International Journal of Systematic and Evolutionary Microbiology (2002), 52, 1007–1015
DOI : 10.1099/ijs.0.01917-0
Cellular fatty acids as chemotaxonomic markers of the genera Anabaena, Aphanizomenon, Microcystis, Nostoc and Planktothrix (cyanobacteria) 1
Department of Applied Chemistry and Microbiology, PO Box 56, Biocentre Viikki, 00014 Helsinki University, Finland
2
INRA, Station d’Hydrobiologie Lacustre, BP 511, 74203 Thonon-lesBains cedex, France
Muriel Gugger,1† Christina Lyra,1 Irmgard Suominen,1 Irina Tsitko,1 Jean-Franc: ois Humbert,2 Mirja S. Salkinoja-Salonen1 and Kaarina Sivonen1 Author for correspondence : Kaarina Sivonen. Tel : j358 9 19159270. Fax : j358 9 19159322. e-mail : kaarina.sivonen!helsinki.fi
The cellular fatty acid content of 22 cyanobacterial strains belonging to the genera Anabaena, Aphanizomenon, Calothrix, Cylindrospermum, Nostoc, Microcystis and Planktothrix were analysed. The identities of the major peaks were confirmed by MS. Correspondence analysis of the data revealed three distinct groups formed by the Microcystis strains, the Nostoc/Planktothrix strains and the Anabaena/Aphanizomenon/Cylindrospermum strains. The Calothrix strain did not cluster with the other heterocystous cyanobacteria, supporting its morphological classification separate from the Nostocaceae family. The presence of large amounts of the fatty acids 18 :3ω6,9,12c and 18 :0 iso distinguished the Microcystis strains from the other cyanobacteria studied. The high content of 16 :1ω7c grouped the Nostoc strains with the Planktothrix strains. A free-living strain of Nostoc contained 16 :1ω5c and 16 :1ω7c (about 1 : 1), separating it from the symbiotic Nostoc strain and the Planktothrix strains. the strains of Anabaena, Aphanizomenon and Cylindrospermum grouped tightly and were characterized by the presence of 16 :1ω9c and 16 :0 anteiso fatty acids. Correspondence analysis of Anabaena, Aphanizomenon and Cylindrospermum showed that all hepatotoxic Anabaena strains grouped together, whereas the non-toxic and neurotoxic Anabaena strains grouped with the non-toxic Aphanizomenon strains.
Keywords : cyanobacteria, fatty acid, multivariate analysis
INTRODUCTION
The taxonomy of cyanobacteria has been based on morphological, physiological and ecological characteristics, and recently also on biochemical and molecular approaches (Wilmotte, 1994). The current classification of these organisms group the unicellular cyanobacteria that reproduce by binary fission or budding into subsection I, the unicellular cyanobacteria with multiple fission into subsection II, the filamentous uniseriate cyanobacteria into subsection III, the filamentous heterocystous cyanobacteria into subsection IV and the filamentous heterocystous .................................................................................................................................................
† Present address : Laboratoire de Cryptogamie, Muse! um National d’Histoire Naturelle, 12 rue Buffon 75005 Paris, France. Abbreviation : PUFA, polyunsaturated fatty acid. 01917 # 2002 IUMS Printed in Great Britain
cyanobacteria with branching into subsection V (see Castenholz, 2001). Although most of taxa defined in this classification correspond to the traditionally used botanical classification, certain discrepancies between the morphological data and the genetic analyses are still unresolved. Therefore, other markers have been looked at to complement the current methods. Cellular fatty acid composition has been used as a tool for classifying bacteria at the family, genus and species levels (Vandamme et al., 1996). On investigating the lipids in five cyanobacterial species, Holton et al. (1968) found that the level of the fatty acid unsaturation correlated with the morphological complexity of the organisms. Kenyon & Stanier (1970) observed that unsaturated fatty acid content was correlated with the morphology- and physiology-based classification of the cyanobacteria. Unicellular and filamentous cyano1007
M. Gugger and others
bacteria have been grouped in five clusters depending on the number and position of double bonds counted from the carboxyl terminus (∆) or from the methyl terminus (ω) of 16 carbon (C ) and 18 carbon (C ) ") ; fatty acids (Kenyon, 1972 ; "'Kenyon et al., 1972 Murata et al., 1992 ; Cohen et al., 1995). Four groups were first found by Kenyon et al. (1972) and confirmed by Murata et al. (1992). The fifth group described by Cohen et al. (1995) was positioned according to the Kenyon–Murata classification system between groups 1 and 2. The strains of group 1 are devoid of polyunsaturated fatty acid (PUFA) and contain only saturated and monounsaturated fatty acids. The strains of the group defined by Cohen et al. (1995) contain 18 : 2∆9,12 (18 : 2ω6, linoleic acid) as the only C PUFA. Group 2 consists of cyanobacterial strains ") containing 18 : 3∆9,12,15 (18 : 3ω3, α-linolenic acid) as the only C PUFA. The strains of group 3 have 18 : ") : 3ω6, γ-linolenic acid) as the major C 3∆6,9,12 (18 ") PUFA, but no or only traces of 18 : 3∆9,12,15. Strains of group 4 contain either 18 : 3∆9,12,15 or 18 : 3∆6,9,12 or both, but also produce 18 : 4∆6,9,12,15 (18 : 4ω3, octadecatetraenoic acid). The double bonds in the hydrocarbon chains of PUFA are introduced by fatty acid desaturases, which play an important role in the acclimation of the various organisms to changes in environmental temperatures (Murata & Wada, 1995 ; Nishida & Murata, 1996). The DesA (∆12), DesB (ω3 or ∆15) and DesC (∆9) acyl-lipid desaturases and their corresponding genes, desA, desB and desC, have been identified in Synechococcus PCC 7002 (Sakamoto et al., 1994a, b, c, 1998 ; Sakamoto & Bryant, 1997), and DesD (∆6), encoded by desD, was found in Synechocystis PCC 6803 (Reddy et al., 1993). Cyanobacteria have also been classified by analysing the composition of whole-cell fatty acids. Caudales & Wells (1992) showed that the filamentous heterocystous Nostoc and Anabaena genera could be distinguished on the basis of their cellular fatty acids. Vargas et al. (1998) found that the nitrogen-fixing genera are also differentiated by their 3-hydroxy and non-polar fatty acid content. Recently, Caudales et al. (2000) showed that the fatty acid composition of unicellular cyanobacterial strains of subsections I and II generally agreed with their morphological distinctions. The study also highlighted the differences between the fatty acid compositions of Chroococcidiopsis and the other baeocytes-forming strains of the order Pleurocapsales. The fatty acid analysis agreed well with the 16S rRNA analysis. In both analyses Chrooccocidiopsis PCC 7203 was positioned distantly from all other strains of the Pleurocapsales cluster (Wilmotte & Herdman, 2001). We investigated the whole-cell fatty acid content of 22 cyanobacterial strains of seven genera belonging to three different orders in the botanical code, Chroococcales (Microcystis), Oscillatoriales (Planktothrix) and Nostocales (Anabaena, Aphanizomenon, Calothrix, Cylindrospermum and Nostoc). The 22 strains represented three different subsections of the Stanier– Rippka classification (unicellular subsection I, fila1008
