vance and Choy, 1979). The enzyme has been purified (Feldman and Weinhold, 1987), and the cDNA coding for its protein identified and expressed (Vance, 1990). The equivalent enzyme, CDPethanalominephosphate cytidylyltransferase (EC 2.7.7.14), has been partly purified from rat liver (see Vance, 1991). CDP-choline cytidylyltransferase is normally thought to be the most important enzyme for regulation of phosphatidylcholine formation (Vance, 2002). Choline phosphotransferase (EC 2.7.8.2) and ethanolamine phosphotransferase (EC 2.7.8.1) are located in the endoplasmic reticulum and catalyse the final step in the CDP-base pathway. Their complete purification has yet to be achieved from animals. The two enzymes seem to have somewhat different properties and appear to be distinct proteins (see Vance, 1991). Phosphatidylcholine can also be synthesized by the stepwise methylation of phosphatidylethanolamine (Figure 10.23). Methyl groups are transferred from S-adenosyl-Lmethionine, and this pathway is a minor one in animals (20% of liver phosphatidylcholine synthesis, but undetectable in other tissues; Bell and Coleman, 1980). In liver, the three transmethylation reactions are catalysed by a single enzyme with a molecular mass of about 18 kDa. The rate of conversion of phosphatidylethanolamine to phosphatidylcholine appears to be regulated by substrate supply (Vance, 1991). A review of phosphatidylethanolamine methylation has been published (Vance and Ridgway, 1988) and Vance (2002) has updated information and speculated on the role of the methylation pathway in forming phosphatidylcholine. Phosphatidylserine accounts for 5 to 15% of the total phospholipids in animal cells. It is made by a base exchange reaction (see Figure 10.23) in which the head group of a preexisting phospholipid is exchanged for serine. The enzyme has been purified, and its base exchange activity is not due to phospholipase D activity (see Vance, 1991). Studies in CHO cells showed that there were two phosphatidylserine synthases, one using phosphatidylcholine and a second, phosphatidylethanolamine (Kuge and Nishijima, 1997). Instead of being dephosphorylated, phosphatidate can be converted to CDP-diacylglycerol. This reaction, catalysed by phosphatidate:CDP-diacylglycerol cytidylyltransferase (EC 2.7.7.41), is important for the synthesis of negatively charged phospholipids, such as phosphatidylglycerol, diphosphatidylglycerol, and the inositol phospholipids (see Figure 10.23). The cytidylyltransferase has been purified from animals (Bell and Coleman, 1980) and cDNAs identified for two isoforms (Vance, 2002). Phosphatidylglycerol can be made by mitochondrial and microsomal fractions from most animal tissues. However, the mitochondrial enzymes are probably 670
Lipid Metabolism
R1
ATP
R1
ADP
R2
ATP
ADP
R2
R2
I
P
R1
4P
I
P
P
I
4,5P2
CMP I
I
P
I
P2
I
P3
R1 2
R
P
CMP R1
PPi
R1
ATP R2
R CTP
FIGURE 10.24
ADP
2
OH
P
The phosphatidylinositol cycle. I = inositol. (From Vance 1991.)
PI
PI-5-P PI-3,5-P2
PI-3-P PI-4-P
PI-3,5-P2
PI-4,5-P2 PI-3,4-P2
PI-3,4,5-P3
FIGURE 10.25 Interconversion of phosphoinositides. PI=phosphatidylinositol. The connecting recactions are catalysed by 3-, 4-, or 5- kinases and 3-, 4-, and 5- phosphatases.
their role in the production of both acyl and alkyl glycerolipids has been reviewed (Hajra, 1995). The coregulation of sterol and phospholipid metabolism is covered by Ridgeway et al. (1991). The overall pathways for phosphoglyceride synthesis in plants have been reviewed (Mudd, 1980; Moore, 1982; Harwood, 1979, 1989). For a recent update, see Dormann (2005). In general, the formation of these lipids is similar to that in animals, but there are some points of difference (Figure 10.26). First, there are two systems for the initial acylation of glycerol 3-phosphate. In the chloroplast-located pathway, acyl-ACPs are used and the substrate specificity of the enzymes results in the preferential formation of sn-1-oleoyl, sn-2-palmitoyl molecular species. Subsequent utilisation of such phosphatidate for lipid synthesis retains this (sn-1 C18, sn-2 C16) distribution of fatty acids and gives rise to what have been called “prokaryotic” molecular species by analogy with cyanobacterial lipid metabolism (see Roughan and Slack, 1984; Harwood, 1989; Joyard and Douce, 1987). A second glycerol 3-phosphate acylation system is located in the endoplasmic reticulum, uses acyl-CoAs, and has a substrate specificity such that more saturated fatty acyl groups
reactions. The latter are extremely important in determining the final molecular species of lipid that are synthesised as well as allowing the participation of specific fatty acids in particular metabolic processes (e.g., release of arachidonate for prostaglandin formation (see van den Bosch, 1982). The entire subject of reacylation and molecular species turnover has been well reviewed by one of the major research workers in this field, Bill Lands (Hills and Lands, 1970). More recent discoveries are covered by Bell and Coleman (1980) and by Hawthorne and Ansell (1982) (see also MacDonald and Sprecher, 1991). Remodelling can occur on either the sn-1 or sn-2 positions (Vance, 2002). The above discussion has centred on the formation of diacylglycerophospholipids, but many phospholipids in nature have O-alkyl bonds. These include such important molecules as the plasmalogens and platelet-activating factor (PAF). For reviews on the metabolism, regulation and function of ether-linked glycerolipids, refer to Synder et al. (1985, 2002). Enzymes of the acyl dihydroxyacetone phosphate pathway for glycerolipid synthesis are located in peroxisomes as well as the endoplasmic reticulum. Their physiological function in peroxisomes as well as 671
10.2
Glycerophospholipids
S-Adenosylmethionine (1)
P-Etn P-Choline OH (DAG) (PtdEtn) (PtdCho) CMP-Phosphorylethanolamine CMP-Phosphorylcholine Serine Pi PPi PPi CTP CTP (2) Phosphorylcholine ATP
ADP
Choline
Phosphorylethanolamine ADP
P (PtdOH)
ATP Ethanolamine
CTP PPi Acyl-CoA (Acyl-ACP)(3) Glycerol 3-phosphate Acyl-CoA (Acyl-ACP)
Ethanolamine
P-Ser (PtdSer) ATP
Inositol Ptdlns P-CMP
PtdlnsP
CMP PtdlnsP2 P OH
P
Pi OH
CDP-DAG
P OH
OH P
CMP
(PtdGly)
P (DiPtdGly)
FIGURE 10.26 Phospholipid metabolism in plants. The relative thickness of arrows indicates the carbon flux along individual reactions. Notes: (1) Other pathways for incorporating ethanolamine into PtdCho exist (see text); (2) other phospholipids can serve for base exchange; (3) acyl-ACPs are used by chloroplasts, but acyl-CoAs by the endoplasmic reticulum.
on various water-soluble intermediates, such as ethanolaminephosphate, which are then incorporated into phosphatidylcholine via phosphatidyldimethyl- (or methyl-) ethanolamine (see Williams and Harwood, 1994). Phosphatidylserine, phosphatidylglycerol, and phosphatidylinositol seem to be synthesized in plants by pathways similar to those discussed above for animals (Mudd, 1980; Moore, 1982; and see Figure 10.23 and Figure 10.26). Phosphatidylglycerol synthesis has been demonstrated in chloroplasts (see Harwood, 1989) in keeping with the major role of this lipid in thylakoid membranes. The metabolism of inositol lipids and their function in cellular regulation in plants has been reviewed recently (Drobak, 2005). N-Acylphosphatidylethanolamine is a minor phospholipid in plants (Dormann, 2005), but there is some interest in it as a source of signalling metabolites (see Chapman, 2004).
are usually located at the sn-1 position, just as in animals (see Harwood, 1989). The endoplasmic reticulum acylations are responsible for extrachloroplastic phospholipid synthesis and triacylglycerol formation (Harwood and Page, 1993), as well as some chloroplast lipid synthesis (see Roughan and Slack, 1984; Browse and Somerville, 1991). The major plant extrachloroplastic lipid, phosphatidylcholine, is mainly made by the CDP-base pathway (Harwood, 1979; Moore, 1982). Separate choline and ethanolamine kinases have been purified from soybean (Harwood, 1979), and the cytidylyltransferase step appears to be rate-limiting (Price-Jones and Harwood, 1986). Separate base transferases for choline and for ethanolaminephosphate appear to be present in soybean (see Harwood, 1979). Although evidence has been produced that plants can convert phosphatidylethanolamine to phosphatidylcholine by stepwise methylation (Mudd, 1980; Moore, 1982), some recent experiments suggest that methylation occurs 672
Lipid Metabolism
reviewed the pathways of synthesis in S.cerevisiae in detail with information on the genes encoding individual enzymes. Particularly informative sections deal with the genetic and biochemical regulation of phospholipid synthesis and its interaction with other metabolic pathways. Phospholipid biosynthesis in microorganisms has been well reviewed by Pieringer (1989). Major pathways for Escherichia coli are shown in Figure 10.28. Several points are of note. First, the acylation of glycerol 3-phosphate uses acyl-ACPs and, thus, phospholipid synthesis is coupled directly to fatty acid formation (Jackowski et al., 1991). Secondly, as in yeasts (see Figure 10.27), phosphatidylserine is made from CDP-diacylglycerol and can be decarboxylated to phosphatidylethanolamine. Thirdly, the
Pathways for the biosynthesis of phospholipids in yeast are generally similar to those demonstrated in other eukaryotes, except that phosphatidylserine is made exclusively from CDP-diacylglycerol (Figure 10.27) and not by base exchange. Phosphatidylserine is also converted to phosphatidylethanolamine by decarboxylation. Work with yeast has been particularly useful as a model system for other eukaryotes because of a large knowledge base in classical genetics and the ability to elucidate and manipulate individual genes coding for enzymes in the pathways. Purification and properties of CDP-diacylglycerol synthetase, phosphatidylserine synthetase, phosphatidylinositol synthetase, and phosphatidylinositol kinase are described by Carman and Henry (1989). Carman and Henry (1999)
Glycerol 3-phosphate Acyl CoA 1 CoA 1-Acylglycerol 3-phosphate
Glucose 6-phosphate
Acyl CoA
15
CoA Inositol-1-P
Phosphatidic acid
CTP
Pi 14
CoA Acyl CoA
Diacylglycerol
Triacylglycerol Ethanolamine ATP 11
Ethanolamine-P
Inositol CMP 16
3 Phosphatidylinositol
Phosphatidylserine
CDP-Ethanolamine
12
Pi
CDP-Diacylglycerol Serine
CMP CTP PPi
ADP
PPi
2
13
CMP
4
Glycerol 3-phosphate
CO2 Phosphatidylethanolamine AdoMet
17 CMP
5 AdoHcy Phosphatidylmonomethylethanolamine AdoMet
Phosphatidylglycerophosphate 18
6 AdoHcy
ATP ADP Choline
8
CTP PPi
Choline-P
9
Phosphatidyldimethylethanolamine AdoMet 7 10 CDP-Choline AdoHcy CMP
Phosphatidylcholine
Phosphatidylglycerol 19 Cardiolipin
FIGURE 10.27 Phospholipid biosynthetic pathways in Saccharomyces cerevisiae. The indicated reactions are catalysed by the following enzymes: (1) glycerol 3-phosphate acyltransferase; (2) CDP-DG synthetase; (3) PS synthetase; (4) PS decarboxylase; (5) PE methyltransferase; (6) and (7) phospholipid methyltransferase; (8) choline kinase; (9) cholinephosphate cytidylyltransferase; (10) cholinephosphotransferase; (11) ethanolamine kinase; (12) ethanolaminephosphate cytidylyltransferase; (13) ethanolaminephosphotransferase; (14) PA phosphatase; (15) I-1-P synthetase; (16) PI synthetase; (17) PGP synthetase; (18) PGP phosphatase; (19) CL synthetase. (From Carman and Henry (1989).)
673
10.2
Glycerophospholipids
O O
OCR
RCO OH Diacylglycerol
OH HO OH Glycerol
Pi
ATP ADP
OH 2RC ACP
HO
H2O
O
ATP O
ADP
O O
O
HO
OCR
OCR CTP RCO RCO P PP CYT PPi P Glycerol 3-phosphate Phosphatidic acid CDP-diacylglycerol
OH
OCR P OH RCO CMP P Phosphatidylglycerophosphate P
NAD+
H2O
L-Serine Pi
OH NADH (NADPH) O O CMP P O + OCR NH 3 Dihydroxyacetone RCO phosphate PCH2CHCO2– Phosphatidylserine
O O
OCR
RCO
PG
OCR
CO2 + PCH2CH2NH3 Phosphatidylethanolamine RCO
O O
OCR
RCO P
P OH O
Oligosaccharide Diacyglycerol
Glycerol O OCR
RCO Diphosphatidylglycerol (cardiolipin)
FIGURE 10.28
OH OH
P Phosphatidylglycerol
O O
O O
Membrane-derived oligosaccharide(glycerol-1-P)n
Phospholipid metabolism in Escherichia coli. (From Pieringer (1989).)
Methods for the isolation and study of numerous enzymes of phospholipid metabolism are given in Dennis and Vance (1992). A summary of phospholipid biosynthesis in different organisms is found in Gurr et al. (2002) and in various chapters of Vance and Vance (2002).
CDP-base pathway for phosphatidylethanolamine is not used. And, finally, diphosphatidylglycerol is made by a reaction involving two phosphatidylglycerol molecules (Figure 10.28). See also Gurr et al. (2002) and Heath et al. (2002). Most research in bacteria has concentrated on E.coli, a Gram-negative bacterium with a rather restricted habitat. It cannot be emphasized too strongly that pathways and enzymes studied in E.coli may not be typical (or even present) in all bacteria (Harwood and Russell, 1984). The overall pathways for phospholipid synthesis in a Gram-positive bacterium, Streptococcus falcium, are shown in Figure 10.29. There has been recent interest in the biosynthesis and occurrence of phosphatidylcholine in bacteria. The occurrence of this lipid in bacteria is thought to be underestimated and, indeed, in Acetobacter aceti represents 73% of the phospholipids. Although it was always thought to be made by methylation of phosphatidylethanolamine, many bacteria use a phosphatidylcholine synthase, which is a member of the CDP-alcohol phosphatidyltransferase group of enzymes. Moreover, the CDP-choline pathway may be present in some bacteria (Sohlenkamp et al., 2003). Individual features of lipid metabolism in bacteria and other prokaryotes or microbes are covered by Pieringer (1989), Jackowski et al. (1991), and Cronan (2003). Smith (1993) has reviewed phospholipid synthesis in protozoa.
10.2.2
Breakdown
The breakdown of glycerophospholipids is catalysed by a series of phospholipases designated A, B, C and D depending on their positions of attack (Figure 10.30). In addition, phosphatidic acid phosphatase is important, but this has already been dealt with (Section 10.2.1). Two phospholipase A specificities are recognized, and these are named A1 and A2, depending on the position of the ester hydrolysed in the diacylphosphoglyceride. Phospholipase B is the name often used for an enzyme that is a monoacylphosphoglyceride acylhydrolase and can be active at either position. However, they are active on diacylphosphoglycerides, whereas lysophospholipases remove the remaining acyl group from a monoacyl phospholipid (see below for further discussion). For reviews of the basic features of phospholipase action, the reader is referred to Hill and Lands (1970), Dawson (1973), van den Bosch (1982), Dennis (1983), Waite (1987), and Wilton and Waite (2002).
674
Lipid Metabolism
O RC ACP
O H2O OCR
O
OH HO
Pi
O
O
RCO
RCO
P P Glycerol-3-phosphate Phosphatidic acid CTP
OH ADP ATP Diacyglycerol
O
CH2OH O O RCO O
OCR UDP- UDP glucose
Monoglucosyldiacylglycerol
O
OCR
RCO P
O
Kojibiosyldiacylglycerol
OCR
RCO Diphosphatidylglycerol (cardiolipin)
OCR
RCO PP cyt CDP-diacylglycerol
Diacylglycerol
Glycerol
Glycerophosphate Phosphatidyglycerol CMP O O
OCR
RCO
O CH2OH P O
O
H2O
O
P OH
OCR
RCO
P Phosphatidylglycerophosphate
O OH OH
O RCO
OCR
O
O
O RCO
OCR
Phosphatidylkojibiosyldiacylglycerol
O
tRNA
OCR
O
CH2OH O
P Phosphatidylglycerol
Pi
HO
OOCR OOCR O
RCO O
AminoacyltRNA O
OCR
O
O
P OH O
O
CH2OH O O RCO O CH2OH O
O
PPi O
UDP- UDP OCR glucose
CH2OH P O
OAA OH
O
P Aminoacylphosphatidylglycerol
HO
O
O OCR
RCO O HO
P
OH P
CH2OH O O n
O
Lipoteichoic acid
FIGURE 10.29 Lipid metabolism in Streptococcus faecium (= faecalis) ATCC9790. For simplicity not all hydroxyl groups on glucose are shown. (From Pieringer (1989).)
B
A2 R2
C O
been published. The interpretation of the kinetics of phospholipase inhibition has been the subject of considerable attention (see Harwood and John, 1990) and suicideinhibitory bifunctionally-linked substrates have been developed as phospholipase A2 inhibitors (Washburn and Dennis, 1991). Phospholipase A enzymes — particularly those from snake venoms or digestive secretions — have been widely studied. Phospholipase A1 is found in microsomal and liposomal fractions (cf. Newkirk and Waite, 1971; Gatt, 1968). It specifically deacylates phosphatidylcholine or phosphatidylethanolamine at the 1-position. Both of these substrates are hydrolysed at the same rate by the adrenal medulla lysosomal enzyme, but that from brain prefers phosphatidylcholine as substrate. Detergents will increase phosphatidylethanolamine hydrolysis by the brain enzyme. A phospholipase A1 (which is relatively specific for phosphatidylglycerol) has been reported from the spores of some bacteria (Raybin et al., 1972), but most bacterial enzymes are unspecific for either the 1- or the 2-positions. Examples of phospholipase A (unspecific) enzymes from bacteria include the so-called detergent-resistant
A1
O
CH2O
C
CH
O O
CH2O
P
R1
O
X
O– C
D
FIGURE 10.30 Position of phosphoglyceride hydrolysis for different phospholipases.
In common with many other lipid-metabolising enzymes, the assay of phospholipases deserves careful consideration. Principally, the potential difficulties are due to the water-insoluble nature of the substrate, the effects of various incubation additions (e.g., solvents, detergents), which can make a big difference to activity, and the problems in kinetic interpretation. Various assay methods have been discussed (Waite, 1987) and an important volume of Methods in Enzymology (Dennis, 1991) has
675
10.2
Glycerophospholipids
e.g., the human phospholipase differs from the others in only a single amino acid residue out of 123 to 125 in each case (Verheij et al., 1983). Further details are given in Waite (1987). A gene encoding the cobra venom phospholipase A2 has been expressed in E. coli and the recombinant enzyme shown to have the expected activity characteristics (Kelley et al., 1992). Lysophospholipases are probably distributed as widely as the phospholipases A (Waite, 1987). A true definition of lysophospholipases is difficult, as discussed by van den Bosch (1982). Indeed, a number of enzymes in this group also have activity against diacyl phospholipids and, therefore, are more properly referred to as phospholipase-B or -A1 enzymes (van den Bosch et al., 1974). Recent examples are a human phosphatidylserine — specific phospholipase A1 that also has lysophosphatidylserine lysophospholipase activity (Nagai et al., 1999) and two phospholipase B lysophospholipases from S. cerevisiae (Merkel et al., 1999). Also, there are several lysophospholipases that show esterase activity and whose classification is difficult. For more discussion, see Waite (1987) and specific chapters on individual lysophospholipases or the so-called phospholipase B from Penicillium notatum in Dennis (1991). Phospholipase C enzymes are secreted by several bacteria, particularly pathogens, such as Clostridium spp. These proteins are zinc metalloenzymes and some also require Ca2+ for activity. The phospholipase C enzymes vary widely in substrate specificity (van den Bosch, 1982). The enzyme from C. perfringens attacks choline-containing lipids most readily. Such hydrolysis requires a positive zetapotential brought about by long-chain cations or by the addition of Ca2+ or Mg2+. Phosphatidylcholine breakdown in the presence of Ca2+ is greatly increased by the addition of specific concentrations of sodium deoxycholate. The phospholipase C from Bacilus cereus attacks a large number of phospholipids (e.g., PG, DPG), which are poorly hydrolysed by the C. perfringens enzyme. It will also hydrolyse short-chain phosphoglycerides, which are poor substrates for the latter enzyme (Dawson, 1973). The use of phospholipase C enzymes in the preparation of diacylglycerols for molecular species examination is discussed by Christie (1982). Mammalian phospholipase C enzymes are now known to be central in regulatory processes (Waite, 1987). Most, if not all, of these enzymes are specific for phosphatidylinositol or its phosphorylated derivatives. Specific enzymes are discussed in Dennis (1991) and their role in transmembrane signaling in various organisms (Berridge and Irvine, 1989) and in higher plants (Hetherington and Drobak, 1992) has been reviewed (see also Section 10.6). Specific examples of important functional aspects of such phospholipase C activity are given by Turk et al. (1987) and by Cockcroft (1992) in relation to insulin secretion by pancreatic islets and in neutrophils, respectively. Plant phospholipase C enzymes can be divided into three groups: (1) enzymes acting on the phosphoinositides,
enzyme from the outer membrane of E. coli and a cytoplasmic enzyme from the same bacterium (Nakagawa et al., 1991). These enzymes differ considerably in their properties (Harwood and Russell, 1984). A number of Gram-negative bacteria have phospholipase A activity in their outer membranes, but it is often not determined whether they are phospholipase A1 or A2. Several other lipases have been noted to have phospholipase (usually phospholipase A1) activity. These include rat hepatic lipase (Waite et al., 1991) and milk lipoprotein lipase (BengtssonOlivecrona and Olivecrona, 1991). The specificity of phospholipase A2 enzymes varies considerably even among those from snake venoms (see Dawson, 1973). The enzymes remove a fatty acyl group on the glycerol carbon adjacent to the phosphoryl substituent — i.e., position 2 in a natural phosphoglyceride. If the phosphoryl substituent is at the 2-position, then only a fatty acid with the correct steric configuration is hydrolysed. Substrate specificities also vary markedly with the chain length and degree of unsaturation. For the snake venom enzyme hydrolysing phosphatidylcholine, the rate of hydrolysis is in the order: (1-unsat., 2-sat.) > (1-unsat., 2unsat.) > (1-sat., 2-polyunsat.) > (1-sat., 2-mono-unsat.) > (1-sat., 2-sat.). In contrast, the pancreatic enzyme exhibits no preference for chain length or degree of unsaturation. The hydrolysis of certain phospholipids can be increased in different ways for the various phospholipase A2 enzymes, e.g., by detergent or diethyl ether addition. This aspect and comments on the underlying physicochemical mechanisms are discussed by Dawson (1973), Slotboom et al. (1982), and Waite (1987), where further references will be found. Specific details of purification and characterization of individual phospholipases are given in Dennis (1991). Receptors for phospholipase A2 enzymes of various types have been discovered and studied. They are thought to play crucial roles in the physiological actions of phospholipases A2 (Ohera et al., 1995). Based on sequence data, phospholipase A2 enzymes from animals are classified into 10 groups, which can be simplified into three major types, based on their physiological properties and function. These are (1) the secretory, low-molecular weight PLA2; (2) cytosolic Ca2+-dependent PLA2; and (3) the intracellular Ca2+-independent PLA2 (Balsinde et al., 1991). Groups (1) and (3) are also reported from plants (Wang, 2001). Enzymes belonging to (1), secretory phospholipase A2 enzymes from plants and their regulatory and catayltic properties, have been reviewed by Lee et al., (2005). Because of the ready availability of the pancreatic phospholipase A2, this enzyme has been thoroughly examined by enzymologists. The phospholipase is secreted as a zymogen and activated by cleavage of a peptide of three to seven amino acids. The amino acid sequence of the active enzyme has been determined for a number of different preparations (pig, ox, horse, and human). The different preparations show a high degree of homology, 676
Lipid Metabolism
tissues the enzyme may control the overall rate of lipid (triacylglycerol) deposition. A comprehensive review of the enzyme in different organisms and tissues has been published by Brindley (1988) (see also Dennis, 1991). Many plant tissues possess high activities of nonspecific lipid acyl hydrolases, which can also give rise to problems with the solvent extraction of plants. Acyl hydrolases will hydrolyse fatty acids from a large number of lipids including phosphoglycerides (cf. Galliard, 1980) The enzymes have been purified from several tissues and their substrate specificities and other properties determined. Potato tubers and leaves from Phaseolus spp. are particularly rich sources (cf. Galliard, 1980) and the position of bond cleavage has been determined for the latter enzyme to be on the fatty acid side of the oxygen ester bond (Burns et al., 1980). An important example of a plant acyl hydrolase is patatin, which is a storage protein from potato tubers. Patatin-like proteins are produced in a number of plants in response to stresses, such as virus or fungal infection (Meijer and Munnik, 2003). For an overall review of the functions of phospholipases in such areas as the phosphatidylinositol cycle, the archidonate cascade, in the digestion of dietary fat in lipoprotein metabolism and in snake venoms, see Waite (1987) and also Vance (1991). Phospholipase A2 has also been invoked in the protection of membranes from lipid peroxidation damage (van Kuijk et al., 1987).
(2) nonspecific phospholipase Cs that act on phosphatidylcholine and some other phospholipids, and (3) a phospholipase C that hydrolyses glycosylphosphatidylinositol anchors on proteins. The first type of phospholipase C has many important regulatory functions, including responses to stimuli, in plants (Wang, 2001). There are few examples of phospholipase D (EC 3.1.4.4) in bacteria (Harwood and Russell, 1984; Cockcroft, 1996), but many plant tissues exhibit high activity. Plant phospholipase D is also a transferase enzyme that catalyses transphosphatidylation to an acceptor molecule. If phospholipase D enzymes are not fully inactivated, then this reaction will lead to the artifactual formation of phosphatidylmethanol in plant tissues extracted with methanolic solutions (Harwood, 1980). A variety of water-soluble primary alcohols can be used as acceptors (e.g., propanol, ethanol, ethylene glycol, glycerol). Most enzymes are stimulated in the presence of linear aliphatic ethers (Dawson, 1973; Galliard, 1980). A possible function in plants for phospholipase D in synthesizing phospholipids has now been largely discounted because the phosphatidylglycerol thus made is a racemic mixture, unlike the naturally occurring phospholipid (Yang et al., 1967). However, Batrakov et al. (1975) found evidence for stereospecificity in the transfer. Roughan and Slack (1976) have claimed that phospholipase D is a structural protein that, under certain nonphysiological conditions, possesses enzymic activity. In plants, phospholipase D (PLD) enzymes can be divided into three groups, based on their properties: (1) a conventional PLD that is most active at 20 to 100 mM Ca++, (2) a polyphosphoinositide-dependent PLD that is active at micromolar Ca++, and (3) a phosphatidylinositolspecific PLD that is independent of Ca++. For Arabidopsis, five groups based on gene sequence were identified. PLDs in plants present interesting properties for activation and regulation and appear to play physiological roles in a wide range of stress conditions. These include freezing, drought, wounding, pathogen infection, nutrient deficiency and air pollution (Wang, 2001). There has been some recent interest in the possible role of mammalian phospholipase D in signal transduction, in that a wide variety of agents (including hormones, neurotransmitters and growth factors) have been shown to activate a phospholipase D to hydrolyse phosphatidylcholine (see e.g., Hurst et al., 1990; Xie and Dubyak, 1991). Various GTPases are known to activate the enzyme in mammals and its physiological relevance has been discussed (Cockcroft, 1996). This article also provides information on the enzymology of phospholipase Ds from different organisms and earlier reviews on the subject. Yamane et al. (1989) discuss the commercial use of phospholipase D enzymes for transphosphatidylation purposes. Phosphatidate phosphohydrolase, although a phospholipid degradative enzyme, is mainly involved in the biosynthesis of lipids (Gurr and Harwood, 1991). In mammalian
References Amidon, B. et al. (1995). Biosynthetic conversion of phosphatidylglycerol to sn-1:sn-1 bis(monoacylglycero) phosphate in a macrophage-like cell line. Biochemistry. 34, 5554–5560. Ansell, G.B. and Spanner, S. (1982). Phosphatidylserine, phosphatidylethanolamine and phosphatidylcholine. In Phospholipids, Eds. J.N. Hawthorne and G.B. Ansell, Elsevier, Amsterdam, pp. 1–49. Balsinde, J. et al. (1999). Regulation and inhibition of phospholipase A2. Annu. Rev. Pharmacol. Toxicol. 39, 175–189. Batrakov, S.G. et al. (1975). Steric analysis of glycerophospholipids by circular dichroism. Stereospecificity of phospholipase D catalyzed transesterification. Biochem. Biophys. Res. Commun., 66, 755–762. Bell, R.M. and Coleman, R.A. (1980). Enzymes of glycerolipid synthesis in eukaryotes. Annu. Rev. Biochem. 49, 459–487. Bengtsson-Olivecrona, G. and Olivecrona, T. (1991). Phospholipase activity of milk lipoprotein lipase. In Methods in Enzymology, vol. 197, Ed. E.A. Dennis, Academic Press, New York, pp. 345–356. Berridge, M.J. and Irvine, R.F. (1984). Inositol triphosphate, a novel second messenger in cellular signal transduction. Nature, 312, 315–321. Berridge, M.J. and Irvine, R.F. (1989). Inositol phosphates and cell signalling. Nature, 341, 197–205. Bloch, K. and Vance, D. (1977). Control mechanisms in the synthesis of saturated fatty acids Annu. Rev. Biochem. 46, 263–298.
677
10.2
Glycerophospholipids
Brindley, D.N. Ed. (1988) Phosphatidate Phosphohydrolase, vols. I and II, CRC Press, Boca Raton, FL. Brindley, D.N. (1991). Metabolism of triacylglycerols. In Biochemistry of Lipids, Lipoprotein and Membranes, Eds. D.E. Vance and J. Vance, Elsevier, Amsterdam, pp. 171–203. Brindley, D.N. and Sturton, R.G. (1982). Phosphatidate metabolism and its relation to triglycerol biosynthesis. In Phospholipids, Eds. J.N. Hawthorne and G.B. Ansell, Elsevier, Amsterdam, pp. 179–213. Brophy, P.J. and Vance, D.E. (1976). Copurification of choline kinase and ethanolamine kinase from rat liver by affinity chromatography. FEBS Lett., 62, 123–125. Browse, J. and Somerville, C.R. (1991). Glycerolipid synthesis: biochemistry and regulation. Annu. Rev. Plant Physiol., 42, 467–506. Burns, D.D. et al. (1980). Properties of acyl hydrolase enzymes from Phaseolus multiflorus leaves. Phytochemisty. 19, 2281–2285. Carman, G.M. and Henry, S.A. (1989). Phospholipid biosynthesis in yeast. Annu. Rev. Biochem., 58, 635–669. Carman, G.M. and Henry, S.A (1999). Phospholipid biosynthesis in the yeast, Saccharomyces cerevisiae and interrelationship with other metabolic processes. Prog. Lipid Res. 38, 361–399. Chapman, K.D. (2004). Occurrence, metabolism and prospective functions of N-acylethanolamines in plants. Prog. Lipid Res. 43, 302–327. Christie, W.W. (1982) Lipid Analysis, 2nd ed., Pergamon, Oxford. Cockcroft, S. (1992). G-protein-regulated phospholipases C, D and A 2 -mediated signalling in neutrophils. Biochim. Biophys. Acta, 1113, 135–160. Cockcroft, S. (1996) Phospholipase D: regulation by GTPases and protein kinase C and physiological relevance. Prog. Lipid Res. 35, 345–370. Cronan, J.E. (2003). Bacterial membrane lipids: where do we stand? Ann. Rev. Microbiol. 57, 203–224. Dawson, R.M.C. (1973). Specificity of enzymes involved in the metabolism of phospholipids. In Form and Function of Phospholipids, Eds. G.B. Ansell, R.M.C. Dawson and J.N. Hawthorne, Elsevier, Amsterdam, pp. 97–116. Dennis, E.A. (1983). Phospholipases. In The Enzymes, vol. XVI, Ed. P. Boyer, Academic Press, New York, pp. 307–353. Dennis, E.A. Ed. (1991) Methods in Enzymology, vol. 197, Academic Press, New York. Dennis, E.A. and Vance, D.E. Eds. (1992) Methods in Enzymology, vol. 209, Academic Press, New York. Dormann, P. (2005). Membrane lipids. In Plant Lipids: biology Utilisation and Manipulation, Ed. D.J. Murphy, Blackwell Publishing, Oxford, U.K., pp. 123–161. Downes, C.P and MacPhee, C.H. (1990). Myo-inositol metabolites as cellular signals. Eur. J. Biochem. 193, 1–18. Drobak, B.K. (2005) Inositol-containing lipids: roles in cellular signalling. In Plant Lipids: Biology, Utilisation and Manipulation, Ed. D.J. Murphy, Blackwell Publishing, Oxford, U.K.. pp. 303–328. Feldman, D.A. and Weinhold, P.A. (1987). CTP: phosphorylcholine cytidylyltransferase from rat liver. Isolation and characterization of the catalytic subunit. J. Biol.Chem. 262, 9075–9081.
