Jérôme FEIGE Biochemistry Department

Graduation Internship Thesis

Functional Genomics Nutrient transport in

Escherichia coli

2 Functional characterization of the MetD methionine transporter and phylogenetic analysis of its homologs 2 Frx, a PTS transporter involved in melibiose metabolism

Internship in Dr Milton Saier’s laboratory University of California, San Diego

March – August 2002

Biochemistry Department

Acknowledgements

I am particularly grateful to Dr Saier to have welcomed me in his lab, making therefore this stay abroad possible. These 6 months in San Diego have been a very enriching experience concerning both scientific and cultural matters. His kindness and availability and his enthusiasm for science are a great source of motivation. I would also like to thank Dr Graciela Lorca and Dr Zhongge Zhang for their help, support and explanations. Their devotion to their work is a great example for the students in the lab. All the graduate and undergraduate students who work in the lab were also of great help and assistance and contribute to the liveliness of the lab. Above all, I wish to thank Abe Chang for his help with the phylogenetic analyses of the MUT (Methionine Uptake Transporter) family. I am also grateful to Iain Anderson (Integrated Genomics, Chicago) for our collaboration on the methionine transporter project. Valuable discussions have been of great help. Finally I wish to thank Agnès Rodrigue for being my INSA tutor and for following my work. This work was supported by NIH grants GM64368 and GM55434 and by a Région RhôneAlpes grant for studies abroad.

Functional genomics in E. coli

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

Table of Contents Saier Laboratory presentation…..……..……….………...…………...……….……A Introduction……………………..…………………………………...………………..1 Common Methods……..………..…………………………………...………………..2 Gene inactivation……………………………………………..…………………………...2 Transport assay...……………………………………………..…………………………...3

The MetD Methionine Transporter: Functional and Phylogenetic Characterization.5 Introduction…………………………………………………..…………………………...5 Material & Methods..…………….…………………………..…………………………...7 - Experimental methods (p. 7) - Computer methods (p. 8) Results..…………….…………………………..………………………………..………...9 - Growth studies (p. 9) - Transport studies (p. 10) - Inhibition studies (p. 11) - Phylogenetic studies (p. 12) Discussion………….…………………………..………………………………..…….....17

Frx, a PTS transporter involved in melibiose metabolism……………..………...21 Introduction…………………………………………………..………………………….21 Material & Methods..…………….…………………………..………………………….22 Results..…………….…………………………..………………………….……..………24 Discussion………….…………………………..……………………….………..………27

Conclusion……………………………………………………………………………29 References………………………….………………………………………………...30 Functional genomics in E. coli

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

Saier Laboratory Presentation Dr Milton Saier’s laboratory is part of the Division of Biological Sciences at the University of California, San Diego (UCSD). Founded four decades ago, UCSD has risen rapidly to its status as one of the top institutions in the United States for education and scientific exploration. UCSD is an active research center in biology and medical sciences and plays an important role in the region's biotechnological industry. Dr Saier’s laboratory has two major research interests which are evolution of transport proteins through phylogenetic analysis and transcriptional and metabolic regulation in bacteria. Although the methods of study are different, these themes are related as transport systems influence major regulatory pathways in bacteria. Dr Saier has proposed a comprehensive classification of membrane transporters based both on function and phylogeny [54, 55]. The Transport Commision (TC) system proposed and approved by the IUBMB* is inspired by the Enzyme Commision (EC) 4-digit classification system for enzymes. However, the TC system takes into account phylogenetic data which complements the functional information with structural and mechanistic features whereas the EC system is based only Figure A. Scheme illustrating the currently recognized primary types of transporters found in nature. These proteins are initially divided into channels and carriers. Channels are subdivided into α-helical protein channels, β-barrel protein porins (mostly in the outer membranes of gram-negative bacteria and eukaryotic organelles), toxin channels, and peptide channels. Carriers are subdivided into primary active carriers, secondary active carriers (including uniporters), and group translocators that modify their substrates during transport. Primary sources of chemical energy that can be coupled to transport include pyrophosphate bond (i.e., ATP) hydrolysis, decarboxylation, and methyl transfer. Oxidation-reduction reactions, light absorption, and mechanical devices can also be coupled to transport (see text). Secondary active transport is driven by ion and other solute (electro)chemical gradients created by primary active transport systems. The only well-established group-translocating system found in nature is the bacterial phosphoenolpyruvate:sugar PTS, which phosphorylates its sugar substrates during transport. Reproduced from [54].

Transporters

*

International Union of Biochemistry and Molecular Biology

Functional genomics in E. coli

A

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

on function. Transport systems are classified on the basis of five criteria, and each of these criteria corresponds to one of the five numbers or letters within the TC number for a particular type of transporter. Thus a TC number normally has five components as follows: V.W.X.Y.Z. V corresponds to the transporter class (figure A) ; W corresponds to the transporter subclass which in the case of primary active transporters refers to the energy source used to drive transport ; X corresponds to the transporter family or superfamily ; Y corresponds to the subfamily in which the transporter is found, and Z corresponds to the substrate or range of substrates transported. A TC data-base was designed in the lab to index functionally characterized transporters (available online at http://tcdb.ucsd.edu/tcdb/background.php). An important effort of the bioinformatics group is therefore to maintain and update both the structure and the content of this database. Specific software has been developed or adjusted in order to perform analyses directly from the website; such software includes search tools, TC-blast to blast a protein or a nucleotide sequence against the TC-database and TC-tree to generate an alignment of a TC family. A new program has recently been developed in order to classify a list of membrane proteins within TC families or sub-families ; each protein is blasted against a representative member of the putative families to which it can belong and is asigned to the family for which it has the highest similarity. Another part of in-silico studies of the lab is the analysis of transport families or superfamilies in order to integrate data from all the available sequenced genomes to construct phylogenies, understand transport protein evolution and infer putative function by relating the results to functionally and in some cases structurally characterized transporters. Current studies concern the Multi-drug/Oligosaccharidyl-lipid/Polysaccharide Flippase superfamily (MOP - 2.A.66), the Voltage-gated Ion Channel superfamily (VIC - 1.A.1), the Large Conductance Mechanosensitive Ion Channel Family (MscL – 1.A.22) and the Tripartite Tricarboxylate Transporter family (TTT – 2.A.80 - [72]). For these studies, specific software concerning transmembrane segment (TMS) analysis, sequence alignments and phylogenetic trees is also developed depending on needs and are available at http://saier-144-37.ucsd.edu/biotools/. Recent efforts concerned developing a program which combines TMS prediction with multiple alignments and several other TMS handling tools [78]. ScreenTransporter (ST) is an automated program which screens non-redundant databases for potential members of recognized transporter families that has also been lately developed [77]. The Grasp-DNA program (available at http://www2.genomatica.com/grasp-dna/) has also been developed in the laboratory in order to identify putative DNA-protein binding sites and repeat motifs in completely sequenced genomes [60]. Computational and in-vitro analysis of secretory Functional genomics in E. coli

B

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

pathways of Pseudomonas aeruginosa are also performed in collaboration with the Dow Chemical Company [44]. The global interest of the “wet” lab is the characterization of the function and the mechanisms of transport and metabolic and transcriptional regulations in both Gram-positive and Gram-negative bacteria. The main bacterial models are Escherichia coli and Salmonella typhimurium for Gram-negatives and Bacillus subtilis and Lactobacillus brevis for Gram-positives. The regulatory studies concern different PhosphoTransferase Systems (PTS) and catabolite repression and its implication in carbon, nitrogen, sulfur and phosphorus metabolisms. The PTS was first described in 1964 [41] and is primarily responsible for the uptake of certain sugars and their phosphorylation in position 6 during the cytoplasmic transmembrane transport. The phosphate added has been successively transfered by different enzymes that play regulatory roles and was ultimately generated during the transformation of PhosphoEnolPyruvate (PEP) to pyruvate (figure B). PhosphoEnol Pyruvate

Enzyme I (His-P)

HPr

Enzyme IIA (His-P)

(His-P)

Enzyme IIB (His-P)

Sugar-6P Enz IIC

Pyruvate

Enzyme I (His)

Enzyme IIA (His)

HPr

(His)

Enzyme IIB (His)

Sugar

Figure B – The major PTS pathway

Enzyme I and HPr are cytosolic energy-coupling proteins whereas enzymes IIA, IIB and IIC are sugar-specific membrane proteins or protein domains that form the transport system. There are various possible combinations of genetic organization of the encoding genes of enzymes IIA, IIB and IIC that can result in different polypeptidic structures (ie a specific protein for each enzyme, two proteins containing respectively one and two enzymes or a single polypeptide for all three). For most PTS systems, enzyme IIC is the actual permease which transports the sugar across the membrane; however, PTS systems of the mannose family also exhibit an enzyme IID that is part of the permease. The PTS is also involved in a variety of secondary functions including metabolic and transcriptional regulation. It therefore plays an important role in different nutrient metabolisms. Catabolite repression in carbon metabolism has been extensively characterized, and the mechanisms involved differ in phylogenetically distant bacteria. In gram-negative enterobacteria such as E. coli, Functional genomics in E. coli

C

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

catabolite repression is mediated by the cAMP Receptor Protein (CRP) which is a transcriptional activator when it binds cAMP (figure Ca). The presence of glucose or any other catabolite-repressing sugar represses many genes involved in metabolism of other carbon sources by lowering the concentration of cAMP. Levels of cAMP are controlled by adenylate cyclase which is allosterically activated by the histidyl-phosphorylated form of enzyme IIA glucose. The phosphorylated form of enzyme IIA is present in maximum concentrations when the PTS is not transporting sugars. On the other hand, catabolite repression in gram-positive bacteria such as B. subtilis is mediated by the transcription repressor CcpA activated by the binding of the phosphorylated form of HPr (figure Cb).

Figure C. Models illustrating the proposed phosphotransferase system (PTS)-mediated control of catabolite repression in (a) E. coli and (b) B. subtilis. In (a) adenylate cyclase is believed to be allosterically activated by the histidylphosphorylated form of the IIAglc protein of the PTS, but it is inactive in the absence of this phosphorylated protein. When active, it synthesizes cyclic AMP (cAMP), which binds to the cAMP receptor protein (CRP). The cAMP-CRP complex binds directly to DNA to activate transcription. Catabolite repression results when glucose in the medium causes dephosphorylation of IIAglc-P, thereby deactivating adenylate cyclase, causing the cytoplasmic cAMP concentration to diminish, and thus causing the cAMP-CRP complex to dissociate from DNA (catabolite repression). In (b) the HPr kinase is activated when intracellular metabolites, such as fructose 1,6-biphosphate (FBP) or gluconate 6-phosphate, accumulate, resulting in the phosphorylation of HPr on Ser46. Phosphorylated HPr (HPr-P) then may form a ternary complex with the CcpA DNA-binding protein and FBP. This ternary complex then binds to CREs (Catabolite Repression Elements) in the regulatory regions of catabolite-sensitive operons. Catabolite repression is presumed to be the direct consequence of HPr(Ser-P)-CcpA-FBP binding to the CRE. Reproduced from [56]. Functional genomics in E. coli