mentous non-heterocystous subsection III and filamentous heterocystous subsection IV). We examined the effect of the culture medium and culturing conditions on the whole-cell fatty acid composition. In addition, we investigated the relationships among planktic cyanobacteria by correspondence analysis and found that the composition of cellular fatty acid content differentiated unicellular cyanobacterial strains from filamentous ones, and certain filamentous strains from each other. METHODS Cyanobacterial strains and growth conditions. The 22 studied
axenic cyanobacterial strains are listed in Table 1. The purity of the cultures was checked microscopically and tested routinely on TYG plates (tryptone\yeast extract\glucose, 5 : 2n5 : 1, by weight). The non-heterocystous strains were grown at 20 mC in Z8 medium and the heterocystous strains at the same temperature in Z8 medium without a nitrogen source other than air (Kotai, 1972). Growth was measured as OD . (&! The effects of the culture volume and aeration on fatty acid content were examined with cultures of 3 l and 100 ml. The cultures were grown with aeration (filter-sterilized air, 0n2 µm pore-size filters ; Millipore) under continuous illumination at 11 µE m−# s−" (3 l) or without aeration under continuous light at 12 µE m−# s−" (100 ml). The fatty acids of the cyanobacterial cells were extracted from the 3 l cultures at the exponential phase and from 100 ml cultures at the lag, exponential and stationary phases. The effect of different culture conditions on fatty acid composition of Aphanizomenon sp. strain TR183 was tested with five different combinations of nitrogen [as NaNO and Ca(NO ) ] and phosphorous (as K HPO ) (see Table 3).$ The $# # % five modified Z8 media contained 0n05, 0n275 or 5n5 mg − phosphorous l " and 0n05, 1 or 30 mg nitrogen l−". The reproducibility of the results was estimated with five replicates of Aphanizomenon sp. TR183 independently cultured in the medium with 0n275 mg phosphate l−" and 1 mg nitrate l−". Two replicates were extracted from 3 l cultures of strains 2, 90, 123, 152 and 271, and from 100 ml cultures of strains TR183, 199 and 265, to assess the reproducibility of the extraction procedure. Identification of fatty acids. Biomass was collected by
centrifugation and washed twice with the sterile Z8 growth medium. Whole-cell fatty acids from 150 to 200 mg wet weight cells were extracted and analysed as described by Va$ isa$ nen et al. (1994). The extracts were filtered through a Nylon Acrodisc filter (0n45 µm ; Gelman Sciences). The fatty acids were identified and quantified using the MIDI Microbial Identification System, library version 3.9 (Microbial ID). Major unidentified peaks were analysed by GC-MS as described by Tsitko et al. (1999). Multivariate analysis. The data were analysed by corre-
spondence analysis using the ADE-4 software package (Thioulouse et al., 1997). The axes of the graphs were chosen to show the largest fraction of variability of the multidimensional data matrix. Thus, the most informative axes are in the direction of maximum variance of the scatter of points (Legendre & Legendre, 1998). The inertia value of an axis is the proportion of variance represented by it. For the graphic representation, the scatter module with a star option was International Journal of Systematic and Evolutionary Microbiology 52
Fatty acids of cyanobacteria Table 1. The cyanobacterial strains used in this study Strain* Anabaena sp. 277 Anabaena sp. 37 Anabaena sp. 123 Anabaena sp. 66A Anabaena sp. 90 Anabaena sp. 202A1 Anabaena sp. 202A1\35 Anabaena sp. 202A2 Anabaena lemmermannii NC83\1 Aphanizomenon sp. TR183 Aphanizomenon sp. 202 Aphanizomenon sp. PCC 7905 Aphanizomenon gracile 219 Aphanizomenon cf. gracile 271 Nostoc sp. 152 Nostoc punctiforme PCC 73102 Cylindrospermum stagnale PCC 7417 Calothrix marchica PCC 7714 Microcystis sp. 199 Microcystis sp. 265 Planktothrix sp. 2 Planktothrix sp. NC128\R
Geographic origin
Year of isolation
Toxicity
Reference†
River Pernio$ njoki, Finland Lake Sa$ a$ skja$ rvi, Finland Lake Sa$ yhteenja$ rvi, Finland Lake Sa$ a$ skja$ rvi, Finland Lake Vesija$ rvi, Finland Lake Vesija$ rvi, Finland Lake Vesija$ rvi, Finland Lake Vesija$ rvi, Finland Lake Edlandsvatnet, Norway The Baltic Sea Lake Vesija$ rvi, Finland Lake Brielse Meer, The Netherlands Lake Madesø, Denmark Lake Nørre, Denmark Lake Sa$ a$ skja$ rvi, Finland Root section, Australia Soil, greenhouse, Sweden Small pool, Aldabra Atoll, India Lake Rusutja$ rvi, Finland Lake Tuusulanja$ rvi, Finland Markusbo$ lefja$ rden, Finland Lake Vesija$ rvi, Finland
1991 1985 1986 1986 1986 1987 1987 1987 1981 1993 1987 ? 1994 1994 1986 1973 1972 1972 1987 1990 1985 1984
Non-toxic Neurotoxic Neurotoxic Hepatotoxic Hepatotoxic Hepatotoxic Hepatotoxic Hepatotoxic Hepatotoxic Non-toxic Non-toxic Non-toxic Non-toxic Non-toxic Hepatotoxic Non-toxic Non-toxic Non-toxic Hepatotoxic Non-toxic Non-toxic Hepatotoxic
5, 6, 8 1, 5, 6, 8 1, 5, 6, 8 2, 4, 5, 6, 8 2, 4, 5, 6, 8 2, 4, 5, 6, 8 2, 8 2, 4, 5, 6, 8 2, 4, 5, 6, 7, 8 6, 8 6, 8 6, 7, 8 8 8 1, 5, 6, 8 6, 8 8 8 5, 8 5, 8 3, 5, 8 3, 5, 8
* NC (l NIVA-CYA), Norwegian Institute for Water Research, Oslo, Norway ; PCC, Pasteur Culture Collection, Paris, France. Aphanizomenon gracile 219 and Aphanizomenon cf. gracile 271 were kindly provided by P. Henriksen, National Environmental Research Institute, Roskilde, Denmark. † 1, Sivonen et al. (1989) ; 2, Sivonen et al. (1990) ; 3, Sivonen (1990) ; 4, Sivonen et al. (1992) ; 5, Rouhiainen et al. (1995) ; 6, Lyra et al. (1997) ; 7, Lehtima$ ki et al. (2000) ; 8, Lyra et al. (2001).
chosen to bind each individual sample to the centre of its category.