Fruman, D.A. et al. (1998). Phosphoinositide kinases. Annu. Rev. Biochem. 67, 481–507. Galliard, T. (1980). Degradation of acyl lipids: hydrolytic and oxidative enzymes. In Biochemistry of Plants, vol. 4, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 85–116. Gatt, S. (1968). Purification and properties of phospholipase A-1 from rat and calf brain. Biochim. Biophys. Acta. 159, 304–316. Groot, P.H.E. et al. (1976). Fatty acid activation: specificity, localization, and function. Adv. Lipid Res. 14, 75–126. Gurr, M.I. and Harwood, J.L. (1991). Lipid Biochemistry, 4th ed., Chapman & Hall, London. Gurr, M.I. et al. (2002). Lipid Biochemistry, 5th ed., Blackwell Scientific, Oxford, U.K. Hajra, A.K. (1995). Glycerolipid biosynthesis in peroxisomes (microbodies). Prog. Lipid Res. 34, 343–364. Harwood, J.L. (1979). The synthesis of acyl lipids in plant tissues. Prog. Lipid Res. 18, 55–86. Harwood, J.L. (1980). Plant acyl lipids: structure, distribution and analysis. In Biochemistry of Plants, vol. 4, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 1–55. Harwood, J.L. (1989). Lipid metabolism in plants. Crit. Rev. Plant Sci. 8, 1–43. Harwood, J.L. and John, R.A. (1990). Evaluation of inhibitors of lipolytic enzymes. Trends Biochem. Sci. 15, 409–410. Harwood, J.L. and Page, R. (1993). Biochemistry of oil synthesis. In Designer Oilseed Crops, Ed. D.J. Murphy, VCH Press, Weinheim, Germany, pp. 165–194. Harwood, J.L., and Russell, N.J. (1984) Lipids in Plants and Microbes, Allen and Unwin, Hemel Hempstead, U.K. Hawthorne, J.N. and Ansell, G.B. Eds. (1982) Phospholipids, Elsevier, Amsterdam. Heath, R.J. et al. (2002). Fatty acid and phospholipid synthesis in prokaryotes. In Biochemistry of Lipids, Lipoproteins and Membranes, Eds. D.E. Vance and J.E. Vance, 4th ed., Elsevier, Amsterdam, pp. 55–92. Heller, M. (1978). Phospholipase D. Adv. Lipid Res., 16, 267–326. Hetherington, A.M. and Drobak, B.K. (1992). Inositol-containing lipids in higher plants Prog. Lipid Res. 31, 53–63. Hill, E.E. and Lands, W.E.M. (1970). Phospholipid metabolism. In Lipid Metabolism, Ed. S.J. Wakil, Academic Press, New York, pp. 185–275. Huang, K.-S., Li, S. and Low, M.G. (1991). Glycosylphosphatidylinositol-specific phospholipase D. In Methods in Enzymology, vol. 197, Ed. E.A. Dennis, Academic Press, New York, pp. 567–575. Hurst, K.M. et al. (1990). The roles of phospholipase D and a GTP-binding protein in guanosine 5'-[gamma-thio]triphosphate-stimulated hydrolysis of phosphatidylcholine in rat liver plasma membranes. Biochem. J. 272, 749–753. Jackowski, S. et al. (1991). Lipid metabolism in prokaryotes. In Biochemistry of Lipids, Lipoproteins and Membranes, Eds. D.E. Vance and J. Vance, Elsevier, Amsterdam, pp. 43–85. Jackson, R.L. and McLean, L.R. (1991). Human postheparin plasma lipoprotein lipase and hepatic triglyceride lipase. In Methods in Enzymology, vol. 197, Ed. E.A. Dennis, Academic Press, New York, pp. 339–345.
678
Lipid Metabolism
Joyard, J. and Douce, R. (1987). Galactolipid synthesis. In The Biochemistry of Plants, vol. 9, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 215–274. Kelley, M.J. et al. (1992). Renaturation of cobra venom phospholipase A2 expressed from a synthetic gene in Escherichia coli. Biochim. Biophys. Acta. 1118, 107–115. Kent, C. (1995). Eukaryotic phospholipid biosynthesis. Annu. Rev. Biochem. 64, 315–343. Kobayashi, M. and Kanfer, J.N. (1991). Solubilization and purification of rat tissue phospholipase D. In Methods in Enzymology, vol. 197, Ed. E.A. Dennis, Academic Press, New York, pp. 575–583. Kuge, O. and Nishijima, M. (1997). Phosphatidylserine synthases I and II of mammalian cells. Biochim. Biophys. Acta. 1348, 151–156. Lee, H.Y. et al. (2005). Multiple forms of secretory phospholipase A2 in plants. Prog. Lipid Res. 44, 52–67. MacDonald, J.I.S. and Sprecher, H. (1991). Phospholipid fatty acid remodeling in mammalian cells. Biochim. Biophys. Acta. 1084, 105–121. Martin, T.F.J. (1998). Phosphoinositide lipids as signalling molecules: common themes for signal transduction, cytoskeletal regulation and membrane trafficking. Annu. Rev. Cell Dev. Biol.14, 231–264. McPhail, L.C. (2005). Glycerolipids in signal transduction. In Biochemistry of Lipids, Lipoproteins and Membranes, Ed. D.E. Vance and J.E. Vance, 4th ed., Elsevier, Amsterdam, pp. 315–340. Meijer, H.J.G. and Munnik, T. (2003). Phospholipid-based signalling in plants. Annu. Rev. Plant Biol. 54, 265–306. Merkel. O. et al. (1999). Characterisation and function in vivo of two novel phospholipases B lysophospholipases from S a ch a ro myc e s c e revi s i a e. J. B i o l . C h e m . 2 7 4 , 28121–28127. Moore, T.M. (1982). Phospholipid biosynthesis. Annu. Rev. Plant Physiol. 33, 235–259. Mudd, J.B. (1980). Phospholipid biosynthesis. In Biochemistry of Plants, vol. 4, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 249–282. Mudd, S.H. and Datko, A.H. (1986). Phosphoethanolamine bases as intermediates in phosphatidylcholine synthesis by Lemna. Plant Physiol. 82, 126–135. Munnik, T. et al. (1998). Phospholipid signalling in plants. Biochim. Biophys. Acta. 1389, 222–272. Nagai, Y. et al. (1999). An alternative splicing form of phosphatidylserine-specific phospholipase A1 that exhibits lysophosphatidylserine-specific lysophospholipase activity in humans. J. Biol. Chem. 274, 11053–59. Nakagawa, Y. et al. (1991). Detergent-resistant phospholipase A1 from Escherichia coli membranes. In Methods in Enzymology, vol. 197, Ed. E.A. Dennis, Academic Press, New York, pp. 309–315. Newkirk, J.D. and Waite, M. (1971). Identification of a phospholipase A1 in plasma membranes of rat liver. Biochim. Biophys. Acta. 225, 224–233. O’Doherty, P.J.A. (1978). Metabolic studies with natural and synthetic fatty acids and enantiomeric acylglycerols. In Handbook of Lipid Research, vol. 1, Ed. A. Kuksis, Plenum, New York, pp. 289–339.
Ohara, O. et al. (1995). Structure and function of phospholipase A2 receptor. Prog. Lipid Res. 34, 117–138. Pieringer, R.A. (1989). Biosynthesis of non-terpenoid lipids. In Microbial Lipids, vol. 2, Eds. C. Ratledge and S.G. Wilkinson, Academic Press, London, pp. 51–114. Price-Jones, M.J. and Harwood, J.L. (1986). The control of CTP:choline phosphate cytidylyltransferase activity in pea (Pisum sativum L.). Biochem. J. 240, 837–842. Rameh, L.E. and Cantely, L.C. (1999). The role of phosphoinositide 3-kinase lipid products in cell function. J. Biol. Chem. 274, 8347–8350. Raybin, D.M. et al. (1972). A phospholipase in Bacillus megaterium unique to spores and sporangia. Biochemistry. 11, 1754–1760. Ridgeway, N.D. et al. (1999). Integration of phospholipid and sterol metabolism in mammalian cells. Prog. Lipid. Res., 38, 337–360. Roughan, P.G. and Slack, C.R. (1976). Is phospholipase D really an enzyme? A comparison of in situ and in vitro activities. Biochim. Biophys. Acta. 431, 86–95. Roughan, P.G. and Slack, C.R. (1984). Glycerolipid synthesis in leaves. Trends Biochem. Sci. 9, 383–386. Sambanthamurthi, R. et al. (2002). Chemistry and biochemistry of palm oil. Prog. Lipid Res., 39, 507–558. Schlame, M. et al. (2000). The biosynthesis and functional role of cardiolipin. Prog. Lipid Res. 39, 257–288. Slotboom, A.J. et al. (1982). On the mechanism of phospholipases A2. In Phospholipids, Eds. J.N. Hawthorne and G.B. Ansell, Elsevier, Amsterdam, pp. 359–434. Smith, J.D. (1993). Phospholipid synthesis in protozoa. Prog. Lipid Res., 32, 47–60. Snyder, F. (1991). Metabolism, regulation, and function of etherlinked glycerolipids and their bioactive species. In Biochemistry of Lipids, Lipoproteins and Membranes, Eds. D.E. Vance and J. Vance, Elsevier, Amsterdam, pp. 241–267. Snyder, F., et al. (1985). Ether-linked glycerolipids and their bioactive species: enzymes and metabolic regulation. In The Enzymes of Biological Membranes, Ed. A.N. Martinosi, Plenum, New York, pp. 1–58. Snyder, F. et al. (2002). Ether-linked lipids and their bioactive species. In Biochemistry of Lipids, Lipoproteins and Membranes, Eds. D.E. Vance and J.E. Vance, 4th ed., Elsevier, Amsterdam, pp. 233–262. Sohlenkamp, C. et al. (2003). Biosynthesis of phosphatidylcholine in bacteria. Prog. Lipid Res., 42, 115–162. Taki, T. et al. (1978). The “base-exchange” reaction: the serine enzyme. Adv. Exp. Med. Biol. 101, 301–318. Turk, J. et al. (1987). The role of phospholipid-derived mediators including arachidonic acid, its metabolites, and inositoltrisphosphate and of intracellular Ca2+ in glucose-induced insulin secretion by pancreatic islets. Prog. Lipid Res. 26, 125–181. Ulane, R.E. et al. (1977). A rapid accurate assay for choline kinase. Anal. Biochem. 79, 526–534. van den Bosch, H. (1982). Phospholipases. In Phospholipids, Eds. J.N. Hawthorne and G.B. Ansell, Elsevier, Amsterdam, pp. 313–357. van den Bosch, H. et al. (1974). Isolation and properties of a phospholipase A1 activity from beef pancreas. Biochim. Biophys. Acta. 348, 197–209.
679
10.3 Glyceride metabolism
van Kuijk, F.J.G.M. et al. (1987). A new role for phospholipaseA2 — protection of membranes from lipid-peroxidation damage. Trends Biochem. Sci. 12, 31–34. Vance, D.E. (1990). Phosphatidylcholine metabolism: masochistic enzymology, metabolic regulation and lipoprotein assembly. Biochem. Cell Biol. 68, 1151–1165. Vance, D.E. (1991). Phospholipid metabolism and cell signalling in eukaryotes. In Biochemistry of Lipids, Lipoproteins and Membranes, Eds. D.E. Vance and J. Vance, Elsevier, Amsterdam, pp. 205–240. Vance, D.E. (2005). Phospholipid biosynthesis in eukaryotes. In Biochemistry of Lipids, Lipoproteins and Membranes, Eds. D.E. Vance and J.E. Vance, 4th ed., Elsevier, Amsterdam, pp. 205–232. Vance, D.E. and Choy, I.C. (1979). How is phosphatidylcholine biosynthesis regulated? Trends Biochem.Sci. 4,145–148. Vance, D.E. and Ridgway, N.D. (1988). The methylation of phosphatidylethanolamine. Prog. Lipid Res. 27, 61–79. Vance, D.E. and Vance, J.E. Eds. (2002). Biochemistry of Lipids, Lipoproteins and Membranes. Elsevier, Amsterdam. Vance, J.E. (1998). Eukaryotic lipid biosynthetic enzymes: the same but not the same. Trends Biochem. Sci. 23, 423–428. Vanhaesebroeck, B. et al. (2001). Synthesis and function of 3phosphorylated inositol lipids. Annu. Rev. Biochem.70, 535–602. Verheij, H.M. et al. (1983). The complete primary structure of phospholipase A2 from human pancreas. Biochim. Biophys. Acta. 747, 93–99. von Wettstein-Knowles, P. (1979). Genetics and biosynthesis of plant epicuticular waxes. In Advances in the Biochemistry and Physiology of Plant Lipids, Eds. L-A Appelqvist and C. Liljenberg, Elsevier, Amsterdam, pp. 1–26. Waite, M. (1987). The phospholipases. In Handbook of Lipid Research, vol. 5, Ed. D.J. Hanahan, Plenum, New York, pp. 111–133. Waite, M. et al. (1991). Purification and substrate specificity of rat hepatic lipase. In Methods in Enzymology, vol. 197, Ed. E.A. Dennis, Academic Press, New York, pp. 331–339. Waku, (1992). Origins and fates of fatty acyl-CoA esters. Biochim. Biophys. Acta. 1124, 101–111. Wang, X. (2001). Plant phospholipases. Annu. Rev. Plant Physiol. Plant Mol. Biol. 52, 211–231. Washburn, W.N. and Dennis, E.A. (1991). Suicide-inhibitory bifunctionally linked substrates (SIBLINKS) as phospholipase A2 inhibitors. Mechanistic implications. J. Biol. Chem. 266, 5042–5048. Webber, K.O. and Hajra, A.K. (1992). Dihydroxyacetone phosphate acyltransferase. In Methods in Enzymology, vol. 209, Eds. E.A. Dennis and D.E. Vance, Academic Press, New York, pp. 92–98. Williams, M. and Harwood, J.L. (1994). Alternative pathways of phosphatidylcholine synthesis in olive (Olea europeea L.) callus cultures. Biochem. J. 304, 463–468. Wilton, D.C. and Waite, M. (2002). Phospholipases. In Biochemistry of Lipids, Lipoproteins and Membranes, Eds. D.E. Vance and J.E. Vance, 4th ed., Elsevier, Amsterdam, pp. 291–314. Xie, M. and Dubyak, G.R. (1991). Guanine-nucleotide- and adenine-nucleotide-dependent regulation of phospholipase D in electropermeabilized HL-60 granulocytes. Biochem. J. 278, 81–89.
Yamane, T. et al. (1989). Conversion of phospholipids by phospholipase D. In Fats for the Future (Ed. R.C. Cambie), Ellis Horwood, Chichester, U.K., pp. 128–138. Yang, S.F. et al. (1967). Transphosphatidylation by phospholipase D. J. Biol. Chem. 242, 477–484.
10.3
Glyceride metabolism
10.3.1
Triacylglycerol synthesis
The lipid stores of animals and almost all plants are triacylglycerols. These compounds, therefore, are of very great importance in physiology as well as to the food industry. One difference between animals and plants is in the relative amounts of de novo synthesized acyl chains in the triacylglycerols. Plants from necessity must make all of the glyceride molecules from simple starting materials (ultimately from photosynthetically fixed CO2). In contrast, animals make use of dietary fatty acids, which to a large degree determine the fatty acid composition of the triacylglycerol stores (e.g., copepods consuming phytoplankton; Bauermeister and Sargent, 1979). In oil seeds the triacylglycerol stores are located in oil bodies. These intracellular structures are surrounded by a half-unit phospholipid membrane containing unique proteins called oleosins. The latter have been isolated and sequenced from a number of plants and probably serve both to stabilize the oil bodies as well as to provide binding sites for lipases to function during germination (for reviews, see Murphy, 1990; Huang, 1992). In animals most triacylglycerol is stored in adipose tissue, though, in times of metabolic stress, significant amounts may build up in liver, heart and skeletal muscle. Also, it should be noted that the variable distribution of adipose tissue stores and the metabolic properties of such tissue in different compartments of the body has important implications for health (Abate and Garg, 1995). The formation of triacylglycerols removes the potentially harmful effects of fatty acids or fatty acyl-CoAs (Brindley, 1991). A major function of animal triacylglycerols is to allow the transport of acyl moieties about the body, in the form of the serum lipoproteins. Two major classes of lipoprotein are relevant. Chylomicrons carry absorbed dietary fat from the intestine to other organs, while very low density lipoproteins carry triacylglycerol from the liver to other tissues (Section 7.5 and Section 11.2). There has been considerable recent interest in the use of oleaginous microorganisms as sources of triacylglycerols, particularly those with unusual fatty acids. In general, yeasts and moulds offer the best possibilities for industrial exploitation, and few bacteria accumulate appreciable amounts of triacylglycerols (actinomycetes are the most noticeable exceptions). The whole topic of the biotechnology of oils and fats in microorganisms is reviewed by Ratledge (1989) and Cohen and Ratledge (2005). See also Davies and Holdsworth (1992). 680
Lipid Metabolism
the endoplasmic reticulum. Evidence for the monoacylglycerol pathway in plants is poor, and it seems probable that diacylglycerol arises from phosphatidate phosphatase in these phyla (Gurr, 1980). For comments on the extensive literature concerning the substrate specificities and other properties of mammalian monoacylglycerol acyltransferases, refer to O’Doherty (1978) and Dircks and Sul (1999). Phosphatidate phosphohydrolase has already been briefly discussed in relation to glycerophospholipid synthesis (Section 10.2.1). The enzyme was first discovered in plants (Kates, 1955) and, subsequently, identified in animal tissues. It has been purified from a large number of mammalian tissues including liver, kidney, intestinal mucosa, adipose tissue, erythrocyte membranes, and avian salt glands (cf. O’Doherty, 1978). The intracellular distribution of mammalian phosphatidate phosphohydrolase is complicated. It has been found in several particulate fractions — mitochondrial, lysosomal, and microsomal (cf. Sedgwick and Hübscher, 1967). However, quantitatively, the most important of these fractions is the endoplasmic reticulum. In addition, there is significant cytosolic activity in animals. The cytosolic phosphatidate phosphohydrolase is able to translocate to the endoplasmic reticulum and mitochondria. The response of such a Mg2+dependent enzyme to hormones and metabolites and variations in its intracellular distribution between soluble and particulate compartments, are compatible with an important role in controlling triacylglycerol synthesis. For a full discussion of this function, see Brindley (1988, 1991). In plants, phosphatidate phosphohydrolase has a less obvious role in controlling the rate of triacylglycerol synthesis, since the latter occurs usually at a specific developmental period when it is not in competition with other metabolic pathways, such as membrane lipid synthesis (Harwood and Griffiths, 1992). The enzyme has been detected in several subcellular compartments (see Harwood and Price-Jones, 1988), and, moreover, its activity in chloroplast envelopes appears to be important in controlling the flux of lipid carbon between the extrachloroplastic and the chloroplast compartments. This is important in giving rise to different molecular species of lipids in various plants (see Heinz and Roughan, 1987; Ohlrogge et al., 1991). There have been some suggestions that the activity of phosphatidate phosphohydrolase may be important in triacylglycerol accumulation in seeds (Ichihara et al., 1990) even though the enzyme, unlike in animals, is not under acute (hormonal) control. Synthesis of triacylglycerols requires the enzyme diacylglycerol acyltransferase (EC 2.3.1.30). The enzyme is located in the endoplasmic reticulum (Brindley, 1991) and its activity is especially high in lypogenic tissues like adipose and liver. Its properties and those of other acyltransferases involved in the glycerol phosphate pathway for triacylglycerol formation were reviewed by Dircks and Sul (1991). Diacylglycerol acyltransferase is unique to
The major pathways for triacylglycerol synthesis are shown in Figure 10.31. The formation of phosphatidic acid by the glycerol phosphate pathway or the dihydroxyacetone phosphate pathway has already been discussed (Section 10.2.1). In addition, see reviews by Gurr (1980) and O’Doherty (1978). Phosphatidic acid can also be formed from monoacylglycerol or diacylglycerol, but, as will be seen, this is not of relevance for triacylglycerol synthesis. Attempts to estimate the relative contribution of different pathways for phosphatidic acid formation (cf. Hill and Lands, 1970; O’Doherty, 1978) indicate that about 50% of lipid glycerol enters by the dihydroxyacetone pathway (e.g., Manning and Brindley, 1972; Pollock et al., 1975). In some tissues (e.g., Ehrlich ascites tumour cells), the acyldihydroxyacetone pathway is the major route of synthesis (Synder, 1972). The enzymes responsible for the acylation of glycerol phosphate have been purified and studied many times. For reviews of their purification and properties, see Lennarz (1970), Hill and Lands (1970), O’Doherty (1978), Gurr (1980), and Stymne and Stobart (1987). The glycerol 3-phosphate acyltransferase and monoacylglycerol 3-phosphate acyltransferase enzymes in both animals and plants differ in their substrate specificities. Typically, the glycerol 3-phosphate acyltransferase uses more saturated acyl-CoAs than are subsequently attached to the sn-1 position. The major site of both enzymes is the endoplasmic reticulum, though animal mitochondria (particularly those of liver) also contain a second glycerol 3-phosphate acyltransferase (Brindley, 1991). In plant chloroplasts, glycerol 3-phosphate acylating enzymes (which use acyl-ACP substrates and have different substrate specificities to their endoplasmic reticulum counterparts) are present. Although triacylglycerol synthesis can take place in chloroplasts (particularly under adverse environmental conditions; see Sakaki et al., 1985), the chloroplast enzymes are mainly important for the generation of thylakoid acyl lipids (Ohlrogge et al., 1991). For the dihydroxyacetone pathway (see Figure 10.31), acyltransferases have been found in both the endoplasmic reticulum and in peroxisomes (Brindley, 1991). As indicated in Figure 10.31, diacylglycerol can originate from two sources. Either it is formed by phosphatidate phosphohydrolase or it is synthesised from monoacylglycerol. The first demonstration of an enzyme system capable of converting monoacylglycerols to triacylglycerols was the work of Clark and Hübscher (1960) with preparations from rabbit intestine. Indeed, the reactions have been studied most thoroughly in intestine, although activity has been detected in a large number of mammalian tissues including kidney, pancreas, adipose tissue, arterial walls, ascites tumour cells and mammary glands (cf. O’Doherty, 1978; Brindley, 1991). The first enzyme in the pathway is monoacylglycerol acyltransferase (EC 2.3.1.22). The enzyme is particulate and is found in microsomal fractions — probably arising from 681
10.3 Glyceride metabolism
C
H
H2C
O
HO
H2 C
OH
H2 C
H
H2C
O
1
CoA
2
O
C
H
R1
C
3
O
H 2C
O
C
H
H2C
O
NADPH
H2C O P Glycerol phosphate 1-Acylglycerol-3-snpathway phosphate R2
4
R1CoA
O
O H2 C HO
P
Dihydroxyacetone phosphate
Glycerol-3-snphosphate R1
C
O P
OH
C
R1
P
1-Acyldihydroxyacetone phosphate
Dihydroxyacetone phosphate pathway
CoA O
O R2
C
H2C
O
C
H
H2C
O
O
R1
C
P
1,2-Diacylglycerol-3-snphosphate (phosphatidic acid) 5
P O
O R2
C
H2C
O
C
H
O
C
O 1
R
R3
R2 7
OH
H2C
O
CoA
C
O
H2 C
O
C
C
H
O
H2C
O
C
R1
R3
1,2-sn-Diacylglycerol Monoacylglycerol pathway
H2C
O R2
C
R2
6
O
C H2C
CoA
OH H OH
2-sn-monoacylglycerol
FIGURE 10.31 Major pathways of triacylglycerol synthesis. Reactions (1), (4), (5), and (7), glycerol phosphate pathway; reactions (2), (3), (4), (5), and (7), dihydroxyacetone phosphate pathway; reactions (6) and (7), monoacylglycerol pathway. Enzymes: (1) glycerol phosphate 1-acyltransferase; (2) dihydroxyacetone phosphate acyltransferase; (3) acyldihydroxyacetone phosphate reductase; (4) 1acylglycerol phosphate 2-acyltransferase; (5) phosphatidate phosphohydrolase; (6) monoacylglycerol acyltransferase; (7) diacylglycerol acyltransferase.
with its natural 1,2-diacylglycerol substrate than the 1,3isomer (Hill et al., 1968). The diacylglycerol acyltransferase of plants has been studied in several tissues, though not yet purified. In many plants, it seems less specific for its acyl-CoA substrate than are the other acytransferases (Stymne and Stobart, 1987). It is located in endoplasmic reticulum and has low activity when measured in vitro (e.g., Berneth and Frentzen, 1990). The latter fact, together with the significant buildup of
triacylglycerol synthesis and, as such, might be expected to be important for regulation, at least under certain conditions (Mayorek et al., 1989). So far as can be ascertained, it competes with the choline and ethanolamine phosphotransferases (that form zwitteronic phospholipids; Section 10.2.1) for a common pool of diacylglycerol (Brindley, 1991). Diacylglycerol acyltransferase can use a wide range of saturated and unsaturated acyl-CoAs (Dircks and Sul, 1991), but has considerably better activity 682
Lipid Metabolism
lipase is the best known and most investigated of all lipolytic enzymes. It acts on mono-, di-, and triacylglycerols, although the reaction rate is slower with partial glycerides (Brockerhoff and Jensen, 1974). It probably lacks stereospecificity. For a thorough review of pancreatic lipase, see Jensen (1971). Metabolism of triacylglycerols in animals requires the interaction of lipoprotein lipase (involved in uptake of acyl chains from plasma) and hormone-sensitive lipase (involved in release of fatty acids from lipid stores). Some aspects of lipoprotein lipase action are discussed in Section 11.3, and the reader is also referred to Brockerhoff and Jensen (1974), Jensen (1971), and Frayn et al. (1995). The enzyme is also known as clearing factor lipase, requires apo-CII for activity, and may be bound via heparin sulfate proteoglycan at the endothelial surface in vivo (Williams et al., 1983). Considerable work has been carried out on intracellular processing of the enzyme in active tissues (Cryer, 1981) and on the action of hormones in controlling the adipose and heart tissue enzymes (Ashby and Robinson, 1980; de Gasquet et al. 1975). There is now considerable sequence information of lipoprotein lipase. The sequence is extraordinarily conserved (87 to 94% with different mammalian enzymes). Comparisons with other lipases have also been made (see Wang et al., 1992). For further information on the structural and functional domains of lipoprotein lipase, the structure and function of apo-CII, and the reaction kinetics of the enzyme, see Wang et al., (1992). Hormone-sensitive lipase tissue activity is stimulated by adrenaline (epinephrine, glucagons, ACTH (corticotrophin), TSH (thyrotropin) and serotonin (Jensen, 1971). These hormones are presumed to exert their effects on adipose tissue by stimulating adenylate cyclase. Certainly, an increased formation of cAMP (cyclic AMP) is brought about by lipolytic hormones and cAMP has been shown to stimulate lipase activity in cell-free preparations (cf. O’Doherty, 1978). Phosphorylation/dephosphorylation has been implicated (Belfarage et al., 1983). The cDNA for rat hormone-sensitive lipase has been reported (Holm et al., 1988) and increased expression of the mRNA for this enzyme found in the adipose tissue of cancer patients (Thompson et al., 1993). A good review is that of Langin et al. (1996) and Bernlohr et al. (2002), also summarises this topic. One typical feature of the mobilisation of fatty acids from adipose tissue, which has not yet been fully explained, is that their release is selective for both chainlength and unsaturation. Part of the reason is due to selectivity properties of hormone-sensitive lipase, but other factors appear important (Raclot, 2003). Other mammalian lipases that have been studied are milk lipase (cf. Jensen, 1971) and monoacylglycerol lipase. The latter enzyme occurs in several tissues including intestine, liver, and adipose tissue. It has high activity with monoacylglycerols when compared to diacylglycerols or triacylglycerols (see O’Doherty, 1978).
diacylglycerol during triacylglycerol formation in vivo (Perry and Harwood, 1990) and in vitro (Perry and Harwood, 1991), has led to suggestions that diacylglycerol acyltransferase may exert significant flux control over oil accumulation in some crops (see, e.g., Griffiths and Harwood, 1990). Apart from diacylglycerol acyltransferase being used to form triacylglycerol, evidence for acyl-CoA independent pathways has been obtained in animal tissues and oilseeds. Thus, in mammals, a diacylglycerol transacylase has been reported (Lehner and Kuksi, 1996). In oilseeds, the same enzyme activity is also present, but, in addition, phospholipid: diacylglycerol acyltransferase is found (see Weselake, 2005), although the quantitative importance of the latter in many tissues seems to be slight (see Ramli et al., 2005). Mammalian triacylglycerol synthesis is affected by a large number of factors. These include nutritional, hormonal, and pharmacological effects (cf. O’Doherty, 1978, for review). A discussion of the control of triacylglycerol synthesis in animals is given by Brindley (1991). General accounts of triacylglycerol synthesis in plants are given by Gurr (1980), Stymne and Stobart (1987), Harwood and Page (1993), and Weselake (2005). Specialist accounts of lipid synthesis in the two important oil crops, olive and palm, are given in Salas et al. (2000) and Sambanthamurthi et al. (2000), respectively. Weselake and Taylor (1999) detail the use of microspore-derived cultures to examine triacylglycerol biosynthesis in oilseed rape. Another subject worthy of mention are the efforts by breeders to increase oil yields. These have been quite successful (e.g., maize varieties have been changed from an average 4 to 5% oil content to give IHO lines (20%) or ILO (0.5%) lines). However, the complexity of the phenotype precludes elucidation of the exact genetic basis of these changes (Ohlrogge et al., 1991). On the other hand, a 20% increase in seed fatty acid content (per unit seed weight) has been induced in Arabidopsis by a single gene variation (James and Dooner, 1990).