D

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

Intracellular glycolysis or pentose phosphate sugar metabolites such as fructose 1,6-biphosphate and gluconate 6-phosphate can activate a PTS-kinase which can phosphorylate HPr on serine46 and therefore activate the repression of CcpA regulated genes. A detailed comparison of catabolite repression and of regulation of sugar transport in gram-positive and gram-negative bacteria is available in [56]. Connections between nitrogen and carbon utilization were described in 1969 [20] but the mechanisms are just starting to be elucidated. A nitrogen-related enzyme IIA was identified within the rpoN operon in E. coli which also contains an HPr-like protein named NPr [49, 52]. An enzyme Initrogen is also found elsewhere in the genome [51] and these proteins are believed to coordinate nitrogen and carbon metabolisms. Links between carbon and sulfur metabolisms have also been demonstrated by inactivation of the sulfur metabolism transcriptional activator CysB [50]. A recent computational analysis of PTS related systems in E. coli has identified several new candidates for which no functional data are available [65]. The laboratory has therefore been very active in the domain of functional genomics in order to try to characterize these putative PTS systems. One of the projects has led to the identification of an ascorbate transporter [76] with no significant sequence similarity to other characterized PTS systems. Several others are under current investigation. During my internship, I have worked on the frx gene which is homologous to an enzyme IIABC gene; further information will be provided in this thesis. Global regulatory studies try to take an overall view of the function of metabolic transcription regulators such as CRP. However inactivation of these major regulators has wide range effects which can be difficult to analyse by phenotypic data. Transcriptome analyses should therefore allow the identification of the target genes of the regulatory proteins of these systems and provide new functional clues. Biochemical analyses of the PTS in E. coli are also performed in the laboratory. Recent evidence has shown that most of the PTS enzymes II also possess a small cytosoluble fraction [1]. Current investigation concerns the dependency of the PTS system on the membrane phospholipid composition; activity of the different PTS enzymes are examined in several mutants for phospholipid biosynthesis [2]. Studies of the PTS regulatory effects in gram-positive bacteria are also under investigation. Catabolite repression in B. subtilis is mediated by the two transcription factors CcpA and CcpB, the latter which has been identified in the laboratory [16]. Recent transcriptome analysis has allowed the identification of several targets to these transcription factors, and these results are currently being confirmed by measuring transcriptional levels and by immuno-precipitation of the proteins and their Functional genomics in E. coli

E

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

DNA binding sites. The genes responsible for galactose uptake and its regulation by the PTS in Lactobacillus brevis have been cloned [24] and another project which I was involved in was the construction of the knock-out mutants for these genes. This project has not yet generated substantial results therefore it will not be discussed in this thesis. After bioinformatic analyses of transporters of Borrelia burgdorferi and Treponema pallidum following the release of the complete genome of these two species [57], transport and regulation studies have also been undertaken in spirochetes to characterize the PTS regulatory scheme. Functional genomics of transport proteins in bacteria has become a major interest of the laboratory with the availability of entire genomes and facilities for phylogenetic and bioinformatic analyses within the lab. Recent studies have led to the identification of several amino-acid transporters: an L-aspartate transporter with broad specificity in B. subtilis [73] and an L- and Dmethionine transporter in E. coli for which I did the biochemical analyses which will be presented further on. After computational analyses of multidrug resistance pumps [18], functional genomic studies of these exporters have been undertaken. I have also been associated with a project which aimed to characterize the substrates of putative efflux transporters (PET) in E. coli following the identification of a novel PET ubiquitous family [31]. No data about this project are available because knock-out mutants will be tested by wide-scale drug discovery processes. Experiments providing clues to possible substrates given by sequence homology to functionally characterized transporters have so far given negative results.

Functional genomics in E. coli

F

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

Introduction In the past years, the availability of whole-sequenced genomes has considerably modified research approaches in biology. The development of bioinformatic tools has allowed one to predict Open-Reading Frames (ORFs) corresponding to genes which have been annotated according to their similarity to functionally characterized homologs. However, some genes have too little similarity for function to be inferred. Moreover, functions are inferred from similarities to related homologs which do not usually show a sufficient degree of identity to a characterized gene to establish a precise function or the substrate specificity in the case of an enzyme-coding gene. The challenge of functional genomics is therefore to determine the functions of these genes by experimental methods. Escherichia coli, the best characterized living organism at a molecular level, is known to possess about 4400 genes. Out of these, 1924 (44%) are functionally characterized and 1199 (27%) have a putative function assigned. Solving the functions of the genes for which no experimental data are available will certainly yield interesting theoretical explanations to poorly understood mechanisms. Moreover, pharmaceutical and biotechnological applications will surely be derived from those elucidations. As previously mentioned, the main interest of Dr Saier’s laboratory is the study of transport and metabolism regulation in bacteria. Functional genomic studies of putative PTS transporters have been undertaken in E. coli following the description of putative PTS systems [65]. One of my projects during this internship was to study the function of the frx gene homologous to an enzyme IIABC fructose gene. I was also involved in the molecular characterization of the MetD transporter in E. coli and the phylogenetic analysis of its homologs. This project was initiated by a bioinformatic analysis performed by Iain Anderson and coworkers from Integrated Genomics. These two main projects are reported in this thesis preceded by a description of the common methods used for gene inactivation and transport function characterization.

Functional genomics in E. coli

1

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

Common Methods All studies were conducted in the genetic background of E. coli strain BW25113 (lacIq rrnBT14 ∆lacZWJ16 hsdR514 ∆araBADAH33 ∆rhaBADLD78). Bacteria were cultured in Luria-Bertani (LB) broth or M9 minimal medium at 37ºC [59]. When appropriate, ampicillin and kanamycin were added to the medium at 100 and 25 µg/ml, respectively. Unless otherwise stated, chemicals were purchased from Sigma-Aldrich.

Gene Inactivation Deletion mutants were generated using the methods described by Datsenko and Wanner [23] and summarized in figure 1. Genotypes of the plasmids used are detailed in table 1. To

prepare

competent

cells

for

Step 1 – PCR amplify FRT-flanked kanamycin resistance gene from pKD24

transformation, BW25113 containing pKD46,

H1

KanR gene

P1

a temperature sensitive plasmid encoding the

FRT P2

FRT

H2

λ Red recombinase under the arabinose Step 2 – Transform E. coli BW25113 expressing λ Red recombinase (pKD46)

promoter, was cultured at 30ºC in SOB broth [59]

containing

ampicillin

and

H1

1 mM

promoter

Gene B

Gene A H2

arabinose. When the OD600 reached 0.5, the culture was centrifuged at 8000 rpm for 5

Step 3 – Select kanamycin-resistant transformants

min, and the cells were washed three times

P3

K2

KanR gene

FRT

promoter

with

cold

10%

glycerol

before

being

resuspended in a minimal volume of 10%

promoter

competent cells were stored at -80ºC prior to kanamycin

gene

flanked

with

Flp

Recombination Targets (FRT) from pKD4; P1 and P2 refer to the parts of the primers which

P4

Step 4 – Eliminate kanamycin cassette using an Flp expression plasmid (pCP20)

glycerol (1% of the original culture). The use. PCR methods were used to clone the

Gene B Kt

FRT

Gene B FRT

Fig. 1 – Strategy for gene disruption. The arrows indicate different PCR primers and H1 and H2 refer to homologous regions. FRT and KanR are the abbreviations for Flp Recombination Target and kanamycin resistance, respectively.

hybridize on the kanamycin gene and H1 and H2 refer to the overhanging sequences homologous to the beginning and the end of the genetic region to knock-out. The PCR products were purified using a Qiagen kit, treated with DpnI to eliminate methylated template DNA, and re-purified by Functional genomics in E. coli

2

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

electrophoresis. The kanamycin gene was transformed into BW25113 competent cells containing pKD46 by electroporation (Gene Pulser, pulse controller at 200 Ω, capacitance at 250 µFD, and voltage of 25 kV). After electroporation, the cells were grown with shaking in 1 ml SOC [59] at 37ºC for 1 h, and the cultures were plated onto LB agar containing kanamycin. The kanamycin resistant transformants were purified on new LB-kanamycin plates. The mutants in which the target genes were replaced by the kanamycin cassette were verified by PCR using the three following pairs of primers : 1) K2/Kt, 2) P3/Kt and 3) P3/P4. Primer positioning is shown on figure 5’

CGGTGCCCTGAATGAACTGC3’, Kt sequence is

1; K2 sequence is

5’

CGGCCACA

3’

GTCGATGAATCC and primers 3 and 4 are specific to each gene and their sequence are given for each study. To kanamycin chromosome,

delete gene

from

pKD46

the the was

removed from the cells by growing the bacteria at 37ºC. pCP20, a heat-sensitive Flp

Table 1 – Common plamsids Plasmid

Genotype

Reference

pKD46

OriR101 repA101 (ts) araBp-gam-bet-exo Apr

Datsenko and Wanner, 2000

pKD4

oriRγ Apr Kmr

Datsenko and Wanner, 2000

pCP20

λ cI857 (ts) ts-rep

Datsenko and Wanner, 2000

recombinase encoding plasmid containing an ampicillin resistance gene [17], was then introduced by transformation. The recombinants containing pCP20 were grown overnight with shaking at 42ºC to eliminate the plasmid, and the cultures were plated on LB agar without antibiotic. Colonies were tested for sensitivity to kanamycin and ampicillin.

Transport assay For transport assays, cells grown in M9 minimal media were harvested in the exponential growth phase, washed once in Tris-maleate (TM) buffer, pH 7.0, and resuspended in the same buffer. Studies were conducted at 37°C in 1 ml and uptake of [14C] radioactive substrate (American Radiolabeled Chemicals) was followed during time. Cell density, length of study and substrate concentration were adjusted according to the speed of uptake and the affinity of the transporter(s). Reactions were initiated by adding 100 µl of ten-fold concentrated substrate to 900 µl of cells. Unless otherwise stated, the ten-fold concentrated substrate solution contained 106 cpm/ml of the radiolabelled form. 100 µl aliquots were then periodically removed. Cells were transferred to 10 ml of ice-cold TM buffer, filtered (0.45µm Millipore filters) and washed twice Functional genomics in E. coli

3

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

with the same buffer. After drying the filters, radioactivity was measured by scintillation counting using 10 ml of Bio-safe NA™ scintillation fluid (Research Products International Corp., Mt. Prospect, IL).

Functional genomics in E. coli

4

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

The MetD Methionine Transporter Functional and Phylogenetic Characterization

Introduction Methionine is an important amino acid for cells as a protein constituent and as a precursor of S-AdenosylMethionine (SAM), a cofactor used for numerous methylation reactions and polyamine biosynthesis [61]. Methionine transport has been characterized in several bacteria, but the genes encoding these transporters remain largely unknown. Recently, an ABC (ATP-Binding Cassette) transporter for methionine and methionine sulfoxide was identified in Bacillus subtilis [62]. Methionine transport and its regulation have been extensively studied both in Escherichia coli and the phylogenetically related bacterium Salmonella typhimurium. E. coli was shown to have two transport systems for L-methionine [35], but only one system for D-methionine was detected [34]. Spontaneous mutants selected for their capacity to grow on toxic methionine analogues were generated. A metD mutant was shown to lack both the high-affinity system for Lmethionine [35] and the system for D-methionine [34, 36]. MetD has therefore been commonly used to describe the transport system affected by the metD mutation. Specificity of the transport systems for L- or D-methionine and related compounds was examined [34, 35] but no inhibitory studies by other amino acids were reported. A metD (formerly metP) mutation was also isolated in S. typhimurium and exhibits characteristics similar to those of the mutation in E. coli [5, 6, 9, 48]. Studies of energy coupling for methionine uptake in E. coli suggested that transport is driven by phosphate bond energy, presumably ATP [37]. Moreover, the metD transport system was shown to be sensitive to osmotic shock and to inhibition by arsenate both in E. coli and S. thyphimurium [21, 37], suggesting it is encoded by a multi-genic region. It was therefore likely for the major uptake system to be an ABC transporter since bacterial ABC transporters are systems composed of a transmembrane protein, a cytoplasmic ATP-binding cassette (ABC) protein and at least one substrate-binding receptor. In Gram-positive bacteria the ABC receptors are lipoproteins whereas they are generally periplasmic proteins in gram-negative bacteria [22]. Functional genomics in E. coli The MetD methionine transporter