RESULTS AND DISCUSSION Fatty acid composition of the 22 cyanobacterial strains
In the 22 axenic cyanobacterial strains, 60 fatty acids containing 9–20 carbons were detected by GC. Of these, 26 were found in large amounts ( 1 %) representing more than 93n8 % of the summed peak area of the fatty acids in the strains studied (Tables 2–5). Compounds 14 : 0, 14 : 0 anteiso, 16 : 0, 16 : 0 10-methyl, 16 : 1ω7c, 16 : 1ω9c, 18 : 2ω6,9c and 18 : 3ω3,6,9c were confirmed by MS (Tables 2–5). The growth phase of harvesting (lag, exponential and stationary) did not significantly affect the overall cellular fatty acid composition of the strains of the different genera (within the variation of the replicate analyses). However, quantities of some of the minor fatty acids varied : for example, in Nostoc PCC 73102 there was up to a fourfold difference in 14 : 0 anteiso, ranging from undetectable up to 9 % of the total cellular fatty acid content for 18 : 1ω5c (see Table 4). In general we observed that different culture volumes, aeration, nitrate and phosphate concentrations in the http://ijs.sgmjournals.org
medium mainly had an effect on the quantity of the minor fatty acids. Originally, the studies of Kenyon (1972) and Kenyon et al. (1972) showed that temperature affected the fatty acid composition of cyanobacteria, but growth medium and light conditions did not. Later, Murata et al. (1992) demonstrated that the growth temperature changed the relative proportions of the fatty acids, but did not eliminate the fatty acids characteristic of the metabolic groups of the Kenyon– Murata classification system. Anabaena. The fatty acid compositions of two neurotoxic, six hepatotoxic and one non-toxic Anabaena strains are shown in Table 2. The proportions of the major fatty acids 16 : 0 and 18 : 3ω3,6,9c were 20n2– 34n7 % and 18n7–38n1 % of the total cellular fatty acid contents of the Anabaena samples. The 16 : 1ω9c, 14 : 0 anteiso, 16 : 0 anteiso, 16 : 0 10-methyl and 18 : 2ω6,9c fatty acids were also present in noticeable amounts in all Anabaena strains. Hepatotoxic Anabaena strains 90, 66A, 202A1, 202A1\35, 202A2 and NC83\1 had a fatty acid composition similar to those of neurotoxic Anabaena strains 37 and 123, and non-toxic Anabaena strain 277, but the hepatotoxic strains contained 5–7 % 14 : 0, which was virtually absent ( 1 %) in the nontoxic or neurotoxic isolates. Aphanizomenon. The fatty acids of five non-toxic 1009
M. Gugger and others Table 2. Relative content (%) of fatty acids in Anabaena strains (filamentous heterocystous cyanobacteria) under different growth conditions .................................................................................................................................................................................................................................................................................................................
M, 3 l culture ; L, lag phase, E, exponential phase, S, stationary phase from 100 ml culture ; , trace amount (less than 1 %) ; –, not detected. Fatty acid*
Neurotoxic 37 M
L
Saturated 14 : 0 16 : 0 31n1 20n8 18 : 0 3n1 Total 32n5 24n5 Unsaturated 16 : 1ω7c 3n3 – 16 : 1ω9c 7n7 11n7 18 : 2ω6,9c 6n5 6n8 18 : 3ω3,6,9c 20n2 32n5 Total 37n8 51 Branched and hydroxy-substituted 14 : 0 anteiso 7n6 9n3 16 : 0 anteiso 3n4 3n2 16 : 0 10-methyl 5n1 16 : 0 iso 3-OH 1n5 1n2 17 : 0 10-methyl 5n5 2n5 Total 23n2 16n9
Hepatotoxic 123
66A
90
Non-toxic 202A1
202A1/35
202A2
NC 83/1
277
E
M†
L
E
S
M
L
E
S
M†
L
E
M
M
M
M
M
–
21n1 22n2
34n7 (11n3) 36n3 (11n7)
25n8 27n1
26n9 28n4
26n4 27n8
6n6 22n4 29n4
5n4 21n6 1n1 28n0
5n9 20n2 1n6 27n7
5n1 24 2n8 31n9
6n1 (0n1) 25n9 (1n6) 32n5 (1n9)
6n3 23n1 2n4 31n8
6n2 20n8 1n5 28n5
5n9 22n2 28n6
6n9 27n2 34n8
6n7 25n2 32n4
4n8 21n5 1n3 27n6
23n5 24n3
29n2 30n1
– 12n3 8n3 38n1 58n7
4n2 (0n0) 7n3 (1n0) 5n4 (5n1) 18n7 (10n3) 35n8 (16n7)
– 12n8 6n7 34n3 53n8
3n0 10n5 7n7 31n2 52n4
2n8 10 8n0 32n1 52n8
3n5 6n3 14n4 23 47n2
– 10n1 9n8 32n1 51n9
– 10n4 11n5 34n6 56n5
– 7n3 13n7 29n9 50n9
2n6 (0n1) 9n5 (0n8) 6n1 (0n3) 24 (4n1) 42n4 (5n3)
– 15n2 4n1 27n1 46n5
– 15n8 3n8 28 47n6
2n7 8n4 8n5 30n7 50n6
3n2 8n5 6n8 25n6 44
3n2 8n8 6n8 27n5 46n2
– 10n2 7n9 32n9 51n5
3n5 11n7 4n1 34n2 53n5
– 16n6 2n9 31n3 50n9
3n8 3n9 3n2 3n1 14n7
6n6 (3n4) 3n6 (1n4) 3n3 (0n1) 1n9 (1n2) 6n2 (3n1) 20n6 (9n2)
3n9 2n8 2n8 1n1 3n4 14
3n6 4n0 2n8 1n0 3n3 14n8
4n2 3n7 2n4 1n1 3n4 14n7
3 6n2 2n3 3n2 15n5
3n5 3n4 2n6 1n3 3n9 14n7
3n1 4n2 1n2 2n7 12n1
3n6 3n6 1n4 2n8 12n1
4n1 (0n9) 4n1 (0n6) 2n5 (1n2) 1n7 (0n2) 5n1 (1n1) 17n4 (4)
7n0 3n2 1n1 1n5 – 12n8
4n9 3n3 3n4 1n5 4 17n1
2n9 4n2 1n8 3 12n8
3n1 3n5 1n2 1n4 3n9 13n1
3n3 3n8 1n4 1n1 3n4 13n1
5n1 3n4 3n7 3n6 16n7
4n9 4 3n6 1n2 4n3 18
6n6 2n9 4n5 1n1 – 15n1
* The fatty acids in bold have been confirmed by GC-MS. † Mean of two replicates ; numbers in parentheses are .