10.3.2
Triacylglycerol breakdown
The breakdown of triacylglycerol is catalysed by lipases. A large number of such enzymes have been purified from animals, plants, and microbes (cf. Brockerhoff and Jensen, 1974). It should be noted that the term “lipase” is frequently misused. A true lipase is one that attacks triacylglycerols and acts only at an oil–water interface. This definition, therefore, excludes enzymes acting on watersoluble esters (esterases) or those preferentially hydrolysing other lipids (acyl hydrolases). Since triacylglycerols are important dietary constituents (Section 11.1), there is interest in digestive lipases. Although pharyngeal lipases have been found and studied, their importance during overall digestion remains to be established (cf. O’Doherty, 1978). In contrast, pancreatic 683
10.3 Glyceride metabolism
Lipase activity has been found in a wide range of plant materials with most work on oil seeds and cereals. One of the best characterized enzymes is the castor bean acid lipase. Ory and coworkers (Ory, 1969) have characterized the enzyme, which is associated with the spherosome membrane. Huang and Moreau (1978) studied lipolytic activity in a range of germinating oil seeds. The enzyme from peanuts seems to be mainly associated with microsomal fractions (Theimer and Rosnitschek, 1978) and is clearly distinguished from the glyoxysomal monoacylglycerol lipase (Huang and Moreau, 1978). Several lipases have been purified from cereal seeds, such as wheat, oat, and rice. Galliard (1980) has reviewed this work (cf. also Jensen, 1971) and has discussed the role of plant growth regulators in controlling activity. For a more recent review, see Huang (1987). Although bacteria do not store energy as triacylglycerol, a number of bacterial lipases have been discovered and studied. An extensive review of microbial lipases (and esterases) has been made by Lawrence (1967). It should be emphasized also that many of the microbial “lipases” (e.g., the widely studied Rhizopus arrhizus enzyme) are, in fact, acyl hydrolases. More recent information is available in Finnerty (1989). The features of lipases useful for industry in lipid modifications have been discussed by Eigtved (1992), and the monograph by Alberghina et al. (1991) contains many specific examples of individual lipases, their characteristics and potential (or actual) industrial use.
(L) involved in triacylglycerol biosynthesis. Plant Sci. 67, 21–28. Bernlohr, D.A. et al. (2002). Adipose tissue and lipid metabolism. In Biochemistry of Lipids, Lipoproteins and Membranes, Ed. D.E. Vance, 4th ed., Elsevier, Amstedam, p. 257–281. Brindley, D.N. (1978). Some aspects of the physiological and pharmacological control of the synthesis of triacylglycerols and phospholipids. Int. J. Obes. 2, 7–16. Brindley, D.N. Ed. (1988). Phosphatidate Phosphohydrolase, vols. I and II, CRC Press, Boca Raton, FL. Brindley, D.N. (1991). Metabolism of triacylglycerols. In Biochemistry of Lipids, Lipoprotein and Membranes, Eds. D.E. Vance and J. Vance, Elsevier, Amsterdam, pp. 171–203. Brokerhoff, H. and Jensen, R.G. (1974). Lipolytic Enzymes, Academic Press, New York. Clark, B. and Hübscher, G. (1960). Biosynthesis of glycerides in the mucosa of the small intestine. Nature. 183, 35–37. Cohen, Z. and Ratledge, C (2005). Single cell oils. Am. Oil Chem. Soc. Champaign, IL. Coleman, R.A. et al. (2000). Physiological and nutritional regulation of enzymes of triacylglycerol synthesis. Annu. Rev. Nutr. 20, 77–103. Cryer, A. (1981). Tissue lipoprotein lipase activity and its action in lipoprotein metabolism. Int. J. Biochem. 13, 525–541. Davies, R.J. and Holdsworth, J. E. (1992). Synthesis of lipids in yeasts, biochemistry, physiology and production. Adv. Appl. Lipid Res. 1, 119–159. de Gasquet, P. et al. (1975). Effect of glucocorticoids on lipoprotein-lipase activity in rat-heart and adipose-tissue. Horm. Metab. Res. 7, 152–157. Dircks, L. and Sul H.S. (1999). Acyltransferases of de novo glycerolphospholipid biosynthesis. Prog. Lipid Res. 38, 461–479. Douce, R. and Joyard, J. (1980). Plant galactolipids. In Biochemistry of Plants, vol. 4, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 321–362. Eigtved, P. (1992). Enzymes and lipid modification. Appl. Lipid Res. 1, 1–64. Finnerty, W.R. (1989). Microbial lipid metabolism. In Microbial Lipids, vol. 2, Eds. C. Ratledge and S.G. Wilkinson, Academic Press, London, pp. 525–566. Frayn, K.N. et al. (1995). Coordinated regulation of hormonesensitive lipase and lipoprotein lipase in human adepose tissue in vivo. Adv. Enz. Regul. 35, 163–178. Galliard, T. (1980). Degradation of acyl lipids: hydrolytic and oxidative enzymes. In Biochemistry of Plants, vol. 4, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 85–116. Gibbons, G.F. et al. (2000). Mobilisation of triacylglycerol stores. Biochim. Biophys. Acta. 1483, 37–57. Griffiths, G. and Harwood, J.L. (1990). Triacylglycerol synthesis in maturing cotyledons of cocoa. In Plant Lipid Biochemistry, Structure and Utilization, Eds. P.J. Quinn and J.L. Harwood, Portland, London., pp. 216–218. Gurr, M.I. (1980). The biosynthesis of triacylglycerols. In Biochemistry of Plants, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 205–248. Harwood, J.L. and Griffiths, G. (1992). Biochemistry of plant lipids. In Advances in Plant Cell Biochemistry and Biotechnology, Ed. I.M. Morrison, JAI Press, London, pp. 1–52.
References Abate, N. and Garg, A (1995). Heterogeneity in adipose tissue metabolism: causes, implications and management of regional adiposity. Prog. Lipid Res. 34, 53–70. Akesson, B. et al. (1970). Initial incorporation into rat liver glycerolipids of intraportally injected (9,10-3H2) palmitic acid. Biochim. Biophys. Acta. 18, 44–56. Alberghina, L. et al., Eds. (1991). Lipases: Structure, Mechanism and Genetic Engineering, VCH Press, New York. Appelqvist, L.A. (1975). Biochemical and structural aspects of storage and membrane lipids in developing oil seeds. In Recent Advances in Chemistry and Biochemistry of Plant Lipids, Eds. T. Gallaird and E.I. Mercer, Academic Press, London, pp. 247–254. Ashby, P. and Robinson, D.S. (1980). Effects of insulin, glucocorticoids and adrenaline on the activity of rat adipose-tissue lipoprotein lipids. Biochem, J. 188, 185–192. Bauermeister and Sargent, J. (1979). Wax esters: major metabolites in the marine environment. Trends Biochem. Sci. 4, 209–211. Belfrage, P. et al. (1983). Hormonal regulation of adipose tissue lipolysis by reversible phosphorylation of hormone-sensitive lipase. In Cell Function and Differentiation, Part C, Eds. G. Akoyunoglou, et al., Alan R. Liss, New York, pp. 213–223. Berneth, K. and Frentzen, M. (1990). Utilization of erucoyl-CoA by acyltransferases from developing seeds of Brassica-napus
684
Lipid Metabolism
Harwood, J.L. and Page, R. (1994). Biochemistry of oil synthesis. In Designer Oilseed Crops, Ed. D.J. Murphy, VCH Press, New York, pp. 165–194. Harwood, J.L. and Price-Jones, M.J. (1988). Phosphatidate phosphohydrolase in plants and microorganisms. In Phosphatidate Phosphohydrolase, vol. II, Ed. D.N. Brindley, CRC Press, Boca Raton, FL, pp. 1–37. Heinz, E. (1977). Enzymatic reactions in galactolipid biosynthesis. In Lipids and Lipid Polymers in Higher Plants, Eds. M. Tevini and H.K. Lichtenhaler, Springer-Verlag, Berlin, pp. 102–120. Heinz, E. and Roughan, P.G. (1983). Similarities and differences in lipid-metabolism of chloroplasts isolated from 18:3 and 16:3 plants. Plant Physiol. 72, 273–279. Hill, E.E. and Lands, W.E.M. (1970). Phospholipid metabolism. In Lipid Metabolism, Ed. S.J. Wakil, Academic Press, New York, pp. 185–275. Hill, E.E. et al. (1968). The selective incorporation of 14C-glycerol into different species of phosphatidic acid, phosphatidylethanolamine, and phosphatidylcholine. J. Biol. Chem. 243, 4440–4451. Hitchcock, C. and Nichols, B.W. (1971). Plant Lipid Biochemistry, Academic Press, London. Holm, C. et al. (1988). Hormone-sensitive lipase: sequence, expression, and chromosomal localization to 19 centq13.3. Science, 241, 1503–1506. Huang, A.H.T. (1987). Lipases. In The Biochemistry of Plants, Eds. P.K. Stumpf and E.E. Conn, vol. 9, Academic Press, New York, pp. 91–119. Huang, A.H.C. (1992). Oil bodies and oleosins in seeds. Annu. Rev. Plant Physiol. 43, 177–200. Huang, A.H.C. and Moreau, R.A. (1978). Lipases in storage tissues of peanut and other oil seeds during germination. Planta. 141, 111–116. Ichihara, K. et al. (1990). Intracellular translocation of phosphatidate phosphatase in maturing safflower seeds — a possible mechanism of feed forward control of triacylglycerol synthesis by fatty acids. Biochim. Biophys. Acta. 1043, 227–234. James, D.W. and Dooner, H.K. (1990). Isolation of ems-induced mutants in arabidopsis altered in seed fatty-acid composition. Theor. Appl. Genet. 80, 241–245. Jensen, R.G. (1971). Lipolytic enzymes. Prog. Chem. Fats Other Lipids. 11, 347–394. Kanoh, H. and Ohno, K. (1973). Studies on 1,2-diglycerides formed from endogenous lecithins by back-reaction of ratliver microsomal CDP-choline-1,2-diacylglycerol cholinephosphotransferase. Biochim. Biophys. Acta. 326, 17–25. Kates, M. (1995). Hydrolysis of lecithin by plant plastid enzymes. Can. J. Biochem. Physiol. 33, 575–589. Langin, D. et al. (1996). Adipocyte hormone-sensitive lipase: a major regulator of lipid metabolism. Proc. Nutr. Soc. 55, 93–109. Lawrence, R.C. (1967). Microbial lipases and related esterases. Part II. Estimation of lipase activity, characterization of lipases — recent work concerning their effect on dairy products. Dairy Sci. Abstr. 29, pp. 59–70. Lehner, R. and Kuksis, A (1996). Biosynthesis of triacylglycerols. Prog. Lipid Res. 35, 169–202. Lennarz, W.J. (1970). Bacterial lipids. In Lipid Metabolism, Ed. S.J. Wakil, Academic Press, New York, pp. 155–184.
Manning, R. and Brindley, D.N. (1972). Tritium isotope effects in the measurement of the glycerol phosphate and dihydroxyacetone phosphate pathways of glycerolipid biosynthesis in rat liver. Biochem. J. 130, 1003–1012. Mayorek, N. et al. (1989). Triacylglycerol synthesis in cultured rat hepatocytes. The rate-limiting role of diacylglycerol acyltransferase. Eur. J. Biochem. 182, 395–400. Mitchell, M.P. et al. (1971). Properties of phosphatidate phosphohydrolase. Eur. J. Biochem. 18, 214–220. Moore, T.J. et al. (1973). Enzymes of phospholipid metabolism in endoplasmic-reticulum of castor bean endosperm. Plant Physiol. 52, 50–53. Murphy, D.J. (1990). Storage lipid bodies in plants and other organisms. Prog. Lipid Res. 29, 299–324. O’Doherty, P.J.A. (1978). Metabolic studies with natural and synthetic fatty acids and enantiomeric acylglycerols. In Handbook of Lipid Research, vol. 1, Ed. A. Kuksis, Plenum Press, New York, pp. 289–339. O’Doherty, P.J. and Kuksis, A. (1975). Stimulation of triacylglycerol synthesis by Z protein in rat liver and intestinal mucosa. FEBS Lett. 60, 256–258. O’Doherty, P.J. et al. (1974). Effect of phosphatidylcholine on triacylglycerol synthesis in rat intestinal mucosa. Can J Biochem. 52, 726–733. Ohlrogge, J.B. et al. (1991). The genetics of plant lipids. Biochim. Biophys. Acta. 1082, 1–26. Ory, R.L. (1969). Acid lipase of the castor bean. Lipids. 4, 177–185. Perry, H.J. and Harwood, J.L. (1990). Studies of lipid metabolism in developing oilseed rape. In Plant Lipid Biochemistry, Structure and Utilization, Eds. P.J. Quinn and J.L. Harwood, Portland, London, pp. 204–206. Perry, H.J. and Harwood, J.L. (1991). Lipid metabolism during seed development in oilseed rape (Brassica napus L. cv. Shiralee). Biochem. Soc. Trans. 19, 243S. Pollock, R.J. et al. (1975). The relative utilization of the acyl dihydroxyacetone phosphate and glycerol phosphate pathways for synthesis of glycerolipids in various tumors and normal tissues. Biochim. Biophys. Acta. 380, 421–435. Quinn, P.J. and Harwood, J.L. Eds. (1990) Plant Lipid Biochemistry, Structure and Utilization, Portland Press, London. Raclot, T. (2003). Selective mobilisation of fatty acids from adipose tissue triacylglycerols. Prog. Lipid Res. 42, 257–288. Ramli, U.S. et al. (2005). Metabolic control analysis reveals an important role for diacylglycerol acyltransferase in olive but not in oil palm lipid accumulation. FEBS J. 222, 5764–5770. Ratledge, C. (1989). Biotechnology of oils and fats. In Microbial Lipids, vol. II, Eds. C. Ratledge and S.J. Wilkinson, Academic Press, London, pp. 567–613. Roughan, P.G. (1987). On the control of fatty acid compositions of plant glycerolipids. In The Metabolism, Structure and Function of Plant Lipids, Eds. P.K. Stumpf, J.B. Mudd and W.D. Nes, Plenum, New York, pp. 247–254. Roughan, P.G. and Slack, C.R. (1982). Cellular organization of glycerolipid metabolism. Annu. Rev. Plant Physiol. 33, 97–132. Sakaki, T. et al. (1985). Polar and neutral lipid changes in spinach leaves with ozone fumigation — triacylglycerol synthesis from polar lipids. Plant Cell Physiol. 26, 253–262.
685
10.4
Glycosylglycerides
Salas, J.J. et al. (2000). Biochemistry of lipid metabolism in olive and other oil fruits. Prog. Lipid Res. 39, 151–180. Sedgwick, B. and Hubscher, G. (1967). Partial purification and properties of a soluble phosphatidate phosphohydrolase from rat liver. Biochim. Biophys. Acta. 144, 397–408. Smith, M.E. et al. (1967). The role of phosphatidate phosphohydrolase in glyceride biosynthesis. Eur. J. Biochem. 3, 70–77. Snyder, F.L. (1972). Ether Lipids: Chemistry and Biology, Academic Press, New York. Stymne, S. and Stobart, A.K. (1987). Triacylglycerol biosynthesis In The Biochemistry of Plants, vol. 9, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 175–214. Sul, H.S. and Wang, D. (1998). Nutritional and hormonal regulation of enzymes in fat synthesis. Studies of fatty acid synthase and mitochondrial glycerol 3-phosphate acyltransferase gene transcription. Annu. Rev. Nutr. 18, 331–351. Theimer, R.R. and Rosnitscheck, I. (1978). Development and intracellular-localization of lipase activity in rapeseed (Brassica-napus-L) cotyledons. Planta. 139, 249–256. Thompson, M.P. et al. (1993). Increased expression of the mRNA for hormone-sensitive lipase in adipose tissue of cancer patients. Biochim. Biophys. Acta.1180, 236–242. Wang, C.-S. et al. (1992). Structure and functional properties of lipoprotein lipase. Biochim. Biophys. Acta. 1123, 1–17. Weselake, R.J (2005). Storage lipids. In Plant Lipids: Biology, Utilisation and Manipulation, Ed. D.J. Murphy, Blackwell Publishing, Oxford, U.K., pp. 162–225. Weselake, R.J. and Taylor, D.C. (1999). The study of storage lipid biosynthesis using microspore-derived cultures of oilseed rape. Prog. Lipid Res. 38, 401–460. Williams, M.P. et al. (1983). Heparan sulphate and the binding of lipoprotein lipase to porcine thoracic aorta endothelium. Biochim. Biophys. Acta, 756, 83–91. Wilson, R.F. and Rinne, R.W. (1976). Involvement of phospholipids in triglyceride biosynthesis by developing soybean cotyledons. Plant Physiol. 57, 556–559.
10.4
Glycosylglycerides
10.4.1
Galactosylglycerides
Kennedy pathway
DAG +UDP-SQD
+UDP-gal 2
1 MDGD SQDG
+MGDG 4
+UDP-gal 3
DGDG
FIGURE 10.32 Biosynthesis of plant glycosylglycerides. Abbreviations: DAG, diacylglycerol; MGDG, monogalactosyldiacylglycerol; DGDG, digalactosyldiacylglycerol; SQD, sulfoquinovose; SQDG, sulfoquinovosyldiacylglycerol. Reactions: (1) UDP-SQD: diacylglycerol sulfoquinovosyltransferase; (2) UDPgalactose: diacylglycerol galactosyltransferase; (3) UDP-galactose:MGDG galactosyltransferase; (4) galactolipid: galactolipid galactosyltransferase.
fractions was due to its envelope content. Plastid envelope membranes from nonphotosynthetic tissues, such as the chromoplasts of Narcissus pseudonarcissus (Leidvogel and Kleinig, 1977) and potato tuber amyloplasts (Fishwick and Wright, 1980) are also active. The first galactosyltransferase is specific for the formation of a β-glycosidic bond while the other forms an α-glycosidic bond (Douce and Joyard, 1980). There are two digalactosyldiacylglycerol synthetases in Arabidopsis and details of the formation of galactosyldiacylglycerol and the galactosyltransferases will be found in Dormann and Benning (2002) and in Dormann (2005). As an alternative to the use of UDP-galactose, digalactosyldiacylglycerol may be formed by the action of galactolipid: galactolipid galactosyltransferase (see Figure 10.32), first demonstrated by Von Besouw and Wintermans (1978). Comparative aspects of these alternative pathways have been discussed (Joyard and Douce, 1987) and methods for the analysis of galactolipids and their metabolites reviewed (Douce et al., 1990). Recent evidence, discussed by Dormann (2005), suggests that the galactolipid transferase is not important for digalactosyldiacylglycerol synthesis in vivo, but can be activated under certain conditions, such as during membrane isolation. The formation of polyunsaturated fatty acids (mainly α-linolenate) associated with the galactosylglycerides has been reviewed (Harwood, 1996). Depending on the plant type, there seem to be differences between the “16:3 species” and the “18:3 species,” as discussed by Heinz and Roughan (1982). The distinctive features of the fatty acid distributions of galactolipids in 16:3 and 18:3 plants is illustrated in Section 2.10 and discussed thoroughly by
The two galactosylglycerides, monogalactosyldiacylglycerol (MGDG), and digalactosyldiacylglycerol (DGDG) are the major lipid components of the photosynthetic membranes of oxygen-evolving organisms. Because of that, they are the most prevalent membrane lipids in the world. They are rare or only found in trace amounts in other organisms. Trigalactosyl and tetragalactosyl derivatives are minor components of some chloroplasts, and in marine algae and in some bacteria the sugar residue(s) may be glucose. The pathway for galactosylglyceride synthesis is outlined in Figure 10.32. UDP-galactose is generated in the cytoplasm of plant leaf cells and is used by two galactosyltransferase enzymes, which are located in the outer half of the envelope membrane. Joyard and Douce (1976) showed that the galactosyltransferase activity was located in the plastid envelope and that any activity in microsomal 686
Lipid Metabolism
appear to be substrates for successive desaturations of oleate and linoleate to α-linolenate (see Harwood and Jones, 1989). Initial breakdown of galactosylglycerides is catalysed by nonspecific acylhydrolase enzymes (see Galliard, 1980, for review). The enzymes from runner bean leaves (Burns et al., 1979) and potato tubers (Hirayama et al., 1975) have been purified. Two enzymes that differed slightly in substrate specificity were isolated from runner bean leaves (Burns et al., 1979) and the position of hydrolysis identified by mass spectrometry as the bond between the fatty acyl carboxy carbon and the oxygen of glycerol (Burns et al., 1980). Acyl hydrolase enzymes are particularly active; homogenization of potato tubers in water at 4°C for a few seconds being sufficient for most of the endogenous
Browse and Somerville (1991). For 18:3 plants the desaturation of oleate in association with phosphatidylcholine is followed by release of the diacylglycerol for galactolipid formation (Joyard and Douce, 1987). Whether diacylglycerol itself or phosphatidylcholine moves from the endoplasmic reticulum to the chloroplast envelope is not known, although Tanaka and Yamada (1982) demonstrated the latter process mediated via a phospholipid transfer protein. After the formation of monogalactosyldiacylglycerol, further desaturation can then take place (Jones and Harwood, 1980; Harwood, 1996; see also Section 10.1.4). The fatty acid combinations (Heinz, 1977; Rullkötter et al., 1975) and turnover (Heinz et al., 1979) in galactosylglycerides have been studied thoroughly. In 16:3 plants and cyanobacteria the galactosylglycerides Epimerase
CH2OH O
CH2OH O HO
O
OH
O’UDP
CH2OH O OH
OH
O’UDP OH
OH UDP-Glc
O
OH
CH2OH O HO O’UDP OH
−H2O
O’UDP OH −H2O
CH2 O O
OH
CH2SO3− O
SO2− 3
O
OH
O’UDP
O’UDP OH
OH
+2H CH2SO3− O HO
OH
O’UDP OH
UDP-sulfoquinovose +Diacylglycerol
SULFOLIPID
FIGURE 10.33 Synthesis of sulfoquinovosyldiacylglycerol. (See Pugh et al. (1995), Harwood and Okanenko, (2003) and Dormann (2005) for details.)
687
10.4
Glycosylglycerides
membrane lipids to be destroyed. Some recent information is given in Dormann (2005) and see Section 10.2.2.
10.4.2
Burns, D.D. et al. (1980). Properties of acyl hydrolase enzymes from Phaseolus multiflorus leaves. Phytochemistry, 19, 2281–2285. Dormann, P. (2005). Membrane lipids. In Plant Lipids:Biology, Utilisation and Manipulation, Ed. D.J. Murphy, Blackwell Publishing, Oxford, U.K., pp. 123–161. Dormann, P. and Benning, C. (2002). Galactolipids rule in seed plants. Trends Plant Sci. 7, 112–118. Douce, R. and Joyard, J. (1980). Plant galactolipids. In Biochemistry of Plants, vol. 4, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 321–362. Douce, R. et al. (1990). Glycolipid analyses and synthesis in plastids. In Methods in Plant Biochemistry, vol. 4, Eds. J.L. Harwood and J.R. Boyer, Academic Press, London, pp. 71–103. Fishwick, M.J. and Wright, A.J. (1980). Isolation and characterization of amyloplast envelope membranes from Solanum tuberosum. Phytochemistry, 19, 5559. Galliard, T. (1980). Degradation of acyl lipids: hydrolytic and oxidative enzymes. In Biochemistry of Plants, vol. 4, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 85–116. Harwood, J.L. (1980). Sulpholipids. In The Biochemistry of Plants, vol. 4, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 301–320. Harwood, J.L. (1996). Recent advances in the biosynthesis of plant fatty acids. Biochim. Biophys. Acta. 1301, 7–56. Harwood, J.L. and Jones, A.L. (1989). Lipid metabolism in algae. Adv. Bot. Res. 16, 1–53. Harwood, J.L. and Okanenko, A.A. (2003). Sulphoquinovosyldiacylglycerol (SQDG) – the sulpholipid of higher plants. In Sulphur in Plants, Eds. Y.P. Abrol and A. Ahmad, Kluwer, Dordrecht, pp. 189–219. Heinz, E. (1977). Enzymatic reactions in galactolipid biosynthesis. In Lipids and Lipid Polymers of Higher Plants, Eds. M. Tevini and H.K. Lictenthaler, Springer-Verlag, Berlin, pp. 102–120. Heinz, E. and Roughan, P.G. (1982). De novo synthesis, desaturation and acquisition of monogalactosyldiacylglycerol by chloroplasts of 16:3 and 18:3 – plants. In Biochemistry and Metabolism of Plant Lipids, Eds. J.F.G.M. Wintermans and P.J.C. Kuiper, Elsevier, Amsterdam, pp. 169–182. Heinz, E. et al. (1979). Investigations on the origin of diglyceride diversity in leaf lipids. In Recent Advances in the Biochemistry and Physiology of Plant Lipids, Eds. L.A. Appelqvist and C. Liljenberg, Elsevier, Amsterdam, pp. 99–120. Heinz, E. et al. (1989). Synthesis of different nucleoside 5´diphosphosulphoquinovoses and their use for studies of sulpholipid synthesis in chloroplasts. Eur. J. Biochem. 184, 445–454. Hirayama, O. et al. (1975). Purification and properties of a lipid acyl-hydrolase from potato tubers. Biochim. Biophys. Acta. 384, 127–137. Jones, A.V.M. and Harwood, J.L. (1980). Desaturation of linoleic acid from exogenous lipids by isolated chloroplasts. Biochem. J. 190, 851–854. Joyard, J. and Douce, R. (1976). Preparation and enzymatic activities of spinach chloroplast envelope. Physiol. Veg. 14, 31–48. Joyard, J. and Douce, R. (1987). Galactolipid synthesis. In The Biochemistry of Plants, vol. 9, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 215–274.
Sulfolipid (diacylsulfoquinovosylglycerol)
Early work on possible pathways for the formation of the plant sulfolipid were reviewed by Harwood (1980) and by Mudd and Kleppinger-Sparace (1987). The immediate precursor, UDP-sulfoquinovose, was shown to transfer its sugar to diacylglycerol using SQDE synthase in the envelope membrane of chloroplasts (Heinz et al., 1989). The UDP-sulfoquinovose itself is synthesised from UDP-glucose and sulfite in a series of reactions whose theory is explained by Pugh et al. (1995). Apparently, a single gene codes for the protein needed to carry out all the conversions to UDP-sulfoquinovose (Sanda et al., 2001) (Figure 10.33). The metabolism, genetics and possible physiological functions of sulfoquinovosyldiacylglycerol have been reviewed by Benning (1998) and by Harwood and Okanenko (2003). Plant tissues are capable of, at least, partial catabolism of sulfolipid. Deacylation appears to be the first step and may take place with two enzymes, such as in green algae (e.g., Yagi and Benson, 1962) or by the action of a single acyl hydrolase (cf. Harwood, 1980). Cleavage of sulfoquinovosylglycerol to sulfoquinovose was reported in leaves of Medicago sativa (Lee and Benson, 1972) where sulfolactaldhyde and, later, sulfolactic acid accumulated. In contrast, catabolism by cell-free preparations from Phaseolus multiforus stopped at sulfoquinovose (Burns et al., 1980). Further aspects of sulfolipid catabolism are discussed by Harwood (1980) and in Harwood and Okanenko (2003). Studies on the turnover of molecular species of sulfolipid have been reported in Vicia faba and Hordeum vulgare. In both these plants, the more saturated species were turned over at high rates, whereas the predominant 1-linolenoyl, 2-palmitoyl species had a low rate of turnover. These differences in metabolism may be related to the function of the sulfolipid — the fast turning over species being involved in a metabolic function, while the trienoic species have a structural role (cf. Harwood, 1980).
References Benning, C. (1998). Biosynthesis and function of the sulfolipid sulfoquinovosyldiacylglycerol. Ann. Rev. Plant Physiol. Plant Mol. Biol. 49, 53–75. Benson, A.A. (1963). The plant sulfolipid. Adv. Lipid Res. 1, 382–394. Browse, J. and Somerville, C. (1991). Glycerolipid synthesis: biochemistry and regulation. Annu. Rev. Plant Physiol. 42, 467–506. Burns, D.D. et al. (1979). Purification of acyl hydrolase enzymes from the leaves of Phaseolus multiflorus. Phytochemistry. 18, 1793–1797.
688
Lipid Metabolism
Lee, R.F. and Benson, A.A. (1972). The metabolism of glyceryl 35S-sulfoquinovoside by the coral tree, Erythrina cristagalli, and alfalfa, Medicago sativa. Biochim. Biophys. Acta. 261, 35–37. Leidvogel, B. and Kleining, H. (1977). Lipid metabolism in chromoplast membranes from daffodil — glycosylation and acylation. Planta, 133, 249–253. Mudd, J.B. and Kleppinger-Sparace, K. (1987). Sulfolipids. In The Biochemistry of Plants, vol. 9, Eds. P.K. Stumpf and E.E. Conn Academic Press, New York, pp. 275–289. Pugh, C.E. et al. (1995). A new pathway for the synthesis of the plant sulpholipid, sulphoquinovosyldiacylglycerol. Biochem. J. 309, 513–519. Rullkötter, J. et al. (1975). Combination and positional distribution of fatty acids in plant digalactosyl diglycerides. Z. Pflanzenphysiol. 76, 163–175. Sanda, S. et al. (2001). Recombinant Arabidopsis SQD1 converts UDP-glucose and sulphite to the sulfolipid head group precursor UDP-sulphoquinovose in vivo. J. Biol. Chem. 276, 3941–3946. Tanaka, T. and Yamada, M. (1982). Properties of phospholipid exchange proteins from germinated castor bean endosperm. In Biochemistry and Metabolism of Plant Lipids, Eds. J.F.G.M. Wintermans and P.J.C. Kuiper, Elsevier, Amsterdam, pp. 99–110. Van Besouw, A. and Wintermans, J.F.G.M. (1978). Galactolipid formation in chloroplast envelopes. I. Evidence for two mechanisms in galactosylation. Biochim. Biophys. Acta 529, 44–53.
Yagi, T. and Benson, A.A. (1962). Plant sulfolipid. V. Lysosulfolipid formation. Biochim. Biophys. Acta 57, 601–603.
10.5
Sphingolipids
10.5.1
Biosynthesis
Sphingolipid chemistry and biosynthesis have been reviewed by Carter et al. (1965), Wiegandt (1971), Stoffel (1971), Kanfer and Hakomori (1983), Sweeley (1991), and Merrill and Sandhoff (2002). Biosynthesis of long-chain bases has been studied in a series of experiments by Snell and coworkers (e.g., Brady et al., 1969) and by Stoffel and coworkers (see Stoffel, 1971). A condensing enzyme, requiring pyridoxal phosphate, is able to condense serine with palmitoyl-CoA to produce 3-oxosphinganine (3-keto sphinganine). The reaction, which is probably the ratelimiting step in sphingoid base biosynthesis, proceeds with overall retention of configuration of the C-2 carbon of serine. L-Cycloserine (4-amino-3-isoazolidinone) is an irreversible inhibitor of the palmitoyltransferase and depresses the level of central nervous system sphingolipids if administered to mice (Sweeley, 1991). Other inhibitors of this and succeeding reactions in sphingolipid metabolism are listed in Merrill and Sandhoff (2002). The next reaction is a reduction using NADPH to produce sphinganine (Figure 10.34). The reductase responsible is probably COO−
O
Palmitoyl-CoA + Serine
SCoA
+ H C CH2OH NH3+ Serine palmitoyltransferase
CO2 O
3-ketosphinganine NADPH + H+
CH2OH
NH3+ 3-Ketosphinganine reductase
NADP OH
CH2OH NH +
Sphinganine
3
Fatty acyl-CoA
Ceramide synthase
CoASH OH Dihydroceramide
CH2OH
NH O Cofactors?