5

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

However, I report in this work a new family of ABC transporters which includes members from Gram-negative bacteria for which one or several lipoproteins serve as substrate-binding proteins. The metD mutation has been mapped to 4.8 minutes on the E. coli chromosome [8] and a structural mapping of the metD locus in S. thyphimurium has revealed that several genes are involved this mutation [28]. Interestingly, an ABC transporter maps very closely to this location and was identified by our collaborators from Integrated Genomics as a putative MetD transporter. Methionine biosynthesis genes in E. coli are regulated by the MetJ repressor [27, 61, 71] and a MetJ DNA binding site consensus sequence was derived [58]. Evidence is available suggesting that methionine transport is also regulated by MetJ since cells grown in medium containing methionine have lower levels of methionine transport [33], and methionine auxotrophs with a metJ mutation have much higher transport activity [33, 34]. The recent identification of a MetJ binding site in the promoter region of the ABC transporter mentioned above [43] corroborated our hypothesis. Recently Gal et al. [25] reported growth studies that led to the tentative molecular identification of this transporter. These authors created a null mutant in the abc - yaeE - yaeC gene cluster (renamed metNIQ). The mutation, encompassing all three genes, was constructed in the genetic background of a methionine auxotroph which could grow in the presence but not the absence of either L- or D-methionine. It was shown that the mutation prevented growth in the presence of D-methionine but not of L-methionine. Growth was restored when the three genes were provided in trans. It was also shown that the MetJ protein, in the presence of methionine, but not in its absence, repressed expression of the operon. Although no transport studies were reported, it was concluded that the 3-gene cluster encodes the methionine transporter characterized physiologically by Kadner and his collaborators. Prior to appearance of the report by Gal et al. [25], we had undertaken the study of this ABC transporter and created the E. coli strains listed in table 2. We had also identified a lipoprotein with striking similarity to the substrate-binding lipoprotein of this transporter (lipoprotein 28 encoded by the nlpA gene). The encoding gene has previously been cloned, and this non-essential lipoprotein was shown to be localized to the inner-membrane [74, 75]. We therefore tested the possibility that this lipoprotein is a substrate-binding receptor for L- or Dmethionine. The isogenic strains we created are (1) wild type, (2) a null mutant of the abc - yaeE genes, (3) a null mutant of the yaeC gene, (4) a null mutant of the nlpA gene and (5) a double mutant lacking both the yaeC and nlpA genes. While the abc gene codes for the ATP-binding Functional genomics in E. coli The MetD methionine transporter

6

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

cassette (ABC) protein, yaeE codes for the membrane constituent and both yaeC and nlpA code for two closely related periplasmic lipoproteins. They are homologous to recognized solute binding receptors of functionally characterized ABC transporters. A permease believed to be capable of transporting S-methylmethionine, encoded by the mmuP (formerly ykfD) gene [66] has not been characterized with respect to its substrate specificity. We therefore constructed the triple mutant lacking the abc, yaeE and mmuP genes in order to test whether this permease might be responsible for the low-affinity methionine transport described by Kadner for the L- isomer [35].

Materials and Methods Experimental Methods Strains created are listed in table 2 and were generated as described in the Common Methods section using the PCR primers listed in table 3. Growth experiments were performed in M9 minimal media with MgCl2 replacing MgSO4 and 10µM L-methionine or 100µM Dmethionine serving as a sole sulfur source. The results presented are the average of two independent experiments. Table 2. Strains

Strain LJ 3001 (BW25113) LJ 3015 LJ 3016 LJ 3017 LJ 3018 LJ 3019

Genotype lacI rrnBT14 ∆lacZWJ16 hsdR514 ∆araBADAH33 ∆rhaBADLD78 BW25113 ∆abc-yaeE (∆MetNI) BW25113 ∆yaeC (∆MetQ) BW25113 ∆nlpA BW25113 ∆yaeC ∆nlpA BW25113 ∆yaeE-abc ∆ykfD (∆MetNI ∆ mmuP) q

Reference Datsenko and Wanner, 2000 This study This study This study This study This study

Transport assays were conducted as previously mentioned over a 20 or 10 minute time interval with the cell densities at 0.09 and 0.36 absorbancy units (A600) and the methionine concentrations at 0.5 µM (55 µCi/µmol) and 12.7 µM (11.8 µCi/µmol), respectively, for the Land D- isomers. For the assay with 55 µCi/µmol of L-methionine, the total count used was only 2.104 cpm/ml since the specific activity of the commercial solution was too low to reach the concentration desired with a total count of 105 cpm/ml. For the comparison of the transport of Land D-methionine between the abc-yaeE and the abc-yaeE/mmuP mutants, methionine Functional genomics in E. coli The MetD methionine transporter

7

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

concentrations of 12.7 and 102.7 µM (11.8 µCi/µmol) were used since the low-affinity Lmethionine transporter was analyzed. Inhibition of the initial uptake rates were performed during a 5 or 3 minute time interval for L- and D- methionine, respectively. Unless otherwise stated, the inhibitory concentration of cold amino acids was ten times the concentration of methionine. All transport assays presented here are the average of three independent repetitions. Table 3. Primers used for generation and verification of BW25113 mutants.

Gene

Primers (5’Æ 3’)

abc-yaeE (metNI)

Generation GATGCGGTCGCCTGCGAACTGAATTAAATAAACCAGAATGACCAGGTGTAGGCTGGAGC TGCTTC (forward) CTTAATGACGATATAAATAATCAATGATAAAACTTTCGAATATCCATATGAATATCCTCC TTAG (reverse) Verification abc-yaeE 3: CGTTACTTGCGAGTGACAGC abc-yaeE 4: GCATGTGACGCTAGTATCGC

yaeC (metQ)

Generation TTACAAATTGTGGAAACAGCCTAAAAATTACCAGCCTTTAACAGCGTGTAGGCTGGAGC TGCTTC (forward) AAGGAATAAGGTATGGCGTTCAAATTCAAAACCTTTGCGGCAGTGCATATGAATATCCT CCTTAG (reverse) Verification yaeC 3: ACAGCCGCTTAGCATGAGTG yaeC 4: AATTCAGTTCGCAGGCGACC

nlpA

Generation ACCGCAGCGACCTTACCGCTATAGTCAGGTAATCATTAATAAAAGGTGTAGGCTGGAGC TGCTTC (forward) TGAGAATTACCAGCCAGGCACCGCGCCACCGTTAAAAATGGTTTCCATATGAATATCCT CCTTAG (reverse) Verification nlpA 3: CGTGGTCAGTAAGAAGTGCC nlpA 4: GCTGCTGATTCTGTCATCGG

ykfD (mmuP)

Generation GGTTGACTTTGCATTCTGTTAACAAACGCGGTATAACAAACCGTGTAGGCTGGAGCTGCT TC (forward) GGTTGAGTAAGGAAATAAGCACCATAGCACAACGCAACAAACCATATGAATATCCTCCT TAG (reverse) Verification ykfD 3: GACTTGTTCGCACCTTCC ykfD 4: GGCTGTCGGCTAAGTTAC

Computer Methods Sequences of the proteins that comprise the 3 elements of the ABC transporters of the Methionine Uptake Transporter (MUT) family were obtained by an initial BLAST search [3] using the sequence of the characterized E. coli member as reference. The resulting hits were Functional genomics in E. coli The MetD methionine transporter

8

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

filtered through a program manipulating the BLAST program to eliminate the sequences more related to other subfamilies in the ABC superfamily (Tran et al., unpublished program). Multiple sequence alignments were constructed using the ClustalX program [67]. The gap penalty and gap extension values used with the ClustalX program were 10 and 0.1 respectively, although other combinations were tried. HMMTOP [68, 69] and TMHMM [64, 39] programs were used to determine the number of transmembrane segments. The phylogenetic trees were derived by the Neighbor-joining method from alignments generated with the ClustalX program using the BLOSUM 62 scoring matrix. The phylogenetic trees were drawn using the TreeView program [45]. Complementary trees were constructed using the phylo_win program [26] with the Neighbor-Joining method and PAM distances as a model of evolution. This study was conducted independently for the three protein constituents of the ABC transporters that comprise the MUT family and the sequences obtained were checked manually to see whether all three ABC elements had been identified for each transporter. G+C content was analyzed with the GeeCee program [53] and codon usage was analyzed with

the

Countcodon

program

from

the

Codon

Usage

Database

website

(http://www.kazusa.or.jp/codon/countcodon.html). Prediction of the lipoproteic structure of the receptors was performed with the Lipop section of the PSORT program (http://psort.nibb.ac.jp).

Results 0.30

Growth studies 0.25

Absorbance (600 nm)

Figure 1 shows the growth of the isogenic strains described previously with 100 µM D-methionine as the sole source of sulfur. The wild type strain and nlpA mutant grew equally well, but the abc-yaeE mutant

0.15

0.10

0.05

and yaeC/nlpA double mutant grew very poorly. The yaeC single mutant also grew

0.00 0

poorly, but slightly better than the two double

5

10

15

20

25

Time (h)

Figure 1. Growth of E. coli as a function of time in M9 minimum media where MgSO4 was replaced by 1mM MgCl2 and D-methionine was added at a concentration of 100 µM as the only sulfur source present. The following strains were examined: wild type BW25113 (), ∆abcyaeE (Y), ∆yaeC (S), ∆nlpA (Ì) and ∆yaeC-∆nlpA (U).

mutants. When 20 µM L-methionine served as the sole source of sulfur, the difference between the wild-type and the double mutants Functional genomics in E. coli The MetD methionine transporter

0.20

9

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

was less pronounced than when D-methionine was used (fig. 1; data not shown). However depressed growth suggested that the transporter could accept both D- and L-methionine as substrates. Transport studies Figures 2A and B show the uptake of [14C]L-methionine (0.5 µM) and [14C]D-methionine (13 µM), respectively. Relative rates of uptake of the two substrates were essentially the same for the wild type and mutant strains. Thus, the wild type and nlpA mutant took up the amino acids at the same rate, the two double mutants (abc-yaeE and yaeC nlpA) took up both substrates poorly, and the yaeC single mutant took up both substrates poorly but slightly better than the double mutants. It was therefore concluded that (1) both L- and D-methionine are substrates, (2) both binding receptors, YaeC and NlpA, can probably activate both D- and L-methionine uptake, and (3) YaeC is much more efficient than NlpA as a receptor. For both isomers, residual uptake is observed in the abc-yaeE mutant, suggesting that a secondary transporter exists both for L- and D-methionine. We tested the transport of L- and D-methionine at concentrations of 10 and 100 µM in the mmuP abc-yaeE triple mutant but no significative difference was observed relatively to the abc-yaeE mutant. This result is consistent with the indistinguishable growth rates

300

250

A D-methionine uptake (pmol / mg dry weight )

200

L-methionine uptake (pmol / mg dry weight )

B

250

150

100

50

200

150

100

50

0

0 0

5

10

15

0

20

4

6

8

10

Time (min)

Time (min) 14

2

14

Figure 2. Uptake of 1-[ C]L-methionine (A) and 1-[ C]D-methionine (B) by wild type BW25113 (), ∆abc-yaeE (Y), ∆yaeC (S), ∆nlpA (Ì) and ∆yaeC-∆nlpA (U) cells. Cells grown in M9 minimal media were prepared as described in the Methods section. Assays were performed at 37°C in 1 ml over a 20 or 10 minute time interval with the cell densities at 0.09 and 0.36 absorbancy units (A600) and the methionine concentrations at 0.5 µM (55 µCi/µmol) and 12.7 µM (11.8 µCi/µmol), respectively, for the L- and D- isomers. Values are expressed in picomoles of L- or D-methionine retained per milligram of bacterial dry weight. Functional genomics in E. coli The MetD methionine transporter

10

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

between the abc-yaeE-yaeC and the abc-yaeE-yaeC mmuP mutants observed by Gal et al. [25] in a methionine auxotrophic strain grown in minimal media supplemented with L-methionine. It can therefore be assumed that the S-methylmethionine permease MmuP does not transport L- or Dmethionine. Inhibition studies Table 3 summarizes the inhibitory effects of several non radioactive amino acids present at 10-fold the concentration of the radioactive amino acid on both L- and D-methionine uptakes. Uptake of L-methionine by the wild type bacteria was strongly inhibited by L-methionine and weakly inhibited by N-formyl L-methionine. However, no other amino acid inhibited appreciably. Uptake by the abc-yaeE mutant was most strongly inhibited by L-methionine and to a lesser extent by L-valine and L-alanine. When uptake of D-methionine was studied in wild type E. coli, L-methionine was most inhibitory followed by N-formyl L-methionine and D-methionine in that order, but no other amino acid inhibited. In the abc-yaeE mutant, the order of inhibition was the same. The increase of transport rate observed in the presence of several amino acids was attributed to an increase of the available energy for the cells in TM buffer. These results are consistent with the conclusion that the ABC-type methionine transporter is specific for L- and Dmethionine as well as their analogues, but that other amino acids can not be recognized or transported. The relative inhibitory effects of L- and D-methionine on uptake of these two radioactive substrates are in line with the relative affinities reported by Kadner who determined Km values of 75 nM for L-methionine [35] and of 1.2 µM for D-methionine [34]. Inhibition A Inhibitor

