HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 30, 26769-26778, July 26, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/30/26769    most recent M202882200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Creuzenet, C. Articles by Lam, J. S. Search for Related Content PubMed PubMed Citation Articles by Creuzenet, C. Articles by Lam, J. S. Social Bookmarking                      What's this?

Structure-Function Studies of Two Novel UDP-GlcNAc C6 Dehydratases/C4 Reductases

VARIATION FROM THE SYK DOGMA^*

Carole Creuzenet^§¶, Robert V. Urbanic^§, and Joseph S. Lam^§

From the  Department of Microbiology and Immunology, University of Western Ontario, London, Ontario N6A 5C1, Canada and the ^§ Department of Microbiology, University of Guelph, Guelph, Ontario N1G 2W1, Canada

Received for publication, March 25, 2002, and in revised form, May 8, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Two subfamilies of UDP-GlcNAc C6 dehydratases were recently identified. FlaA1, a short soluble protein that exhibits a typical SYK catalytic triad, characterizes one of these subfamilies, and WbpM, a large membrane protein that harbors an altered SMK triad that was not predicted to sustain activity, represents the other subfamily. This study focuses on investigating the structure and function of these C6 dehydratases and the role of the altered triad as well as additional amino acid residues involved in catalysis. The significant activity retained by the FlaA1 Y141M triad mutant and the low activity of the WbpM M438Y mutant indicated that the methionine residue was involved in catalysis. A Glu⁵⁸⁹ residue, which is conserved only within the large homologues, was shown to be essential for activity in WbpM. Introduction of this residue in FlaA1 enhanced the activity of the corresponding V266E mutant. Hence, this glutamate residue might be responsible for the retention of catalytic efficiency in the large homologues despite alteration of their catalytic triad. Mutations of residues specific for the short homologues (Asp⁷⁰, Asp¹⁴⁹-Lys¹⁵⁰, Cys¹⁰³) abolished the activity of FlaA1. Among them, C103M prevented dimerization but did not significantly affect the secondary structure. The fact that we could identify subfamily-specific residues that are essential for catalysis suggested an independent evolution for each subfamily of C6 dehydratases. Finally, the loss of activity of the FlaA1 G20A mutant provided evidence that a cofactor is involved in catalysis, and kinetic study of the FlaA1 H86A mutant revealed that this conserved histidine is involved in substrate binding. None of the mutations investigated altered the substrate, product, and function specificity of these enzymes.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

FlaA1 and WbpM are two novel bifunctional UDP-GlcNAc C6 dehydratases/C4 reductases that were recently characterized at the biochemical level (1, 2). Both enzymes catalyze the stereo-specific conversion of UDP-GlcNAc to Qui2NAc¹ via the formation of a 4-keto, 6-deoxy intermediate. They are both highly specific for their substrate, UDP-GlcNAc, and exhibit no activity with closely related substrates, such as UDP-Glc, UDP-Gal, or UDP-GalNAc, or with substrates of other known C6 dehydratases such as GDP-mannose or dTDP-glucose. This is consistent with the fact that, despite their dehydratase activity, FlaA1 and WbpM exhibit very limited sequence similarity with other known C6 dehydratases, such as GDP-D-mannose (3-6) and dTDP- and CDP-D-glucose dehydratases (7-14). In contrast, they are homologous to C4 epimerases such as GalE from Escherichia coli (45% homology) (15) and WbpP from P. aeruginosa (49% homology) (16). The molecular basis for their unexpected functional specificity is not understood to date. Hence, FlaA1 and WbpM are representative members of a larger family of bifunctional dehydratases/reductases that is characterized by the presence of five conserved domains organized in the same pattern (1). One of these motifs corresponds to the signature for nucleotide-binding proteins. Another corresponds to an -helicoidal segment, where the typical SYK catalytic triad that defines enzymes of the short chain dehydrogenase/reductase (SDR) family is found (17-19).