“Desaturase” OH
Ceramide
NH O
FIGURE 10.34
CH2OH
Biosynthesis of ceramide.
689
10.5 Sphingolipids
closely associated with the palmitoyltransferase in the endoplasmic reticulum. Both the palmitoyltransferase and the reductase exhibit chain-length specificity, with C14 to C18 CoA esters and C14 to C20 3-dehydrosphinganines being utilized. The amino acid may form a Shiffbase complex with the pyridoxal phosphate coenzyme and Mg 2+ during the condensation reaction. Some researchers have also suggested that palmitaldehyde is an intermediate in the reaction, although this has not been proven. Stoffel et al. (1968) have studied the direct transformation of sphinganine into 4-hydroxysphinganine in the yeast, Hansenula ciferri. The origin of the hydroxyl group is obscure, since Thorpe and Sweeley (1967) concluded that it did not arise from either molecular oxygen or water. For further discussion of the formation of sphingoid bases, see Sweeley (1991) and for their structures and nomenclature, see Merrill and Sandhoff (2002). Ceramides can be rapidly formed from erythro or threo long-chain bases. The specificity of this acylation by acyl-CoAs has been studied in brain and other tissues (see Stoffel, 1971). The significance of the nature of the fatty acid moiety in ceramides is well seen when the glycosylation of ceramides is considered: hydroxyl fatty acid-containing ceramides accept predominantly galactose, while nonhydroxy fatty acid ceramides accept glucose. Reversal of ceramidase to yield ceramide does not appear to be important (Merrill and Sandoff, 2002). Genes for ceramide synthase have been identified in yeast and animals. Microorganisms can produce inhibitors of the enzyme and fumonisins from Fusaria spp. are notable as causes of a number of human pathologies. The last step in ceramide synthesis is the insertion of a 4,5-trans-double bond into the sphingoid base (see Figure 10.34). Desaturase genes have been identified in plants (Sperling et al., 2000; Dunn et al., 2004) and animals. For the 4-hydroxysphinganines, insertion of the 4-hydroxyl group occurs at the level of sphinganine (see Dunn et al., 2004). Cerebrosides are made from ceramide using glycosyltransferase enzymes, which are specific for UDP-galactose or UDP-glucose. The latter also recognises ceramides with nonhydroxy fatty acids, as mentioned above. Thus, galactocerebrosides (and sulfatides) are enriched in α-hydroxy fatty acids (Merrill and Sandhoff, 2002). As expected, the formation of sulfatides involves transfer of sulfate from phosphoadenosine phosphosulfate (PAPS). Galactosylceramide and lactosylceramide were both acceptors for the sulfate from PAPS and the reaction is catalysed by the Golgi enzyme, galactosylceramide sulfotransferase. Sulfate transfer is preceded by receptormediated translocation of PAPS from the cytosol across the Golgi membrane, and this process can be inhibited by 3′-P-AMP, palmitoyl-CoA, or atractyloside (Sweeley, 1991). Sulfatide synthesis is most rapid in the period 20 to 25 days after birth in rat brain (Stoffel, 1971). The
Man-Man-Glc-Cer (mollu)
Man-Glc-Cer Glc-Cer
GlcNAc-Man-Glc-Cer (arthro) Gal-Glc-Cer GalNAc-Gal-Glc-Cer (ganglio) Gal-Gal-Glc-Cer (globo)
Cer GlcNAc-Gal-Glc-Cer (lacto) Gal-Cer
Gal-Gal-Cer (gala)
FIGURE 10.35 Biosynthesis of different root glycosphingolipids from ceramide. (From Sweeley, 1991.)
cDNA encoding the sulfotransferase has been cloned (Honke et al., 1997) and activity of this enzyme appears important in controlling overall sulfatide formation. See also Vos et al. (1994) for metabolic and functional aspects of sulfogalactolipids. Synthesis of the neutral glycosphingolipids begins from glucosylceramide or galactosylceramide. Specific glycosyltransferases are involved, and the activated forms of the sugar substrates (UDP-Glc, UDP-Gal, UDP-GlcNAc, GDP-Man, and GDP-Fuc) are produced in the cytosol from nucleoside triphosphates and hexose 1-phosphates. The formation of different root glycosphingolipids from ceramide is illustrated in Figure 10.35 and occurs in the lumen of the Golgi apparatus. The active sites of the glycosyltransferases are localized on the lumenal surface of the Golgi membrane and may be organized in several multiglycosyltransferases, which could account for the various types of products that are formed (Sweeley, 1991). This contrasts to glycosylceramide, which is made on the cytosolic face of the endoplasmic reticulum and/or early Golgi membranes. Thus, glucosylceramide must undergo trans-bilayer movement before further metabolism (Merrill and Sandhoff, 2002). Gangliosides are formed by stepwise elongation of the carbohydrate chain by the action of various glycosyltransferases. Several of the individual enzymes have been studied (see Stoffel, 1971) and assay methods are detailed by Basu et al. (1987). The individual sugars are transferred from their UDP derivatives, while sialic acid residues are donated by CMP-N-acetylneuraminic acid (CMP-NANA). CMP-NANA itself is produced by a reaction of NANA with CTP. Initial reactions in the formation of gangliosides are shown in Figure 10.36. Ganglioside synthesis involves the use of relatively few enzymes (Table 10.8), which can 690
Lipid Metabolism
NeuAc
NeuAc Gal (GT3)
NeuAc
CMP-NeuAc
Gal
Glc
Glc
Other c-series gangliosides
Cer
Cer Other b-series gangliosides
NeuAc (GD3)
NeuAc CMP-NeuAC
UDP-GalNAc Gal
CMP-NeuAc
Glc
NeuAc Lactosylceramide
Cer
GalNAc
Gal
Glc
Glc
Cer
Other a-series gangliosides
NeuAc (GM2)
(GM3)
UDP-GalNAac
GalNAc
Gal
Cer UDP-Gal
Gal
GalNAc
Gal
Glc
Cer
CMP-NeuAc
Gal
GalNAc
Gal
Glc
Cer
(GM1b)
NeuAc
Other O-series gangliosides
FIGURE 10.36 Key steps in the initial formation of gangliosides. (From Gurr, M.I., Harwood, J.L. and Frayn, K.N. (2002). Lipid Biochemistry, 5th ed., Blackwell Scientific, Oxford, U.K. With permission.)
TABLE 10.8
Glycosyltransferases catalysing the biosynthesis of gangliosides
Abbreviation
Name
Linkage created
Substrates
GalT-1 SAT-1 SAT-2 SAT-3 SAT-4 SAT-5 GalNAcT GalT-2
β-Galactosyltransferase Sialyltransferase Sialyltransferase Sialyltransferase Sialyltransferase Sialyltransferase β-N-Acetylgalactosaminyltransferase β-Galactosyltransferase
Gal(β14)Glc Neu5Ac(23)Gal Neu5Ac(28)Neu5Ac Neu5Ac(28)Neu5Ac Neu5Ac(23)Gal Neu5Ac(28)Neu5Ac GalNAc(β14)Gal Gal(β13)GalNAc
Glc-Cer Lac-Cer GM3 GD3 GA1, GM1a, GD1b GM1, GD1a, GT1b Lac-Cer, GM3, GD3 GA2, GM2, GD2
From Sweeley (1991).
both the plasma membrane and in the Golgi apparatus with the proportions depending on the cell type. An analogous sphingolipid, ceramide phosphorylehanolamine, is made in a similar way but using phosphatidylethanolamine as the donor. Inositolphosphoceramides are made similarly (Merrill and Sandhoff, 2002). A comprehensive review of sphingomyelin metabolism, intracellular transport and various aspects of its biological functions has been published (Koval and Pagano, 1991). The biological function (Stoffel, 1971) and immunochemistry of the sphingolipids — especially the gangliosides — have been reviewed (Hakomori, 1981). Kanfer
work in various combinations. Thus, the final ganglioside composition can be influenced by the relative activity of these enzymes as well as substrate availability (Muramatsu, 2000). Regulation of ganglioside biosynthesis occurs at transcriptional and post-transcriptional levels (Merrill and Sandhoff, 2002). As with neutral glycosphingolipid synthesis, the glycosyltransferases are membrane-bound and predominantly located in the Golgi apparatus. Sphingomyelin, which is both a phospho- and sphingolipid, is synthesised by transfer of phosphorylcholine from phosphatidylcholine to ceramide, liberating diacylglycerol. De novo sphingomyelin synthesis occurs on 691
10.5 Sphingolipids
further by a neuraminidase to give ceramide lactoside (Figure 10.37) (see Brady, 1978). The cleavage of ceramide lactoside needs a β-galactosidase. Two β-galactosidases have been demonstrated in mammalian tissues, both of which have acidic pH optima (pH 4.2 and 4.8). When assayed in vitro, either of these enzymes catabolizes ceramide lactoside, depending on the detergent used in the experiment. The pH4.2 galactosidase also hydrolyses galactose from galactocerebroside. The pH 4.8 enzyme preferentially catalyses the removal of the terminal galactose of GM1 as well as catabolizing ceramide lactoside. There are also two liver β-galactosidases with more neutral pH optima that catabolize ceramide lactoside, but are inactive with galactocerebroside or GM1 (BenYoseph et al., 1977). For degradation of sphingolipids with 4 or less carbohydrate residues, there is often a requirement for sphingolipid activator proteins (SAPs or saposins) in vivo. Some inherited diseases are caused by mutation of the domains of the exoglycosidases that interact with SAPs (Merrill and Sandhoff, 2002). Glucocerebroside is hydrolysed by glucocerebrosidase to give glucose and ceramide. Catabolism of ceramide is catalysed by acid (Gatt, 1963), neutral (Sugita et al., 1975) or alkaline ceramidases in mammalian tissues. A sphingosine
and Hakomori (1983) have published a comprehensive review on sphingolipid biochemistry and see Bell et al. (1993) for aspects of metabolism. Merrill and Hannun (2000) have edited a useful volume of Methods in Enzymology and Lynch (1993) describes plant sphingolipids.
10.5.2
Breakdown
Breakdown of sphingolipids occurs by stepwise hydrolytic cleavage of the various substituents, starting with the terminal hydrophilic portions of the molecules. As an example, catabolism of ganglioside GM1[Gal(β1→3) GalNAc(β1→4)Gal-(3→2αAcNeu)(β1→4)Glc(β1→1′) Cer] begins by the action of a β-galactosidase that cleaves the terminal galactose to give ganglioside GM2 as the other product. Ganglioside GM2 can be cleaved by two pathways. Either a hexosaminidase removes the molecule of GalNAc or a neuroaminidase hydrolyses AcNeu. These reactions have been demonstrated in vivo, but because more GM2[GalNAc(β1→4) Gal-(3→2AcNeu) (β1→4) Glc(β1→1′)Cer] accumulates in Tay-Sachs disease than GA2[GalNAc(β1→4)-Gal(β1→4)Glc(β1→1′)Cer], the hexosaminidase reaction appears the more important. The latter enzyme attacks GM2 to yield GM3[(AcNeu2→3)Gal(β1→4)Glc(β1→1′)Cer], which is then catalysed β
Cer-Glc(4
1)Gal (4
β
1)GalNAc (3
β
1)Gal
3 2 NANA Ganglioside GM1 Galactosidase Cer-Glc-Gal-GalNAc NANA Tay-sachs ganglioside, GM2 Hexosaminidase
Neuraminidase
Cer-Glc-Gal-NANA
Cer-Glc-Gal-GalNAc Hexosaminidase
Neuraminidase Cer-Glc-Gal
β-Galactosidase Cer-Glc β-Glucosidase Cer Ceramidase Sphingosine + Fatty acid
FIGURE 10.37 Breakdown of ganglioside GMI-galactosidase.
692
Lipid Metabolism
Gangliosides Sphingomyelin Sphingomyelin synthase Neutral glycosphingolipids
Glucosyl ceramide β -glucosidase
Galactosyl ceramide
Ceramide
β -galactosidase
Arylsulphatase A
Phosphatidate phosphohydrolase
Ceramide kinase
Ceramide synthase
Sulphatide
Sphingomyelinase
Ceramidase
Sphingosine
Sphingosine phosphatase Ethanolamine phosphate
Ceramide 1- phosphate
Sphingosine-N-methyl transferase
Sphingosine kinase
N,N-dimethyl Sphingosine
Sphingosine 1- phosphate Sphingosine 1- phosphate lyase
Hexadecenal
FIGURE 10.38
Interconversions between and catabolism of simple sphingolipids.
For a simple review of sphingolipid metabolism and function see Gurr et al. (2002).
base and a nonesterified fatty acid are liberated. The former is phosphorylated to give a 1-phosphate derivative and then hydrolysed to yield a long-chain aldehyde and phosphoethanolamine (Stoffel, 1971). Sulfatides are catabolised by a sulfatase and the resultant galactocerebroside hydrolysed by a β-galactosidase. In addition, sphingomyelin is hydrolysed by sphingomyelinase to yield ceramide and phosphocholine (see Barenholz and Gatt, 1982) (Figure 10.38). Sphingolipid breakdown is reviewed by Kanfer and Hakomori (1983) and deficiency diseases of sphingolipid catabolism are covered in Section 11.5. Sweeley (1991) and Merrill and Sandhoff (2002) discuss some aspects of sphingolipid breakdown and regulation of their turnover. In the latter connection, glycosphingolipids have important functions during cellular differentiation and oncogenic transformation. Recent interest has also included their function as modulators of transmembrane signalling and as mediators for cellular interactions (Hakomori, 1990). Sphingolipids as cell signalling molecules are discussed in Section 10.6 (see also Aue et al., 2000). Apart from the main compounds (ceramide, sphingoid bases, and sphingosine-1-phosphate), a number of other lysosphingolipids have acute activity. Thus, lysosphingomyelin is a potent mitogen while psychosine (lyso-GlcCer or lyso-GalCer) is highly cytotoxic. Ceramide-1-phosphate is an active Ca++-mobilising agent (see Merrill and Sandhoff, 2002).
References Aue, N. et al. (2000). Sphingomyelin metabolites in vascular cell signalling and atherogenesis. Prog. Lipid Res. 39, 207–229. Barenholz, Y. and Gatt, S. (1982). Sphingomyelin: metabolism, chemical synthesis, chemical and physical properties. In Phospholipids (Eds. J.N. Hawthorne and G.B. Ansell), Elsevier, Amsterdam, pp. 129–177. Basu, M. et al. (1987). Glycolipids. In Methods in Enzymology, vol. 138 (Ed. V. Ginsburg), Academic Press, Orlando, FL, pp. 575–607. Bell, R.M. et al. Eds. (1993). Sphingolipids, Part B: regulation and function of metabolism. Adv. Lipid Res. 26 (special issue). Ben-Yoseph, Y. et al. (1977). Purification and properties of neutral beta-galactosidases from human liver. Arch. Biochem. Biophys, 184, 373–380. Brady, R.N. et al. (1969). Biosynthesis of sphingolipid bases. 3. Isolation and characterization of ketonic intermediates in the synthesis of sphingosine and dihydrosphingosine by cell-free extracts of Hansenula ciferri. J. Biol. Chem. 244, 491–496. Brady, R.O. (1978). Sphingolipidoses. Annu. Rev. Biochem. 47, 687–713. Carter, H.E. et al. (1965). Glycolipids. Annu. Rev. Biochem. 34, 109–142.
693
10.6
Lipids as signalling molecules
Dunn, T.M. et al. (2004). A post-genomic approach to understanding sphingolipid metabolism in Arabidopsis thaliana. Ann. Bot.. 93, 483–497. Gatt, S. (1963). Enzymic hydrolysis and synthesis of ceramides. J. Biol. Chem. 238, 3131–3133. Gurr, M.I. and James, A.T. (1980). Lipid Biochemistry, Chapman & Hall, London. Gurr, M.I. et al. (2002). Lipid Biochemistry, 5th ed., Blackwell Scientific, Oxford, U.K. Hakomori, S.-I. (1981). Glycosphingolipids in cellular interaction, differentiation, and oncogenesis. Annu. Rev. Biochem. 50, 733–764. Hakomori, S.-I. (1990). Bifunctional role of glycosphingolipids. Modulators for transmembrane signaling and mediators for cellular interactions. J. Biol. Chem. 265, 18713–18716. Honke, K. et al. (1997). Molecular cloning and expression of cDNA encoding human 3-phosphoadenylsulfate: galactosylceramide 3-sulfotransferase. J. Biol. Chem. 272, 4864–4868. Kanfer, J.N. and Hakomori, S-I. (1983). Sphingolipid Biochemistry, Plenum, New York. Koval, M. and Pagano, R.E. (1991). Intracellular transport and metabolism of sphingomyelin. Biochim. Biophys. Acta. 1082, 113–125. Lynch, D.V. (1993). Sphingolipids. In Lipid Metabolism in Plants, Ed. T.S. Moore, CRC Press, Boca Raton, FL, pp. 285–308. Merrill, A.H. and Hannun, Y.A. Eds. (2000). Sphingolipid metabolism and function Part A, Part B. In Methods of Enzymology, vols. 311 and 312. Merrill, A.H. and Sandhoff, K. (2002). Sphingolipids: metabolism and cell signalling. In Biochemistry of Lipids, Lipoproteins and Membranes, Eds. D.E. Vance and J.E. Vance, 4th ed., Elsevier, Amsterdam, pp. 373–407. Muramatsu, T. (2000). Essential roles of carbohydrate signals in development, immune response and tissue functions, as revealed by gene targeting. J. Biochem (Tokyo). 127, 171–176. Spence, M.W. (1989). Sphingomyelin biosynthesis and catabolism. In Phosphatidylcholine Metabolism, Ed. D.E. Vance, CRC Press, Boca Raton, FL, pp. 185–203. Sperling, P., Blume, A., Zahringer, U. and Heinz, E. (2000). Further characterisation of ∆8-desaturases from higher plants. Biochem. Soc. Trans. 28, 638–641. Stoffel, W. (1971). Sphingolipids. Annu. Rev. Biochem. 40, 57–82. Stoffel, W. et al. (1968). Metabolism of sphingosine bases. IX. Degradation in vitro of dihydrospingosine and dihydrospingosine phosphate to palmitaldehyde and ethanolamine phosphate. Hoppe-Seyler’s Z. Physiol. Chem. 349, 1745–1748. Sugita, M. et al. (1975). Ceramidase and ceramide synthesis in human kidney and cerebellum. Description of a new alkaline ceramidase. Biochim. Biophys. Acta. 398, 125–131. Sweeley, C.C. (1991). Sphingolipids. In Biochemistry of Lipids, Lipoproteins and Membranes, Eds. D.E. Vance and J. Vance, Elsevier, Amsterdam, pp. 327–361. Thompson, G.A. (1973). Phospholipid metabolism in animal tissues. In Form and Function of Phospholipids, Eds. G.B. Ansell, R.M.C. Dawson, and J.N. Hawthorne, Elsevier, Amsterdam, pp. 67–96. Thorpe, S.R. and Sweeley, C.C. (1967). Chemistry and metabolism of sphingolipids. On the biosynthesis of phytosphingosine by yeast. Biochemistry, 6, 887–897. Wiegandt, H. (1971). Glycosphingolipids. Adv. Lipid Res. 9, 249–289.
10.6
Lipids as signalling molecules
The first lipids to be recognised as giving rise to lipid second messengers were the inositol-containing phosphoglycerides. Since that time a multitude of lipids with acute biological activity have been recognised and this general area is one of the most active in biochemistry. Signalling molecules include intact phosphoglycerides (e.g., phosphatidic acid (PA), platelet-activating factor (PAF), membrane-soluble hydrolysis products (e.g., diacylglycerol (DAG)), water-soluble products (e.g., inositol1,4,5-tris-phosphate) and other degradative metabolities (e.g., lysoPA, fatty acids). The role of polyunsaturated fatty acids themselves is described in Section 10.1 and Section 11.1. Two general reviews about lipids as signalling molecules or bioactive lipids are Bell et al. (1996) and Nicolaou and Kokotos (2004). Methods for analysis are described in Christie (2003) and in Laychock and Rubin (1999). Simpler summaries of some aspects of lipid signalling are in Gurr et al. (2002) and Vance and Vance (2002).
10.6.1
Platelet activating factor (PAF)
Intact phospholipids have various important roles in signalling. One of the first such lipids to be recognised was platelet activating factor (PAF; 1-O-alkyl-2-acetyl-snglycerol-3-phosphocholine). PAF is produced by many types of cells in response to stimuli. In inflammatory cells, such as monocytes or macrophages, it can be rapidly synthesised by a deacylation-acetylation pathway (Tokumura, 1995). It can also be produced by de novo synthesis, which seems particularly important for maintaining PAF levels in the central nervous and reproductive systems. Because of its very potent biological activity, PAF levels in cells and in the circulation are strictly regulated (Tokomura, 1995). PAF-acetylhydrolase plays a primary role in inactivation. There are a family of PAF-acetylhydrolases that consist of two intracellular isoforms and one secreted (plasma) isoform. Details of these enzymes, physiological function and role in human disease are given by Karasawa et al. (2003). PAF elicits responses in many cells and organs. Originally it was shown to aggregate platelets (hence, the name), but it is now known to affect vasodilation/constriction, bronchial responsiveness and many acute responses connected to inflammation. PAF acts by binding to a unique G-protein-coupled seven transmembrane receptor that links it to various signalling pathways (Ishii and Shimizu, 2000). A series of PAF-like lipids have been identified, many of which have biological activity and some can also be formed under oxidising conditions such as those produced by smoking (see Tokumura, 1995). A general review of the biochemistry of PAF is that by Synder (1995) and Honda et al. (2002) describe PAF
694
Lipid Metabolism
Lysophosphatidic acid (LPA) was first identified as the active ingredient of Darmstoff (smooth muscle-stimulating substance) in 1957. Tokumura (1995) reviewed the activity of LPAs as a vasopressor, a platelet agonist, growth factor, and putative second messenger. LPA can be generated from phosphatidic acid (phospholipase A1 or A2), from lysophospholipids (phospholipase D), or by oxidative modification of low-density lipoprotein. LPA can bind to receptors, a number of which have been reported (Fukushima et al., 2001). It is thought to have important pathophysiological roles in cancer, cell survival apoptosis, and vascular activity (Tigyi and Parrill, 2003). See Pyne (2004) for an update on the metabolism and function of LPA. The biochemistry of the parent compound for LPA, phosphatidic acid, is discussed thoroughly by Bocckino and Exton (1996), who pointed out that many of the purported physiological effects of phosphatidic acid can be explained by traces of LPA. Phosphatidic acid, itself, has many important functions in plants that are well reviewed by Wang (2006) and by Testerink and Munnik (2005).
Reviews of the roles of different intact inositol phospholipids (as well as their biologically active metabolites) in plants have been made recently (Meijer and Munnik, 2003; Drøbak, 2005). Although plant proteins, such as those with PH- or PX-domains, can bind to certain phosphoinositides, there are unique features of plants compared to mammals. Moreover, the pathways for phosphorylation and, hence, the spectrum of plant phosphoinositides (Drøbak, 2005) differ from animal systems (Vanhaesebroeck et al., 2001). Phosphoinositides can also exert their role in the control of cellular processes by being hydrolysed and giving rise to second messengers. The classic reaction is the hydrolysis of PIP2 to give rise to the dual second messengers diacylglycerol and inositol 1,4,5-trisphosphate (Berridge, 1987). For more recent description of these effectors and general descriptions of the importance of phosphoinositides, refer to Payrstre et al. (2001), Toker (2002), and Payrastre (2004). The product of phosphoinositide hydrolysis with phospholipase C (above) is diacylglycerol (DAG), which has a well-known function with protein kinase C (Takai et al., 1979). While hydrolysis of phosphoinositides produces a quick elevation of DAG, it can be produced from phosphatidylcholine in a more sustained manner via phosphatidic acid (Nishizuka, 1995). Other aspects of DAGs are covered in Becker and Hannun (2004).
10.6.3 Inositol lipids
10.6.4 Plasmalogens
Although phosphatidylinositol-4,5-bis-phosphate (PIP2) is an important precursor of the twin second messengers inositol-1,4,5-trisphosphate and diacyglycerol, it has functions as an intact lipid also. These include functions in ion channel function (e.g., Kobrinsky et al., 2000) and in membrane trafficking (see Hilgemann, 2003). There are also roles for phosphoinositides in cytoskeletal function (Yin and Janmey, 2003) and in the activity and binding of phospholipase D. Phosphatidylinositol 4-phosphate may have a function in binding the cytoskeletal protein, talin (Payrastre, 2004). Recent work has also revealed a number of important signalling roles for phosphoinositides that does not involve their hydrolysis. Thus, in addition to PIP2, there are well-evidenced roles for phosphatidylinositol-3phosphate and phosphatidylinositol-3,4,5-trisphosphate as membrane lipids for the recruitment and/or activation of various proteins. In turn, this can influence the action of a large number of proteins and, hence, signalling pathways that underpin mechanisms for signal transduction, cytoskeletal, and membrane trafficking events (Martin, 1998). Additional information on the synthesis and function of 3-phosphorylated inositol phospholipids is given by Vanhaesebroeck et al. (2001) and Katso et al. (2001). Vivanco and Sawyers (2002) discuss phosphatidylinositol 3-kinases and cancer.
Plasmalogens are thought to have a number of important roles in controlling tissue functions. General reviews are those of Farooqui and Horrocks (2001) and Nagan and Zoeller (2001). Functions include those in ion transport, membrane fusion, protection of membranes against oxidative stress, cholesterol efflux, and cellular differentiation. Other ether lipids have been reported to have various physiological functions and to be involved in different human diseases (Farooqui and Horrocks, 2004).
receptors. Farooqui and Horrocks (2004) discuss general aspects of PAF and provide up-to-date references.
10.6.2 Lysophosphatidic acid (LPA)
10.6.5 Sphingolipids Sphingolipids can have various signalling roles or can act as bioactive lipids. Their relationship with each other is illustrated in Figure 10.39 (see also Gurr et al., 2002). Ceramide was the first such lipid to be described as a second messenger (Hannun et al., 1996). Its biochemistry and signalling actions have been reviewed by Hannun and Obeid (2002) and by de Avalos et al. (2004). Ceramide plays a very important role in apoptosis (Pettus et al., 2002), in oxidative and heat stress, and in various diseases (de Avalos et al., 2004). Ceramides can be hydrolysed to produce sphingosine, which itself has important properties in regulating cellular systems (Merrill et al., 1996). Importantly, sphingosine can be phosphorylated to produce another signalling
695
10.6
Lipids as signalling molecules
Gangliosides, Complex neutral ceramides Sphingomyelin Sphingomyelin synthase Sphingomyelinase
Glucosyl ceramide
Ceramide glucosyl transferase
β - glucosidase
Ceramide kinase
Ceramide
Phosphatidate phosphohydrolase Ceramide synthase
Ceramidase Sphingosine kinase
Sphingosine
Sphingosine 1- phosphatase
FIGURE 10.39
Sphingosine 1- phosphate
Biosynthetic and catabolic pathways of the core sphingolipid signaling molecules.
found was 2-arachidonylglycerol and most other compounds contain the arachidonyl component (Kokotos, 2004). The endocannabinoids can be chemically synthesised (Razdan and Mahadevan, 2002) and have therapeutic possibilities (Goutopoulos and Makriyannis, 2002). A general review is that of Kokotos (2004). N-acylethanolamines that occur in significant amounts in certain plant tissues and which also bind to endocannabinoid receptors appear to have a role in plant defence (for review, see Chapman, 2004).
molecule, sphingosine 1-phosphate. This compound was reviewed by Pyne and Pyne (2000) and has two major intracellular functions. First, it acts in a “sphingolipid rheostat” where the balance between such lipids determines cellular fate (Pyne, 2004). Second, it is involved in calcium homeostatis. Sphingosine 1-phosphate binds to specific G-protein coupled receptors (Kluk and Hla, 2002). The general aspect of sphingolipids and signalling was reviewed by Smith and Merrill (2002). A large proportion of the sphingolipids that are present in cells and tissues are the complex glycosphingolipids. These lipids accumulate in particular patterns in different cell types and species (Hakomori, 1981). The pattern of glycosphingolipids changes with cell growth, differentiation, viral transformation, ontogenesis, and oncogenesis (Kolter, 2004). Because of their widespread functions, there is increasing interest in them as therapeutic agents or targets (Gagnon and Saragovi, 2002). The role of sphingolipids in storage diseases is covered in Section 11.5.
10.6.6
Ceramide 1- phosphate
References Becker, K.P. and Hannun, Y.A. (2004). Diacylglycerols. In Bioactive Lipids (Eds. A. Nicolaou and G. Kokotos), The Oily Press, Bridgwater, U.K., pp. 37–61. Bell, R.M. et al. Eds. (1996). Lipid Second Messengers, Handbook of Lipid Research, Plenum Press, New York. Berridge, M.J. (1987) Inositol triphosphate and diacylglycerol: two interacting second messengers. Annu. Rev. Biochem. 56, 159–193. Bocckino, S.B. and Exton, J.H. (1996). Phosphatidic acid. In Lipid Second Messengers, Eds. R.M. Bell, J.H. Exton and S.M. Prescott, Plenum Press, New York, pp. 1–58. Chapman, K.D. (2004). Occurrence, metabolism and prospective function of N-acylethanolamines in plants. Prog. Lipid Res. 43, 302–327. Christie, W.W. (2003). Lipid Analysis, 3rd ed., The Oily Press, Bridgwater, U.K. De Avalos, S.V. et al. (2004). Ceramides. In Bioactive Lipids, Eds. A. Nicolaou and. G. Kokotos, The Oily Press, Bridgwater, U.K., pp. 135–168.
Endocannabinoids
The identification, in the 1960s, of brain-specific receptors for one of the major components of Cannabis sativa preparations led to the discovery of endogenous ligands for such receptors. These are known as the endocannabinoids. The first determination and characterisation of a cannabinoid receptor in brain was by Devane et al. (1988). The first endocannabinoid identified and the best studied is anandamide (Devane et al., 1992). The second endocannabinoid 696
Lipid Metabolism
Devane, W.A. et al. (1988) Determination and characterization of a cannabinoid receptor in rat brain. Mol. Pharmacol., 34, 605–613. Devane, W.A. et al. (1992) Isolation and structure of a brain constituent that binds to the cannabinoid receptor. Science, 258, 1946–1949. Drøbak, B.K. (2005). Inositol-containing lipids: roles in cellular signalling. In Plant Lipids: Biology, Utilisation and Manipulation, Ed. D.J. Murphy, Blackwell Publishing, Oxford, U.K., pp. 303–328. Farooqui, A.A. and Horrocks, L.A. (2004). Plasmalogens, plateletactivating factor and other ether glycerophospholipids. In Bioactive Lipids, Eds. A. Nicolaou and G. Kokotos, The Oily Press, Bridgwater, U.K., pp. 107–134. Fukushima, N. et al. (2001) Lysophospholipid receptors. Ann. Rev. Pharmacol. Toxicol. 41, 507–534. Gagnon, M. and Saragovi, H.U. (2002). Gangliosides: therapeutic agents or therapeutic targets? Expert Opin. Ther. Patents, 12, 1215–1223. Goutopoulos, A. and Makriyannis, A. (2002) From cannabis to cannabinergics: new therapeutic opportunities. Pharmacol. Ther. 95, 103–117. Gurr, M.I. et al. (2002). Lipid Biochemistry, 5th ed., Blackwell Scientific, Oxford, U.K. Hakomori, S. (1981) Glycosphingolipids in cellular interaction, differentiation and oncogenesis. Annu. Rev. Biochem. 50, 733–764. Hannun, Y.A. et al. (1996). Ceramide: a novel second messenger and lipid mediator. In Lipid Second Messengers, Eds. R.M. Bell, J.H. Exton and S.M. Prescott, Plenum Press, New York, pp. 75–123. Hannun, Y.A. and Obeid, L.M. (2002) The ceramide-centric universe of lipid-mediated cell regulation: stress encounters of the lipid kind. J. Biol. Chem. 277, 25847–25850. Hilgemann, D.W. (2003). Getting ready for the decade of the lipids. Annu. Rev. Physiol. 65, 697–700. Honda, Z. et al. (2002) Platelet-activating factor receptor. J. Biochem. 131, 773–779. Ishii, S. and Shimizu, T. (2000). Platelet-activating factor (PAF) receptor and genetically engineered PAF receptor mutant mice. Prog. Lipid Res. 39, 41–82. Katso, R. et al. (2001) Cellular function of phosphoinositide 3-kinase: implication for development, immunity, homeostasis and cancer. Ann. Rev. Cell Dev. Biol. 17, 615–675. Karasawa, K. et al. (2003). Plasma platelet activating factoracetylhydrolase (PAF-AH). Prog. Lipid Res. 42, 93–114. Kluk, M.J. and Hla, T. (2002) Signalling of sphinosine 1-phosphate via the SIP/EDG-family of G-protein coupled receptors. Biochim. Biophys. Acta. 1582, 72–80. Kobrinsky, E. et al. (2000). Reception-mediated hydrolysis of plasma membrane messenger PIP2 leads to K+ current desensitisation. Nat. Cell Biol. 2, 507–514. Kokotos, G. (2004). Endocannabinoids. In Bioactive Lipids, Eds. A. Nicolaou and G. Kokotos, The Oily Press, Bridgwater, U.K., pp. 245–264. Kolter, T. (2004). Glycosphingolipids. In Bioactive Lipids (Eds. A. Nicolaou and G. Kokotos), The Oily Press, Bridgwater, U.K., pp. 169–196. Laychock, S. and Rubin, R.P. Eds. (1999). Lipid Second Messengers. CRC Press, Boca Raton, FL.