L-methionine D-methionine N-formyl-Lmethionine L-alanine D-alanine L-leucine D-leucine L-valine L-serine D-serine L-threonine D-threonine

B

∆abc-yaeE

WT pmol/(min x mgdry weight)

percentage of control

pmol/(min x mgdry weight)

percentage of control

16,4 ± 3,1 0,7 ± 0,6 19,0 ± 7,3

100 5 116

6,4 ± 2,4 1,8 ± 0,4 7,3 ± 3,4

100 28 114

14,2 ± 0,9

86

7,6 ± 0,0

119

24,9 ± 2,5 23,8 ± 1,0 18,1 ± 3,2 16,7 ± 0,0 24,6 ± 4,2 28,1 ± 3,5 24,9 ± 2,8 24,7 ± 0,8 19,9 ± 3,5

152 145 110 102 150 172 152 151 121

4,1 ± 1,3 6,0 ± 1,3 5,2 ± 2,0 5,9 ± 3,8 3,3 ± 1,4 6,7 ± 3,6 10,1 ± 0,6 3,5 ± 1,3 6,4 ± 1,9

64 93 81 91 52 104 157 54 100

D-methionine L-methionine N-formyl-Lmethionine L-alanine D-alanine L-leucine D-leucine L-valine L-serine D-serine L-threonine D-threonine

∆abc-yaeE

WT

Inhibitor

pmol/(min x mgdry weight)

percentage of control

pmol/(min x mgdry weight)

percentage of control

41,8 ± 8,1 14,4 ± 0,6 9,8 ± 5,3

100 34 23

12,0 ± 1,5 10,5 ± 1,9 6,7 ± 2,1

100 87 56

11,9 ± 0,5

29

10,0 ± 1,7

83

57,9 ± 43,1 ± 43,1 ± 46,9 ± 44,9 ± 48,4 ± 51,5 ± 53,1 ± 48,6 ±

138 103 103 112 107 116 123 127 116

13,7 ± 1,9 13,8 ± 0,3 14,0 ± 0,2 16,4 ± 0,4 11,4 ± 5,1 12,9 ± 1,4 23,5 ± 4,9 13,9 ± 5,0 18,8 ± 6,1

114 115 117 136 95 108 196 116 156

10,4 3,8 3,8 1,2 16,1 4,7 6,1 12,2 4,4

Table 3. Inhibition of L-methionine (A) and D-methionine (B) uptake by the L- and D- isomers of several amino acids. Assays were performed as previously described for 5 and 3 minutes, respectively, for L- and D-methionine. Non-radioactive inhibitory amino acids were added at concentrations of 5 µM for L-methionine uptake and 100 µM for D-methionine uptake. Functional genomics in E. coli The MetD methionine transporter

11

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

studies of D-methionine transport with 100 and 1000 fold excess of L- and D-methionine were consistent with the existence of a secondary low-affinity transporter for methionine (data not shown). Surprisingly, the uncharacterized methionine transporter present in the mutant is also specific for methionine, but it is not clear that the same transporter catalyzed uptake of both Land D-methionine. Phylogenetic studies Our preliminary results suggested that the E. coli ABC methionine transporter identified by Gal et al. [25] and in the work described here belongs to a novel family within the ABC superfamily. Surprisingly, this family is more closely related to the Polar Amino Acid Uptake Transporter (PAAT) family (TC#3.A.1.3) than to the Hydrophobic Amino Acid Uptake Transporter (HAAT) family (TC#3.A.1.4). We have termed this family the Methionine Uptake Transporter (MUT) family (TC#3.A.1.23). We identified the three constituent proteins that comprise each member of the MUT family (table 4). Most organisms having representation in the MUT family have only one homologue within this family, but a few have two. Organisms with two paralogues include Salmonella thyphimurium, Yersinia pestis and Pseudomonas aeruginosa, all γ-proteobacteria, and four low G+C Gram-positive bacteria, Bacillus anthracis, Staphylococcus aureus and two species of Listeria. No organism has three or more paralogues within the MUT family. With the exception of the truncated membrane protein of Providencia stuastii, all members of the family are of similar size for each component. All membrane proteins exhibit 5 putative transmembrane α-helical segments (TMSs), suggesting uniform topology. We constructed phylogenetic trees (figure 3A-C; see Computer Methods). The phylogenetic analyses revealed several major subfamilies, each stemming from points near the center of the tree. The phylogenetic trees shown in figure 3A, B and C for the ABC, membrane and receptor proteins, were analysed according to phylogenetic cluster. Cluster 1 includes the E. coli methionine uptake transporter, and only γ-proteobacteria are represented. All the members have lipoproteins as receptors. The phylogenies of the proteins follow those of the organisms thereby suggesting orthology. One system, Vch, has its ABC protein and its receptor in cluster 1, but its membrane protein is loosely clustered with the α-proteobacterial proteins in clusters 7 and 13.

Functional genomics in E. coli The MetD methionine transporter

12

Saier Lab – UC San Diego March – August 2002

Table 4 – Members of the Methionine Uptake Transporter (MUT) family Source

Abbreviat ion

Gene Name

Salmonella typhimurium

γ-Proteobacteria

Sty1

Escherichia coli 1

γ-Proteobacteria

Eco

Membrane proteins

Acc. #

Size (aa)

Gene Name

ABC or STM0247

AAL19210

343

YAEE or STM0246

ABC or B0199

AAC73310

343

YaeE

-

Acc. #

Substrate-binding receptors Acc. #

Size (aa)

Lipoprotein ?

YAEC or STM0245

AAL19208

271

Yes

1) YAEC or B0197

AAC73308

271

Yes

2) NLPA or B3662

AAC76685

272

Yes

Size (aa)

Gene Name

AAL19209

217

AAC73309

217

-

Yersinia pestis

γ-Proteobacteria

Ype1

ABC or YPO1073

CAC89916

343

YPO1072

CAC89915

217

YPO1071

CAC89914

271

Yes

Vibrio cholerae

γ-Proteobacteria

Vch

VC0907

AAF94069

344

VC0906

AAF94068

225

VC0905

AAF94067

269

Yes

γ-Proteobacteria

Pmu

PM1728

AAK03812

344

PM1729

AAK03813

229

PLPB or PM1730

AAK03814

276

Yes

γ-Proteobacteria

Hin

ABC or HI0621

AAC22280

345

HI0620.1

P46492

198

HLPA or HI0620

AAC22279

273

Yes

Lla

-

CAB59828

368

YDCC or LL0323

AAK04421

231

-

CAB59827

286

No

Spn

ABC-NBD

AAK98953

353

ABC-MSD

AAK98954

230

ABC-SBP

AAK98951

284

Yes

Pasteurella multocida Haemophilus influenzae Lactococcus lactis

2

ABC proteins

Organism

Streptococcus pneumoniae Streptococcus pyogenes Streptococcus mutans Fusobacterium nucleatum

Firmicutes; Bacillales Firmicutes; Bacillales Firmicutes; Bacillales Firmicutes; Bacillales

Spy

atmD

AAM78841

354

atmE

AAM78842

230

atmB

AAM78840

281

Yes

Smu

ATMD

AAL04079

354

ATME

AAL04080

229

ATMB

AAL04077

280

Yes

Fusobacteria

Fnu

FN0660

AAL94856

335

FN0659

AAL94855

233

FN0658

AAL94854

261

No

Helicobacter pylori

ε-Proteobacteria

Hpy

HP1576

AAD08616

327

HP1577

AAD08617

215

HP1564

AAD08604

271

No

4

Salmonella typhimurium

γ-Proteobacteria

Sty2

SFBB or STM0511

AAL19465

338

SFBC or STM0512 or STY0560

AAL19466

219

SFBA or STM0510

AAL19464

276

No

5

Streptomyces coelicolor

Firmicutes; Actinobacteria

Sco

SCO1559 or SCL11.15C

CAB76078

368

SCO1558 or SCL11.14C

CAB76077

240

SCO1557 or SCL11.13C

CAB76076

275

Yes

6

Yersinia pestis

γ-Proteobacteria

Ype2

YPO1318

CAC90148

328

YPO1319

CAC90149

223

YPO1317

CAC90147

274

No

α-Proteobacteria

Sme

CAC47469

358

R02889 or SMC03158

CAC47468

221

CAC47467

258

No

7

Sinorhizobium meliloti Agrobacterium tumefaciens

α-Proteobacteria

Atu

AAL45281

346

ATU4488

AAK88954

222

AAK88953

259

No

Thermus-Deinococcus group

Dra

AAF10928

325

DR1357

AAF10929

218

1) DR1359

AAF10931

256

No

2) DR1358

AAF10930

256

No

Spirochaetales

Tpa

TPN32 or TP0821

AAC65789

268

Yes

γ-Proteobacteria

Pae1

3

8 9 10

Deinococcus radiodurans Treponema pallidum Pseudomonas aeruginosa Brucella melitensis

R02890 or SMC03159 ATU4487 or AGR_L_765 DR1356

TP0120

AAC65110

269

PA2350

AAG05738

BMEII0337

AAL53579

12 13

14

Neisseria meningitidis Ralstonia solanacearum Caulobacter crescentus Pseudomonas aeruginosa Xylella fastidiosa Xanthomonas axonopodis Xanthomonas campestris Bacillus anthracis

15

Listeria innocua Listeria monocytogenes Staphylococcus aureus Bacillus anthracis Bacillus halodurans

16 Bacillus subtilis Listeria innocua

17 18 19

20

Listeria monocytogenes Staphylococcus aureus Corynebacterium glutamicum

TP0119

AAC65109

219

369

PA2351

AAG05739

217

PA3931

AAG07318

259

No

369

BMEII0336

AAL53578

230

1) BMEII0338

AAL53580

278

No

2) BMEI1954

AAL53135

268

No

-

NP_105584

284

No

CAB83799

287

Yes

α-Proteobacteria

Bme

α-Proteobacteria

Mlo

-

NP_105583

365

-

NP_105582

218

β-Proteobacteria

Nme

NMA0504

CAB83797

245

NMA0505

CAB83798

228

β-Proteobacteria

Rso

RSC0920 or RS04495

CAD14622

350

RSC0921

CAD14623

217

CAD14624

266

No

α-Proteobacteria

Ccr

CC2669

AAK24636

332

CC2668

AAK24635

224

CC2664

AAK24631

268

Yes

γ-Proteobacteria

Pae2

PA5503

AAG08888

335

PA5504

AAG08889

225

PA5505

AAG08890

260

No

γ-Proteobacteria

Xfa

XF0875

AAF83685

334

XF0874

AAF83684

235

XF0873

AAF83683

261

No

γ-Proteobacteria

Xax

abc

AAM38512

335

yaeE

AAM38511

231

XAC3667

AAM38510

269

No

Xca

abc

AAM42900

335

yaeE

AAM42899

232

XCC3628

AAM42898

266

No

Ban1

-

NP_654117

346

-

NP_654116

222

-

NP_654118

270

Yes Yes

-

11 Mesorhizobium loti

-

R02888 or SMC03157 ATU4489 or AGR_L_761

γ-Proteobacteria Firmicutes; Bacillales Firmicutes; Bacillales Firmicutes; Bacillales Firmicutes; Bacillales Firmicutes; Bacillales Firmicutes; Bacillales Firmicutes; Bacillales Firmicutes; Bacillales Firmicutes; Bacillales Firmicutes; Bacillales Firmicutes; Actinobacteria