Aphanizomenon strains are listed in Table 3. The Aphanizomenon strains contained 16 : 0 (22n6–36n3 % of total) and the 18 : 3ω3,6,9c (19–38n8 %) as major fatty acids. The 14 : 0 anteiso, 16 : 1ω9c, 16 : 0 anteiso, 17 : 0 10-methyl and 18 : 2ω6,9c fatty acids were found in all Aphanizomenon strains. Nostoc. In Table 4, the fatty acids of hepatotoxic Nostoc sp. 152 (Sivonen et al., 1990) and Nostoc sp. PCC 73102, isolated from a symbiotic association (Rippka et al., 1979), are shown. The Nostoc samples contained 16 : 0 (12n7–18n4 %), 18 : 3ω3,6,9c (18n2– 41n4 %) and 18 : 2ω6,9c (7n5–20n8 %) as the major fatty acids. The fatty acid 16 : 1ω7c was also found in relatively high amounts (8n9–29n3 %) in both Nostoc strains. The free-living Nostoc sp. 152 contained the two isomers 16 : 1ω5c and 16 : 1ω7c in a ratio of approximately 1 : 1, whereas the symbiotic strain PCC 73102 only contained 16 : 1ω7c. The two Nostoc strains were distinguished from the other Nostocales strains (Anabaena, Aphanizomenon, Cylindrospermum and Calothrix) by the absence of 16 : 1ω9c and 16 : 0 anteiso fatty acids. Other Nostocales species. The fatty acid composition of Cylindrospermum stagnale PCC 7417 and Calothrix marchica PCC 7714 is presented in Table 4. The fatty acid profile of Cylindrospermum stagnale PCC 7417 resembled that of the Anabaena and Aphanizomenon strains, except for the unsaturated fatty acids. In Cylindrospermum stagnale the ratio of 18 : 3ω3,6,9c to 18 : 2ω6,9c was approximately 1 : 1, whereas in Anabaena and Aphanizomenon the 18 : 3 acid dominated over 18 : 2 in ratios varying from 2 to 10, indicating the 1010
higher ω3 desaturase activity in these strains. Calothrix marchica differed from the Anabaena, Aphanizomenon, Cylindrospermum and Nostoc strains by the extreme prevalence of 16 : 0 ( 48 % of total), the low content of 18 : 3ω3,6,9c and the presence of 18 : 3ω6,9,12c. The fatty acids of non-toxic Microcystis strain 265 and hepatotoxic Microcystis strain 199 are shown in Table 5. Both contained 16 : 0 as the major fatty acid ( 30 % of total). The two isomers 18 : 3ω6,9,12c and 18 : 3ω3,6,9c were observed in Microcystis strains. A unique signature fatty acid, 18 : 0 iso, was only found in Microcystis and represented 9–25 % of the total.
Microcystis.
The fatty acids of non-toxic Planktothrix strain 2 and hepatotoxic Planktothrix sp. NC128\R are presented in Table 5. The major fatty acids found in Planktothrix cells were 16 : 0 (14n3– 22n7 %), 16 : 1ω7c (15n5–25n4 %) and 18 : 3ω3,6,9c (27– 48n6 %). The fatty acids 14 : 0 anteiso (1n1–6n8 %) and 18 : 2ω6,9c (5n6–13n8 %) were found in noticeable amounts in the two Planktothrix strains. The fatty acid profiles of the two Planktothrix strains were similar. Planktothrix.
According to the fatty acid classification (Kenyon, 1972 ; Kenyon et al., 1972 ; Murata et al., 1992, Cohen et al., 1995), the filamentous strains belonging to the genera Anabaena, Aphanizomenon, Cylindrospermum, Nostoc and Planktothrix can be classified in group 2 on the basis of their content of 18 : 2ω6,9c and 18 : 3ω3,6,9c. Despite the absence of 18 : 4, the two Microcystis strains and Calothrix marchica PCC 7714 can be assigned to group 4. The latter strains contain 18 : International Journal of Systematic and Evolutionary Microbiology 52
Fatty acids of cyanobacteria Table 3. Relative content (%) of fatty acids in non-toxic Aphanizomenon strains (filamentous heterocystous cyanobacteria) under different growth conditions .................................................................................................................................................................................................................................................................................................................
M, 3 l culture ; L, lag phase, E, exponential phase, S, stationary phase from 100 ml culture ; a, P l 0n275 mg l−", N l 1 mg l−" ; b, P l 0n005 mg l−", N l 1 mg l−" ; c, P l 5n5 mg l−", N l 1 mg l−" ; d, P l 0n275 mg l−", N l 0n005 mg l−" ; e, P l 0n275 mg l−", N l 30 mg l−" ; , trace amount (less than 1 %) ; –, not detected. Fatty acid*
202 M
L
PCC 7905 E
Saturated 14 : 0 2 2n5 2n5 16 : 0 25n2 36n3 29n4 18 : 0 2n7 Total 27n5 39n6 34n6 Unsaturated 16 : 1ω7c 4n6 5n7 5n1 16 : 1ω9c 9n5 5n7 7n8 18 : 2ω6,9c 8 4n9 5n8 18 : 3ω3,6,9c 28n2 19 24n7 Total 50n3 35n3 43n4 Branched and hydroxy-substituted 14 : 0 anteiso 4n3 5n3 6n8 16 : 0 anteiso 3n2 1n2 2n4 16 : 0 10-methyl 4n3 3n3 1n0 16 : 0 iso 3-OH 2n1 1n3 17 : 0 10-methyl 3n9 5n8 1n5 Total 16n5 17n7 13
219
271
TR183
M
L
E
M
M†
–†
a‡
b
c
d
e
23n9 25
22n6 2n1 25n6
23n3 2n5 25n9
27n8 28n7
29n0 (0n6) 30n2 (0n7)
27n9 (0n9) 28n9 (1n0)
28n3 (2n4) 29n6 (2n5)
25n8 1n0 27n3
24n9 25n8
27n3 1n8 29n9
27n6 28n6
3 10n5 5n8 32n7 51n9
– 14n8 3n6 38n8 57n2
– 14 3n3 36n9 54n2
3n2 8n2 10n6 30n1 52n1
4n7 (0n2) 6n7 (0n0) 11n7 (0n1) 28n5 (0n7) 51n6 (1)
3n8 (0n2) 9n1 (0n1) 8n1 (0n4) 32n0 (0n3) 53 (0n9)
2n7 (0n1) 9n1 (1n7) 4n7 (0n2) 34n5 (2n5) 51 (4n4)
– 13 4n3 38n8 56n0
– 13n2 4n8 37n8 55n7
2n6 8n7 4n4 33n7 49n4
2n7 9n4 5n1 35n8 53n1
3n6 2n6 3n2 3n6 14
4n3 1n8 1n8 1n2 3n3 12n5
5n7 1n8 1n18 1n5 3n4 13n3
3 2n7 2n8 1 3n8 13n2
3n2 (0n1) 3n0 (0n1) 1n8 (0n6) 1n2 (0n2) 3n9 (0n2) 13n2 (1n1)
3n4 (0n5) 2n9 (0n2) 2n8 (0n1) 1n2 (0n1) 3n1 (1n2) 13n4 (2n2)
4n5 (0n2) 1n9 (0n4) 1n1 (0n8) 1n2 (0n2) 3n4 (1n0) 12 (2n6)
5 2 1n5 2n5 12n1
4n6 2n2 2n9 2n5 12n8
7n3 1n9 – 1n3 2n2 13
3n9 2n1 2 1 2n9 11n9
* The fatty acids in bold have been confirmed by GC-MS. † Mean of two replicates ; numbers in parentheses are . ‡ Mean of five replicates ; numbers in parentheses are .