Maijer, H.J.G. and Munnik, T. (2003). Phospholipid-based signalling in plants. Annu. Rev. Plant Biol. 54, 265–306. Martin, T.F.J. (1998). Phosphoinositide lipids as signalling molecules: common themes for signal transduction, cytoskeletal regulation and membrane trafficking. Annu. Rev. Cell Dev. Biol. 14, 231–264. Merrill, A.H. et al. (1996). Bioactive properties of sphingosine and structurally related compounds. In Lipid Second Messengers, Eds. R.M. Bell, J.H. Exton and S.M. Presscott, Plenum Press, New York, pp. 205–237. Nagan, N. and Zoeller, R.A. (2001) Plasmalogens: biosynthesis and functions. Prog. Lipid Res. 40, 199–229. Nicolaou, A. and Kokotos, G. Eds. (2004). Bioactive Lipids, The Oily Press, Bridgwater, U.K. Nishizuka, Y. (1995) Protein kinase C and lipid signalling for sustained cellular responses. FASEB J. 9, 484–496. Payrastre, B. et al. (2001) Phosphoinositides: key players in cell signalling, in time and space. Cell Signal. 13, 377–387. Payrastre, B. (2004). Phosphoinositides. In Bioactive Lipids, Eds. A. Nicolaou and G. Kokotos, The Oily Press, Bridgwater, U.K., pp. 63–84. Pettus, B.J. et al. (2002) Ceramide in apoptosis: an overview and current perspectives. Biochim. Biophys. Acta. 1585, 114–125. Pyne, S. and Pyne, N.J. (2002) Sphingosine 1-phosphate signalling and termination at lipid phosphate receptors. Biochim. Biophys. Acta. 1582, 121–131. Pyne, S. (2004). Lysolipids: sphingosine 1-phosphate and lysophosphatidic acid. In Bioactive Lipids, Eds. A. Nicolaou and G. Kokotos, The Oily Press, Bridgwater, U.K., pp. 85–106. Razdan, R.K. and Mahadevan, A. (2002). Recent advances in the synthesis of the endocannabinoid-related ligands. Chem. Phys. Lipids. 121, 21–33. Rubin, R.P. and Laychock, S. Eds. (1999). Lipid Second Messengers. CRC Press, Boca Raton, FL. Smith, W.L. and Merrill, A.H. (2002) Sphingolipid metabolism and signalling, mini review series. J. Biol. Chem. 277, 25841–25842. Snyder, F. (1995) Platelet-activating factor: the biosynthetic and catabolic enzymes. Biochem. J. 305, 689–705. Takai, Y. et al. (1979) Unsaturated diacylglycerol as a possible messenger for the activation of calcium-activated, phospholipid-dependent protein kinase system. Biochem. Biophys. Res. Commun. 91, 1218–1224. Testerink, C. and Munnik, T. (2005). Phosphatidic acid: a multifunctional stress signalling lipid in plants. Trends Plant Sci. 10, 368–375. Tigyi, G. and Parrill, A.B. (2003). Molecular mechanisms of lysophosphatidic acid action. Prog. Lipid Res. 42, 93–114. Toker, A. (2002) Phosphoinositides and signal transduction. Cell Mol. Life Sci. 59. 761–779. Tokumura, A. (1995). A family of phospholipid autacoids: occurrence, metabolism and bioactions. Prog. Lipid Res. 34, 151–184. Vance, D.E. and Vance, J.E. Eds. (2002). Biochemistry of Lipids, Lipoproteins and Membranes, 4th ed., Elsevier, Amsterdam. Vanhaesebroceck, B. et al. (2001). Synthesis and function of 3-phosphorylated inositol lipids. Annu. Rev. Biochem. 70, 535–602. Vivanco, I. and Sawyers, C.L. (2002) The phosphatidylinositol 3-kinase-Akt pathway in human cancer. Nature Rev. 2, 489–501.
697
10.7 Sterol esters
protein (Gu et al., 2000). Selective uptake from HDL needs SR-B1 binding. The receptor also promotes cholesterol efflux from the plasma membrane by an unknown mechanism. SR-B1 has been localised to cholesterol-rich microdomains called calveoli (Fielding and Fielding, 2002). LCAT may also participate in the movement of cholesterol out of cells by esterifying excess cholesterol in the intravascular circulation (cf. Marcel, 1982). Schneider (2002) has written a useful review on lipoprotein receptors and their importance in plasma lipid metabolism. The purification and properties of LCAT, together with a discussion of its mechanism of reaction, are given by Marcel (1982). A number of disease states involve LCAT activity. Two LCAT deficiencies have been found. In one, no cholesteryl esters are formed in plasma and cholesterol accumulates as droplets in peripheral tissues. In a second disease, LCAT can transesterify cholesterol from VLDL and LDL, but not from exogenous HDL (Fielding and Fielding, 2002). A discussion of cholesteryl ester metabolism in relation to other liver diseases and dyslipoproteinaemia has been reported (Marcel, 1982). Similarly, the metabolism of cholesteryl esters in relation to arteries and arterial disease has been fully discussed (Kritchevsky and Kothari, 1978). Mammalian steroid sulfates have been reviewed by Farooqui (1981). Sterol esters and acylated sterol glycosides have been detected in a number of plant tissues and, in some cases, can be quite significant components (Mudd, 1980). For sterol esters any one of palmitate, oleate, linoleate, or α-linoleate could be the principal fatty acyl component depending on the tissue. In acylated sterol glycosides, palmitate or linoleate are the most abundant fatty acids. Synthesis of sterol esters by preparations from spinach leaves have been studied by Mudd’s group (Mudd, 1980). Enzyme activity may have been associated with mitochondria and diacylglycerol was found to be the best acyl donor, although other lipids including phosphatidylcholine could also serve. The acyl donor for sterol ester synthesis in Phycomyces blakesleanus was also found to be phosphatidylcholine, as in animal tissues (Bartlett et al., 1974). For a summary of plant sterol ester biochemistry, refer to Goad et al. (1987). The acylation of sterol glucosides in plant tissues has been studied by several workers. The research has been reviewed by Eichenberger (1977) and by Mudd (1980). The acylating enzymes are usually particulate and have been partly purified following solubilization. The best purification has been with an enzyme from Gossipium spp. (Forsee et al., 1976). Soluble preparations have been studied from carrot roots (cf. Eichenberger, 1977) and bean leaves (Heinz et al., 1975). The acyl donor for acylated sterol glucoside synthesis seems to vary with the preparation being studied. The particulate enzymes tend to use various phosphoglycerides, whereas the soluble enzymes utilise diacyldigalactosylglycerol much better. The bean leaf
Wang, X. et al. (2006). Signalling functions of phosphatidic acid. Prog. Lipid Res. 45, 250–278. Yin, H.L. and Janmey, P.A. (2003). Phosphoinositide regulation of the actin cytoskeleton. Annu. Rev. Physiol. 65, 761–789.
10.7
Sterol esters
The esterification of cholesterol in animals has attracted considerable research because of the possible involvement of cholesterol and its ester in various disease states (see Glomset and Norum, 1973; Section 11.1, Section 11.3 and Section 11.6). Cholesteryl esters are formed by the action of lecithin cholesterol acyltransferase (LCAT, EC 2.3.1.43), which is particularly active in plasma; see Sabine (1977) for a review of cholesterol metabolism. The reaction involves transfer of a fatty acid from position-2 of the lecithin (phosphatidylcholine) to the 3-hydroxyl group of cholesterol with the formation of monoacylphosphatidylcholine. LCAT is a glycoprotein of mass 60 kDa (Fielding, 1990). LCAT consumes unesterified cholesterol and the cholesteryl ester is retained in the high-density lipoprotein (HDL) core while lysophosphatidylcholine is transferred to albumin. LCAT plays a critical role in the genesis of HDL. In addition, it may be able to directly reactivate lipid-poor HDL (Kendrick et al., 2001). The enzyme is a 416-amino acid serine hydrolase with rather limited sequence homology to other plasma lipases. Aspects of its structure and regulation are discussed by Fielding and Fielding (2002). ApoA1 is needed for both its acyltransferase and phospholipase activities. When LCAT interacts with low-density lipoprotein (LDL), it can catalyse phosphatidylcholine acyl exchange. LCAT is reviewed by Jonas (2000). Cholesterol ester transfer protein (CETP) catalyses the movement of cholesteryl esters, triacylglycerols, and nonpolar lipids (such as retinyl esters) between plasma lipoproteins. CETP expression in hepatocytes is PPARdependent (Luo et al., 2001). Physiologically, the main effect of CETP may be to promote the transfer of LCATderived cholesteryl esters out of HDL (where they were formed) into VLDL and LDL, in exchange for triacylglycerol (see Fielding and Fielding, 2002); CETP is a glycoprotein of mass 53 kDa (Drayna et al., 1987). Its physiological function and mechanism of activity are discussed by Fielding and Fielding (1991). Activity of CETP in plasma is regulated by an inhibitor protein, which acts by displacing it from its lipoprotein binding sites (Morton and Zilverschmidt, 1981). It is structurally related to phospholipid transfer protein (PLTP), which transfers phospholipids between serum lipoprotein classes. Cholesteryl esters, in contrast to free cholesterol, are taken up by cells mostly via specific receptor pathways (Brown et al., 1981), are hydrolysed by lysosomal enzymes and eventually re-esterified and stored within cells. Scavenger receptor B1 (SR-B1) is the important trans-membrane 698
Lipid Metabolism
enzyme would also use diacylgalactosylglycerol efficiently. The overall pathway seems to involve glucosylation of the sterol before acylation (see Mudd, 1980). Work on sterol ester metabolism in insects has been reviewed by Thompson et al. (1973).
Marcel, Y.L. (1982). Lecithin: cholesterol acyltransferase and intravascular cholesterol transport. Adv. Lipid Res. 19, 85–136. Morton, R.E. and Zilverschmidt, D.B. (1981). A plasma inhibitor of triglyceride and cholesteryl ester transfer activities. J. Biol. Chem. 256, 11992–11995. Mudd, J.B. (1980). Sterol interconversions. In Biochemistry of Plants, vol. 4, Eds. P.K. Stumpf and E.E. Conn, Academic Press, New York, pp. 509–534. Sabine, J. R. (1977). Cholesterol, Marcel Dekker, New York. Schneider, W.J. (2002) Lipoprotein receptors. In Biochemistry of Lipids, Lipoproteins and Membranes, Eds. D.E. Vance and J.E. Vance, 4th ed., Elsevier, Amsterdam, pp. 553–572. Thompson, M.J. et al. (1973). Metabolism of steroids in insects. Adv. Lipid Res. 11, 219–265.
References Bartlett, K. et al. (1974). Biosynthesis of sterol esters in Phycomyces blakesleanus. Phytochemistry, 13, 1107–1113. Brown, M.S. et al. (1981). Regulation of plasma cholesterol by lipoprotein receptors. Science. 212, 628–635. Drayna, D. et al. (1987). Cloning and sequencing of human cholesteryl ester transfer protein cDNA. Nature. 327, 632–634. Eichenberger, W. (1977). Steryl glycosides and acylated steryl glycosides. In Lipids and Lipid Polymers from Higher Plants, Eds. M. Tevini and H.K. Lichtenthaler, SpringerVerlag, Berlin, pp. 167–179. Farooqui, A.A. (1981). Metabolism of sulfolipids in mammalian tissues. Adv. Lipid Res. 18, 159–202. Fielding, C.J. (1990). Lecithin: cholesterol acyltransferase. In Advances in Cholesterol Research, Eds. M. Esfahani and J.B. Swaney, Telford, Caldwell, NJ, pp. 271–314. Fielding, P.E. and Fielding, C.J. (1991). Dynamics of lipoprotein transport in the circulatory system. In Biochemistry of Lipids, Eds. D.E. Vance and J. Vance, Elsevier, Amsterdam, pp. 427–459. Fielding, P.E. and Fielding, C.E. (2002). Dynamics of lipoprotein transport in the human circulatory system. In Biochemistry of Lipids, Lipoproteins and Membranes, Eds. D.E. Vance and J.E. Vance, 4th ed., Elsevier, Amsterdam, pp. 527–552. Forsee, W.T. et al. (1976). Acylation of steryl glucosides by phospholipids. Solubilization and properties of the acyl transferase. Arch. Biochem. Biophys., 172, 410–418. Goad, L.J. et al. (1987). The steryl esters of higher plants. In The Metabolism, Structure and Function of Plant Lipids, Eds. P.K. Stumpf, J.B. Mudd and W.D. Nes, Plenum, New York, pp. 95–102. Glomset, J.A. and Norum, K.R. (1973). The metabolic role of lecithin: cholesterol acyltransferase: perspectives from pathology. Adv. Lipid Res. 11, 1–65. Gu, X. et al. (2000) Scavenger receptor, class B type 1-mediated 3H-cholesterol efflux to high and low density lipoproteins is dependent on lipoprotein binding to the receptor. J. Biol. Chem. 275, 29993–30001. Heinz, E. et al. (1975). Enzymatic acylation of steryl glycoside. Z. Pflanzenphysiol., 75, 78–87. Jonas, A. (2000) Lecithin cholesterol acyltransferase. Biochim. Biophys. Acta. 1529, 245–256. Kendrick, J.S. et al. (2001) Superior role of apolipoprotein B-48 over apolipoprotein B100 in chylomicron assembly and fat absorption: an investigation of apobec-1 knockout and wild type mice. Biochem. J. 356, 821–827. Kritchevsky, D. and Kothari, H.V. (1978). Arterial enzymes of cholesteryl ester metabolism. Adv. Lipid Res. 16, 221–266. Luo, Y. et al. (2001) The orphan nuclear receptor LRH-1 potentiates the sterol-mediated induction of the human CETP gene by liver X receptor. J. Biol. Chem. 276, 24767–24773.
10.8
Control mechanisms
It is clearly important for organisms to be able to control the activity of individual enzymes and, hence, the rate of flux down a pathway. This may be as an adaptation or response to an environmental stress, during development or because of disturbances in the normal status quo. One important point to note is that there is (contrary to what is said in many textbooks) no such thing as a rate-controlling enzyme in a pathway. All enzymes can contribute to control, but their contribution can vary with circumstances. An excellent review of the control of metabolism is that by Fell (1997). General comments on the regulation of lipid metabolism will be found in Gurr et al. (2002). Individual aspects of control mechanisms are detailed in the earlier sections of this chapter, but a few additional remarks will be made here. For a recent summary of the regulation of different aspects of lipid metabolism, see Vance and Vance (2002); for plant metabolism, refer to Browse and Somerville (1991), Quinn and Harwood (1990), and Murphy (2005); and for microbial aspects, see Ratledge and Wilkinson (1988, 1989). The overall control of lipogenesis has been reviewed by Saggerson (1980) and the mechanisms by which carbohydrates regulate expression of lipogenic genes, by Girard et al. (2002). So far as saturated fatty acid synthesis is concerned, the activity of both acetyl-CoA carboxylase and fatty acid synthase can be altered in various ways (Gurr et al., 2002). Short-term or acute control involves metabolic or allosteric regulation and the covalent modification of enzymes. Long-term control involves alterations in the amounts of enzyme protein (Wakil et al., 1983). Because acetyl-CoA carboxylase catalyses the first committed step in lipid synthesis and because its substrate lies at a crossroads between carbohydrate and lipid metabolism, its regulation is clearly important. Mammalian acetyl-CoA carboxylase is regulated both acutely (by phosphorylation/dephosphorylation, by acyl-CoAs, and by tricarboxylic acids) and chronically due to changes in enzyme amounts (see Gurr et al., 2002). Citrate not only 699
10.8 Control mechanisms
so that alterations in saturated/unsaturated fatty acid synthesis have to be controlled via fatty acid synthase. At lower temperatures the amount of cis-vaccenic acid is rapidly increased due to increased activity of β-ketoacyl-ACP synthase II. Overexpression of β-ketoacyl-ACP I can also increase the amount of vaccinate, but in a temperatureindependent manner. At low temperatures, vaccenate is also transferred to the sn-1 position of glycerolipids (where it competes with palmitate), whereas it is usually concentrated at the sn-2 position. However, the mechanism for the apparent change in acyltransferase selectivity is unknown (Heath et al., 2002). Growth temperature has also been reported to alter the fatty acid products of other fatty acid synthases. For example, the Saccharomyces cerevisiae synthase produces more palmitate rather than stearate at lower growth temperatures (Okuyama et al., 1979). There has been a considerable advance in our understanding of the control of aerobic fatty acid desaturases in recent years. In the simple protozoa, Tetrahymena and Acanthamoeba, low temperatures induce an increase in desaturase activity in order that the organisms can maintain membrane fluidity (Gurr et al., 2002). In the case of Acanthamoeba, the desaturase concerned with temperature adaptation is a ∆12 (ω6) oleate desaturase which produces linoleate (Avery et al., 1995). It is also induced independently by oxygen (Rutter et al., 2002). Another class of organism where there has been considerable study of temperature adaptation is the cyanobacteria. Induction of a number of individual desaturases and their mechanism of control have been well reviewed (Murata and Wada, 1995; Mikami and Murata, 2003). General aspects of temperature adaptation in different organisms will be found in Cossins (1994). For animals, considerable attention has been paid to the effects of diet on desaturase induction. In particular, there has been much study of stearoyl-∆9-desaturase, which can show extreme responses (100-fold changes in activity) to dietary manipulation. A thorough review of the earlier work in this area was made by Ntambi (1995), who discussed tissue-specific expression and the ability of carbohydrate or unsaturated fatty acids to regulate the stearoyl-CoA desaturase genes. The molecular mechanisms by which dietary polyunsaturated fatty acids could regulate genes (including those of glucose or fatty acid metabolism) is discussed by Jump et al. (1996) and developed further by Clarke (2000). Recently Ntambi and Miyazaki (2004) have revisited the topic of the mammalian stearoyl-CoA desaturases including aspects of the specific functions of isoforms and contribution of the enzyme activity to the regulation of metabolism. In addition to fatty acid synthesis, the assembly of complex acyl lipids is also under careful metabolic control. For general remarks, see Gurr et al. (2002) and Vance and Vance (2002).
causes polymerisation of acetyl-CoA carboxylase, but also can overcome inhibition caused by the enzyme product malonyl-CoA or the overall products of fatty acid synthesis, acyl CoAs (Allred and Reilly, 1997). See also Kim (1997) for a general discussion. The overall regulation of fatty acid synthesis in plants has been reviewed (Ohlrogge and Jaworski, 1997). The role of acetyl-CoA carboxylase in leaves was specifically addressed by flux control experiments and it was shown that this enzyme could exert up to 60% of the total control of flux towards lipid synthesis in the light (Page et al., 1994). The mechanism of regulation may involve changes in the enzyme’s redox state (Harwood, 1996; Rawsthorne, 2002). Cronan and Waldrop (2002) have given a very good recent survey of multisubunit acetyl-CoA carboxylases with particular emphasis on that from E. coli. They discuss the physiology, catalytic mechanism and function of the enzyme. Heath et al. (2003) also describe the regulation of E. coli acetyl-CoA carboxylase. Mammalian fatty acid synthase is also subject to adaptive changes in enzyme content. Any short-term metabolic control is ill defined (Wakil et al., 1983) and generally thought to be unimportant (Semenkovich, 1997). However, diet, triiodothyronine, hydrocortisone and insulin effects have been noted on the amount of synthase protein. Hydrocortisone and triiodothyronine have no effect alone, but potentiate the insulin induction of synthase (Wakil et al., 1983). The increases in synthase activity on refeeding or insulin administration are due to an increase in transcription of mRNA, which is elevated 70-fold (Morris et al., 1982). The mechanisms underlying fatty acid synthase regulation are reviewed by Semenkovich (1997). Rangan and Smith (2002) also discuss the regulation of fatty acid synthesis. A thorough discussion of the regulation of fatty acid synthase in plants has been made by Ohlrogge and Jaworski (1997) and recently updated (Harwood, 2005). Rawsthorne (2002) also covered some general aspects of fatty acid synthesis in relation to seed oil production. Heath et al. (2002) have described the overall regulation of fatty acid formation in E. coli from a quantitative and qualitative viewpoint. This article mentions the acetyl-CoA carboxylase accBC operon as well as the fab (fatty acid synthase) cluster. The role of the FadR protein in transcriptional control through the regulation of acyl-CoA concentrations is important in altering the balance between fatty acid synthesis and oxidation. It is well known that changes in growth temperature lead frequently to a modification in the pattern of fatty acids made and in those accumulated in the membranes of poikilotherms. Typical changes include an increase in unsaturated or of shorter chain-length fatty acids at lower growth temperatures. The adaptation has been studied in a large number of organisms. In anaerobic bacteria, such as E. coli, it is not possible for desaturases to be induced, 700
Lipid Metabolism
The regulation of the enzymes of triacylglycerol synthesis has been well reviewed by Coleman and Lee (2004) with reference to earlier work and to the use of sterol regulatory element-binding protein (SREBP), the liver X receptors and PPARs (peroxisome proliferator-activated receptors) in the regulation. SREBP is, of course, also involved in the relative rates of fatty acid and cholesterol biosynthesis (Gibbons, 2003). Interactions between phospholipids and sterol metabolism in mammalian cells are reviewed by Ridgway et al. (1999), while a comprehensive account of the regulation of phosphoglyceride synthesis and degradation in different organisms will be found in Hawthorne and Ansell (1982); see also Kent (1995) and Vance (1998); and see also various chapters in Vance and Vance (2002). Flux control analysis has been applied to the study of lipid accumulation in oil crops. Experiments with olive and palm have shown that fatty acid synthesis exerts more control than lipid assembly (Ramli et al., 2002), but that, within the Kennedy pathway, diacylglycerol acyltransferase may be important in some crops (Ramli et al., 2005). The regulation of storage oil accumulation has been discussed by Voelker and Kinney (2001), Rawsthorne (2002) and, recently, by Weselake (2005). The regulation of plant lipid metabolism was reviewed by Harwood (1989) and, with respect to environmental stress, by Harwood (1994, 1998). For a review of phospholipid metabolism in yeast and its interrelationship with other metabolic processes, see Carman and Henry (1999), and for metabolic regulation of phospholipids in E. coli consult Shibuya (1992) and Cronan (2003).
Cronan, J.E. (2003) Bacterial membrane lipids: where do we stand? Annu. Rev. Microbiol. 57, 203–224. Cronan, J.E. and Waldrop, G.L. (2002) Multi-subunit acetylCoA carboxylases. Prog. Lipid Res. 41, 407–435. Fell, D.A. (1997) Understanding the Control of Metabolism. Portland Press, London. Gibbons, G.F. (2003) Regulation of fatty acid and cholesterol synthesis: co-operation or competition? Prog. Lipid Res. 42, 479–497. Girard, J. et al. (1997) Mechanisms by which carbohydrates regulate expression of genes for glycolytic and lipogenic enzymes. Annu. Rev. Nutr. 17, 325–352. Gurr, M.I. et al. (2002) Lipid Biochemistry, 5th ed., Blackwell Scientific, Oxford, U.K. Hardie, D.G. (1992) Regulation of fatty acid and cholesterol metabolism by the AMP-activated protein kinase. Biochim. Biophys. Acta. 1123, 231–238. Harwood, J.L. (1989) Lipid metabolism in plants. Crit. Rev. Plant Science. 8, 1–44. Harwood, J.L. (1994) Environmental factors affecting lipid metabolism. Prog. Lipid Res., 33, 193–202. Harwood, J.L. (1996) Recent advances in the biosynthesis of plant fatty acids. Biochim. Biophys. Acta. 1210, 369–372. Harwood, J.L. (1998) Involvement of chloroplast lipids in the reaction of plants submitted to stress. In Lipids in Photosynthesis: Structure, Function and Genetics, Eds. P-A Siegenthaler and N. Murata, Kluwer, Dordrecht, pp. 287–302. Harwood, J.L. (2005) Fatty acid biosynthesis. In Plant Lipids: Biology, Utilisation and Manipulation, Ed. D.J. Murphy, Blackwell Publishing, Oxford, U.K., pp. 27–66. Hawthorne, J.N. (1982). Inositol phospholipids. In Phospholipids, Eds. J.N. Hawthorne and G.B. Ansell, Elsevier Biomedical Press, Amsterdam, pp. 263–278. Hawthorne, J.N. and Ansell, G.B. Eds. (1982). Phospholipids, Elsevier, Amsterdam. Heath, R.J. et al. (2002) Fatty acid and phospholipid metabolism in prokaryotes. In Biochemistry of Lipids, Lipoproteins and Membranes, Eds. D.E. Vance and J.E. Vance, 4th ed., Elsevier, Amsterdam, pp. 55–92. Jump, D.B. et al. (1996) Dietary polyunsaturated fatty acid regulation of gene transcription. Prog. Lipid Res. 35, 227–241. Kent, C. (1995) Eukaryotic phospholipid biosynthesis. Annu. Review Biochem. 64, 315343. Kim, K-H. (1997) Regulation of mammalian acetyl-CoA carboxylase. Annu. Rev. Nutr. 17, 77–99. Lane, M.D. et al. (1979). Hormonal regulation of acetyl-CoA carboxylase activity in the liver cell. CRC Crit. Rev. Biochem. 2, 121–141. Mikami, K. and Murata, N. (2003) Membrane fluidity and the perception of environmental signals in cyanobacteria and plants. Prog. Lipid Res. 42, 527–543. Morris, S.M. et al. (1982). Molecular cloning of gene sequences for avian fatty acid synthase and evidence for nutritional regulation of fatty acid synthase mRNA concentration. J. Biol. Chem. 257, 3225–3229. Murata, N. and Wada, H. (1995) Acyl-lipid desaturases and their importance in the tolerance and acclimatization to cold of cyanobacteria. Biochem. J. 308, 1–8. Murphy, D.J. Ed. (2005) Plant Lipids: Biology, Utilisation and Manipulation. Blackwell Publishing, Oxford, U.K.
References Allred, J.B. and Reilly, K.E. (1997). Short-term regulation of acetyl-CoA carboxylase in tissues of higher animals. Prog. Lipid Res. 35, 371–385. Avery, S.V. et al. (1995). Temperature dependent changes in plasma membrane lipid order and phagocytotic activity of the amoeba Acanthamoeba castellanii are closely related. Biochem. J. 312, 811–816. Browse, J. and Somerville, C.R. (1991). Glycerolipid synthesisbiochemistry and regulation. Annu. Rev. Plant Physiol. 42, 467–506. Carman, G.M. and Henry, S.A. (1999). Phospholipid synthesis in the yeast Saccharomyces cerevisiae and interrelationships with other metabolic processes. Prog. Lipid Res. 38, 361–399. Clarke, S.D. (2000). Polyunsaturated fatty acid regulation of gene transcription: a mechanism to improve energy balance and insulin resistance. Brit. J. Nutr. 83 (Supplement 1) S59–S66. Coleman, R.A. and Lee, D.P. (2004) Enzymes of triacylglycerol synthesis and their regulation. Prog. Lipid Res. 43, 134–176. Cossins, A.J. Ed. (1994). The Temperature Adaptation of Biological Membranes, Portland Press, London.
701
10.8 Control mechanisms
Ntambi, J.M. (1995) Regulation of stearoyl-CoA desaturase. Prog. Lipid Res. 34, 139–150. Ntambi, J.M. and Miyazaki, M. (2004) Regulation of stearoylCoA desaturases and their role in metabolism. Prog. Lipid Res. 43, 91–104. Numa, S. (1991). Two long-chain acyl-coenzyme A synthetases: their different roles in fatty acid metabolism and its regulation. Trends Biochem. Sci. 6, 113–115. Ohlrogge, J.B. and Jaworski, J.G. (1997) Regulation of fatty acid synthesis. Annu. Rev. Plant Physiol. Plant Mol. Biol. 48, 109–136. Okuyama, H. et al. (1979). Regulation by temperature of the chain length of fatty acids in yeast. J. Biol. Chem. 254, 12281–12284. Quinn, P.J. and Harwood, J.L. Eds. (1990) Plant Lipid Biochemistry, Structure and Utilization, Portland, London. Ramli, U.S. et al. (2002) Control analysis of lipid biosynthesis in tissue cultures from oil crops shows that flux control is shared between fatty acid synthesis and lipid assembly. Biochem. J. 364, 393–401. Ramli, U.S. et al. (2005) Metabolic control analysis reveals an important role for diacylglycerol acyltransferase in olive but not in oil palm lipid accumulation. FEBS J. 272, 5764–5770. Rangan, V.S. and Smith, S. (2002) Fatty acid synthesis in eukaryotes. In Biochemistry of Lipids, Lipoproteins and Membranes, Eds. D.E. Vance and J.E. Vance, 4th ed., Elsevier, Amsterdam, pp. 151–179. Ratledge, C. and Wilkinson, S.G. Eds. (1988) Microbial Lipids, vol. 1, Academic Press, London. Ratledge, C. and Wilkinson, S.G. Eds. (1989) Microbial Lipids, vol. 2, Academic Press, London. Rawsthorne, S. (2002) Carbon flux and fatty acid synthesis in plants. Prog. Lipid Res. 41, 182–196.