-

NMA0506 or GNA1946 RSC0922 or RS04493

Lin2

LIN0312

CAC95545

338

LIN0311

CAC95544

220

LIN0313

CAC95546

273

Lmo2

LMO0284

CAD00811

338

LMO0283

CAD00810

220

LMO0285

CAD00812

273

Yes

Sau1

SAV0837 or SA0769

BAB56999

341

SAV0838 or SA0770

BAB57000

231

SAV0839 or SA0771

BAB57001

273

Yes

Ban2

-

NP_653454

341

-

NP_653453

221

-

NP_653451

268

Yes

Bha

BH3481

BAB07200

338

BH3480

BAB07199

218

BH3479

BAB07198

286

Yes

YUSC

CAB15264

341

YUSB

CAB15263

222

1) YUSA

CAB15262

274

Yes

2) YHCJ

CAB12739

263

Yes Yes

Bsu

-

-

Lin1

LIN2514

CAC97741

340

LIN2513

CAC97740

224

LIN2512

CAC97739

276

Lmo1

LMO2419

CAD00497

340

LMO2418

CAD00496

224

LMO2417

CAD00495

276

Yes

Sau2

SAV0462 or SA0420

BAB56624

341

SAV0463

BAB56625

219

SAV0464 or SA0422

BAB56626

280

Yes

Cgl

-

NP_599870

360

-

NP_599869

225

-

NP_599871

299

Yes

AAD18427

272

No

AAK78962

272

Yes

Chlamydophila pneumoniae

Chlamydiales

Cpn

DPPF_1 or CPN0280 or CP0478

AAD18429

341

CPN0279 or CPJ0279 or CP0479

AAD18428

221

CPN0278 or CPJ0278 or CP0480

Clostridium acetobutylicum

Firmicutes; BacillusClostridium group

Cac

CAC0984

AAK78960

320

CAC0985

AAK78961

218

CAC0986

CJ0774C

CAB73039

336

CJ0773

CAB73038

303

Campylobacter jejuni

ε-Proteobacteria

Cje

Providencia stuartii*

γ-Proteobacteria

Pst

Mannheimia haemolytica*

γ-proteobacteria

Mha1 Mha2 Mha3 Legionella pneumophila* Neisseria gonorrhoeae*

γ-proteobacteria

Lpn

β-proteobacteria

Ngo

-

YAEE

AAF71397

-

CJ0772C

CAB73037

257

No

CJ0771C

CAB73036

256

No

CJ0770C

CAB73035

258

No

Yes

-

146 PLPA

AAA25538

277

PLPB

AAA25547

277

Yes

PLPC

AAA25540

263

Yes

-

CAA06664

259

Yes

GNA1946

AAF44768

288

Yes

i) All full length membrane protein homologues of the membrane protein were predicted to exhibit 5 TMS’s with either TMHMM or HMMTOP. The database entry for the Hin membrane protein was found to be truncated at both the N and C termini due (1) to an incorrect initiation codon assignment and (2) to a frameshift mutation in the structural gene. The reconstructed protein was 198 residues long. This frameshift mutation has been shown to be authentic and not due to a sequencing error. Hin may therefore be a pseudogene. The database entries for the Vch and Bha receptors were also found to be erroneous due to wrong initiation codon assignments.* indicates an organism for which the complete genome sequence is not available in genbank.

13

15 Ban1

14

16

Ban2

Xca

Xax

Xfa

Sau1

Lin2 Lmo2

Biochemistry Department

Bha Bsu

Pae2

Lin1

Lmo1

Ccr

13

Rso

12

Nme

Figure 3A. ABC proteins

Sau2

17

Cgl

18 Cpn

Mlo

11 Bme

Cac

20 10

Cje

Pae1 Vch

9

Ype1 Eco Sty1

Tpa

Pmu

1

Hin Dra

Atu

8

19

7

Cpn

Sme

Lla Spn Ype2

1

Hin

6 10

Pmu Ype1 Eco Sty1 Pae1

6

Cje

17

Sty2

5

4

Fnu

Hpy

Spy Smu

2

3

11 Mlo

Ype2

Sco

Bme

8

Dra

Cac

5 Sco

1’ Vch

13

Hpy

3

Rso

Atu

7

Fnu

Figure 3B. membrane proteins

Sme

9

Tpa

Lla

4

Sty2 Spn

12

2

Nme Spy

Pae2

Ccr

14

Smu Sau1

Xca Xax

Lmo1 Lin1 Xfa

Bsu

Pst* Sau2

0.1

17

Ban2 Lin2 Lmo2

15

Bha

16

Ban1 Cgl

18 Functional genomics in E. coli The MetD methionine transporter

14

Saier Lab – UC San Diego March – August 2002

19

Biochemistry Department

6

14

19 Cpn

Xfa

Xax Xca

Dra2* Sco

Figure 3C. Receptors

5 3

Fnu

Ban2

Bsu1

16

10 14’ 8

Pae1Pae2 Dra1

Ccr

Ype2

Hpy

Cac Sau1

20 Cje1 Cje2*

Lin1 Lmo1

Cje3*

Bha

9

Tpa

Lin2

15

Lla

Lmo2 Spn

Ban1 Sty1 Eco1

2

Spy

Ype1

Smu

Vch Mha3

1

Bsu2*

Mha2 Sau2

Pmu Mha1

Atu

Hin

Sme Eco2* Lpn*

0.1

7 Bme2*

Ngo* Nme

12

Rso Bme

11

17

13 Cgl

18

Mlo

11’

Sty2

4

Figure 3. Phylogenetic trees of the three constituents of the MUT family of ABC transporters: A, ATP-binding cassette (ABC) constituents; B, membrane constituents; C, solute binding receptors. The multiple alignments were generated with the ClustalX program [67] using the BLOSUM 62 scoring matrix. The trees are based on the Neighbor-Joining method and were drawn with the TreeView program [45]. On tree C, * refers to a receptor which is not encoded with an ABC and a membrane protein.

This may represent an unusual shuffling of constituents between systems [40]. Only the membrane protein is not as expected. However, using PAM distances as a model of evolution with the Phylo_win program [26], the membrane constituent Vch clusters together with other cluster 1 proteins. Subfamily 2 consists of proteins from closely related lactic and Gram-positive bacteria. These proteins are also probably orthologous to each other. Cluster 3 proteins from Helicobacter pylori and Fusobacterium nucleatum, are always clustered together in spite of the extreme phylogenetic diverge of these two organisms. It seemed Functional genomics in E. coli The MetD methionine transporter

15

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

possible that horizontal transfer had occurred. The G+C content and codon usage of the genes encoding the ABC transporters of these two organisms were compared but no significative difference was observed relatively to those of the whole genome. This observation can not lead to any conclusion as horizontal transfer may have occurred from a bacterium with similar G+C content and codon usage. Clusters 4-13 consist of only one or two proteins per cluster. In clusters 7 and 11, the two proteins in each cluster are from α-proteobacteria, suggesting orthology. However, some of the distantly related proteins belong to closely related organisms (eg, Sty2 and Ype2; Nme and Rso). This clearly suggests that sequence divergent primordial proteins resulted from early gene duplication events, and these early paralogues were not transmitted to most of the organisms. Moreover, this is confirmed by the presence of several lipoproteic receptors from Gram-negative organisms (Ccr, Nme, Tpa) whereas most are periplasmic (table 4). Cluster 14 consists of γ-proteobacteria proteins except for Ccr which is from an αproteobacterial species. The clustering pattern in figures 3A and C are consistent with orthology but the clustering of Ccr in figure 3B is anormalous and the same topology was obtained using PAM distances as a model of evolution. Clusters 15 and 16 clearly represent two sequence divergent groups of paralogues, both represented in Bacillus anthracis and two Listeria species. Staphylococcus aureus as well as B. subtilis and B. halodurans encode within their genomes only the second of these homologues (cluster 16). The clustering of Bha in figure 3C is anormalous and this was also the case on the tree drawn using PAM distances. Cluster 19 includes a single chlamydial protein although three chlamydial species have been sequenced. In spite of their close phylogenetic relationship, the other two close relatives of C. pneumoniae must have lost the corresponding orthologues. Finally, in cluster 20, Cac and Cje are always together (figs. 3A, B and C). Because of the great distance between C. acetobutylicum (a Gram-positive bacterium) and C. jejuni (a Gramnegative ε-proteobacteria), we suggest that horizontal transfer has occurred. However, this assumption can not be confirmed as both organisms have similar overall G+C content and the G+C content and codon usage of the MUT genes are not different from those of the complete genomes. With the exception of Lla, all receptors from Gram-positive organisms were predicted to be lipoproteins. For Lla, a signal peptide with high similarity with those of Spn, Spy and Smu Functional genomics in E. coli The MetD methionine transporter

16

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

was found. However, the conserved cysteine residue essential for lipid ancrage was replaced by a glycine residue. This might be explained by a sequencing error or by a single-point mutation since the codons for these two residues only differ by one nucleotide.

Discussion The results presented here confirm and extend the molecular characterization [25] of the MetD transporter identified earlier in the laboratories of Kadner and Ayling. abc and yaeE encode respectively the ATP-Binding Cassette (ABC) and the membrane protein of the MetD ABC methionine transporter. YaeC, the receptor encoded with the other components of the transporter, is the major binding-protein for both L- and D-methionine. However, its related paralogue NlpA (lipoprotein 28) may also exhibit the capacity to bind the two isomers of methionine. The abc - yaeE - yaeC genes were renamed metNIQ by Gal et al. [25]. Inhibition studies revealed that the transporter characterized is specific for methionine. However, N-formyl methionine can be accepted as substrate. This observation is consistent with the fact that MetD is the major methionine transporter since MetD also transports this methionine analogue for bacterial translation initiation [63]. Moreover, this observation is also consistent with the fact that other L- and D-methionine analogues can be transported [34, 35]. We also confirm Kadner’s observation that L-methionine effectively competes for Dmethionine transport, but that D-methionine does not strongly compete with L-methionine. This led Kadner to suggest that the metD locus encodes a component of at least two transport systems, and that metD may not represent the initial methionine binding site [34]. However, the data presented here shows that MetD is a single transport system which includes the major methionine binding receptor. The difference in inhibition observed between the two isomers is a consequence of the difference in affinity of the transporter for the two substrates. The Km determined for MetD transport of L-methionine (75 nM) [35] was 15 fold lower than the Km determined for the transport of D-methionine (1.2 µM) [34]. Phylogenetic analyses led to the conclusion that the MetD methionine transporter characterized in E. coli belongs to a new ABC family which we named Methionine Uptake Transporter (MUT) family (TC#3.A.1.23). The MUT family is widely represented among bacterial subdivision. All members are of a similar size for each component, and all membrane proteins exhibit 5 putative transmembrane α-helical segments (TMSs). The overall topology of Functional genomics in E. coli The MetD methionine transporter

17

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

the trees presented does not follow the phylogeny of the organisms since several duplications of the system have occurred early during the evolutionary history of the family. The existence of several sequence divergent primordial paralogues is likely to explain the topology of the phylogenetic trees. However, no more than two of these paralogues have been transmitted to any currently sequenced organism since none of them has more than two MUT systems. A clustering which takes into account data from the three trees is proposed. It is surprising that the YaeC receptor as well as the NlpA receptor are putative lipoproteins since these occur rarely in Gram-negative bacteria. However, we have noted that other receptors from Gram-negative bacteria of the same ABC family are also putative lipoproteins. All of cluster 1 receptors are predicted to be lipoproteins therefore suggesting orthology and possibly uniform function. It is interesting to note that Yersinia pestis and Salmonella typhimurium, two close relatives of E. coli, possess two paralogues within the MUT family whereas only the receptor is present twice in E. coli. The possibility that the common ancestor had the two paralogues and that E. coli lost one does not seem likely as Ype2 and Sty2 do not cluster together. Moreover, Eco2 clusters with the proteins of cluster 1 which include Eco1. We therefore propose that the nlpA gene (Eco2) arose by duplication of yaeC (Eco1). The fact that both Sty2 and Ype2 are not lipoproteins whereas Eco2 is corroborates this idea. Cluster 1 is a group of γ-proteobacterial proteins conserved in all three trees. For this cluster, the phylogenies of the proteins are consistent with those of the organisms. Moreover, collaborators from Integrated Genomics identified MetJ binding sites in the promoter region of all the members of this cluster. The functional data provided here for E. coli and this observation strongly suggest that all cluster 1 members are methionine transporters. This is corroborated by the presence of the MetD transporter in S. typhimurium [5, 6, 9, 48] which maps in the region of the ABC transporter described in this study [28]. Several members of the MUT family from Gram-positive bacteria are likely to be involved in the transport of sulfur compounds. The B. subtilis yusCBA operon is preceded by an S box, and is therefore probably involved in transport of methionine or another sulfur compound [29]. S-boxes are consensus sequences of termination control systems involved in sulfur metabolism regulation and especially in the regulation of methionine and cysteine biosynthesic genes in Gram-positive bacteria [29].In the phylogenetic trees, YusC, B and A cluster with proteins from Staphylococcus aureus , Bacillus halodurans and two Listeria species. Our Functional genomics in E. coli The MetD methionine transporter