3ω3,6,9c and 18 : 3ω6,9,12c, indicating that both ω3 and ∆6 desaturases are present. As discussed by Cohen et al. (1995), organisms in which the two desaturases coexist are potentially able to produce 18 : 4ω3. Previously, strains of both genera were shown to contain this 18 : 4 fatty acid (Kenyon et al., 1972 ; Ahlgren et al., 1992 ; Li et al., 1998). Therefore, the desaturation step from 18 : 3 to 18 : 4 might be regulated by temperature in different strains of Microcystis and Calothrix, and could be examined by growing the cells at lower temperatures than those used in this study (for example 15 versus 20 mC). Correspondence analysis of the fatty acid content in cyanobacterial strains
The correspondence analysis was performed on the 60 fatty acids detected in the 60 independent cultures obtained from 22 cyanobacterial strains. This analysis permitted the evaluation of the contribution of the major and minor fatty acids in the characterization of the cyanobacterial strains and the determination on this basis of the taxonomic relationships between these strains. Projection of the samples in the plane defined by the two most informative axes, 1 and 2 (representing 30n7 and 12n3 % of the variance, respectively), showed http://ijs.sgmjournals.org
the existence of the following groups ; Microcystis strains (group A), Nostoc and Planktothrix strains (group B), and Anabaena, Aphanizomenon and Cylindrospermum strains (group C) (Fig. 1a). The variation in the minor fatty acid content among the samples of a strain grown under different culture conditions was less than the variation observed between the different groups of strains (Fig. 1). Fatty acids such as 16 : 0, 18 : 2ω6,9c, 18 : 3ω3,6,9c and 14 : 0 anteiso contained in all strains did not separate the different groups of strains, but 18 : 3ω6,9,12c, 18 : 0 iso, 16 : 1ω5c, 16 : 1ω7c and 16 : 1ω9c were clearly associated with the definition of the major axes of the correspondence analysis and thus can be used as chemotaxonomic markers to discriminate groups of strains or genera (Fig. 1b). According to the more discriminative axis (axis 1), the Microcystis strains (group A) were distinguished from the filamentous strains of groups B and C, and from Calothrix marchica PCC 7714. Fatty acids 18 : 3ω6,9,12c and 18 : 0 iso had the most important contribution to the clustering of Microcystis (Fig. 1a, b) and can be used as chemotaxonomic markers for Microcystis. No difference in fatty acid content was observed between the hepatotoxic and non-toxic Microcystis strains. The filamentous strains formed two separate groups along the second axis. The Nostoc 1011
M. Gugger and others Table 4. Relative content (%) of fatty acids in Nostoc, Cylindrospermum and Calothrix strains (filamentous heterocystous cyanobacteria) under different growth conditions .................................................................................................................................................................................................................................................................................................................
M, 3 l culture ; L, lag phase, E, exponential phase, S, stationary phase from 100 ml culture ; , trace amount (less than 1 %) ; –, not detected. Fatty acid*
Hepatotoxic
Non-toxic
Nostoc sp. 152
Saturated 14 : 0 16 : 0 18 : 0 Total Unsaturated 16 : 1ω5c 16 : 1ω7c 16 : 1ω9c 18 : 1ω5c 18 : 2ω6,9c 18 : 3ω3,6,9c 18 : 3ω6,9,12c Total Branched 12 : 0 anteiso 14 : 0 anteiso 16 : 0 anteiso 16 : 0 10-methyl 17 : 0 10-methyl 17 : 1 iso\anteiso Total
Nostoc sp. PCC 73102
Cylindrospermum stagnale PCC 7417
Calothrix marchica PCC 7714
M†
L
E
S
M
L
E
S
3n4 (0n2) 18n4 (4n0) 22n0 (4n2)
3n1 18n1 21n8
4 14n6 19n1
4n2 12n7 16n9
16 16n9
– 15 2n8 17n8
14n6 1n6 16n5
– 17n2 1n9 19n1
26n2 27n7
48n4 49n2
10n4 (0n1) 10n8 (0n1) – – 20n5 (4n8) 25n5 (2n2) – 67n3 (7n5)
13n1 8n9 – 3n2 9n9 33n8 – 68n7
14 9n5 – 1n1 8n6 37n8 – 71n0
15n5 9n2 – – 7n5 41n4 – 73n5
– 25n7 – – 17n8 28n9 – 72n5
– 18n8 – 9 15n3 18n2 – 61n3
– 29n3 – 4 20n8 22n6 – 76n6
– 26n8 – – 16n6 28n9 – 72n2
4n1 14n2 – 15n2 13n4 – 47n3
– – 9 – 6n6 4n9 4n3 24n8
– 2n2 (0n0) – 2n3 (0n8) 1n8 (1n2) – 6n2 (2)
1n8 1n5 – – 1n7 5n5
1n1 – 1n2 1n6 5n5
1n1 – 1n4 4n8
1n2 2n6 – 1n7 7
4n3 9n6 – – – 5 18n9
1 2n4 – – 6n5
2n4 2n8 – – – 2n5 6
– 3n1 4 2 6 – 15
– 3n6 5n3 2n1 4n6 – 16
* The fatty acids in bold have been confirmed by GC-MS. † Mean of two replicates ; numbers in parentheses are .