Ridgway, N.D. et al. (1999) Integration of phospholipid and sterol metabolism in mammalian cells. Prog. Lipid Res. 38, 337–360. Rutter, A.J. et al. (2002) Oxygen induction of a novel fatty acid n-6 desaturase in the soil protozoan, Acanthamoeba castellanii. Biochem. J. 368, 57–67. Saggerson, E.D. (1980). Regulation of lipid metabolism in adipose and liver cells. Biochemistry of Cellular Regulation, vol. 2, CRC Press, Boca Raton, FL, pp. 207–256. Semenkovich, C.F. (1997) Regulation of fatty acid synthase. Prog. Lipid Res. 36, 43–53. Shibuya, I. (1992) Metabolic regulation and biological functions of phospholipids in E. coli. Prog. Lipid Res. 31, 245–300. Stubbs, C.D. and Smith, A.D. (1984). The modification of mammalian membrane polyunsaturated fatty acid composition in relation to membrane fluidity and function. Biochim. Biophys. Acta. 779, 89–137. Vance, D.E. and Vance, J.E. Eds. (2002) Biochemistry of Lipids, Lipoproteins and Membranes. 4th ed., Elsevier, Amsterdam. Vance, J.E. (1998) Eukaryotic lipid biosynthetic enzymes: the same but not the same. Trends Biochem. Sci. 23, 423–428. Voelker, T. and Kinney, A.J. (2001) Variations in the biosynthesis of seed storage lipids. Annu. Rev. Plant Physiol. Plant Mol. Biol. 52, 335–361. Volpe, J.J. and Vagelos, P.R. (1976). Mechanisms and regulation of biosynthesis of saturated fatty acids. Physiol. Rev. 56, 339–417. Wakil, S.J. et al. (1983). Fatty acid synthesis and its regulation. Annu. Rev. Biochem. 52, 537–539. Yeh, L.A. et al. (1981). Coenzyme A activation of acetyl-CoA carboxylase. J. Biol. Chem. 256, 2289–2296.
702
11 MEDICAL AND AGRICULTURAL ASPECTS OF LIPIDS
J.L. Harwood, M. Evans, D.P. Ramji, D.J. Murphy and P.F. Dodds
11.1
Human dietary requirements
11.1.1
Introduction
for intakes of >3000 kcal/day, it should represent at least 30% of the calories. This gives minimum intakes of 56 to 140 g/day. Of course, there are circumstances when dietary fat should be limited (see below and Section 11.2 and Section 11.3). In a survey in the U.S., it was found that intake of fat increased with higher incomes. This correlation may reflect a greater consumption of food or, alternatively, the selection of more expensive foods, such as meat or dairy products instead of cereals (Rizek et al., 1974). Pearson and Dutson (1990) give information on meat lipids, including several chapters dealing with dietary lipids and health, while Gregory et al. (1990) provide data on fat intakes and blood lipid levels for U.K. subjects. Major food types that contribute to fat consumption in the U.K. are indicated in Table 11.1. For the U.S., the intake from fats given by Rizek et al. (1974) was 35% for salad and cooking oil, 23% for shortening, 15% for margarine, 8% for butter, and the remaining 19% from animal sources (meat, lard, etc.). It will be clearly seen from Table 11.1 that dairy products, meats, spreads, and cooking oils are the major contributors. Although not reflected in the table, there has been a significant shift in the sources of fats consumed in recent years. Thus, vegetable fats (margarine, cooking oils) are increasingly important when compared to animal fats (dairy products, lard). For example, the contribution of vegetable fats to the total in the U.S. increased from 17 to 38% in the period 1900 to 1950. This shift has raised the contribution of n-6 polyunsaturated fatty acids from 3 to 6% of the total dietary calories. In addition, the reduction in animal fats has lowered cholesterol consumption (cf. Section 11.3). Gold et al. (1992) have discussed the possible connections between cholesterol and coronary heart disease and Gurr (1992a) has
There are two aspects to lipid requirements in the human diet — qualitative and quantitative. First, certain lipids are needed for good health — essential fatty acids and fat-soluble vitamins are good examples. Secondly, it is usually considered that, in the normal diet, some 25 to 30% of the total calories are conveniently supplied as fat (Jones, 1974). Such lipids (in reasonable amounts) also usually make food more palatable. Some comprehensive sources of information on the role of fats in nutrition are Gurr (1992b, 1999), Vergroesen and Crawford (1989), and Akoh and Lai (2005). The human diet has always contained fat, but the amounts and types vary. Typical intakes in Europe and North America are between 80 and 150 g/day, which represents 30 to 40% of dietary calories. The nutrient fat per capita has been maintained, with a slight overall rise since 1900. However, the same foods are not always responsible for the fat consumed (Rizek et al., 1974). Salad and cooking oils have always been major contributors, followed by dairy products and shortening in the period 1910 to 1930, but by margarine, shortening, and meat in the period 1930 to 1960. Since that time, the increase in dietary fat has been due almost entirely to a rise in meat consumption. Moreover, current consumption of fat in different parts of the world varies markedly. In Asian countries, there is minimal dietary fat, whereas Inuits (North American Eskimos) consume 300 g/day, an amount that would nauseate a European. Because a diet lacking in fat tends to be bulky, some rough rules can apply. If total intake is > phospholipid = glycolipid > sterols. In processed food, there is often a small amount of monoacylglycerol and antioxidants, such as α-tocopherol, may be added. Fatsoluble vitamins may be present naturally or added. The process of digestion is beyond the scope of this review, but a clear, simple description of it will be found in Gurr and Harwood (1991). The use of medium-chain triacylglycerols in patients with digestive problems is discussed in Babayan (1974), Vergroesen (1975) and Thomas and Holub (1994). Triacylglycerols are usually 97 to 100% digested and the products of this process are absorbed into intestinal cells where resynthesis takes place. Unsaturated triacylglycerols are hydrolysed faster than saturated ones. Cholesterol is absorbed as the free sterol and then re-esterified, as necessary, for transport in the lymph. Bile may act as a cofactor for cholesterol esterase and improves cholesterol absorption, but plant sterols inhibit this absorption. Indeed, plant sterols (sitosterol, stigmasterol) and ergosterol are themselves rather poorly absorbed (Boyd, 1975). Thus, commercial margarines containing plant sterols and/or their derivatives are increasingly important as a means of lowering blood cholesterol (see Section 11.2 and Section 11.3). Part of their action is to lower cholesterol absorption (see Moreau et al. (2002) for review).
11.1.2
Food processing
The fatty acid of composition of foods can be affected by agricultural practice (see Section 11.8), but, usually, the biggest effects are those produced by industrial processing. Catalytic hydrogenation (see Section 4.2) is carried out to improve oxidative stability and physical properties. If oxidative stability is increased, then there is less chance of oxidation creating poor flavour and colour and giving rise to toxic compounds. Physical properties are mainly changed by “hardening” so that there are better textural
704
Medical and Agricultural Aspects of Lipids
TABLE 11.2 Maldigestion
Disorders leading to poor assimilation of dietary fats Pancreatic insufficiency
Hepatic insufficiency
Malabsorption
Gastric disturbances Ileum abnormality Intestinal defects
Steatorrhea
Pancreatitis Pancreatic tumour Malnutrition (e.g., Kwashiorkor) Cystic fibrosis side-effect Pancreatic lipase mutation Liver disease Bilary obstruction Abnormal acid secretion Poor reabsorption of bile Bacterial invasion (Tropical spruce, Whipple’s disease) Sensitisation (e.g., gluten sensitivity in coeliac disease) Impared chylomicron formation (e.g., Anderson’s disease) Bacterial invasion
properties (Gurr et al., 2002). The analysis of lipid oxidation is covered in Kamal-Eldin and Pokorny (2005). Chemically, there are three main results of hydrogenation: (1) the total number of double bonds are reduced, (2) some of the cis double bonds are isomerised to trans, and (3) the double bonds may be shifted from their original positions. Of these effects, it is the increase in trans-unsaturated fatty acids that has attracted the most attention. Although trans fatty acids are found naturally and are consumed at the rate of 5 to 7 g/day in the U.K. (British Nutrition Foundation, 1987), there is considerable evidence for adverse effects on health (Gurr, 1996). This has led to the increasing use of interesterification for triacylglycerol modification (Gunstone, 1998) or to the use of different dietary lipid sources (see Section 11.8). It has also led to a requirement for labelling food packaging with the trans fatty acid content and to attempts to minimise their intake (Hunter, 2004). Other processes that can cause lipid changes in food include heating and irradiation. Heating, where there is little contact with air (e.g., deep fat fryer), gives rise to a gradual accumulation of polymeric products (see, e.g., Varela and Ruiz-Rozo, 2000). Provided this oil is not reused excessively, these polymers do not cause problems (Gurr et al., 2002). Heating in the presence of oxygen and, particularly, if there are trace metal catalysts (e.g., iron, copper) can cause lipid peroxidation and production of reactive oxygen species. A major harmful side effect is the loss of antioxidant nutrients, such as vitamin E or carotenes (Gurr, 1988). Little peroxidised lipid is thought to be absorbed intact, but there is some evidence for liver toxicity and gut damage (Gurr, 1999). Any absorbed oxidised cholesterol may have health implications (Gurr et al., 2002). Lipid oxidation may be reduced by the presence of lipid-soluble antioxidants, either natural (e.g., carotenoids, vitamin E) or synthetic (butylated hydroxytoluene, BHT or butylated hydroxyanisole, BHA). Irradiation is used to kill pathogens in some types of food and may generate lipid radicals. Vitamins E and K are particularly susceptible to radiation damage, but not, apparently, carotenoids (Gurr et al., 2002).
11.1.3
Specific dietary lipids that may be harmful
The possible deleterious effects of high quantities of medium- or long-chain saturated fatty acids or cholesterol in the diet is touched on in several sections of Chapter 11; however, there are also specific lipids that are not known to have any nutritional benefit and that may be present in the diet. Trans unsaturated fatty acids were mentioned above (Section 11.1.2) and epidemiological studies of human population and controlled dietary experiments with human subjects have been reported. Some, but not all, of these studies show a correlation of the intake of certain types of trans fatty acids and increased risk of coronary heart disease (see Recommendations of the European Atherosclerosis Society, 1992 and Willett et al., 1993). In general, though, the evidence for harmful effects of trans unsaturated fatty acids (at least, at normal dietary concentrations) is not very persuasive (Gurr, 1999; Gurr et al., 2002). Despite this, in 2006 the U.S. Food and Drug Administration (FDA) made it mandatory to declare the amount of trans fat present in foods. These concerns have led to research on solutions or alternatives to trans fatty acids in foods (see Kodali and List, 2005). Cottonseed oil is the only important oil in the human diet that contains cyclopropene fatty acids. Because such acids (as sterculic acid) fed at 5% of dietary energy to rats caused death and at the 2% level caused disturbances of reproduction, there has been concern about the effect in humans. However, their concentration is low (0.6 to 1.2%) in cottonseed oil and reduced to 0.1 to 0.5% by processing. There has been no evidence that consumption of cottonseed oil in manufactured products has any harmful nutritional effects (Gurr et al., 2002). Very long-chain monounsaturated fatty acids (such as erucic acid, 22:1n-9), when fed to rats at 5% or more of their total energy requirements, caused a buildup of triacylglycerols in heart muscle. Other pathological changes were also noticeable (Gurr et al., 2002). Despite lack of evidence for harmful effects in man, breeding programmes were intitiated to replace older varieties of oilseed rape (up to 45% erucate in its oil) with zero erucate or canola
705
11.1
Human dietary requirements
varieties (see Section 11.8.2.3). The use of such varieties in most industrial countries is now mandatory; however, high erucate varieties are still used extensively in China. Furthermore, although fish oils are consumed for their desirable enrichment with the n-3 PUFAs (polyunsaturated fatty acids), eicosapentaenoic acid, and decosahexaenoic acid, they often contain high concentrations of 20:1n-9 and 22:1n-9. Moreover, certain fat spreads that incorporate hardened fish oils may also have significant concentrations of such acids. The long-term consequences (beneficial or otherwise) of consuming marine long-chain monoenes have been poorly researched (Gurr et al., 2002). Conjugated linolenic acids (CLAs) are a group of geometric and positional isomers of linolenic acid and can be formed by biohydrogenation and oxidation processes in nature. They are significant components of dairy products where they are produced in the rumen by microbiological action. The main form of CLA is usually cis-9, trans-1118:2, with trans-10, cis-12-18:2 the next most abundant. Adverse effects of CLAs have been demonstrated in animal studies, but it is not clear whether similar actions are applicable to humans. On the other hand, a surprising number of health benefits have been attributed to CLAs. This topic is well reviewed by Wahle et al. (2004). We should not close this section without referring to the huge amount of evidence in the literature that claims (or not) to correlate dietary saturated fatty acids or cholesterol with an increase in plasma cholesterol (particularly LDLcholesterol; see Section 11.2) and, then, a rise in cardiovascular disease. This work began with the classic study of Keys et al. (1957). Part of the difficulty in interpreting the various studies is that, often, an individual paper concludes that there are rather poor (and/or statistically not significant) correlations between the various components. This has led to the use of meta-analyses with all the inherent problems of selectivity and trying to compare studies where the measurements are not strictly comparable. Gurr (1992) discussed this subject very comprehensively and for an update of his views, see Gurr (1999). He “expresses scepticism for a major role for dietary lipids in the development of ischaemic heart diesease.” Such a view is, of course, contrary to dietary advice as practiced currently by physicians in North America and most of Europe, even though Gurr’s view is based on a careful scientific analysis of data produced over the past 25 years. Nevertheless, a recent prospective meta-analysis of data from 90,056 participants in 14 randomised trials of statins (which reduce LDL cholesterol) showed a lowering (20%) of major coronary events and stroke (Cholesterol Treatment Trialists (2005)). The reader is also referred to Tholstrup et al. (1994) for information on saturated fatty acids, to Riemersma (1994) for a review of antioxidants in coronary heart disease prevention and to Wald (1994) for lipoprotein: CHD relationships.
11.1.4
Specific fat requirements
There is no doubt that PUFAs are necessary for good health, and both α-linolenic (18:3n-3) and linolenic (18:2n-6) are regarded generally as essential. However, in a thought-provoking article, Cunnane (2003) has argued that PUFAs should be regarded as conditionally indispensable or dispensable, depending on the development stage. Thus, during pregnancy, lactation, infancy, and childhood, he regards linoleate, α-linolenate, arachidonate, and docosahexaenoiate (22:6n-3) as conditionally indispensable. During adulthood, α-linolenate is described as conditionally indispensable, but eicosapentaenoic (20:5n-3) and 22:6n-3 could be described thus, depending on the geographical area and lifestyle of individuals (Cunnane, 2003). Current recommendations for PUFAs in the diet are an adequate intake of linolenic acid at 2% energy, a healthy intake of α-linolenic acid at 0.7% energy and, for cardiovascular health, a minimum intake of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) combined of 500 mg/day. It was also recognised that there may be a healthy upper limit for linolenic acid (ISSFAL, 2004). Full references and justification will be found in this publication. The role of PUFAs in human nutrition and metabolism has been reviewed (Neuringer et al., 1988; Galli and Simpoulos, 1989; Heird and Lapillonne, 2005). Polyunsaturated fatty acids may also be involved in a number of other disease states. These include acrodermatis enteropathica, biliary tract disease, kinky hair disease, hepatoma, kwashiorkor, mycobacterial infections and multiple sclerosis (Soderhjelm et al.,1971). Essential fatty acid deficiency leads to problems in practically every tissue of the body (Holman, 1977). Classic symptoms include dermatitis, growth retardation, and infertility (Table 11.3). There are also biochemical changes, such as in mitochondrial efficiency, in various tissues (Holman, 1977; Gurr et al., 2002). Essential fatty acid deficiency is easily recognised because tissue fatty acids of the n-6 group are partly replaced by n-9 unsaturated fatty acids. In particular, arachidonic acid is reduced and eicosatrienoic acid (20:3n-9, the “Mead” acid) increased (Gurr et al., 2002) (Figure 11.1). Many of the effects of essential fatty acids are due to their conversion to eicosanoids. There are three types of enzyme-catalysed conversions — cyclooxygenase, lipooxygenase and oxidations involving cytochrome P450. The reactions give rise to prostaglandins, thromboxanes, prostacyclin, lipoxins, leukotrienes and other important biologically active molecules (Gurr et al., 2002). The n-3 and n-6 PUFAs compete with each other at a number of levels (Table 11.4). First, the main dietary PUFAs, linoleate and α-linolenate, are converted to the 20C eicosanoid precursors (arachidonate, eicosapentaenoate) using the same enzymes (see Figure 11.1). Secondly, the eicosanoids produced from arachidonate are proinflammatory, whereas those from eicosapentaenoate are mildly or 706
Medical and Agricultural Aspects of Lipids
TABLE 11.3
Major effects of n-6 essential fatty acid deficiency in rats
Skin
Dermatosis, water permeability increased Sebum secretion decreased Epithelial hyperplasia
Body weight
Decreased
Circulation
Heart enlargement Capillary resistance decreased
Kidney
Enlargement, intertubular haemorrhage
Lung
Cholesterol accumulation
Endocrine glands
Adrenals: weight decreased in females, increased in males Thyroid: weight increased
Reproduction
Females: irregular oestrus; impaired lactation, reproduction Males: degeneration of the seminiferous tubules
Metabolism
Changes in tissue fatty acid composition Reduced cholesterol concentration in plasma Increased cholesterol concentration in liver, adrenals and skin Mitochondrial swelling and uncoupled oxidative phosphorylation Increased triacylglycerol output by liver
Diet or 18:1n-9 endogenous (oleic acid) synthesis
∆6D
Diet only
18:2n-6 (linoleic acid)
∆6D
Diet only
18:3n-3 (α-linolenic acid)
∆6D
E
18:2n-9
E
18:3n-6 (γ-linolenic acid)
E
18:4n-3
20:2n-9
20:3n-6
20:4n-3
∆5D
∆5D
∆5D
20:3n-9 (‘Mead acid’)
20:4n-6 (arachidonic acid)
20:5n-3 (eicosapentaenoic acid)
FIGURE 11.1 Competition between different fatty acids for production of 20 carbon PUFAs. Abbreviations: ∆5D, delta5-desaturase; E, elongase; ∆6D, delta6-desaturase.
EFA-deficiency is seldom seen unless very unusual dietary conditions prevail. However, there is special interest now in elucidating the health benefits of substantial intakes of n-3 and n-6 PUFAs such as recommended in a healthy diet (see above). In particular, n-3 PUFAs are needed for the development and proper function of brain and retina (Lauritzen et al., 2001) and there is increasing evidence that they are of benefit in reducing senile dementia (Morris, et al., 2003; Lim et al., 2005) and other cognitive problems (Harwood and Caterson, 2006). There is also a good deal of interest in the role of PUFAs in cancer (Guthrie and Carroll, 1999; Diggle, 2002) and in cardiovascular disease (NHFA, 1999;
noninflammatory. Thirdly, the n-3 and n-6 PUFAs have effects on the expression of many different proteins (Jump, 2002; Sampath and Ntambi, 2005) and their actions are often different (Harwood and Caterson, 2006). It must also be noted that it has been recently discovered that the n-3 PUFAs, eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), can be converted to new types of biologically active and mainly anti-inflammatory metabolites. Thus, EPA can yield resolvins of the E series, while DHA can produce resolvins of the D series and neuroprotectins (Serhan et al., 2004). While small amounts of essential fatty acids are needed to prevent deficiency syndromes, in practice, TABLE 11.4
How n-3 and n-6 PUFAs can compete with each other
Effect on enzymes during their metabolism Derived eicosanoids have opposing effects
The n-3 and n-6 PUFAs affect gene expression
∆5 and ∆6-desaturases and the 18C PUFA elongase show substrate competion. In general, metabolites from arachidonic acid (AA) are proinflammatory (e.g., PGE2, LTB4), whereas those from eicosapentadecanoic acid (EPA, 20:5 n-3) are non- or antiinflammatory (e.g., PGE3, LTB5) The effects of the two sets of PUFAs are often opposite. For example, COX-2 expression and activity are reduced by EPA, whereas AA has no effect or increases activity in a variety of tissues
707
11.1
Human dietary requirements
important of which are vitamin D2 (ergocalciferol) and vitamin D3 (cholecalciferol). Again, as with vitamin A, the diet is often able to supply provitamins (e.g., ergosterol), which are converted to the active compounds. All of the provitamins require ultraviolet light for their conversion. Like vitamin A, vitamin D is toxic in high doses. Even amounts only five times normal intake can be toxic and cause more calcium to be absorbed than can be excreted, resulting in excessive deposition in and damage to the kidneys (Gurr et al., 2002). Vitamin E activity is possessed by eight tocopherols and tocotrienols (Gurr et al., 2002). The most potent and abundant form is α-tocopherol. Vitamin E is a natural antioxidant, although it may have other roles, such as structural functions in membranes (Wang and Quinn, 1999). The vitamin is needed for mitochondrial electron-transport function and it prevents oxidation of various compounds, including polyunsaturated fatty acids and vitamin A. The dependence of vitamin E requirement on the amount of dietary polyunsaturated fatty acids has been fully discussed by Jager (1975). Traditionally, vitamin E has been known as the “fertility” vitamin. However, deficiency leads to serious changes in skeletal muscle, the blood system and other tissues before reproduction is impaired. The effects on different animals are described by Jager (1975). Vitamin E is discussed by Scott (1978) and Packer and Fuchs (1993). Vitamin K is the generic name given to a group of compounds, having in common a naphthoquinone ring system (menadione) with different side chains (Gurr et al., 2002). Vitamin K1 is made by plants and found in their green tissues, while K2 is synthesised by microorganisms. The best-studied role for vitamin K is in relation to blood clotting where four of the procoagulant proteins of the clotting cascade depend on vitamin K. It also functions as an enzyme cofactor and plays a role in bone metabolism. It is rare to see vitamin K deficiency in adults except where fat absorption is impaired (see Section 11.1.1 and Sickinger, 1975). Vitamin K is discussed by Suttie (1978). Unlike some other fat-soluble vitamins, there is little evidence of harmful effects from high doses of vitamin K. For a general account of vitamin requirements and overdose symptoms, refer to Wilson (1994) and references therein.
Wijendran and Hayes, 2004). Many of these important diseases have chronic inflammation as a major causative factor (Harwood and Caterson, 2006) and the relative roles of n-3 and n-6 PUFAs in this regard is fundamental (Calder, 1997; Calder et al., 1998). Some further sources of information on dietary PUFAs are British Nutrition Foundation (1992), Chow (1992), Forsyth (1998), Garrow et al. (2000), Innes (1991), Lermer and Mattes (1999) and Lands (2005).
11.1.5
Vitamins
Although somewhat outside the scope of this book, it would be wrong to describe the role of lipids in diets without mentioning fat-soluble vitamins. Vitamin A is alltrans-retinol, which is only found in animal fats. However, plant materials often contain abundant quantities of β-carotene (provitamin A). This can be easily converted to alltrans-retinol in the body. Ritinyl esters, mainly retinol palmitate, are stored principally in the liver from where the latter is released by hydrolysis and transported to target tissues bound to retinal-binding protein. Vitamin A has a number of important functions of which its role in vision is the best understood at the molecular level. An early sign of vitamin A deficiency is “night blindness.” When severe deficiency occurs, it can lead to blindness in young children. This tragic disease, xerophthalmia, is one of the four most common preventable diseases in the world (Gurr et al., 2002; see also Section 11.8.2.5 for efforts to prevent this with “golden” rice). The 9-cis-retinoic acid analog is involved importantly in differentiation. It can bind to two high-affinity receptor proteins, called RAR and RXR. Each of these may be present in one of three isoforms. Retinoic acid isomers are known to have extensive effects on gene expression and to interact (via RXR) with the vitamin D receptor or the PPAR (peroxisomal proliferator activated receptor) system. By these means, they have important effects on cellular differentiation (Gurr et al., 2002). Vitamin A is important in preventing degenerative changes in epithelial surfaces, such as keratinization of skin. It is needed for normal bone development and a deficiency in young animals can lead to secondary nervous problems due to compression of the brain and spinal cord (de Luca, 1978). Vitamin A is involved in the immune response, mainly through the T-helper cell. About 750 µg of vitamin A is needed for the average person daily but, like other fat-soluble vitamins, excessive intakes lead to accumulation, particularly in the liver. Chronic overconsumption may cause not only liver necrosis, but also permanent damage to bones, joints, muscles, and vision. Vitamin D is needed for calcium homeostatis and has various other functions for tissue development (Gurr et al., 2002). It is needed to prevent rickets and its deficiency is involved in other pathological states (de Luca, 1978). A number of different structures have activity, the two most
References Akoh, C.C. and Lai, O.M. Eds. (2005). Healthful Lipids. ACOS, Champaign, IL. Babayan, V.K. (1974). Modification of food to control fat intake. J. Am. Oil Chem. Soc. 51, 260–264. Boyd, G.S. (1975). Cholesterol absorption. In The Role of Fats in Human Nutrition, Ed. A.J. Vergroesen, Academic Press, London, pp. 353–380. British Nutrition Foundation (1987). Report of the Task Force on Trans Fatty Acids, Brit. Nutr. Foundation, London. British Nutrition Foundation (1992). Task force report on unsaturated fatty acids: nutritional and physiological significance. Chapman and Hall, London.
708
Medical and Agricultural Aspects of Lipids
Calder, P.C. (1997). n-3 Polyunsaturated fatty acids and immune cell function. Adv. Enzyme Regul. 37, 197–237. Calder, P.C. et al. (1998). Symposium of the Nutrition Society comprising twelve reviews on ‘Lipids and the Immune System.’ Proc. Nutr. Soc. 57, 487–585. Cholesterol Treatment Trialists (2005). Efficiency and safety of cholesterol-lowering treatment: prospective meta-analysis of data from participants in 14 randomised trials of statins. Lancet, 336, 1267–1278. Chow, C.K. Ed. (1992). Fatty Acids in Foods and Their Health Implications. Marcel Dekker, New York. Cunnane, S.C. (2003). Problems with essential fatty acids: time for a new paradigm? Prog. Lipid Res. 42, 544–568. De Luca, H.F. (1978a). Vitamin D. In Handbook of Lipid Research, vol. 1, The Fat-Soluble Vitamins, Ed. H.F. De Luca, Plenum, New York, pp. 69–132. De Luca, L.M. (1978b). Historical developments in vitamin A research. In Handbook of Lipid Research, vol. 2, The Fat Soluble Vitamins, Ed. H.F. De Luca, Plenum, New York, pp. 1–67. Diggle, C.P. (2002). In vitro studies on the relationship between polyunsaturated fatty acids and cancer: tumour or tissue specific effects? Prog. Lipid Res. 41, 240–253. Forsyth, J.S. (1998). Lipids and infant formulas. Nutr. Res. Rev. 11, 255–278. Galli, C. and Simpoulos, A.P. Eds. (1989). Dietary ω3 and ω6 fatty acids. In Biological Effects and Nutritional Essentiality. Plenum Press, New York. Garrow, J.S. et al. Eds. (2000). Human Nutrition and Dietetics, Churchill Livingston, Edinburgh, U.K. Gold, P. et al. (1992). Cholesterol and Coronary Heart Disease – The Great Debate, Panthenon, Carnforth and Park Ridge. Gregory, J. et al. (1990). The Dietary and Nutritional Survey of British Adults, HMSO, London. Gunstone, F.D. (1998). Movements towards tailor-made fats. Prog. Lipid Res. 37, 277–305. Gurr, M.I. (1988). Lipids: products of industrial hydrogenation, oxidation and heating. In Nutritional and Toxicological Aspects of Food Processing, Eds. R. Walker and E. Quattrucci, Taylor & Francis, London, pp. 133–155. Gurr, M.I. (1992a). Dietary lipids and coronary heart disease: old evidence, new perspective. Prog. Lipid Res. 31, 195–243. Gurr, M.I. (1992b). The Role of Fat in Food and Nutrition, Elsevier, London. Gurr, M.I. (1996). Dietary fatty acids with trans unsaturation. Nutr. Res. Rev. 9, 259–279. Gurr, M.I. (1999). Lipids in Health and Nutrition: A Reappraisal. The Oily Press, Bridgwater, U.K. Gurr, M.I. and Harwood, J.L. (1991). Lipid Biochemistry, 4th ed., Chapman & Hall, London. Gurr, M.I., Harwood, J.L. and Frayn, K.N. (2002). Lipid Biochemistry. 5th ed., Blackwell, Oxford, U.K. Guthrie, N. and Carroll, K.K. (1999). Specific versus non-specific effects of dietary fat on carcinogenesis. Prog. Lipid Res. 38, 261–271. Harwood, J.L. and Caterson, B. (2006). Dietary omega-3 polyunsaturated fatty acids and inflammation. Lipid Technol. 18, 7–10.
Heird, W.C. and Lapillonne, A. (2005). The role of essential fatty acids in development. Annu. Rev. Nutr. 25, 549–571. Holman, R.T. (1977). The deficiency of essential fatty acids. In Polyunsaturated Fatty Acids, Eds. W.H. Kunau and R.T. Holman, American Oil Chemists’ Society, Champaign, IL, pp. 163–182. Hunter, J.E. (2004). Alternatives to trans fatty acids in foods. INFORM, 15, 510–512. Hwang, D. (2000). Fatty acids and immune responses — a new perspective in searching for clues to mechanism. Annu. Rev. Nutr. 20, 431–456. Innes, S.M. (1991). Essential fatty acids in growth and development. Prog. Lipid Res. 30, 39–103. ISSFAL (2004). Recommendations for intake of polyunsaturated fatty acids in healthy adults. ISSFAL Newsl. 11, 12–25. Jager, F.C. (1975). Linoleic acid intake and vitamin E requirement. In The Role of Fats in Human Nutrition, Ed. A.J. Vergroesen, Academic Press, London, pp. 381–432. Jones, R.J. (1974). Role of dietary fat in health. J. Am. Oil Chem. Soc. 51, 251–254. Jump, D.B. (2002). Dietary polyunsaturated fatty acids and regulation of gene transcription. Curr. Opin. Lipidol. 13, 155–164. Kamal-Eldin, A. and Pokorny, J. (Eds.) (2005). Analysis of Lipid Oxidation. AOCS, Champaign, IL. Keys, A., Anderson, J.T. and Grande, F. (1957). Prediction of serum cholesterol responses in man to changes in fats in the diet. Lancet (ii), 959–966. Kodali, D.R. and List, G.R. Eds. (2005) Trans Fats Alternatives. AOCS, Champaign, IL. Lands, W.E.M. (2005). Fish, Omega-3 and Human Health, 2nd ed., AOCS, Champaign, IL. Lauritzen, L., Hansen, H.S., Jorgensen, M.H. and Michaelsen, K.F. (2001). The essentiality of long chain n-3 fatty acids in relation to development and function of the brain and retina. Prog. Lipid Res. 40, 1–94. Lermer, C.M. and Mattes, R.D. (1999). Perception of dietary fat: ingestive and metabolic implications. Prog. Lipid Res. 38, 117–128. Lim, G.P. et al. (2005). A diet enriched with the omega-3 fatty acid docosahexaenoic acid reduces amyloid burden in an aged Alzheimer mouse model. J. Neurosci. 25, 3032–3040. Morris, M.C. et al. (2003). Consumption of fish and n-3 fatty acids and risk of incident Alzheimer disease. Arch. Neurol. 60, 940–946. Neuringer, M. et al. (1988). The essentiality of n-3 fatty acids for the development and function of the retina and brain. Annu. Rev. Nutr. 8, 517–541. NHFA (National Heart Foundation of Australia) (1999). A review of the relationship between dietary fat and cardiovascular disease. Aust. J. Nut. Diet. 56, 55–522. Packer, L. and Fuchs, J. (1993). Vitamin E in Health and Disease, Marcel Dekker, New York. Pearson, A.M. and Dutson, T.R. (1990). Meat and health. Advances in Meat Research, vol. 6, Elsevier, New York. Petersdorf, R.G. et al. (Eds.) (1983). Principles of Internal Medicine, 10th ed., McGraw-Hill, New York. Recommendations of the European Atherosclerosis Society (1992) Prevention of coronary heart disease: scientific background and new clinical guidelines. Nutri. Metab. Cardiovasc. Dis. 2, 113–156.