18

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

collaborators have identified potential S boxes upstream of the S. aureus and Listeria operons. The regulatory evidence and the homology to the characterized E. coli member suggest these may be methionine transporters; however, they might also encode transporters for other sulfur compounds possibly used as methionine precursors. Two other transporters from this family are likely to transport sulfur compounds based on the presence of predicted S-boxes. One of these was previously identified by Grundy and Henkin [29] in Clostridium acetobutylicum. An S-box was also identified upstream of the operon encoding Lmo2 proteins. The existence of proteins likely to be involved in sulfur acquisition in positions throughout the phylogenetic tree suggests that this entire family may be involved in the transport of organic sulfur compounds. Further experimental studies will be required to determine the substrate ranges of the transporters in this family. The MUT family is of pharmaceutical interest since several members are involved in pathogenicity. The S. typhimurium sfbA gene is found in a pathogenicity islet and is essential for infection in a mouse model [47], although its specific contribution to pathogenicity is unknown. It was predicted to encode the binding protein of an ABC transporter for iron because its expression was increased under iron-limiting conditions. Despite its regulation by iron, a Fur regulatory binding site was not found close to this operon [46]. Based on the analysis presented here, we suggest the possibility that the Sfb transporter is involved in the transport of an amino acid or sulfur compound during infection. The H. influenzae hlpA gene is not essential for infection, but a mutation in this gene results in reduced invasion in rats [15]. Helicobacter pylori also contains a transporter of the MUT family (AbcBCD). Mutation of the membrane protein (AbcD) or ATPase (AbcC) of this transporter reduces the activity of urease which is essential for colonization [32], but the function of the transporter has not yet been determined. Since H. pylori urease requires nickel, it was thought to function as a nickel transporter, but this was not demonstrated. Methionine transport is probably nonessential in most organisms since it can be synthesized from other amino acids, but it may have greater importance for Yersinia pestis. Many strains of Y. pestis will not grow in culture without the addition of methionine [12], and methionine transporters, therefore, may represent a drug target for this organism. Finally,

a

methionine/methionine

sulfoxide

transporter

has

been

functionally

characterized in Bacillus subtilis [62]. Based on initial Blast results, we believe that this ABC transporter belongs to a new family within the ABC superfamily. However, only members from high G+C Gram-positive bacteria were represented in the initial analysis. Suprisingly, this new Functional genomics in E. coli The MetD methionine transporter

19

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

family of ABC transporters is more closely related to the Polar Amino Acid Uptake Transporter (PAAT) family (TC #3.A.1.3) than to the Methionine Uptake Transporter (MUT) family (TC #3.A.1.23) characterized in this work. A complete phylogenetic characterization will be required to confirm this observation.

Functional genomics in E. coli The MetD methionine transporter

20

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

Frx, a PTS transporter involved in melibiose metabolism Introduction A recent computational analysis of PTS systems (see Laboratory presentation section, pages B-C) of Escherichia coli identified several functionally uncharacterized systems [65]. The Frx transporter is one of those for which some experimental data are available. Frx belongs to the fructose-like PTS permease family [65]. Enzymes IIA, B and C are expressed in a single polypeptide encoded by a unique gene [70]. No information is available regarding substrate specificity. However, the frx gene is located on a bicistronic operon with ybgG, a gene which encodes an α-mannosidase homologue. It can therefore be inferred that Frx may transport the substrate of YbgG, possibly an α-glycoside. Prior to the availability of the fully sequenced genome of E. coli, Utsumi et al. isolated a fragment of the frx (hrsA) gene during a shotgun cloning experiment and then published its sequence [70]. These authors isolated mutants of the micF and ompR regions which lacked the thermoinduction of the outer-membrane protein ompC in an envZ - genetic background. EnvZ and OmpR are two members of a sensor-regulator system. The sensor EnvZ is an inner-membrane protein with kinase activity which responds to osmolarity to phosphorylate the transcriptional regulator OmpR [14]. The micF RNA has been shown to be essential for thermoregulation of the OmpF and OmpC outer-membrane proteins [4, 70]. When the frx fragment containing the regions homologous to enzymes IIA and B was provided in trans, the original phenotype was restored in the micF but not in the ompR mutants. It was concluded that the frx gene is involved in thermoregulation by activating ompC expression, dependent on OmpR in the absence of EnvZ. However, we believe that the phenotype observed may be a secondary effect due to crosstalk between Frx and OmpR since the PTS and two-component sensor-regulator systems both function by phosphotransfer involving histidine residues. Moreover, conserved histidine residues which are phosphorylated during PTS phosphotransfer are located in enzyme IIA which correspond to the fragment provided in trans. In the work presented here, the frx gene was therefore studied by a functional genomic approach to characterize the substrate transported.

Functional genomics in E. coli The Frx PTS transporter

21

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

Materials and Methods Gene disruptions were performed as described in the Common Methods section with primers listed in table 6. The resulting strains are listed in table 5. Phenotypic screening of the mutants for the frx and ybgG genes was then performed on GN2 Biolog plates [10]. 95 different carbon sources contained in separate wells of microtitle plates allow one to screen for differences in carbon utilization between the wild type and the mutant strains. Wells are inoculated at low bacterial concentration (600nm absorbancy of 0.075) in minimal media without a carbon source. Carbon source oxidization was followed indirectly through cell respiration by measuring the reduction of tetrazolium violet by NADH [10, 11]. This irreversible reaction results in a purple coloration which can be followed by eye. Table 5. Strains and plasmids.

Strain/plasmid Strain LJ 3001 (BW25113) LJ 3030 LJ 3031

Genotype

Reference

lacIq rrnBT14 ∆lacZWJ16 hsdR514 ∆araBADAH33 ∆rhaBADLD78 BW25113 ∆frx (∆hrsA) BW25113 ∆ybgG

Datsenko and Wanner, 2000 This study This study

Plasmid pRK415

OriV pLac-MCS OriT tetR tetA trfA

Keen et al, 1988

Growth experiments were conducted at 37°C in M9 minimal media containing 0.5% melibiose or sodium acetate pH 7 as sole carbon sources. Cultures were inoculated at 1/500th of the volume with cells grown in LB broth. Cells were centrifuged and washed twice with minimal media with no carbon source prior to inoculation. Transport assays with acetate were performed as described in the Common Methods section over a 70 minute time period. The acetate concentration used was 100 µM (1.5 µCi/µmol) and the cell density was 0.5 absorbancy units. For both growth and transport experiments, the results presented are the averages of two repetitions. Melibiose fermentation was tested by growing the bacteria at 37°C on MacConkey agar plates containing 1% melibiose. The frx gene and its promoter were cloned in pRK415, a low-copy number plasmid [38], using the primers listed in table 6. The gene was amplified by PCR and the resulting fragment was purified with a Qiagen kit and cut with BamHI and HindIII restriction enzymes. Ligation was performed at 15°C overnight with the cut PCR fragment and pRK415 vector in a ratio 1:1 and Functional genomics in E. coli The Frx PTS transporter

22

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

was then transformed in the strain DH5α. The plasmid containing the insert was then transferred after extraction to the frx mutant for complementation studies. Cloning of the gene in a high-copy number plasmid (pBluescript) was also attempted unsuccessfully, possibly because of lethality problems due to high expression levels of the transporter. Cloning in the low-copy number plasmid pACYC184 was also tried but resulted twice in genetic recombination of the vector with the inserted fragment. Table 6. Primers used for cloning and for generation and verification of BW25113 mutants.

Gene

Primers (5’Æ 3’) Mutant generation and verification

frx (hrsA)

Generation GATGCGGTCGCCTGCGAACTGAATTAAATAAACCAGAATGACCAGGTGTAGGCTGGAGC TGCTTC (forward) CTTAATGACGATATAAATAATCAATGATAAAACTTTCGAATATCCATATGAATATCCTCC TTAG (reverse) Verification abc-yaeE 3: CGTTACTTGCGAGTGACAGC abc-yaeE 4: GCATGTGACGCTAGTATCGC

ybgG

Generation TGGCGTAATGCCATAAACAAAAAGGAAACGACGATGAAAGCAGTAGTAGGCTGGAGCT GCTTC (forward) CATATGTGGCTTCGGTTTGATTGCTATTCAGGCAAGCCGGTAACTCATATGAATATCCTC CTTAG (reverse) Verification ybgG 3: GCGGTTAAGCATGGCAACTA ybgG 4: GCCCTTTGTCAACAATCTGG

Cloning AAGCTTCCGGTATAAGGGCTTGTGTC (forward) frx GGATCCCAGTCGGCACAGGATCTCTT (reverse)

α-galactosidase activity was assayed in whole cells since the characterized enzymes are unstable in solution and require addition of NADH for stabilization when isolated in cell extracts. Cells grown in M9 minimal media at 30 and 37°C with 0.5% melibiose as a sole carbon source were harvested in the mid-exponential phase and washed three times at room temperature with M9 media without any carbon source. The cells were concentrated ten times prior to assay and used extemporaneously. Assays were performed in triplication at 37°C with 0.08% p-nitrophenyl α-galactoside as substrate and a cell density of 1 absorbancy unit (final volume of 1ml). Reactions were stopped in the linear range by adding 500 µl of 1M sodium carbonate. Cells were centrifuged and nitrophenol absorption was measured in the supernatant at 420nm. The results presented are the average of two independent studies. Functional genomics in E. coli The Frx PTS transporter

23

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

Cloning of the ybgG gene and of the entire frx/ybgG operon was also attempted in order to overexpress the proteins and increase the activity of the YbgG enzyme relatively to that of the wild-type. Disruption of the melAB genes encoding an α-galactosidase and a sodium/melibiose symporter was also attempted in the frx and ybgG mutants as well as in the wild-type to check possible YbgG α-galactosidase activity without background. Both attempts were unsuccessful.

Results Comparative analysis of the metabolic patterns of the frx and ybgG mutants relative to the wild type strain on Biolog GN2 plates [10] revealed two reproducible differences. Formation of the purple color corresponding to the reduction of tetrazolium violet during cellular respiration [11] was delayed for the frx and ybgG mutants relatively to the wild type strain in the well containing melibiose. No oxidization of acetate was observed for the frx mutant suggesting that the frx mutation also affected acetate utilization. Since Biolog plates only measure cellular respiration and the correlation to bacterial growth is indirect, complementary experiments were undertaken. Fermentation of melibiose was tested on MacConkey agar plates, but no significative difference was observed between the wild type and the frx mutant. Growth of the frx mutant was analyzed relative to that of the wild type in M9 minimal media using melibiose (figure 4A) or acetate (figure 4B) as sole carbon sources. When melibiose was the sole carbon source (figure 4A), E. coli exhibited a lag phase of 40 hours with the growth conditions described in figure 4. A lag phase remained when the cells were induced in minimal media with melibiose or when they were inoculated at higher concentration (data not shown). Growth of the frx mutant was significantly different from that of the wild type strain even though the growth rates during the exponential phase do not exhibit a striking difference. The possibility of occurrence of spontaneous mutations which could explain the long lag phase followed by a high growth rate is not likely since the lag phase is still observed when the cells are re-inoculated to fresh media. Growth of the frx mutant complemented with the frx gene cloned on pRK415 gave irreproducible results. When acetate was used as a carbon source (figure 4B), growth of the frx mutant was slightly delayed relatively to the wild type. However, exponential growth rates were the same.

Functional genomics in E. coli The Frx PTS transporter

24

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

3,0

2,0

A

B Absorbance (600nm)

Absorbance (600nm)

2,5

2,0

1,5

1,0

1,5

1,0

0,5 0,5

0,0 0

30

40

50

0,0

60

0

10

20

30

40

Time (hours)

Time (hours)

Figure 4 - Growth of wild type BW25113 (Y) and the isogenic ∆frx mutant () in M9 minimal media containing 0.5% melibiose (A) or sodium acetate pH 7 (B) as the sole carbon source. Cultures were inoculated at 1/500th of the volume with cells grown in LB media and washed twice prior to inoculation.