and Planktothrix strains (group B) were characterized by 16 : 1ω7c, while Anabaena, Aphanizomenon and Cylindrospermum strains (group C) were strongly associated with the presence of 16 : 1ω9c and 16 : 0 anteiso (Fig. 1a, b). Calothrix marchica PCC 7714 was positioned apart from the groups (Fig. 1a, b). The distinction of Calothrix marchica PCC 7714 from the tight group C of heterocystous cyanobacteria and from Microcystis is based mainly on the respective presence of 18 : 3ω6,9,12c and absence of 18 : 0 iso in the Calothrix strain. Cellular fatty acid composition (Kenyon et al., 1972 ; this study) and molecular data (Lyra et al., 2001) for Calothrix strains support the morphological classification of the genus Calothrix to a separate family from the Nostocaceae (Koma! rek & Anagnostidis, 1989). The high amount of 16 : 1ω7c in the two Nostoc strains was congruent with the high quantity of the same fatty acid reported in Nostoc muscorum (Ahlgren et al., 1992) and in seven other Nostoc strains (Vargas et al., 1998). The free-living Nostoc strain 152 was separated 1012
by the high amount of the signature fatty acid 16 : 1ω5c from Nostoc punctiforme PCC 73102. The fatty acid analysis (this study) and molecular methods (Lyra et al., 1997, 2001) are congruent in differentiating Nostoc strains 152 and PCC 73102, although they are morphologically similar. The presence of 16 : 1ω5c and 16 : 1ω7c, and the absence of 16 : 1ω9c and 16 : 0 anteiso in Nostoc strain 152 allowed the separation of this strain from the other Nostocales strains. The differentiation of planktic Nostoc strain 152 from the planktic Anabaena strains is in agreement with a previous fatty acid study of free-living species of these two genera (Caudales & Wells, 1992). The fatty acid composition of the two Planktothrix strains was similar to the one found in Planktothrix (Oscillatoria) agardhii PCC 6507 (Kenyon et al., 1972) and in the Planktothrix (Oscillatoria) agardhii strains studied by Ahlgren et al. (1992). The fatty acid content of hepatotoxic and non-toxic Planktothrix strains was similar. Surprisingly, the two Planktothrix strains had a fatty acid composition similar to the two Nostoc strains. Although the fatty acids of the Nostoc and International Journal of Systematic and Evolutionary Microbiology 52
Fatty acids of cyanobacteria Table 5. Relative content (%) of fatty acids in Microcystis strains (unicellular cyanobacteria) and Planktothrix strains (filamentous cyanobacteria) under different growth conditions .................................................................................................................................................................................................................................................................................................................
M, 3 l culture ; L, lag phase, E, exponential phase, S, stationary phase from 100 ml culture ; , trace amount (less than 1 %) ; –, not detected. Fatty acid*
Hepatotoxic
Non-toxic
Non-toxic
Hepatotoxic
Microcystis sp. 199
Microcystis sp. 265
Planktothrix sp. 2
Planktothrix sp. NC128/R
M Saturated 16 : 0 36n7 18 : 0 Total 37n2 Unsaturated 16 : 1ω7c 2n7 16 : 1ω9c 18 : 2ω6,9c 3n7 18 : 3ω3,6,9c 6n1 18 : 3ω6,9,12c 21n6 Total 36n6 Branched and alcohol 12 : 0 anteiso – 14 : 0 anteiso 4n1 16 : 0 10-methyl 3n1 16 : 0 N alcohol – 17 : 0 10-methyl 3n2 17 : 1 iso\anteiso – 18 : 0 iso 9n3 Total 20n1
–†
–†
L
E
M†
38n8 (1n8) 1n2 (0n5) 39n9 (2n3)
33n5 (3n2) 34n4 (3n4)
31n9 1n2 33n1
34n4 3n9 38n3
22n7 (0n8) 1 (0) 23n7 (0n8)
2n5 (0n1) 3n3 (0n8) 5n2 (1n2) 19n6 (0n6) 33n2 (3)
2 (0n2) 7n4 (0n2) 14n5 (1n2) 8n3 (1) 34n3 (3n1)
2 1n2 3n5 18n5 5n7 32n1
2n4 1n6 3n1 15n2 4n4 27n5
– 4n1 (0n1) 1n1 (0n7) – 3n4 (1) – 10n9 (4n3) 20n2 (6n7)
– 3n6 (0n7) 3n4 (0n4) – 3 (0n8) – 16n7 (3n8) 26n8 (5n8)
1n3 1n5 1n1 1n3 – 24n7 29n5
1n6 7n1 – 1n9 1n4 16n9 29n7
L
E
S
M
L
E
17 – 17
17n2 – 17n2
18n8 1n4 20n2
22 22n6
20 8n3 28n3
14n3 14n8
20n3 (1n7) 13n8 (0n7) 27 (2n7) – 61n6 (5n3)
21n5 – 8n1 32 1n4 62n9
25n4 – 9n8 39n9 – 75n1
22n8 – 10n8 37 – 71n6
19n7 – 12n2 36 68n1
15n5 – 5n7 32n1 – 54
23n1 – 5n6 48n6 – 78n5
– 6n8 (5n8) – 3n2 (0n1) – – 11n2 (5n7)
2n1 3n8 – 5n6 4 3n2 – 13n1
2n2 2n8 – – – 1n9 – 6n9
1n1 1n3 – 1n9 – 6n1
– 1n8 1n7 2 – – 6
2n4 3n7 – 6n4 – 2n2 – 8n3
– 1n1 1n2 – 1n3 – 5n5
* The fatty acids in bold have been confirmed by GC-MS. † Mean of two replicates ; numbers in parentheses are .
.....................................................................................................
Group C
Group A
(a)
Group B
(b)
Planktothrix strains were similar, the morphology, the whole-cell protein pattern and the genetic data (Lyra et al., 1997, 2001) clearly separated these strains. http://ijs.sgmjournals.org
Fig. 1. Correspondence analysis of wholecell fatty acids from 60 independent cultures of 22 cyanobacterial strains belonging to the genera Anabaena, Aphanizomenon, Calothrix, Cylindrospermum, Microcystis, Nostoc and Planktothrix. (a) Projection of the 66 cyanobacterial cultures in the plane defined by the two main axes of the correspondence analysis. The inertia of axes 1 and 2 are 30n67 and 12n30 %, respectively. The independent cultures are indicated by the strain number and the growth conditions at the time of fatty acid extraction (see Tables 2–5). The star representation binds each individual sample to the centre of its category (genus). Symbols : a, Anabaena ; ap, Aphanizomenon ; m, Microcystis; n, Nostoc ; p, Planktothrix. (b) Projection of the 60 cellular fatty acids detected from the cyanobacterial cultures in the plane defined by the same two main axes of the correspondence analysis. The major fatty acids detected are indicated.
To obtain a better representation of the strains in group C (Fig. 1), the dataset of Anabaena, Aphanizomenon and Cylindrospermum strains was reinvestigated 1013
M. Gugger and others
.................................................................................................................................................