709
11.2
Lipids and cardiovascular disease
Organisation (WHO), an estimated 17 million people die from CVD each year, with heart attacks and stroke responsible for the majority of deaths. It has been predicted that the total number of deaths from CVD may rise to 20.5 million by 2020 and 24.2 million by 2030 as developing countries acquire westernised habits. CVD is clearly a major economic burden due to expenses incurred for hospital care and medication for patients and days lost from work because of illness, death and looking after relations with the disease. According to the American Heart Association and the National Heart, Lung and Blood Institute, the economic cost of CVD in 2005 alone is estimated to be 393.5 billion dollars.
Riemersma, R.A. (1994). Epidemiology and the role of antioxidants in preventing coronary heart disease: a brief overview. Proc. Nutr. Soc. 53, 59–65. Rizek, R.L., Friend, B. and Page, L. (1974). Fat in todays food supply — level of use and sources. J. Am. Oil Chem. Soc. 51, 244–250. Sampath, H. and Ntambi, J.M. (2005). Polyunsaturated fatty acid regulation of genes of lipid metabolism. Annu. Rev. Nutr., 25, 317–340. Scott, M.L. (1978) Vitamin E2. In Handbook of Lipid Research, The Fat-Soluble Vitamins, Ed. H.F. De Luca, Plenum, New York, pp. 133–210. Serhan, C.N. et al. (2004). Resolvins, decosatrienes and neuroprotectins, novel omega-3-derived mediators and their aspirin-triggered epimers. Lipids, 39, 1125–1132. Sickinger, K. (1975). Clinical aspects and therapy of fat malassimilation with particular reference to the use of medium-chain triglycerides. In The Role of Fats in Human Nutrition, Ed. A.J. Vergroesen, Academic Press, London, pp. 116–209. Söderhjelm, L. et al. (1971). The role of polyunsaturated acids in human nutrition and metabolism. Prog. Chem. Fats Lipids, 9, 555–585. Suttie, J.W. (1978). Current advances in vitamin K research. In The Fat-Soluble Vitamins, Ed. J.W. Suttie, William Heinemann, Ltd., London, pp. 211–220. Varela, G. and Ruiz-Rosa, B. (2000). Some nutritional aspects of olive oil. In Handbook of Olive Oil, Eds. J.L. Harwood and R. Aparcio, Aspen Publishers, Gaithersburg, MD, pp. 565–582. Vergroesen, A.J. (Ed.) (1975). The Role of Fats in Human Nutrition, Academic Press, London. Vergroesen, A.J. and Crawford, M. (Eds.) (1989). The Role of Fats in Human Nutrition, Academic Press, London. Wahle, K.W.J. et al. (2004). Conjugated linolenic acids: are they beneficial or detrimental to health? Prog. Lipid Res. 43, 553–587. Wald, N.J. et al. (1994). Apolipoproteins and ischemic-heartdisease — implications for screening. Lancet, 343, 75–79. Wang, X. and Quinn, P.J. (1999). Vitamin E and its function in membranes. Prog. Lipid Res. 38, 225–248. Wijenderan, V. and Hayes, K.C. (2004). Dietary n-6 and n-3 fatty acid balance and cardiovascular health. Annu. Rev. Nutri. 24, 597–615. Willett, W.C. et al. (1993). Intake of trans fatty acids and risk of coronary heart disease among women. Lancet, 341, 581–585. Wilson, J.D. (1994). Vitamin deficiency and excess. In Principles of Internal Medicine, Eds. K.J. Isselbacher, E. Braunwald, J.D. Wilson, J.B. Martin, A.S. Fanci and D.L. Kasper, 13th ed., McGraw-Hill, New York, pp. 472–480.
11.2
Lipids and cardiovascular disease
11.2.1
Cardiovascular disease
11.2.2
Atherosclerosis is the underlying cause of cardiovascular disease
Atherosclerosis, which comes from the Greek words “athero” (meaning gruel or paste) and “sclerosis” (hardness), is the principal cause of CVD. A normal artery consists of three layers, the intima lining the lumen, the middle layer called the media and an outermost layer termed the adventitia. The intima consists of a single layer of endothelial cells that regulate vascular tone. The media consists predominantly of smooth muscle cells, whereas the adventia contains smooth muscle cells, fibroblasts, and a looser connective tissue. Atherosclerosis, which develops during the life span of an individual, causes a buildup of plaques, consisting of cholesterol, other lipids, and debris from cellular death, in the inner lining of the arteries. Although continued growth of such plaques may impede blood flow, the major problem arises when it becomes fragile and ruptures. This leads to a clot that can block blood flow or can break off and get trapped in another part of the body. Heart attack (also called myocardial infarction, coronary occlusion, or coronary thrombosis) occurs when the clot blocks a coronary artery and, thus, deprives the heart of oxygen and nutrients. On the other hand, blockage of a blood vessel to the brain leads to stroke. Atherosclerosis is initiated by damage to the endothelial cells by a number of risk factors (see Section 11.2.3 below), which then triggers a series of changes in the arterial wall (see Figure 11.2 for a summary; and Ross, 1999; Lusis, 2000; Glass and Witztum, 2001; and Lusis et al., 2004 for a detailed description of these changes). First of all, the permeability of the vascular wall is increased because of the synthesis of cell surface adhesion molecules, such as intercellular adhesion molecule-1, E-selectin, P-selectin and vascular cell-adhesion molecule-1. In addition, the damaged endothelial cells secrete chemo-attracting cytokines (chemokines), such as monocyte-chemoattractant protein-1, which in turn attracts monocytes and T-lymphocytes from circulation and stimulates their migration into the intima of the arterial wall. These monocytes then differentiate into macrophages, a process that is associated with the
Cardiovascular disease (CVD) is a major cause of morbidity and mortality in the western world, with the number of individuals with the disease in developing countries increasing all the time. The clinical manifestations of CVD include heart attacks, stroke, and gangrene of the extremities. According to the statistics from the World Health 710
Medical and Agricultural Aspects of Lipids
Normal artery
Fatty streak
Fibrous plaque
Accumulation of LDL and its oxidation to oxLDL Initial damage to ECs by oxLDL and other atherogenic factors Expression of adhesion proteins and cytokines by ECs Infiltration of monocytes and T-lymphocytes Macrophage differentiation and expression of SRs Uptake of oxLDL by macrophages to form foam cells A chronic inflammatory response due to cytokines produced by macrophages and T-lymphocytes Migration of smooth muscle cells from the media to the intima Proliferation of smooth muscle cells Formation of smooth muscle cell-derived foam cells Secretion of extracellular matrix proteins Apoptosis and necrosis of macrophage and smooth muscle cell-derived foam cells Accumulation of extracellular cholesterol Secretion of matrix metalloproteinases by macrophages leading to instability of the fibrous plaque
Plaque rupture
FIGURE 11.2 A schematic representation of the major steps during the pathogenesis of atherosclerosis (see text for more details). Abbreviations: ECs, endothelial cells; LDL, low density lipoproteins; oxLDL, oxidized low density lipoproteins; SRs, scavenger receptors.
SMC, T-lymphocytes, extracellular matrix and debris from dying foam cells. Instability of such advanced lesions may lead to plaque rupture and thrombus formation, which may ultimately result in the occlusion of the artery. An inflammatory response is a major contributor to plaque instability and the development of acute CVD has been found to be associated with elevated circulating levels of markers of inflammation, such as C-reactive protein.
expression of so-called scavenger receptors (SRs), such as SR-A and CD36. Although the normal function of these SRs is believed to be in the uptake of pathogens and apoptotic cells, they also take up modified lipoproteins (see Section 11.2.7 below). This latter property causes macrophages to transform into lipid-loaded foam cells. Such accumulation of foam cells in the vascular wall is a critical early step in the pathogenesis of atherosclerosis and is responsible for the formation of a so-called fatty streak. The macrophages also secrete cytokines, which lead to a local inflammatory response that recruits further monocytes and T-lymphocytes and, thereby, amplify foam cell formation. The importance of macrophages in atherogenesis can be gauged by the finding that diet-induced atherosclerosis in a murine model of the disease is significantly reduced in mice bred to have severely reduced monocyte levels (Smith et al., 1995; see Osterud and Bjorklid, 2003 for an in-depth review on the role of monocytes/macrophages in atherogenesis). This fatty streak then progresses to form a fibrous plaque via a number of cellular and biochemical changes. First, an intermediate fibro-fatty lesion is formed, which consists of foam cells, smooth muscle cells (SMC), T-lymphocytes and a relatively poorly developed extracellular matrix. The transition from such fibro-fatty lesions to more complex lesions is associated with the migration of SMC from the media to the intima, where they proliferate and also take up lipoproteins, thereby contributing to further foam cell formation. In addition, such SMC synthesise extracellular matrix proteins leading to the development of a fibrous cap. More advanced lesions contain a dense fibrous cap, which protrudes into the lumen of the artery and covers a core of macrophages,
11.2.3
Risk factors for CVD
More than 300 risk factors for CVD have been identified from laboratory- and clinical-based research. The major risk factors that have a high prevalence in many different populations are age, gender, hypertension, smoking, physical inactivity, obesity, diabetes, socioeconomic status, Chlamydia pneumoniae infection, hyperhomocysteinemia and high levels of circulating lipids (see Stoker and Keaney, 2004 for a more detailed discussion of the various risk factors). A number of such risk factors often co-exist in atherosclerotic patients where they may act in a synergistic manner. For example, hypercholesterolemia, obesity, hypertension and physical inactivity are often associated in a number of male patients. The risk for the development of CVD increases with age, with the average risk in males aged 65 being about sevenfold greater than those who are 35. In addition, males have a much higher risk for CVD compared to age-matched women (Barrett-Conner and Bush, 1991). It has been suggested that estrogen provides protection in premenopausal women because such gender-specific effects are not seen in postmenopausal women. However, this speculation has not been substantiated further as estrogen treatment does not decrease 711
11.2
Lipids and cardiovascular disease
homozygous individuals. Furthermore, randomised clinical trials of lipid lowering therapy have shown a greater than 30% reduction in CVD (Maron et al. 2000; Grundy et al. 2004). From the different classes of plasma lipids, high levels of fasting cholesterol and LDL and low levels of HDL have been identified as the most proatherogenic. High concentration of fasting serum triacylglycerols also represents an independent risk factor. The different classes of serum lipids are considered below in more detail.
the incidence of the disease in postmenopausal women. Other factors are thus likely to contribute for such gender differences (see Mendelsohn and Karas, 2005); for example, women generally have higher levels of the protective highdensity lipoprotein (HDL) (see Section 11.2.4) than agematched males. Hypertension, obesity and diabetes often coexist; for example, hypertension and diabetes are relatively common in obese individuals. According to figures from the WHO, 6.3% of individuals aged 20 or above in developed countries and 4.1% in developing countries suffer from diabetes. This figure is expected to increase in the future because sedentary lifestyle, intake of “convenience food” containing high levels of fat and salt, and obesity are increasing at an alarming rate. More than 60% of adults in the U.S. are overweight or obese and even in China, a population that has been previously classed as slim and physically active, there are 70 million overweight individuals. A risk between smoking and CVD was first suggested in 1940 and this has been substantiated by numerous studies. Smoking promotes CVD via several mechanisms, including damage to the endothelial lining of the arterial wall, increase in the circulating levels of the proatherogenic lowdensity lipoproteins (LDL), decrease in the concentration of HDL (see Section 11.2.4) and stimulation of blood clotting. Cigarette smoking also raises the concentration of plasma carbon monoxide, which has a number of additional detrimental effects, such as promotion of both endothelial hypoxia and thrombus formation. The nicotine in cigarettes also accelerates heart rate and raises blood pressure. From the different risk factors identified, hyperlipidemia, particularly high levels of serum cholesterol, has been the subject of intense research for a number of decades and caused the most debate. A number of epidemiological studies have shown a direct correlation between high plasma cholesterol levels and CVD (Martin et al., 1986; Anderson et al., 1987; Gurr, 1992). Additional support for a proatherogenic role of plasma cholesterol is provided by studies on individuals with the autosomal dominant disorder familial hypercholesterolemia, which is characterised by a two- to five-fold increase in plasma LDL cholesterol (see Section 11.2.6). About 80% of patients that are heterozygous for this disorder experience CVD by the age of 60 and this is reduced to 15 years in TABLE 11.5
11.2.4
Serum lipids: classification and metabolic roles
All lipids are carried in the plasma as complexes with proteins. Although long-chain fatty acids circulate bound to plasma albumin, the other lipids are carried by a number of lipoprotein particles, which are responsible for the transport of endogenously produced and dietary lipids. Such lipoprotein particles are spherical in shape, with diameter between 5 µm and 1200 µm, and comprised of a hydrophobic core, containing triacyglycerols and cholesterol esters, surrounded by a hydrophilic shell of phospholipids, unesterified (free) cholesterol and apolipoproteins. There are five major classes of lipoproteins: chylomicrons, very low-density (also called pre-β) lipoproteins (VLDL), intermediate density lipoproteins (IDL), low-density (β-) lipoproteins, and high-density (α-) lipoproteins. The classification is based on the hydrated density of the lipoproteins. The different lipoproteins differ in size, electrophoretic mobility, and composition of lipids and apolipoproteins (see Table 11.5). Although each lipoprotein is synthesised with a characteristic set of apolipoproteins, considerable exchange of these apolipoproteins with other lipoprotein particles occur during their metabolism (see Table 11.5). Chylomicrons are synthesised by the mucosal cells of the small intestine and act as a vehicle for the transport of dietary triacylglycerols and cholesterol. VLDL, which is synthesised and secreted by the liver, is also rich in triacylglycerols, but these are derived from endogenous sources. IDL is formed as triacylglycerols are removed from VLDL (see Section 11.2.5). LDL is the main carrier of cholesterol in the plasma and is derived primarily from the catabolism of VLDL. HDL, on the other hand, is formed mainly in the liver as a lipid-poor particle that becomes modified
The composition of plasma lipoproteins Composition (% total)
Lipoprotein
Density (g/ml)
Chylomicrons VLDL IDL LDL HDL
< 0.95 0.95–1.006 1.006–1.019 1.019–1.036 1.063–1.210
a
Diameter (µm)
TAG
Cholesterol
Phospholipid
75–1200 30–80 25–35 18–25 5–12
90 60 27 10 5
5 12 34 50 20
3 18 27 15 25
Protein 2 10 12 25 50
Apolipoproteinsa B48 (A, C, E) B100 (A, C, E) B100, E B100 AI, AII (C, E)
The main apolipoproteins in each of the lipoproteins are shown first, with those that are exchanged with other lipoprotein particles indicated in parenthesis (see text for more details). TAG, triacylglycerol.
712
Medical and Agricultural Aspects of Lipids
The apolipoproteins play a crucial role in the regulation of lipoprotein metabolism and the stabilisation of the lipoprotein particles. Additionally, they modulate the activity of several key enzymes in lipoprotein metabolism and some of them also bind to specific cell surface receptors and, thereby, regulate the metabolic fates of lipoproteins (Table 11.6). There are at least four receptors for lipoproteins and remnant particles: the VLDL receptor, the remnant receptor(s), the LDL receptor (LDL-R) (also called the apo B/E receptor), and the LDL receptorrelated protein (LRP). Chylomicrons synthesised by the intestine are secreted into the lymphatic system and reach the plasma through the thoracic duct. Chylomicrons contain apoB48, which is synthesised in the intestine, along with apolipoproteins-A, -C, and -E. ApoB48 is found exclusively in chylomicrons and is derived from the apoB100 gene by RNA editing in the intestinal epithelium. ApoB48 lacks the LDL-R-binding domain present in apoB100. The triacylglycerol core of chylomicrons is hydrolysed by LPL present on the surface of capillary endothelial cells. This is accompanied by transfer of phospholipids, via phospholipid transfer protein (PLTP), and apolipoproteins-A and -C to HDL. The loss of apoCII, an activator of LPL, prevents further hydrolyses of the smaller, chylomicron remnant particles. These chylomicron remnants, which contain primarily cholesterol, apoE and apoB48, are then taken up by the liver principally via a chylomicron remnant receptor. The VLDL synthesised and secreted by the liver contains apoB100 and acquires cholesteryl esters and apolipoproteins-A, -C, and -E from circulating HDLs. Similar to chylomicrons, the triacylglycerol component of VLDL is subjected to LPL-mediated hydrolysis. The action of LPL, along with the transfer of phospholipids and apolipoproteins-A and -C to circulating HDLs, converts VLDLs to IDLs (also called VLDL remnants). Approximately half of these VLDL remnants are removed from the circulation by high affinity binding to the LDLR of liver cells due to the presence of both apoB100 and apoE. The remaining remnants transform into cholesterol-rich LDL particles by losing more triacylglycerols, via hydrolysis by HL, and shedding all of their lipoproteins except for apoB100.
during the catabolism of VLDL. The major function of HDL is to aid the transport of cholesterol from peripheral tissues to the liver where it can be excreted as bile acids (a process called reverse cholesterol transport). In addition to the lipoprotein classes detailed above, Lipoprotein (a) [Lp (a)] is also found in the plasma where its concentration may range from 0.2 mg/dl to 120 mg/dl. LP (a), which is synthesised in the liver, has generated substantial interest because of its presence in atherosclerotic lesions and its identification as a major risk factor for CVD (see Berglund and Ramakrishnan, 2004; Boffa et al., 2004). LP (a) contains an LDL particle with apoB100 to which a genetically polymorphic form of apo (a) is attached by a disulfide bond. The precise mechanism(s) for a proatherogenic role of LP (a) remain to be determined. However, LP (a) is known to be taken up by macrophages and, thus, contributes to foam cell formation. Apo (a) also shows striking homology with the fibrinolytic proenzyme plasminogen. Although LP (a) does not possess the protease activity associated with plasminogen, it may still interfere with its action and, thereby, impair fibrinolysis (clot resolution).
11.2.5
Metabolic fates of circulating lipoproteins
At least four enzymes play crucial roles in the metabolism of lipoproteins: lipoprotein lipase (LPL), hepatic lipase (HL), lecithin:cholesterol acyltransferase (LCAT), and cholesteryl ester transfer protein (CETP). LPL and HL are involved in the hydrolysis of triacylglycerol-rich lipoproteins. LPL, which requires apoCII as a specific co-activator, interacts with heparin sulfate proteoglycans (HSPG) on the surface of vascular endothelial cells, whereas HL is associated with the plasma membrane in the liver. LPL hydrolyses triacylglycerols in chylomicrons and VLDL to produce nonesterified fatty acids and 2-monoacylglycerol, which are either re-esterified for storage in the adipose tissue or used as a source of energy in the muscle. On the other hand, HL acts on particles that have already been partially digested by LPL and facilitates the conversion of IDL to HDL. LCAT, which is activated by apoAI, esterifies cholesterol acquired by HDL, whereas CETP catalyses the transfer of cholesterol esters from HDL to triacylglycerol-rich lipoproteins. TABLE 11.6
Properties of major apoliproteins
Apoliprotein
Major lipoprotein
Effect on enzyme activity
AI AII B48 B100 CI CII CIII E
HDL HDL Chylomicrons VLDL, IDL and LDL Chylomicrons, VLDL and HDL Chylomicrons, VLDL and HDL Chylomicrons, VLDL and HDL Remnants
Activates LCAT Activates HL LCAT cofactor – – LCAT? Activates LPL Inhibits LPL –
Putative receptor SR-BI ? LRP LDL-R – – – LDL-R
Note: HL, hepatic lipase; LCAT, lechithin:cholesterol acyltransferase; LDL-R, LDL receptor; LPL, lipoprotein lipase; LRP, LDL receptor-related protein; SR-BI, scavenger receptor-BI.
713
11.2
Lipids and cardiovascular disease
HDL plays a crucial role in reverse cholesterol transport (Figure 11.3). Nascent HDL is formed in the liver and the intestine as a disk-like particle containing apoAI and some phospholipids. This HDL particle acts as a potent acceptor of cholesterol derived from peripheral cells. The ATPbinding cassette transporter (ABC)-A1, along with a number of other such transporters, play a key role in the efflux of cholesterol from cells using ATP as a source of energy. The importance of ABCA1 is shown by patients with Tangier disease, which lack this transporter and suffer premature CVD because of a massive accumulation of cholesteryl esters in a number of tissues. The cholesterol taken up by HDL becomes esterified to cholesteryl esters through the action of LCAT. The cholesteryl esters then move deeper into the HDL particle, which now assumes a small, spherical shape and is called HDL3. HDL transfers part of its cholesteryl esters to triacylglycerol-rich lipoproteins and acquires phospholipids and apolipoproteins from them to form larger HDL2 particles. The return of HDL cholesterol to the liver occurs via three pathways (see Figure 11.3). First, the entire HDL particle is taken up by the liver through the LDL-R. Secondly, the cholesterol ester in HDL is transferred to other lipoproteins, via the action of CETP, and these lipoproteins are then taken up by the liver through the LDL-R. The importance of this pathway is supported by several lines of evidence. For example, premature CVD is common in individuals with CETP-deficiency despite the presence of high HDL levels (Bruce et al., 1998). In addition, overexpression of CETP in mice, which normally lack this enzyme, is anti-atherogenic (Bruce et al., 1998). Thirdly, selective delivery of cholesteryl esters in HDL to the liver can occur via scavenger receptor class BI (SR-BI) (Acton et al., 1996). SR-BI binds HDL avidly and mediates the selective delivery of cholesteryl esters to the cell membrane in the liver without internalisation and degradation of the HDL particle. This direct uptake of cholesteryl ester from HDL is facilitated by the binding of HDL to cell surface HSPG because of the presence of apoE in the lipoprotein
LDL particles are taken up by the liver or peripheral tissues via LDL-R-mediated endocytosis. The endocytosed membrane vesicles fuse with lysosomes where the apolipoproteins are digested and the cholesterol esters are hydrolysed to yield free cholesterol. The excess intracellular cholesterol is then re-esterified for storage by the action of the enzyme acyl-CoA-cholesterol acyltransferase (ACAT), whose activity is enhanced by free cholesterol. The cellular uptake of LDL via LDL-R is under negative feedback inhibition by the concentration of intracellular cholesterol. This regulation occurs through a family of membrane-bound transcription factors called sterol regulatory element-binding proteins (SREBPs). A low level of intracellular cholesterol leads to the activation of SREBPs, which then stimulate the transcription of the LDL-R gene along with a number of other genes implicated in cholesterol biosynthesis (see Brown and Goldstein, 1999; Eberle et al., 2004). SREBPs are produced as an integral membrane-bound precursor form in the endoplasmic reticulum and are activated by a SREBP cleavage-activating protein (SCAP), which contains a sterol-sensing domain. When the intracellular concentration of cholesterol is low, SCAP and the SREBP precursors move from the endoplasmic reticulum to serine proteases located in the Golgi apparatus. The proteases cleave the SREBP precursors and produce the active transcription factor, which can translocate to the nucleus and bind to its recognition sequences in the regulatory regions of target genes. As discussed in detail in Section 11.2.7 below, lipoproteins, such as LDL, are subject to several types of modifications, particularly oxidation. Such modified lipoproteins are readily taken up by scavenger receptors (SRs). Macrophages express at least six different forms of SRs for modified LDL. Unlike the LDL-R, such SRs are not subject to feedback inhibition by intracellular cholesterol concentration and can, therefore, take up cholesterol in an uncontrolled manner, thereby contributing to foam cell formation (see Greaves and Gordon, 2005, for details on SRs).
SR-BI
HDL2
LCAT Bile
LCAT
HL HDL3 CETP
Preβ-HDL
ABCA1
Cholesteryl esters
VLDL/IDL HL LDL-R
LDL
Liver
LDL-R Peripheral cell
FIGURE 11.3 A schematic representation of reverse cholesterol transport (see text for more details). Abbreviations: ABCA1, ATPbinding cassette transporter 1; CETP, cholesteryl ester transfer protein; HDL, high density lipoproteins; HL, hepatic lipase; IDL, intermediate density lipoproteins; LCAT, lecithin-cholesterol acyltransferase; LDL-R, low density lipoprotein receptor; SR-B1, scavenger receptor B1; VLDL, very low density lipoproteins.
714
Medical and Agricultural Aspects of Lipids
There are a number of rare inborn errors of lipid metabolism. Those that comprise the familial hyperlipoproteinemias are probably the most widely studied (Table 11.7; Levy and Fredrickson, 1968; Fredrickson and Breslow, 1973; Genest, 2003). These are divided into five main types according to the major changes in plasma lipoprotein profiles. Type I hyperlipoproteinemia, also known as familial hyperchylomicronemia, familial exogenous hypertriglyceridemia, or familial fat-induced lipemia, is a rare recessive condition due to a deficiency of the enzyme LPL or, more rarely, by the absence of its activator apoCII. Chylomicron and VLDL metabolism, therefore, is defective and their accumulation results in very high plasma concentration of triacylglycerols. Clinical symptoms include eruptive xanthomas, hepatosplenomegaly, lipemia retinalis, abdominal pain and pancreatitis. Indeed, pancreatitis rather than atherosclerosis is the major reason for early deaths in these patients. Type II hyperlipoproteinemia includes several genetic conditions, including familial hypercholesterolemia, familial combined hyperlipidemia, familial defective apolipoprotein B and polygenic hypercholesterolemia. These disorders are associated with a very high concentration of plasma LDL and, in certain cases, VLDL. Some classify type II hyperlipoproteinemias into type IIA and type IIB depending on whether hypertriglyceridemia is absent or present, respectively (see Table 11.7). The other major symptoms include xanthomas, particularly on the hand and knee tendons. Familial hypercholesterolemia is caused by the absence of fully functional LDL-R, resulting in delayed clearance of LDL. Homozygous individuals have extremely high levels of plasma LDL-cholesterol irrespective of diet or life style and exhibit severe CVD, usually resulting in death at an early age. Heterozygous individuals are a more diverse group with a 50% probability of death from CVD before the age of 50. A form of familial hypercholesterolemia with
(Arai et al., 1999). The importance of SR-BI can be gauged by the finding that its inactivation in mice is associated with a dramatic increase in serum cholesterol levels and the size of HDL particles without any significant effect on the HDL protein concentration (Rigotti et al., 1997). In addition, liver-specific expression of SR-BI in mice results in reduced HDL levels and an increase in reverse cholesterol transport (Wang et al., 1998). Cholesterol delivered to the liver by HDL is excreted as a free sterol or as bile acids. The HDL particle shrinks as a result of the transfer of cholesteryl ester to the liver and some of the particles become nascent HDL for another round of transport of cholesterol.
11.2.6
Factors affecting the composition and metabolism of circulating lipids
When factors that affect the composition of circulating lipids in humans are considered, the nature of the “normal” subject must be defined as genetic prepositions, age, sex, diet, exercise, and overt or hidden disease may all contribute to discernible differences. Where there is no evidence of malnutrition, the dietary components that most affect the lipid profiles are fats and carbohydrates. In relation to fats, the nature of the fatty acid components has attracted the most attention. When conventional diets are changed by an increased intake of polyunsaturated fats, significant reductions in total plasma cholesterol concen-trations are seen, with a major reduction in the LDL fraction. The beneficial effects of polyunsaturated fatty acids are seen at virtually all stages of the disease, including the control of overall lipid metabolism and transport, regulation of nuclear receptors (see Section 11.2.8), modulation of adhesion protein and cytokine expression by endothelial cells, and control of platelet function (De Caterina et al., 2004; Mori and Beilin, 2004; Vanden Heuvel, 2004; Mori, 2004; Jump, 2004). TABLE 11.7
Hyperlipoproteinemias
Typea
Other designations
Elevated lipoprotein class
I
Familial Hyperchylomicronemia Familial Exogenous Hypertriglyceridemia Familial Fat-Induced Lipemia Familial Hypercholesterolemia Familial Defective Apolipoprotein B Polygenic Hypercholesterolemia Familial Combined Hyperlipidemia Broad Beta Disease Familial Dysbetalipoproteinemia Endogenous Hypertriglyceridemia Hyperprebetalipoproteinemia Mixed Hyperglyceridemia Mixed Hyperlipidemia Hyperprebetalipoproteinemia with Chylomicronemia
II-a
II-b III IV V
a b
[Cholesterol]b
[Triacylglycerols]b
Chylomicrons
↑
↑
LDL
↑
−
LDL, VLDL β-VLDL
↑ ↑
↑ ↑
VLDL
−
↑
VLDL Chylomicrons
↑
↑
The designation is as proposed by Fredrickson and adopted by the WHO. ↑, increase; ↓, decrease; −, no change.