To confirm the phenotypes

700

observed by Biolog plate and growth 600

performed with [14C]acetate. No significative difference in uptake was observed between the wild type and the frx mutant (figure 5). It was therefore concluded that the Frx

Acetate uptake (pmol / mg dry weight )

experiments, a transport assay was

500

400

300

200

100

transporter is not responsible for uptake

of

acetate.

The

effect

observed on growth is probably an indirect regulatory effect. Utilization

of

acetate

is

dependent on the glyoxylate bypass, a shunt of the Tri-Carboxylate Acid

0 0

10

20

30

40

50

60

70

80

Time (min)

Figure 5. Uptake of 1,2-[14C]acetate by wild type BW25113 (Y) and ∆frx (U) cells. Cells grown in M9 minimal media were prepared as described in the Methods section. Assays were performed at 37°C in 1 mL over a 70 minute time interval. The acetate concentration used was 100 µM (1.5 µCi/µmol) and the cell density was at 0.5 absorbancy units (A600). Values are expressed in picomoles of acetate retained per milligram of bacterial dry weight.

(TCA) cycle [19]. The ace operon which encodes the two enzymes of the glyoxylate bypass (isocitrate lyase and malate synthase) and a transcriptional regulator is therefore a potential regulatory target. However, many mutations which stress the cells cause slower growth on acetate since acetate is a poor carbon source for E. coli (David Laporte, personal communication). It is Functional genomics in E. coli The Frx PTS transporter

25

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

therefore not likely that the frx gene has a transcriptional effect on the ace operon even though it has not been confirmed by reporter gene lacZ fusion. Transport of melibiose was not assayed since radioactive melibiose is not commercially available. Experiments were also focused on solving the function of ybgG since it would reveal the substrate transported by Frx. Figure 6 shows that α-galactosidase activity was decreased by about two fold when cells were grown at 37°C relatively to when they were grown at 30°C. This observation is consistent with the thermorepression of the melibiose operon [13, 30] which encodes an α-galactosidase (melA), a sodium/melibiose symporter (melB) and a transcriptional regulator (melR) [42]. Results have been examined for the cells grown at 37°C to minimize the effect of the background due to the MelA activity. The frx mutant has an α-galactosidase activity equal to that of the wild type, therefore suggesting that frx has no regulatory effect on the transcription of the melibiose operon. However, the activity of the ybgG mutant is decreased of 25% relatively to the wild type. This observation suggests that YbgG has α-galactosidase activity. Such a function would be consistent with prior results which suggested that frx is involved in melibiose metablism since melibiose is an α-galactoside.

5

Activity

(Miller units)

4 3 2 1 0

WT

∆YbgG ∆Frx 1

WT

∆YbgG ∆Frx 2

30°C 37°C Figure 6 – Whole-cell α-galactosidase activity of wild type BW25113 strain and the isogenic frx and ybgG mutants. Cells were cultured in M9 minimal media at the temperatures indicated with 0.5% melibiose serving as sole carbon source. Assays were performed with a final cell density of 1 absorbancy unit. Activity was measured by following nitrophenol formation resulting from the cleavage of p-nitrophenyl-α-galactopyranoside. Reactions were stopped in the linear range by adding sodium carbonate; cell suspensions were then centrifuged and the absorbancy of the supernatant was measured at 420 nm.

Functional genomics in E. coli The Frx PTS transporter

26

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

Discussion No molecular characterization of an acetate transporter has been reported [19]. Transport studies revealed that Frx is not responsible for acetate uptake. Despite the growth difference observed, a direct regulatory role on transcription of the ace operon is unlikely. The effect observed is probably due to a secondary regulatory consequence of the frx mutation which remains to be determined. Global results suggested that the Frx / YbgG system is involved with melibiose metabolism. The inactivation of the Frx PTS transporter was shown to reduce growth on melibiose therefore suggesting that melibiose might be transported. However, the MelB sodium/melibiose permease [7] is still active in the strains studied. It therefore causes high background activity which might reduce the relative effect of the frx mutation. Moreover, αgalactosidase activity was also depressed relative to the wild type in the ybgG mutant at 37°C even though background levels were high due to MelA α-galactosidase activity. The functional scheme proposed for the Frx / YbgG system is therefore 1) transport of melibiose across the cytoplasmic membrane by Frx and phosphorylation in position 6 of the galactosyl moiety and 2) cleavage of melibiose-phosphate by YbgG in the cytoplasm to form monosaccharides which can catabolised. The Frx / YbgG system might be a complementary system to the one encoded by the mel operon responsible for uptake and degradation of melibiose under different conditions. Induction at high temperature (above 37°C) seems an interesting idea to test since the melibiose operon is thermorepressed [13, 30]. However, cells will have to be grown with a different carbon source because E. coli can not grow at 42°C with melibiose as a sole carbon source. Glycerol is probably a good choice as it is not transported and metabolised via PTS systems. Further experiments are required to solve the function of the Frx / YbgG system with certainty. The most important control is the repetition of the studies presented here (growth and α-galactosidase activity) after disrupting the melAB genes in the wild type and the frx and ybgG mutant strains in order to eliminate the background. Transport and trans-phosphorylation studies using [3H]melibiitol could also be undertaken if the system also functions with the reduced sugar as is the case for the melibiose operon. [3H]melibiitol can be obtained by reduction of melibiose by tritiated sodium borohydride. Specificity of the Frx / YbgG system for other disaccharides will also need to be examined using the corresponding p-nitrophenyl susbstrates. Comparison of the results will allow one to determine the catalytic specificity of the YbgG enzyme. Functional genomics in E. coli The Frx PTS transporter

27

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

Interestingly, no system with high similarity to the frx / ybgG operon is found in E. coli’s close relative Salmonella typhimurium. Although further phylogenetic analyses are required, this observation suggests that the Frx / YbgG system is not essential. Considering the presence of the melibiose operon in both organisms , this assumption is consistent with the substrate proposed.

Functional genomics in E. coli The Frx PTS transporter

28

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

Conclusion The work presented here concerns two independent studies of functional genomics in Escherichia coli. The molecular identification of an ABC transporter specific for L- and Dmethionine which is the prototype for a novel family of the ABC superfamily is reported. Data which suggests that the Frx PTS transporter is involved in melibiose metabolism is also detailed in this thesis although further experiments are required to solve the function and the specificity with certainty. The common theme to these studies which is also one of the main research interests of Dr Saier’s laboratory is nutrient transport in bacteria. Understanding transport mechanisms in bacteria is of major interest to the scientific community because of their major pharmaceutical and biotechnological applications. Transport proteins are important targets for drug discovery since certain transporters are responsible for the uptake of essential nutrients for virulence and pathogenicity. Moreover, efflux pumps have been characterized as drug and toxic substance expellers and are potentially responsible for bacterial multi-drug resistance. Understanding those transport mechanisms and their regulation and specificity in pathogenic bacteria is therefore a step towards the isolation of new antibiotics for which limited resistance occurs. Bacterial transport mechanisms also have biotechnological importance since drugs and proteins are produced by genetic engineering in bacteria. One of the problems of protein overexpression is the extraction of the substance of interest out of the cell without disturbing the production level. The understanding of transport physiology is therefore a key to strain improvement and to better yields. This 6-month internship which finalizes my studies in the Department of Biochemistry of INSA Lyon was a perfect initiation to a long-term research program. Acquisition of theoretical and practical competence in microbiology and molecular biology is a great complement to the education dispensed by the INSA Biochemistry Department. Undoubtedly, this specialization will be very helpful in the future.

Functional genomics in E. coli

29

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

References 1. Aboulwafa, M., and M. H. Saier, Jr. 2002. Analyses of soluble activities of the glucose and mannitol Enzymes II of the phosphotransferase system in Escherichia coli: Evidence for two physically distinct forms. under publication. 2. Aboulwafa, M., and M. H. Saier, Jr. 2002. Dependency of sugar transport and phosphorylation by the phosphoenolpyruvate-dependent phosphotransferase system on membranous phosphatidyl glycerol in Escherichia coli: studies with a pgsA mutant lacking phosphatidyl glycerophosphate synthase. under publication. 3. Altschul, S. F., T. L. Madden, A. A. Schaffer, J. Zhang, Z. Zhang, W. Miller, and D. J. Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25:3389-402. 4. Andersen, J., S. A. Forst, K. Zhao, M. Inouye, and N. Delihas. 1989. The function of micF RNA. micF RNA is a major factor in the thermal regulation of OmpF protein in Escherichia coli. J Biol Chem 264:17961-70. 5. Ayling, P. D., and E. S. Bridgeland. 1972. Methionine transport in wild-type and transportdefective mutants of Salmonella typhimurium. J Gen Microbiol 73:127-41. 6. Ayling, P. D., T. Mojica-a, and T. Klopotowski. 1979. Methionine transport in Salmonella typhimurium: evidence for at least one low-affinity transport system. J Gen Microbiol 114:227-46. 7. Bassilana, M., T. Pourcher, and G. Leblanc. 1988. Melibiose permease of Escherichia coli. Characteristics of co- substrates release during facilitated diffusion reactions. J Biol Chem 263:9663-7. 8. Berlyn, M. K. 1998. Linkage map of Escherichia coli K-12, edition 10: the traditional map. Microbiol Mol Biol Rev 62:814-984. 9. Betteridge, P. R., and P. D. Ayling. 1975. The role of methionine transport-defective mutations in resistance to methionine sulphoximine in Salmonella typhimurium. Mol Gen Genet 138:41-52. 10. Bochner, B. 1993. Advances in the identification of bacteria and yeast. Am Clin Lab 12:6. 11. Bochner, B. R., and M. A. Savageau. 1977. Generalized indicator plate for genetic, metabolic, and taxonomic studies with microorganisms. Appl Environ Microbiol 33:434-44. 12. Brubaker, R. R. 1972. The genus Yersinia: biochemistry and genetics of virulence. Curr Top Microbiol Immunol 57:111-58. 13. Burstein, C., and A. Kepes. 1985. The melibiose permease system of Escherichia coli K12. Biochimie 67:59-67. Functional genomics in E. coli References

30

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

14. Cai, S. J., and M. Inouye. 2002. EnvZ-OmpR interaction and osmoregulation in Escherichia coli. J Biol Chem 277:24155-61. 15. Chanyangam, M., A. L. Smith, S. L. Moseley, M. Kuehn, and P. Jenny. 1991. Contribution of a 28-kilodalton membrane protein to the virulence of Haemophilus influenzae. Infect Immun 59:600-8. 16. Chauvaux, S., I. T. Paulsen, and M. H. Saier, Jr. 1998. CcpB, a novel transcription factor implicated in catabolite repression in Bacillus subtilis. J Bacteriol 180:491-7. 17. Cherepanov, P. P., and W. Wackernagel. 1995. Gene disruption in Escherichia coli: TcR and KmR cassettes with the option of Flp-catalyzed excision of the antibiotic-resistance determinant. Gene 158:9-14. 18. Chung, Y. J., and M. H. Saier, Jr. 2001. SMR-type multidrug resistance pumps. Curr Opin Drug Discov Devel 4:237-45. 19. Clarck, D. P., and J. E. Cronan. 1996. Two-carbon compounds and fatty acids as carbon sources. In F. C. Neidhardt (ed.), Escherichia coli and Salmonella: Cellular and Molecular Biology, vol. 1. ASM Press, Washington DC. 20. Contesse, G., M. Crépin, F. Gros, A. Ullmann, and J. Monod. 1969. On the mechanism of catabolite repression, p. 401-415. In J. Beckwith and D. Zipser (ed.), The Lactose Operon. Cold Spring Harbor, NY. 21. Cottam, A. N., and P. D. Ayling. 1989. Genetic studies of mutants in a high-affinity methionine transport system in Salmonella typhimurium. Mol Gen Genet 215:358-63. 22. Dassa, E., and P. Bouige. 2001. The ABC of ABCS: a phylogenetic and functional classification of ABC systems in living organisms. Res Microbiol 152:211-29. 23. Datsenko, K. A., and B. L. Wanner. 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci U S A 97:6640-5. 24. Djordjevic, G. M., J. H. Tchieu, and M. H. Saier, Jr. 2001. Genes involved in control of galactose uptake in Lactobacillus brevis and reconstitution of the regulatory system in Bacillus subtilis. J Bacteriol 183:3224-36. 25. Gal, J., A. Szvetnik, R. Schnell, and M. Kalman. 2002. The metD D-Methionine Transporter Locus of Escherichia coli Is an ABC Transporter Gene Cluster. J Bacteriol 184:4930-2. 26. Galtier, N., M. Gouy, and C. Gautier. 1996. SEAVIEW and PHYLO_WIN: two graphic tools for sequence alignment and molecular phylogeny. Comput Appl Biosci 12:543-8.