Fig. 2. Correspondence analysis of whole-cell fatty acids from 42 independent cultures of the filamentous heterocystous Anabaena, Aphanizomenon and Cylindrospermum genera. Projection of the 42 cyanobacterial cultures in the plane defined by the two main axes of the correspondence analysis. The inertia of axes 1 and 2 were 23 and 12 %, respectively. The independent cultures are indicated by the strain numbers and the growth conditions at the time of fatty acid extraction (see Tables 2, 3 and 4). The star representation binds each individual sample to the centre of its category. Symbols : an, neurotoxic Anabaena ; ah, hepatotoxic Anabaena ; ap, Aphanizomenon.
by correspondence analysis (Fig. 2). According to the more discriminative axis (axis 1), the hepatotoxic Anabaena strains were distinguished from the neurotoxic and non-toxic Anabaena. This distinction was due mainly to the relative proportion of the fatty acid 14 : 0 within the hepatotoxic Anabaena samples 66A, 90, 202A1, 202A1\35, 202A2 and NC83\1. The neurotoxic and non-toxic Anabaena strains were not separated from the non-toxic Aphanizomenon strains. Cylindrospermum stagnale PCC 7417 was distinguished from the Anabaena and Aphanizomenon strains. The distribution of the samples along axis 2 demonstrated the effect of the culture volume on fatty acid composition. The fatty acid composition of cyanobacterial biomass grown in 3 l cultures was separated from the samples grown in 100 ml cultures. The fatty acid compositions of Aphanizomenon strain TR183 grown with different nitrate and phosphate concentrations were closely grouped. Although the fatty acid content of differently cultured Aphanizomenon cells varied, the variation did not affect their taxonomic grouping. The hepatotoxic Anabaena strainsweredistinguishablefromtheneurotoxicandnontoxic Anabaena, Aphanizomenon and Cylindrospermum strains. This reveals the robustness of the fatty acids as chemotaxonomic markers. The genera Anabaena and Aphanizomenon are morphologically different (Koma! rek & Anagnostidis, 1989), but genetically similar (Lyra et al., 1997, 2001). The grouping of Anabaena and Aphanizomenon strains 1014
based on whole-cell fatty acid composition disagreed with their morphological classification, but was coherent with the whole-cell protein pattern, wholegenome fingerprinting (REP and ERIC) and RFLP, and sequencing of the 16S rRNA gene (Lyra et al., 1997, 2001). Congruently, a study by Li et al. (1998) based on non-polar and 3-hydroxy fatty acid composition also showed the clustering of 13 planktic Anabaena strains and a strain of Aphanizomenon flosaquae f. gracile. The fatty acid content of the 22 studied strains permitted classification of the strains into groups 2 and 4, according to the Kenyon–Murata classification system. The correspondence analysis of the whole-cell fatty acids classified these strains into three groups and demonstrated that certain fatty acids were useful chemotaxonomic markers for cyanobacteria, such as 14 : 0 for the hepatotoxic Anabaena strains, 16 : 0 for Calothrix strain, 16 : 1ω5c and 16 : 1ω7c for the Nostoc and Planktothrix strains, 16 : 1ω9c for Anabaena and Aphanizomenon strains, and 18 : 0 iso and 18 : 3ω6,9,12c for Microcystis strains. We have shown that the taxonomic differences remained even though the growth conditions were varied. We found that fatty acid composition is a useful tool to distinguish strains within the filamentous cyanobacteria and to separate strains within the Nostocales genera. The separation between Anabaena and Nostoc strains, depending on the presence or absence of fatty acids with 16 carbons (16 : 0 anteiso, 16 : 1ω5c), is congruent with the morphological and genetic data (Lyra et al., 1997, 2001). Based on the presence of 14 : 0 in hepatotoxic Anabaena and its absence in neurotoxic and non-toxic strains, we outlined a clear dichotomy among Anabaena strains. These results are incongruent with the morphology of the strains but are in agreement with the genetic data of these Anabaena strains (Lyra et al., 1997, 2001). We have confirmed that the analysis of fatty acid content could be used to complement other approaches to establish a polyphasic classification of cyanobacteria. The results obtained for group B (Nostoc\Planktothrix) are incongruent with the morphologic and genetic data, while the results for group C (Anabaena\ Aphanizomenon) are in agreement with molecular phylogeny (Lyra et al., 2001), in contrast to morphological classification. ACKNOWLEDGEMENTS This investigation was supported by grants from Helsinki University, from the Centre for International Mobility (CIMO) and the Academy of Finland. We thank Matti Wahlsten for technical assistance and Raimo Mikkola for expertise at the GC-MS. We are grateful to Dr Peter Henriksen for kindly providing two Danish Aphanizomenon strains and to Dr Sari Repka for her help.
REFERENCES Ahlgren, G., Gustafsson, I.-B. & Boberg, M. (1992). Fatty acid content and chemical composition of freshwater microalgae. J Phycol 28, 37–50. International Journal of Systematic and Evolutionary Microbiology 52
Fatty acids of cyanobacteria Castenholz, R. W. (2001). Oxygenic photosynthetic bacteria. In Bergey’s Manual of Systematic Bacteriology, 2nd edn, vol. 1, pp. 473–600. Edited by D. R. Boone & R. W. Castenholz. New York : Springer. Caudales, R. & Wells, J. M. (1992). Differentiation of free-living Anabaena and Nostoc cyanobacteria on the basis of fatty acid composition. Int J Syst Bacteriol 42, 246–251. Caudales, R., Wells, J. M. & Butterfield, J. E. (2000). Cellular fatty acid composition of cyanobacteria assigned to subsection II, order Pleurocapsales. Int J Syst Evol Microbiol 50, 1029–1034. Cohen, Z., Margheri, M. C. & Tomaselli, L. (1995). Chemotaxonomy of cyanobacteria. Phytochemistry 40, 1155–1158. Holton, R. W., Blecker, H. H. & Stevens, T. S. (1968). Fatty acids in blue-green algae : possible relation to phylogenetic position. Science 160, 545–547. Kenyon, C. N. (1972). Fatty acid composition of unicellular strains of blue-green algae. J Bacteriol 109, 827–834. Kenyon, C. N. & Stanier, R. Y. (1970). Possible evolutionary significance of polyunsaturated fatty acids in blue-green algae. Nature 227, 