715
11.2
Lipids and cardiovascular disease
and caused by mutations of ABCG5 and ABCG8, is characterised by an accumulation of both animal and plant sterols in the body. This is thought to be due to abnormal absorption of plant sterols, cholesterol hyperabsorption and reduced secretion of sterols into bile. Another recessive form of hypercholesterolemia is associated with deficiency of cholesterol 7α-hydroxylase, a key enzyme in the synthesis of bile acids. Finally, Tangier disease, first identified in the island of Tangier in the Chesapeake Bay in the U.S., is an autosomal recessive disorder because of mutations in the ABCA1 gene that results in hypertriglyceridemia and extremely low levels of HDL and apoA1. The most common form of dyslipidemias, however, has multifactorial origins, such as defects in several genes implicated in cardiovascular disease, environmental and life-style influences and other pathological conditions. Hyperlipidemia is known to arise because of medication (e.g., estrogen, oral contraceptives, steroids), excessive alcohol consumption, chronic and uncontrolled diabetes, nephrosis and endocrine disorders (e.g., hyperthyroidism). Infection and inflammatory responses associated with certain pathological conditions can also cause hyperlipidemias, changes in lipoprotein profile and premature CVD. The elevated levels of circulating cytokines are mainly responsible for such changes (see Mead and Ramji, 2002; Mead et al., 2002; Daugherty et al., 2005; Greenow et al., 2005; Harvey and Ramji, 2005). For example, pro-inflammatory cytokines reduce the expression of LPL in the adipose tissue, thereby causing an accumulation of chylomicrons and VLDL because of their defective clearance. In addition, pro-inflammatory cytokines inhibit the expression of apoE and ABCA1 by macrophages, thereby suppressing cholesterol efflux and accelerating foam cell formation. The most frequent cause of hyperlipidemia is the Metabolic Syndrome (Section 11.3.7) seen in obese individuals. This is clearly a major problem at the moment as the number of obese individuals, including young children, is increasing throughout the world. Lack of physical activity, a diet rich in saturated fats and refined sugars, high intake of calories compared to expenditure, and a sedentary life style all contribute to the proatherogenic lipid and lipoprotein profile in these individuals. In addition, they frequently have elevated blood pressure, peripheral insulin resistance, reduced HDL-cholesterol levels, increased plasma triacylglycerols, and onset of type II diabetes. The combination of these factors often acts in a synergistic manner to promote premature CVD.
a similar clinical phenotype, called familial defective apolipoprotein B, arises because of mutations in apoB100 in regions that represent binding sites for the LDL-R. Familial combined hyperlipidemia is characterised by excess circulating levels of LDL, VLDL, or both. Excessive hepatic production of apoB100, a major protein constituent of VLDL and LDL, is common. It is transmitted in a dominant manner, but does not often manifest until after adolescence. Defects in a number of genes implicated in lipoprotein metabolism and transport is the most frequent cause for this disorder. Finally, polygenic hypercholesterolemia is a heterogenous group of disorders that accounts for the largest number of patients in type II hyperlipoproteinemias. Most patients have elevated levels of LDL due to its impaired clearance. Type III hyperlipoproteinemia, also known as broad beta disease or familial dysbetalipoproteinemia, is associated with elevated levels of triacyglycerols and cholesterol because of abnormalities in VLDL (i.e., presence of an abnormal form of β-migrating VLDL), xanthomas, and premature CVD (Mahley et al., 1999). The major defect is the presence of an abnormal apoE, which does not bind efficiently to hepatic receptors that require this apolipoprotein for interaction with lipoproteins. The patients have the so-called apoE-2/2 phenotype or genotype. There are three common alleles for apoE, apoE-2, -E3, and E-4, with the apoE-2 allele associated with a marked decrease in binding to the LDL-R and premature CVD (Greenow et al., 2005). Type IV hyperlipoproteinemia, also called endogenous hypertriglyceridemia or hyperprebetalipoproteinemia, is a common disorder characterised by elevated levels of circulating VLDL and increased predisposition to CVD. This disorder is frequently associated with insulin resistance and obesity and is particularly common in American middle-aged men. Type V hyperlipoproteinemia, also called mixed hypertriglyceridemia, mixed hyperlipidemia, or hyperprebetalipoproteinemia with chylomicronemia, is a relatively uncommon disorder associated with defective clearance of exogenous and endogenous triacylglycerols. Symptoms include eruptive xanthomas, lipemia retinalis, hepatosplenomegaly, and abdominal pain. Similar to type I hyperlipoproteinemia, acute pancreatitis rather than CVD is the major reason for early deaths. In addition to these relatively common forms of hyperlipoproteinemias, a number of other rare genetic disorders exist that are characterised by marked hypercholesterolemia and/or abnormal lipid and lipoprotein profile (Rader et al., 2003). For example, autosomal recessive hypercholesterolemia (ARH) is caused by mutations of the ARH gene, which codes for a novel adapter protein involved in the internalisation of the LDL-R:LDL complex. Homozygous ApoA1 deficiency results in the virtual absence of HDL and early CVD. Individuals with LCAT deficiency also exhibit extremely low levels of HDL. Sitosterolemia, which is also associated with premature CVD
11.2.7
Circulating lipids and the pathogenesis of atherosclerosis
The transformation of macrophages into foam cells is clearly a key step in the pathogenesis of atherosclerosis and a major target for therapeutic intervention of the disease (Li and Glass, 2002). Native LDL is not taken up by macrophages rapidly enough to form foam cells and numerous studies have shown that modification of the lipid 716
Medical and Agricultural Aspects of Lipids
of epitopes, resulting in not only a cellular immune response, but also a humoral response (Horkko et al., 2000). For example, oxLDL is known to stimulate endothelial cells to secrete a range of pro-atherogenic cytokines. In addition, modified LDL is able to activate nuclear factor-κB (NF-κB), a master transcription factor implicated in the induced expression of a battery of genes associated with an inflammatory response. Modified LDL has also been shown to act as a chemoattractant for circulating monocytes, modulate vascular tone and cause aggregation of platelets (Stoker and Keaney, 2004).
and apoB100 moiety drives the formation of fatty streaks (Navab et al., 1996). Oxidation of LDL occurs in the arterial wall and becomes prevalent when levels of circulating LDL are raised. LDL diffuses through the endothelial cell junctions to the subendothelial matrix and its retention in the vessel wall may involve interaction with matrix proteoglycans (Stoker and Keaney, 2004). Although the precise mechanisms responsible for the oxidation of LDL remain to be fully elucidated, lipooxygenases, myelo-peroxidases, NADPH oxidases, and inducible nitric oxide synthase are major contributing enzymes (Stoker and Keaney, 2004). For example, there is diminished atherosclerosis in mice lacking 12/15 lipoxygenase. The inducible nitric oxide synthase also contributes to LDL oxidation in vivo and inhibitors of this enzyme have been shown to decrease atherosclerosis in rabbits. The precise action of oxidised LDL depends on the extent of its modification, which ranges from minimal to extensive. Minimal modification allows LDL to be recognised by the LDL receptor as normal, whereas extensive modification results in the particle being bound by SRs expressed on the surface of macrophages and SMCs, primarily SR-A and CD36. The importance of scavenger receptors in atherogenesis can be gauged by the observation that inactivation of SR-A or CD36 leads to reduced atherosclerosis in murine models of the disease (Greaves and Gordon, 2005). OxLDL taken up by macrophage SRs is delivered to lysosomes where its cholesteryl ester content is hydrolysed to free cholesterol and fatty acids. This free cholesterol has a number of potential metabolic fates, including esterification and storage as lipid droplets in foam cells (see Li and Glass, 2002). The cholesteryl ester stores of macrophages have been shown to undergo a continuous cycle of hydrolysis and reesterification. The hydrolytic step of this cycle is carried out by a neutral cytoplasmic cholesteryl ester hydrolase and the reesterification step is mediated by ACAT-1. As intracellular cholesterol levels increase, the proteolytic activation of SREBPs required for cholesterol biosynthesis and LDL-R expression is inhibited. Although this prevents the further accumulation of cholesterol via these pathways, cholesterol is still taken up via the SRs and, therefore, cholesterol homeostasis cannot be maintained. The macrophage can dispose of excess cholesterol by either enzymatic modification to more soluble forms or efflux to acceptors, such as HDL. The enzyme 27-hydroxylase is expressed in macrophages at relatively high levels and may play a role in cholesterol excretion by converting it to the more soluble 27-oxygenated steroid, which can be readily accepted by albumin (Babiker et al., 1997). The efflux of cholesterol involves the ABC family of membrane transporters, particularly ABCA1, and apoA1 and apoE in HDL as major acceptors of the steroid. LDL also undergoes other types of modifications, such as nonenzymatic glycation, enzymatic degradation, and aggregation. All such modifications generate a wide range
11.2.8
Therapeutic approaches in cardiovascular disease based on the control of lipid metabolism
Various health organisations are strongly recommending dietary and other lifestyle changes to help to slow down the development of CVD and decrease the pro-atherogenic parameters. These include cessation of smoking; reduced intake of diets rich in fats, particularly saturated fats, and salt; substituting saturated fatty acids in the diet with polyunsaturated fatty acids; eating the recommended daily five servings of fruit and vegetables; at least 30 minutes of moderate exercise on 5 or more days of the week by adults; moderate alcohol consumption (one or two drinks per day); and combating detrimental psychosocial factors, such as stress, depression and anxiety. A number of other dietary supplements, such as antioxidants, which inhibit oxidation of LDL at least in vitro, have also been recommended for the prevention of atherosclerosis. However, the results from clinical studies have not always identified a positive antiatherogenic effect. Dietary and life-style changes are clearly not sufficient when a “clinical horizon” has been reached. Intake of drug(s) that limit the levels of the pro-atherogenic agent(s) in patients is necessary to prevent premature death. Because atherosclerosis is associated with dramatic changes in lipid metabolism and transport, it is not surprising that several drugs, that are either in current use or being developed, target key proteins or enzymes implicated in the maintenance of lipid homeostasis (Choy, 2004; Wierzbicki, 2004). As detailed above, numerous studies in the past 4 decades have supported a strong link between high circulating levels of LDL-cholesterol and atherosclerosis. About 75% of the total cholesterol pool in the body is derived from de novo synthesis with the remainder obtained from dietary intake. Inhibition of cholesterol synthesis, therefore, represents the main approach for reducing levels of circulating LDL-cholesterol. The conversion of 3-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) to mevalonate, which is catalysed by the enzyme HMG-CoA reductase, is the rate-limiting step in the biosynthesis of cholesterol. The activity of this enzyme is inhibited by the statin class of drugs, such as lovastatin, 717
11.2
Lipids and cardiovascular disease
being studied intensely for therapeutic intervention of CVD are peroxisome proliferators-activated receptors (PPARs), liver-X-receptors (LXRs), and the farnesoid X receptor (FXR) (see Li and Glass, 2004; Marx et al., 2004; Ory, 2004; Barish and Evans, 2004; Berger et al., 2005; Claudel et al., 2005; Lehrke et al., 2005). These receptors form obligate heterodimers with retinoid X receptor, another nuclear receptor, and interact with recognition sequences in the regulatory regions of target genes. The PPAR family contains three members; PPAR-α, -γ, and -δ (also called PPAR-β). Fibrates, such as fenofibrate and gemfibrozil, are PPAR-α agonists, which lower concentration of circulating triacylglycerols by stimulating the expression of numerous genes implicated in the cellular uptake and catabolism of lipids. Many natural compounds have also been shown to act as PPAR-α agonists, including polyunsaturated fatty acids, such as eicosapentanoic acid, dodecahexanoic acid and linoleic acid. The action of PPAR-α on lipid homeostasis is mediated, at least in part, through the activation of genes implicated in the uptake, metabolism and β-oxidation of fatty acids. The channeling of fatty acids to the β-oxidative pathway reduces the availability of substrate for the synthesis of triacylglycerols and, thereby, ultimately leads to a decrease in the synthesis and secretion of VLDL by the liver. In addition, PPAR-α agonists stimulate the hydrolysis of circulating lipoproteins by increasing the synthesis of LPL and decreasing the levels of apoCIII, an inhibitor of LPL. Furthermore, PPAR-α agonists elevate HDL levels because of their ability to increase the synthesis of apoA1, thereby enhancing the formation of new HDL particles and inhibiting the expression of SR-BI, thus decreasing the clearance of HDL. Moreover, PPAR-α agonists have been shown to stimulate cholesterol efflux from foam cells. The PPAR-γ gene gives rise to two isoforms, PPARγ1 and γ2, by alternative use of promoters. PPAR-γ2 is expressed specifically in the adipose tissue, whereas PPAR-γ1 is the predominant isoform in other tissues, such as the liver and muscle. PPAR-γ2 is essential for the differentiation of adipocytes and the maintenance of normal glucose metabolism and promotes lipid accumulation by these cells. A number of naturally occurring fatty acid metabolites can activate PPAR-γ, including 15-deoxy∆12,14-prostaglandin J2 and oxidised linoleic acid (9- and 13-hydroxyoctadecadienoic acids). Pharmacological activators of PPAR-γ, such as glitazones (e.g., rosiglitazone, pioglitazone), have been used widely to improve insulin sensitivity in type II diabetes. Several actions of PPAR-γ are likely to contribute to improved insulin sensitivity, including induced expression of insulin-dependent glucose transporter GLUT4, stimulation of insulin signalling, inhibition of lipolysis and increased uptake of fatty acids and synthesis of triacylglycerols. In addition, PPAR-γ agonists regulate the secretion of several proteins by the adipose tissue, such as adiponectin, which then affects insulin signalling
rosuvastatin, atorvastatin, pravastatin and simvastatin (Maron et al., 2000; Grundy et al., 2004). By decreasing cholesterol synthesis, the statins also increase the expression of the LDL-R by relieving the feedback inhibition, thereby further decreasing circulating LDL-cholesterol levels. Statins also cause a moderate decrease in serum triacylglycerol levels, a slight increase in HDL-cholesterol and inhibit the inflammatory response (Linsel-Nitschke and Tall, 2005; Elrod and Lefer, 2005). Statins are generally well tolerated except when taken at high doses where they may affect liver function. Although a reduction of LDL-cholesterol of 50 to 60% can be achieved by statins, doubling of doses produces only a marginal added decrease in LDL-cholesterol levels, but at the expense of increased side-effects. The inhibition of intestinal uptake of dietary cholesterol, which accounts for approximately 25% of this lipid in the body, is necessary to achieve further reduction of circulating LDL-cholesterol. Consumption of products rich in plant sterols and stanols, such as the Pro-ActivTM range of products from FloraR and similar products from BenecolR, which act as inhibitors of intestinal cholesterol uptake, can lower serum LDL cholesterol levels by 10% to 14% (Plat and Mensink, 2005; Cater et al., 2005). Bile acid sequestrants (e.g., colesevelam), which bind to bile acids in the intestine, also cause a 10 to 15% decrease in plasma LDL-cholesterol levels. Such agents interrupt the enterohepatic circulation of bile acids through which most of them are recycled back to the liver and reabsorbed and, thereby, promote their excretion in the faeces (Norata and Catapano, 2004). The use of such sequestrants is, however, limited because of their severe side effects. Direct inhibition of intestinal cholesterol absorption, therefore, offers the most promise. Ezetimibe is a new drug that acts in this manner and can often be taken with statins to achieve maximal reduction of LDLcholesterol (Gagne et al., 2002; Clader, 2005). Increasing circulating levels of HDL represents another, important avenue to limit atherosclerosis and its complications. Low circulating levels of HDL are common in obese individuals and this is likely to be a major problem in the future because of a dramatic worldwide increase in obesity. Nicotonic acid has been used to raise HDL levels indirectly by inhibiting hepatic VLDL synthesis and peripheral lipolysis and stimulating ABCA1 expression (Carlson, 2005). More recently, major advances in devising potential approaches for increasing circulating HDL levels and reducing other pro-atherogenic changes have been made from studies on nuclear receptors, which represent “hot” therapeutic targets for CVD at the moment. Such nuclear receptors are transcription factors that regulate the expression of a battery of genes implicated in the control of triacylglycerol and cholesterol homeostasis and, additionally, have potent antiinflammatory properties. A number of such receptors were originally identified from studies on the action of antidiabetic and lipid-lowering drugs. The major nuclear receptors that are currently 718
Medical and Agricultural Aspects of Lipids
Thus, selective LXR-β agonists could represent potential therapeutic agents for limiting atherosclerosis. FXR was initially identified as a nuclear receptor that is activated by bile acids and products of cholesterol metabolism. FXR regulates the expression of a number of genes implicated in the synthesis, transport and detoxification of bile acids. In addition, recent studies have suggested a role for FXR in the control of lipid metabolism and implicate its activation as a new means for limiting atherosclerosis. For example, polyunsaturated fatty acids, such as arachidonic, linolenic and docosahexaenoic acid, have all been shown to act as FXR ligands in vitro. Additionally, FXR-deficient mice have increased circulating levels of total triacylglycerols and cholesterol and decreased expression of SR-BI in the liver. Other targets for preventing atherosclerosis by modulating circulating lipid levels include inhibitors of CETP and ACAT and infusion of HDL particles or apolipoproteins that act as acceptors of cholesterol (Choy, 2004). CETP deficiency in humans is characterised by increased HDL levels and slightly reduced LDL levels. Moderate consumption of alcohol can increase circulating HDL levels and reduce atherosclerosis, at least in part, by inhibiting CETP. Inhibitors of CETP (e.g., CP-529/414 (torcetrapib), JTT-705) have been developed and are undergoing advanced clinical trials. The esterification of cholesterol by ACAT is critical for macrophage foam cell formation and lipoprotein synthesis in the liver and the intestine. Inhibitors of ACAT, such as avasimibe and CS-505, offer a further avenue for limiting atherosclerosis. Interest in apoA1 as a therapeutic target has emerged from the finding that its deficiency in humans is associated with premature atherosclerosis and extremely low levels of circulating HDL. Although a search of synthetic molecules that specifically induce apoA1 expression has not been fruitful, some success has been achieved, at least in murine models of atherosclerosis, by infusion of apoAI or synthetic peptides that mimic the actions of this lipoprotein. Overall, it appears that patients with CVD will be treated in the future with combination of drugs that lower cholesterol synthesis and induce the expression of the LDL-R (statins), inhibit the absorption of dietary cholesterol (e.g., ezetimibe) and enhance cholesterol efflux from foam cells and raise circulating HDL levels (e.g., agonists of nuclear receptors).
in the liver and muscles. PPARγ activators have been shown to inhibit atherosclerosis in murine models of the disease. Activators of both PPAR-α and PPAR-γ are also known to inhibit the inflammatory response indirectly by antagonising the actions of key transcription factors, such as NF-κB. As both PPAR-α and -γ have beneficial effects on lipid metabolism, inflammation, and insulin sensitivity, combined use of agonists for both these nuclear receptors promises to be a more effective approach at limiting atherosclerosis than agonists for individual receptors. Studies on the third PPAR member, PPAR-δ (previously called PPAR-β), have generally lagged behind the other two family members, but recent studies have suggested that it also represents another promising therapeutic target for limiting atherosclerosis. PPAR-δ is expressed in a ubiquitous manner and regulates energy homeostasis and fatty acid catabolism. PPAR-δ-deficient mice are susceptible to obesity, whereas transgenic mice that overexpress an activated form of this transcription factor are resistant to genetically or diet-induced hyperlipidemia and obesity. There are two LXRs, LXR-α and LXR-β, with the latter being expressed ubiquitously and the former present at high levels in the liver, intestine, adipose tissue, and macrophages. Both LXRs are activated by oxidised derivatives of cholesterol (e.g., 22(R)-hydroxycholesterol, 24(S)-hydroxycholesterol, 24(S), 25-epoxycholesterol, 27hydroxycholesterol) and, therefore, act as intracellular sensors of cholesterol. Target genes for the action of LXRs include those implicated in the efflux of cellular cholesterol (e.g., apoE, ABCA1, ABCG1, ABCG4), modification of HDL (e.g., CETP, PLTP), secretion of cholesterol into bile (cholesterol 7α-hydroxylase, ABCG5, ABCG8), and fatty acid metabolism (e.g., sterol response element binding protein SREBP1c, fatty acid synthase, sterol-CoA desaturase-1). The expression of the LXR-α gene is also subject to autoregulation. Overall, LXRs restore cellular cholesterol balance and prevent lipotoxicity by activating cholesterol catabolic and efflux pathways along with those involved in de novo lipogenesis. In addition, LXR activators inhibit the expression of genes implicated in inflammation by antagonising the actions of key transcription factors, such as NF-κB and activator protein-1 (AP-1), largely via a mechanism that does not require sequencespecific DNA binding by the LXRs. A number of in vitro and in vivo studies have revealed a potent anti-atherogenic action of LXRs. However, a major current limitation of employing LXRs as therapeutic targets for atherosclerosis is that they increase the synthesis of fatty acids and cause the accumulation of triacylglycerols. Development of synthetic agonists that increase HDL levels without causing hypertriglyceridemia will clearly be necessary. This is a possibility since ongoing research suggests differences in the action of LXR-α and LXR-β, with the former having a dominant effect on hepatic lipogenesis.
References Acton, S. et al. (1996). Identification of scavenger receptor SRBI as a high-density lipoprotein receptor. Science, 271, 518–520. Anderson, K.M. et al. (1987). Cholesterol and mortality: 30 years of follow-up from the Framingham study. JAMA, 257, 2176–2180. Arai, T. et al. (1999). Decreased selective uptake of highdensity lipoprotein cholesteryl esters in apolipoprotein
719
11.2
Lipids and cardiovascular disease
E knock-out mice. Proc. Natl. Acad. Sci. U.S.A. 96, 12050–12055. Babiker, A. et al. (1997). Elimination of cholesterol in macrophages and endothelial cells by the sterol 27-hydroxylase mechanism. Comparison with high density lipoproteinmediated reverse cholesterol transport. J. Biol. Chem. 272, 26253–26261. Barish, G.D. and Evans, R.M. (2004). PPARs and LXRs: atherosclerosis goes nuclear. Trends Endocrinol. Metab. 15, 158–165. Barrett-Connor, E. and Bush, T.L. (1991). Estrogen and coronary heart disease in women. JAMA 265, 1861–1867. Berger, J.P. et al. (2005). PPARs: therapeutic targets for metabolic disease. Trends Pharmacol. Sci. 26, 244–251. Berglund, L. and Ramakrishnan, R. (2004). Lipoprotein (a): an elusive cardiovascular risk factor. Arterioscler. Thromb. Vasc. Biol. 24, 2219–2226. Boffa, M.B. et al. (2004). Lipoprotein (a) as a risk factor for atherosclerosis and thrombosis: mechanistic insights from animal models. Clin. Biochem. 37, 333–343. Brown, M.S. and Goldstein, J.L. (1999). A proteolytic pathway that controls the cholesterol content of membranes, cells, and blood. Proc. Natl. Acad. Sci. U.S.A. 96, 11041–11048. Bruce, C. et al. (1998). Plasma lipid transfer proteins, high-density lipoproteins, and reverse cholesterol transport. Annu. Rev. Nutr. 18, 297–330. Carlson, L.A. (2005). Nicotinic acid: the broad-spectrum lipid drug. A 50th anniversary review. J. Intern. Med. 258, 94–114. Cater, N.B. et al. (2005). Responsiveness of plasma lipids and lipoproteins to plant stanol esters. Am. J. Cardiol. 96, 23D–28D. Choy, P.C. (2004). Lipids and atherosclerosis. Biochem. Cell Biol. 82, 212–224. Clader, J.W. (2005). Ezetimibe and other azetidinine cholesterol absorption inhibitors. Curr. Top. Med. Chem. 5, 243–256. Claudel, T. et al. (2005). The farnesoid X receptor: a molecular link between bile acid and lipid and glucose metabolism. Arterioscler. Thromb. Vasc. Biol. 25, 2020–2030. Daugherty, A. et al. (2005). Thematic review series: the Immune system and atherogenesis. Cytokine regulation of macrophage functions in atherogenesis. J. Lipid Res. 46, 1812–1822. De Caterina, R. et al. (2004). Effects of omega-3 fatty acids on cytokines and adhesion molecules. Curr. Atheroscler. Rep. 6, 485–491. Eberle, D. et al. (2004). SREBP transcription factors: master regulators of lipid homeostasis. Biochimie. 86, 839–848. Elrod, J.W. and Lefer, D.J. (2005). The effects of statins on endothelium, inflammation and cardioprotection. Timely Top. Med. Cardiovasc. Dis. 9, E20. Fredrickson, D.S. and Breslow, J.L. (1973). Primary hyperlipoprotenemia in infants. Annu. Rev. Med. 24, 315–324. Gagne, C., Gaudet, D. and Bruckert, E. (2002). Efficacy and safety of ezetimibe added to ongoing statin therapy for treatment of patients with primary hypercholesterolemia. Am. J. Cardiol. 90, 1084–1091. Genest, J. (2003). Lipoprotein disorders and cardiovascular risk. J. Inherit. Metab. Dis. 26, 267–287. Glass, C.K. and Witztum, J.L. (2001). Atherosclerosis. The road ahead. Cell. 104, 503–516. Greaves, D.R. and Gordon, S. (2005). Recent insights into the biology of macrophage scavenger receptors. J. Lipid Res. 46, 11–20.
Greenow, K. et al. (2005). The key role of apolipoprotein E in atherosclerosis. J. Mol. Med. 83, 329–342. Grundy, S.M. et al. (2004). Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III Guidelines. J. Am. Coll. Cardiol. 44, 720–732. Gurr, M.I. (1992). Dietary lipids and coronary heart disease: old evidence, new perspective. Prog. Lipid Res. 31, 195–243. Harvey, E.J. and Ramji, D.P. (2005). Interferon-γ and atherosclerosis: pro- or anti-atherogenic? Cardiovasc. Res. 67, 11–20. Horkko, S. et al. (2000). Immunological responses to oxidized LDL. Free Radic. Biol. Med. 28, 1771–1779. Jump, D.B. (2004). Fatty acid regulation of gene transcription. Crit. Rev. Clin. Lab. Sci. 41, 41–78. Lehrke, M. et al. (2005). Gaining weight: the Keystone symposium on PPAR and LXR. Genes Dev. 19, 1737–1742. Levy, R.I. and Fredrickson, D.S. (1968). Diagnosis and management of hyperproteinemia. Am. J. Cardiol. 22, 576–583. Li, A.C. and Glass, C.K. (2002). The macrophage foam cell as a target for therapeutic intervention. Nat. Med. 8, 1235–1242. Li, A.C. and Glass, C.K. (2004). PPAR- and LXR-dependent pathways controlling lipid metabolism and the development of atherosclerosis. J. Lipid. Res. 45, 2161–2173. Linsel-Nitschke, P. and Tall, A.R. (2005). HDL as a target in the treatment of atherosclerotic cardiovascular disease. Nat. Rev. Drug Discov. 4, 193–206. Lusis, A.J. (2000). Atherosclerosis. Nature. 407, 233–241. Lusis, A.J. et al. (2004). Genetics of atherosclerosis. Annu. Rev. Genomics Hum. Genet. 5, 189–218. Mahley, R.W. et al. (1999). Pathogenesis of type III hyperlipoproteinemia (dysbetalipoproteinemia). Questions, quandaries, and paradoxes. J. Lipid. Res. 40, 1933–1949. Maron, D.J. et al. (2000). Current perspectives on statins. Circulation. 101, 207–213. Martin, M.J. et al. (1986). Serum cholesterol, blood pressure and mortality: implications from cohort of 361,662 men. Lancet, 2, 933–936. Marx, N. et al. (2004). Peroxisome proliferators-activated receptors and atherogenesis: regulators of gene expression in vascular cells. Circ. Res. 94, 1168–1178. Mead, J.R. and Ramji, D.P. (2002). The pivotal role of lipoprotein lipase in atherosclerosis. Cardiovasc. Res. 55, 261–269. Mead, J.R. et al. (2002). Lipoprotein lipase: structure, function, regulation, and role in disease. J. Mol. Med. 80, 753–769. Mendelsohn, M.E. and Karas, R.H. (2005). Molecular and cellular basis of cardiovascular gender differences. Science, 308, 1583–1587. Mori, T.A. (2004). Effect of fish and fish oil-derived omega-3 fatty acids on lipid oxidation. Redox Rep. 9, 193–197. Mori, T.A. and Beilin, L.J. (2004). Omega-3 fatty acids and inflammation. Curr. Atheroscler. Rep. 6, 461–467. Navab, M. et al. (1996). The yin and yang of oxidation in the development of the fatty streak. A review based on the 1994 George Lyman Duff Memorial Lecture. Arterioscler. Thromb. Vasc. Biol. 16, 831–842. Norata, G.D. and Catapano, A.L. (2004). Lipid lowering activity of drugs affecting cholesterol absorption. Nutr. Metab. Cardiovasc. Dis. 14, 42–51. Ory, D.S. (2004). Nuclear receptor signalling in the control of cholesterol homeostasis: have the orphans found a home? Circ. Res. 95, 660–670.
720
Medical and Agricultural Aspects of Lipids
Osterud, B. and Bjorklid, E. (2003). Role of monocytes in atherogenesis. Physiol. Rev. 83, 1069–1112. Plat, J. and Mensink, R.P. (2005). Plant stanol and sterol esters in the control of blood cholesterol levels: mechanism and safety aspects. Am. J. Cardiol. 96, 16D–22D. Rader, D.J. et al. (2003). Monogenic hypercholesterolemia: new insights in pathogenesis and treatment. J. Clin. Invest. 111, 1795–1803. Rigotti, A. et al. (1997). A targeted mutation in the murine gene encoding the high density lipoprotein (HDL) receptor scavenger receptor class B type I reveals its key role in HDL metabolism. Proc. Natl. Acad. Sci. U.S.A. 94, 12610–12615. Ross, R. (1999). Atherosclerosis — an inflammatory disease. N. Engl. J. Med. 340, 115–126. Smith, J. D. et al. (1995). Decreased atherosclerosis in mice deficient in both macrophage colony stimulating factor (op) and apolipoprotein E. Proc. Natl. Acad. Sci. U.S.A. 92, 8264–8268. Stocker, R. and Keaney, J.F., Jr., (2004). Role of oxidative modifications in atherosclerosis. Physiol. Rev. 84, 1381–1478. Vanden Heuvel, J.P. (2004). Diet, fatty acids, and regulation of genes important for heart disease. Curr. Atheroscler. Rep. 6, 432–440. Wang, N. et al. (1998). Liver-specific overexpression of scavenger receptor BI decreases levels of very low-density lipoprotein apoB, low density lipoprotein apoB and high density lipoprotein in transgenic mice. J. Biol. Chem. 273, 32920–32926. Wierzbicki, A.S. (2004). Lipid-altering agents: the future. Int. J. Clin. Prac. 58, 1063–1072.
11.3
Clinical aspects of lipids with emphasis on cardiovascular disease and dyslipaemia
11.3.1
Objectives of cardiovascular disease prevention in clinical practice
11.3.2
Dyslipidaemia and cardiovascular disease
As concentrations of total and LDL-cholesterol increase, so does the risk of cardiovascular disease (Stamler et al., 2000). The relationship between cholesterol and cardiovascular risk is continuous, while reduction of total and low density lipoprotein-cholesterol (LDL-C) is associated with numerous sequelae (which attenuate atherogenesis) including improved endothelial function, reduced oxidative stress and reduced inflammation (Ross, 1993; Vogel, 1997). In the context of lipids, cardiovascular risk is principally determined by the concentrations of total, LDL- and high density lipoprotein-cholesterol (HDL-C) (inversely) and, to a lesser extent, plasma triacylglycerol concentrations. Indeed, data from various studies including Framingham and MRFIT (Kannel and Larson, 1993; Stamler et al., 1993) have demonstrated that low levels of HDL-C independently predict increased cardiovascular risk, irrespective of total or LDL-C levels (Stamler et al., 1993). The ratio of total or LDL-C to HDL-C has been consistently demonstrated as the strongest determinant of cardiovascular risk in prospective studies. This was further illustrated by the recent INTERHEART study, in which the ratio of apo Bto apo AI-containing lipoproteins was demonstrated to be the single most important determinant of risk for first myocardial infarction from over 10,000 subjects from varying geographical locations and ethnic origins (Yusuf et al., 2004). Meta-analysis of prospective studies has consistently demonstrated that a 1% decrease in HDL-C is associated with a 2 to 3% increase in cardiovascular risk, while the relationship between HDL-C and cardiovascular risk may be stronger in women (Expert Panel, 2001). The reduction of total LDL and LDL-cholesterol is, however, currently the primary goal of lipid-lowering therapy with respect to cardiovascular risk reduction. The reduction of cholesterol whether by diet, drugs, or other means is associated with a reduced risk of CVD (Brady and Betteridge, 2003) and since lipoproteins are only one element of cardiovascular risk, which is determined overall by the presence of other risk factors, the absolute benefit of cholesterol reduction, is a function of baseline cardiovascular risk. The MRC/BHF Family Heart Protection Study (2002) demonstrated that the benefits of cholesterol lowering therapy extend into all forms of atherosclerotic vascular disease including peripheral vascular disease (PVD) and cerebrovascular disease. In a systematic review and meta-analysis evaluating the effects of cholesterol reduction on coronary heart disease (CHD) and stroke, 58 clinical trials of cholesterol reduction by any means were included (Law et al., 2003). Reduction in CHD death and nonfatal myocardial infarction for a 1.0 mmol/l reduction in cholesterol was 11% in the first year, 24% in the second, 33% in the third to fifth, and 36% in the sixth and subsequent years. After standardisation for reduction in LDL-C and duration of treatment, the reduction in risk of fatal and nonfatal events was similar for different
The specific objective of cardiovascular disease (CVD) prevention for all high-risk people in clinical practice is to reduce the risk of the disease and its associated complications. These include the need for percutaneous or surgical revascularisation procedures in any arterial territory, along with improvements in both quality of life and life expectancy. Cardiovascular disease risk reduction in practice involves a multifactorial approach involving appropriate lifestyle changes, such as weight loss, dietary modification, increasing exercise and smoking cessation. Pharmacological interventions aimed at reducing cardiovascular risk also revolve around a multifactorial approach including antiplatelet therapy, anticoagulation and a target-driven approach to blood pressure reduction. Although optimal cardiovascular disease risk reduction requires a multifactorial approach, prospective epidemiological data have consistently demonstrated the preeminence of abnormalities in lipid metabolism in the aetiology of atherosclerotic vascular disease, in particular, coronary heart disease. Hence, the management of dyslipidaemia in clinical practice represents a central pillar in the approach to cardiovascular disease risk reduction. 721
11.3
Clinical aspects of lipids with emphasis on cardiovascular disease and dyslipaemia
substantial LDL-C reductions in all patients at high risk of any type of major vascular event. Furthermore, in a recent study of the safety and efficacy of intensive cholesterol reduction in 1825 patients with acute coronary syndrome, continuing outcome benefits were demonstrated in association with reductions in LDL-C as low as