Based on the size and cellular localization of the enzymes as well as the presence or absence of a SYK catalytic triad, the family of bifunctional dehydratases/reductases can be further divided into two subfamilies. FlaA1 is a representative member of the subfamily of short, soluble enzymes that possess a SYK catalytic triad. It is a 37-kDa soluble protein that is found in Helicobacter pylori (20, 21) and has homologues in Campylobacter jejuni (22, 23) as well as Caulobacter crescentus (24). On the contrary, WbpM is a large (75-kDa) membrane protein (2) that exhibits an altered catalytic triad composed of SMK residues. WbpM is found in Pseudomonas aeruginosa (25-27) and has homologues in numerous medically relevant bacteria, including Bordetella pertussis (WlpL) (28), Yersinia enterocolitica serotype O:3 (WbcP) (29), and Staphylococcus aureus (CapD) (30-32). WbpM is the only enzyme of this subfamily that has been characterized at the biochemical level. In addition, these large homologues have been shown to be essential for the production of surface carbohydrates that contain Qui2NAc or a derivative and are all associated with bacterial virulence. Understanding their mechanism of action is a prerequisite to the rational design of inhibitors with high specificity toward these enzymes and a broad spectrum of activity against all of the bacteria that harbor the homologues.

In the ever expanding SDR protein family, the level of identity between enzymes is rather low (15-30%), and the tyrosine of the catalytic triad has been observed as one of the few absolutely conserved residues (17). Functional and structural studies of C4 epimerases and of other SDR enzymes (15, 33-35) have confirmed that the SYK triad found in enzymes of the SDR family plays an essential role in the formation of the 4-keto intermediate, a reaction step shared by most members of the SDR enzymes, notwithstanding their ultimate biochemical function (4, 9, 33, 34, 36). In particular, the tyrosine residue acts as a catalytic base that allows the transfer of a hydrogen atom from the 4'-hydroxyl of the sugar ring to an acceptor nucleotide cofactor such as NAD⁺ or NADP⁺ (15, 33, 37-39). In agreement with the widely accepted "SYK dogma," alteration of the tyrosine residue in the catalytic triad has resulted in inactive mutants as seen in the UDP-Gal C4 epimerase GalE S124A/Y149F and Y149F mutants (34, 35) or in other SDR enzymes (17, 40). At most, 0.25% of wild-type catalytic activity was reported using a Y/C dehydrogenase mutant (41). Such an essential role of the tyrosine in the catalytic triad was also confirmed in the case of C6 dehydratases (9, 12). Hence, it is intriguing that WbpM and its large homologues have the capacity to perform catalysis via the formation of a 4-keto intermediate, since they contain an altered catalytic triad lacking the central tyrosine. This suggests that, in this subfamily of UDP-GlcNAc C6 dehydratases, another residue must assume the role of the catalytic base and/or that the catalytic mechanism involved might be slightly different. A site-directed mutagenesis approach was used to investigate the role of the tyrosine and methionine residues in the SYK and SMK catalytic triads present in FlaA1 and WbpM, respectively. The catalytic role of charged residues specific for each subfamily was also investigated.

Also, FlaA1 and WbpM exhibit the GXXGXXG signature for nucleotide binding proteins and also share a conserved alternating -helicoidal/-sheeted secondary structure that, in SDR proteins, delineates a Rossmann fold essential for nucleotide cofactor binding. The importance of this fingerprint sequence for nucleotide cofactor binding and activity has been demonstrated previously by the abolition of protein activity by single point mutations (42-44). Previous studies indicated that FlaA1, like C4 epimerases and unlike other C6 dehydratases, might recycle an internal nucleotide cofactor during catalysis, but no evidence for the existence of a bound cofactor could be obtained (1). Site-directed mutagenesis of residues potentially involved in Rossmann fold formation and cofactor binding was performed in FlaA1 to address this question.