1164–1166. Kenyon, C. N., Rippka, R. & Stanier, R. Y. (1972). Fatty acid composition and physiological properties of some filamentous bluegreen algae. Arch Microbiol 83, 216–236. Koma! rek, J. & Anagnostidis, K. (1989). Modern approach to the classification system of Cyanophytes 4 – Nostocales. Arch Hydrobiol Suppl 82, 247–345. Kotai, J. (1972). Instructions for Preparation of Modified Nutrient Solution Z8 for Algae. Publication B-11\69. Blindern, Oslo : Norwegian Institute for Water Research. Legendre, P. & Legendre, L. (1998). Numerical ecology. In Developments in Environmental Modeling, 2nd edn, pp. 451–476. Amsterdam : Elsevier. Lehtima$ ki, J., Lyra, C., Suomalainen, S., Sundman, P., Rouhiainen, L., Paulin, L., Salkinoja-Salonen, M. & Sivonen, K. (2000). Characterization of Nodularia strains, cyanobacteria from brackish waters, by genotypic and phenotypic methods. Int J Syst Evol Microbiol 50, 1043–1053. Li, R., Yokota, A., Sugiyama, J., Watanabe, M., Hiroki, M. & Watanabe, M. (1998). Chemotaxonomy of planktonic cyanobacteria based on non-polar and 3-hydroxy fatty acid composition. Phycol Res 46, 21–28. Lyra, C., Hantula, J., Vainio, E., Rapala, J., Rouhiainen, L. & Sivonen, K. (1997). Characterization of cyanobacteria by SDS-PAGE of whole-cell proteins and PCR\RFLP of the 16S rRNA gene. Arch Microbiol 168, 176–184. Lyra, C., Suomalainen, S., Gugger, M., Ve! zie, C., Sundman, P., Paulin, L. & Sivonen, K. (2001). Molecular characterization of planktic cyanobacteria of Anabaena, Aphanizomenon, Microcystis and Planktothrix genera. Int J Syst Evol Microbiol 51, 513–526. Murata, N. & Wada, H. (1995). Acyl-lipid desaturases and their importance in the tolerance and acclimatization to cold of cyanobacteria. Biochem J 308, 1–8. Murata, N., Wada, H. & Gombos, Z. (1992). Modes of fatty-acid desaturation in cyanobacteria. Plant Cell Physiol 33, 933–941. Nishida, I. & Murata, N. (1996). Chilling sensitivity in plants and cyanobacteria : the crucial contribution of membrane lipids. Annu Rev Plant Physiol Plant Mol Biol 47, 541–568. Reddy, A. S., Nuccio, M. L., Gross, L. M. & Thomas, T. L. (1993). Isolation of a Delta super(6)-desaturase gene from the cyanobacterium Synechocystis sp. strain PCC 6803 by gain-of-function expression in Anabaena sp. strain PCC 7120. Plant Mol Biol 22, 293–300. Rippka, R., Deruelles, J., Waterbury, J. B., Herdman, M. & Stanier, R. Y. (1979). Generic assignments, strain histories and properties of pure cultures of cyanobacteria. J Gen Microbiol 111, 1–61.
http://ijs.sgmjournals.org
Rouhiainen, L., Sivonen, K., Buikema, W. J. & Haselkorn, R. (1995). Characterization of toxin-producing cyanobacteria by using an oligonucleotide probe containing a tandemly repeated heptamer. J Bacteriol 177, 6021–6026. Sakamoto, T. & Bryant, D. A. (1997). Temperature-regulated mRNA accumulation and stabilization for fatty acid desaturase genes in the cyanobacterium Synechococcus sp. strain PCC 7002. Mol Microbiol 23, 1281–1292. Sakamoto, T., Los, D. A., Higashi, S., Wada, H., Nishida, I., Ohmori, M. & Murata, N. (1994a). Cloning of ω3 desaturase from cyanobacteria and its use in altering the degree of membrane-lipid unsaturation. Plant Mol Biol 26, 249–263. Sakamoto, T., Wada, H., Nishida, I., Ohmori, M. & Murata, N. (1994b). ∆9 acyl-lipid desaturases of cyanobacteria : molecular cloning and substrate specificities in terms of fatty-acids sn-position and polar head groups. J Biol Chem 269, 25576–25580. Sakamoto, T., Wada, H., Nishida, I., Ohmori, M. & Murata, N. (1994c). Identification of conserved domains in the ∆12 desaturases of cyanobacteria. Plant Mol Biol 24, 643–650. Sakamoto, T., Shen, G., Higashi, S., Murata, N. & Bryant, D. A. (1998). Alteration of low-temperature susceptibility of the cyanobacterium Synechococcus sp. PCC 7002 by genetic manipulation of membrane lipid unsaturation. Arch Microbiol 169, 20–28. Sivonen, K. (1990). Effects of light, temperature, nitrate, orthophosphate, and bacteria on growth of and hepatotoxin production by Oscillatoria agardhii strains. Appl Environ Microbiol 56, 2658–2666. Sivonen, K., Himberg, K., Luukkainen, R., Niemela$ , S. I., Poon, G. K. & Codd, G. A. (1989). Preliminary characterization of neurotoxic cyanobacteria blooms and strains from Finland. Toxic Assess 4, 339–352. Sivonen, K., Carmichael, W. W., Namikoshi, M., Rinehart, K., Dahlem, A. M. & Niemela$ , S. (1990). Isolation and characterization of hepatotoxic microcystin homologs from the filamentous freshwater cyanobacterium Nostoc sp. strain 152. Appl Environ Microbiol 56, 2650–2657. Sivonen, K., Namikoshi, M., Evans, W. R., Carmichael, W. W., Sun, F., Rouhiainen, L., Luukkainen, R. & Rinehart, K. L. (1992). Isolation and characterization of a variety of microcystins from seven strains of the cyanobacterial genus Anabaena. Appl Environ Microbiol 58, 2495– 2500. Thioulouse, J., Chessel, D., Dole! dec, S. & Olivier, J.-M. (1997). ADE-4 : a multivariate analysis and graphical display software. Stat Comput 7, 75–83. Tsitko, I. V., Zaitsev, G. M., Lobanok, A. G. & Salkinoja-Salonen, M. S. (1999). Effect of aromatic compounds on cellular fatty acids composition of Rhodococcus opacus. Appl Environ Microbiol 65, 853–855. Va$ isa$ nen, O. M., Nurmiaho-Lassila, E.-L., Marmo, S. A. & Salkinoja-Salonen, M. S. (1994). Structure and composition of biological slimes on paper and board machines. Appl Environ Microbiol 60, 641–653. Vandamme, P., Pot, B., Gillis, P., De Vos, P., Kersters, K. & Swings, J. (1996). Polyphasic taxonomy, a consensus approach to bacterial systematics. Microbiol Rev 60, 407–438. Vargas, M. A., Rodriguez, H., Moreno, J., Olivares, H., Del Campo, J. A., Rivas, J. & Guerrero, M. G. (1998). Biochemical composition and fatty acid content of filamentous nitrogen-fixing cyanobacteria. J Phycol 34, 812–817. Wilmotte, A. (1994). Molecular evolution and taxonomy of the cyanobacteria. In The Molecular Biology of Cyanobacteria, pp. 1–25. Edited by D. A. Bryant. Dordrecht : Kluwer. Wilmotte, A. & Herdman, M. (2001). Phylogenetic relationships among the cyanobacteria based on 16S rRNA sequences. In Bergey’s Manual of Systematic Bacteriology, 2nd edn, vol. 1, pp. 487–492. Edited by D. R. Boone & R. W. Castenholz. New York : Springer.
1015