Functional genomics in E. coli References

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

27. Greene, R. C. 1996. Biosynthesis of Methionine, p. 542-560. In F. C. Neidhardt (ed.), Escherichia coli and Salmonella: Cellular and Molecular Biology, 2nd ed, vol. 1. ASM Press, Washington DC. 28. Grundy, C. E., and P. D. Ayling. 1992. Fine structure mapping and complementation studies of the metD methionine transport system in Salmonella typhimurium. Genet Res 60:16. 29. Grundy, F. J., and T. M. Henkin. 1998. The S box regulon: a new global transcription termination control system for methionine and cysteine biosynthesis genes in gram-positive bacteria. Mol Microbiol 30:737-49. 30. Hanatani, M., H. Yazyu, S. Shiota-Niiya, Y. Moriyama, H. Kanazawa, M. Futai, and T. Tsuchiya. 1984. Physical and genetic characterization of the melibiose operon and identification of the gene products in Escherichia coli. J Biol Chem 259:1807-12. 31. Harley, K. T., and M. H. Saier, Jr. 2000. A novel ubiquitous family of putative efflux transporters. J Mol Microbiol Biotechnol 2:195-8. 32. Hendricks, J. K., and H. L. Mobley. 1997. Helicobacter pylori ABC transporter: effect of allelic exchange mutagenesis on urease activity. J Bacteriol 179:5892-902. 33. Kadner, R. J. 1975. Regulation of methionine transport activity in Escherichia coli. J Bacteriol 122:110-9. 34. Kadner, R. J. 1977. Transport and utilization of D-methionine and other methionine sources in Escherichia coli. J Bacteriol 129:207-16. 35. Kadner, R. J. 1974. Transport systems for L-methionine in Escherichia coli. J Bacteriol 117:232-41. 36. Kadner, R. J., and W. J. Watson. 1974. Methionine transport in Escherichia coli: physiological and genetic evidence for two uptake systems. J Bacteriol 119:401-9. 37. Kadner, R. J., and H. H. Winkler. 1975. Energy coupling for methionine transport in Escherichia coli. J Bacteriol 123:985-91. 38. Keen, N. T., S. Tamaki, D. Kobayashi, and D. Trollinger. 1988. Improved broad-hostrange plasmids for DNA cloning in gram-negative bacteria. Gene 70:191-7. 39. Krogh, A., B. Larsson, G. von Heijne, and E. L. Sonnhammer. 2001. Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J Mol Biol 305:567-80. 40. Kuan, G., E. Dassa, W. Saurin, M. Hofnung, and M. H. Saier, Jr. 1995. Phylogenetic analyses of the ATP-binding constituents of bacterial extracytoplasmic receptor-dependent ABC-type nutrient uptake permeases. Res Microbiol 146:271-8. Functional genomics in E. coli References

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

41. Kundig, W., S. Ghosh, and S. Roseman. 1964. Phosphate bound to histidine in a protein as an intermediate as a novel phosphotransferase system. Proc. Natl. Acad. Sci. USA 52:10671074. 42. Lin, E. C. C. 1996. Dissimilatory pathways for sugars, polyols and carboxylates, p. 309-311. In F. C. Neidhardt (ed.), Escherichia coli and Salmonella: Cellular and Molecular Biology, vol. 1. ASM Press, Washington DC. 43. Liu, R., T. W. Blackwell, and D. J. States. 2001. Conformational model for binding site recognition by the E.coli MetJ transcription factor. Bioinformatics 17:622-33. 44. Ma, Q., Y. Zhai, J. C. Schneider, T. M. Ramseier, and M. H. Saier, Jr. 2002. Protein secretion systems of Pseudomonas species. under publication. 45. Page, R. D. 1996. TreeView: an application to display phylogenetic trees on personal computers. Comput Appl Biosci 12:357-8. 46. Panina, E. M., A. A. Mironov, and M. S. Gelfand. 2001. Comparative analysis of FUR regulons in gamma-proteobacteria. Nucleic Acids Res 29:5195-206. 47. Pattery, T., J. P. Hernalsteens, and H. De Greve. 1999. Identification and molecular characterization of a novel Salmonella enteritidis pathogenicity islet encoding an ABC transporter. Mol Microbiol 33:791-805. 48. Poland, J., and P. D. Ayling. 1984. Methionine and glutamine transport systems in Dmethionine utilising revertants of Salmonella typhimurium. Mol Gen Genet 194:219-26. 49. Powell, B. S., D. L. Court, T. Inada, Y. Nakamura, V. Michotey, X. Cui, A. Reizer, M. H. Saier, Jr., and J. Reizer. 1995. Novel proteins of the phosphotransferase system encoded within the rpoN operon of Escherichia coli. Enzyme IIANtr affects growth on organic nitrogen and the conditional lethality of an erats mutant. J Biol Chem 270:4822-39. 50. Quan, J. A., B. L. Schneider, I. T. Paulsen, M. Yamada, N. M. Kredich, and M. H. Saier, Jr. 2002. Regulation of carbon utilization by sulfur availability in Escherichia coli and Salmonella typhimurium. Microbiology 148:123-31. 51. Reizer, J., A. Reizer, M. J. Merrick, G. Plunkett, 3rd, D. J. Rose, and M. H. Saier, Jr. 1996. Novel phosphotransferase-encoding genes revealed by analysis of the Escherichia coli genome: a chimeric gene encoding an Enzyme I homologue that possesses a putative sensory transduction domain. Gene 181:103-8. 52. Reizer, J., A. Reizer, M. H. Saier, Jr., and G. R. Jacobson. 1992. A proposed link between nitrogen and carbon metabolism involving protein phosphorylation in bacteria. Protein Sci 1:722-6.

Functional genomics in E. coli References

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

53. Rice, P., I. Longden, and A. Bleasby. 2000. EMBOSS: the European Molecular Biology Open Software Suite. Trends Genet 16:276-7. 54. Saier, M. H., Jr. 2000. A functional-phylogenetic classification system for transmembrane solute transporters. Microbiol Mol Biol Rev 64:354-411. 55. Saier, M. H., Jr. 1999. A functional-phylogenetic system for the classification of transport proteins. J Cell Biochem Suppl:84-94. 56. Saier, M. H., Jr., S. Chauvaux, J. Deutscher, J. Reizer, and J. J. Ye. 1995. Protein phosphorylation and regulation of carbon metabolism in gram- negative versus gram-positive bacteria. Trends Biochem Sci 20:267-71. 57. Saier, M. H., Jr., and I. T. Paulsen. 2000. Whole genome analyses of transporters in spirochetes: Borrelia burgdorferi and Treponema pallidum. J Mol Microbiol Biotechnol 2:393-9. 58. Saint-Girons, I., N. Duchange, G. N. Cohen, and M. M. Zakin. 1984. Structure and autoregulation of the metJ regulatory gene in Escherichia coli. J Biol Chem 259:14282-5. 59. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular Cloning: A Laboratory Manual., 2nd Edition ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory. 60. Schilling, C. H., L. Held, M. Torre, and M. H. Saier, Jr. 2000. GRASP-DNA: a web application to screen prokaryotic genomes for specific DNA-binding sites and repeat motifs. J Mol Microbiol Biotechnol 2:495-500. 61. Sekowska, A., H. F. Kung, and A. Danchin. 2000. Sulfur metabolism in Escherichia coli and related bacteria: facts and fiction. J Mol Microbiol Biotechnol 2:145-77. 62. Sekowska, A., S. Robin, J. J. Daudin, A. Henaut, and A. Danchin. 2001. Extracting biological information from DNA arrays: an unexpected link between arginine and methionine metabolism in Bacillus subtilis. Genome Biol 2. 63. Snyder, L., and W. Champness. 1997. Introduction to macromolecular synthesis: Gene expression, p. 60-61, Molecular genetics of bacteria. ASM Press, Washington DC. 64. Sonnhammer, E. L., G. von Heijne, and A. Krogh. 1998. A hidden Markov model for predicting transmembrane helices in protein sequences. Proc Int Conf Intell Syst Mol Biol 6:175-82. 65. Tchieu, J. H., V. Norris, J. S. Edwards, and M. H. Saier, Jr. 2001. The complete phosphotranferase system in Escherichia coli. J Mol Microbiol Biotechnol 3:329-46. 66. Thanbichler, M., B. Neuhierl, and A. Bock. 1999. S-methylmethionine metabolism in Escherichia coli. J Bacteriol 181:662-5.

Functional genomics in E. coli References

Saier Lab – UC San Diego March – August 2002

Biochemistry Department

67. Thompson, J. D., T. J. Gibson, F. Plewniak, F. Jeanmougin, and D. G. Higgins. 1997. The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res 25:4876-82. 68. Tusnady, G. E., and I. Simon. 2001. The HMMTOP transmembrane topology prediction server. Bioinformatics 17:849-50. 69. Tusnady, G. E., and I. Simon. 1998. Principles governing amino acid composition of integral membrane proteins: application to topology prediction. J Mol Biol 283:489-506. 70. Utsumi, R., T. Horie, A. Katoh, Y. Kaino, H. Tanabe, and M. Noda. 1996. Isolation and characterization of the heat-responsive genes in Escherichia coli. Biosci Biotechnol Biochem 60:309-15. 71. Weissbach, H., and N. Brot. 1991. Regulation of methionine synthesis in Escherichia coli. Mol Microbiol 5:1593-7. 72. Winnen, B., R. N. Hvorup, and M. H. Saier, Jr. 2002. The Tripartite Tricarboxylate Transporter (TTT) family. under publication. 73. Winnen, B., G. Lorca, and M. H. Saier, Jr. 2002. Identification of a high affinity, broad specificity L-amino acid transporter in Bacillus subtilis. under publication. 74. Yamaguchi, K., and M. Inouye. 1988. Lipoprotein 28, an inner membrane protein of Escherichia coli encoded by nlpA, is not essential for growth. J Bacteriol 170:3747-9. 75. Yu, F., S. Inouye, and M. Inouye. 1986. Lipoprotein-28, a cytoplasmic membrane lipoprotein from Escherichia coli. Cloning, DNA sequence, and expression of its gene. J Biol Chem 261:2284-8. 76. Zhang, Z., M. Smith, and M. H. Saier, Jr. 2002. The ascorbate transporter of Escherichia coli. under publication. 77. Zhou, X., R. N. Hvorup , and M. H. Saier, Jr. 2002. An automated program to screen databases for members of protein families. under publication. 78. Zhou, X., N. Yang, C. Tran, R. N. Hvorup , and M. H. Saier, Jr. 2002. Web-based programs for TMS alignment display and grouping. under publication.

Functional genomics in E. coli References

Saier Lab – UC San Diego March – August 2002

Abstract

The work presented here concerns two functional genomic studies of nutrient transporters in Escherichia coli. I report the molecular identification of an ABC-type transporter which transports both Land D-methionine but not other natural amino-acids. This system is the first functionally characterized member of a novel family of transporters within the ABC superfamily. We have designated this family the Methionine Uptake Transporter (MUT) family. The proteins that comprise the transporters of this family were analysed phylogenetically revealing the probable existence of several sequence divergent primordial paralogues, no more than two of which have been transmitted to any currently sequenced organism. There is evidence that several members of the MUT family are regulated by the MetJ regulator or S-box termination control systems therefore suggesting that many members of this family are involved in sulfur compound, possibly methionine, transport. Evidence suggesting that the Frx PTS transporter and the YbgG enzyme are involved in melibiose metabolism is also presented. This system would be a second operon for melibiose utilization for which the conditions of expression remain to be determined. Although further experiments are required for the complete functional characterization, the proposed scheme is transport and phosphorylation of melibiose by Frx and cleavage of the phosphorylated form to monosaccharides by YbgG in the cytoplasm.