Finally, most C4 epimerases and C6 dehydratases have been shown to exist as dimers and tetramers under physiological conditions, but the importance of oligomerization for activity is not fully understood (10, 12, 17, 45, 46). Previous studies on dehydrogenases, epimerases, and dehydratases showed that the interactions between subunits involved mostly the N-terminal moiety of the proteins and more specifically the conserved long -helices 4 and 5 (17, 44, 47). For clarity, these helices correspond to amino acid residues 97-120 and 140-160 in FlaA1 based on secondary structure prediction using the ExPASy molecular biology software (available on the World Wide Web at www.expasy.ch). Because the N-terminal moiety of SDR enzymes is also devoted to cofactor binding, it is possible that oligomerization and cofactor binding are interdependent. Each subunit possesses its own cofactor-binding site, so that nucleotide cofactor binding should be stoichiometric with the amount of enzyme if it is independent of oligomerization. However, several studies report a stoichiometry of 1 molecule of cofactor per dimer of protein (12, 15, 46, 48), implying that dimerization might promote cofactor binding and hence be essential for activity. Moreover, the helix 5 that is involved in oligomerization contains the tyrosine and lysine residues of the catalytic triad. Hence, it is possible that participation of this helix in oligomer formation results in a slightly different structure at the level of the catalytic site, which could also have a significant effect on the activity. Consistent with this hypothesis is the observation that some SDR proteins crystallize as a dimer in an asymmetric unit, where substantial differences have been observed between the structures of the binding sites of two monomers (39), suggesting that dimerization might be responsible for fine tuning of the active site configuration. Hence, cofactor binding and oligomerization might be interconnected phenomena that affect enzymatic activity. In the case of FlaA1, the formation of a dimer under physiological conditions has been described previously, but the residues essential for the process have not been identified (1). In FlaA1 and other short homologues, the -helix 4, which is potentially involved in dimerization, contains two conserved cysteine residues (Cys¹⁰³ and Cys¹¹⁸). Their potential role for dimerization and stabilization of protein structure was investigated by site-directed mutagenesis.

This report presents the structure-function characterization of mutants of FlaA1 and WbpM with a view to explain their particular biochemical and catalytic properties. The effects of the mutations on protein structure and dimerization were investigated by circular dichroism, fluorescence spectroscopy, and gel filtration chromatography. The substrate specificity and physico-kinetic characteristics of each mutant were determined by capillary electrophoresis analysis of the reaction products. The results reveal novel essential catalytic residues and provide the basis for understanding the mechanism of action of these dehydratases/reductases that are devoid of the usual SYK triad.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- Unless stated otherwise, all chemical reagents used were from Sigma, and all cloning was performed in E. coli DH5. All of the kits or enzymes were used following the manufacturers' instructions. Protein quantitation was performed using the BCA reagent (Pierce).

Site-directed Mutagenesis-- Site-directed mutagenesis was performed using the QuikChange^TM procedure (Stratagene, La Jolla, CA) with modifications according to Wang and Malcolm (49). Complementary oligonucleotide primers were designed according to the manufacturer's instructions, and codons that resulted in minimal mispairing were used to introduce the desired mutations (Table I). The templates for mutagenesis were His-FlaA1-pET and His-S262-pET. S262 corresponds to a soluble truncated version of WbpM where the four N-terminal transmembrane domains have been removed (2). For clarity, WbpM refers to the soluble truncated form S262 throughout this study. In both templates, the gene of interest was cloned into a pET23 derivative (50) in the presence of an N-terminal hexahistidine tag (1, 2). After mutagenesis, DNA sequencing was performed on each mutant construct to confirm the presence of the desired mutations and exclude the existence of any unwanted secondary mutation. DNA sequencing was performed at the Laboratory Services Division of the University of Guelph.

Purification of Recombinant Proteins-- Proper clones carrying the desired mutations were transformed into CaCl₂-competent E. coli BL21(DE3)pLysS, and protein expression was carried out as described previously (1, 2). In some cases, conditions for protein expression were modified to circumvent the production of mutant proteins as inclusion bodies. Incubations were performed at 30 °C instead of 37 °C, and growth was performed in Tryptone/phosphate medium (51) instead of Terrific broth. The concentration of the inducer isopropyl-1-thio--D-galactopyranoside varied from 0.15 to 1 mM. The wild-type and mutant proteins were purified by low pressure nickel chelation chromatography as described elsewhere (1). The fractions containing pure protein were pooled, dialyzed against 50 mM Tris-HCl buffer, adjusted to pH 7, and concentrated using polyethylene glycol 8000 before use for activity assays.

Activity Assays and Capillary Electrophoresis Analysis of the Reaction Products-- Typically, reactions were carried out using 0.75 mM UDP-GlcNAc as a substrate. For WbpM and its mutants, 0.14 mM of NAD⁺ was added to the reactions. The reactions were incubated in 20 mM Tris, pH 7, at 37 °C, for FlaA1 and its mutants, or in bis-Tris-propane, pH 10, at 30 °C for WbpM and its mutants. The specific composition of each reaction mix and the reaction time are indicated in the legend of each figure. For the determination of kinetic parameters, reactions were performed in a total volume of 70 µl with 0.7-3.8 µg of pure enzyme. Incubations were performed at the optimum pH and temperature determined for each mutant as indicated in Table II. To account for the high and low K_m for UDP-GlcNAc of WbpM and FlaA1, respectively, the substrate concentration varied from 0.05 to 3.5 mM for mutants of WbpM and from 0.02 to 1.5 mM for mutants of FlaA1. The reactions were quenched by boiling for 6 min. To investigate the substrate specificity, UDP-GlcNAc was replaced by UDP-Glc, UDP-Gal, UDP-GalNAc, GDP-mannose, or dTDP-Glc.

Capillary electrophoresis (CE) analysis of the reaction products was performed as described elsewhere (2).

Determination of the K_m for NADH for the H86A Mutant of FlaA-- Determination of the K_m for NADH was performed spectrophotometrically as the variation of A₃₄₀ over time as described previously (1).

Analysis of the Secondary Structure of the Mutants by CD-- CD analyses were performed on a Jasco J-600 spectropolarimeter (Easton, MD) using a 0.1-cm cell. Three spectra recorded between 190 and 250 nm were averaged for each measurement. Acquisition was performed with a step of 0.2 nm. The samples were dialyzed in 50 mM sodium phosphate buffer at pH 7 and adjusted to the appropriate pH by the addition of concentrated stock buffer solutions before use. The samples were then centrifuged at 12,000 × g for 20 min at 4 °C, and the supernatant was used for spectroscopic analysis. The protein concentration was adjusted to 0.1 mg/ml as determined by the BCA assay.

Protein Fluorescence Analysis-- Fluorescence spectra were recorded on a Photon Technology International Alphascan fluorimeter (Lawrenceville, NJ) using Felix^TM version 1.21 software. A 75-µl cell with a 3-mm-long path was used for all experiments. The excitation and emission slits were set on 4 and 12 nm, respectively. The excitation wavelength was 295 nm. The proteins were prepared at a concentration of 0.1 mg/ml in 100 mM sodium phosphate buffer, pH 7, 100 mM NaCl.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Protein Expression and Purification-- Since WbpM is a large membrane protein, a short soluble version that was previously engineered and shown to be enzymatically equivalent to full-length WbpM (2) was used to derive all of the WbpM mutants described in this study. As reported previously, the expression of FlaA1 and WbpM as soluble proteins could only be achieved if the bacteria were grown in rich medium (Terrific broth) and if expression was carried out at 37 °C (1, 2). In several cases, the mutation of a single amino acid was enough to trigger the production of totally insoluble protein (Table II). Different conditions for protein expression had to be established to obtain high enough yields of soluble mutant proteins. For example, the H86A and D70A mutants of FlaA1 described below could only be obtained in a soluble form after expression in Tryptone/phosphate medium at 30 °C. Although more closely related to the wild-type enzyme than the H86A mutant, the H86F counterpart could not be obtained as a soluble protein under any of the conditions tested. Similarly, no suitable expression conditions could be found to express the FlaA1 D70N and C118M mutants as soluble proteins, which hampered their characterization. All soluble proteins were purified to homogeneity by metal chelation before use for activity assays or biophysical analysis (data not shown).

Role of Conserved Cysteine Residues Found in the Subfamily of Short Homologues-- FlaA1 and its short homologues exhibit two conserved cysteines that are not found in the large membrane homologues. They are found within an -helix predicted to be part of the dimer interface in wild-type enzyme. Their role for enzymatic activity, protein dimerization, and structural stability was examined by constructing single mutants of FlaA1 where each cysteine was replaced by a methionine residue. Out of the two resulting mutants, C103M and C118M, only C103M could be expressed as a soluble protein. However, it was found to be totally inactive after purification (Table II).

As judged by circular dichroism analysis, the secondary structure of the cysteine mutant was not affected by the mutation (Fig. 1). The CD spectrum was nearly superimposable to that of wild-type FlaA1. Nevertheless, gel filtration chromatography showed that this mutant had lost its ability to form a dimer, since the protein was recovered in a fraction corresponding to an exclusion size of 37 kDa, clearly well resolved from any potential 74-kDa dimer. This suggested a potential role of this cysteine in the dimerization process. However, SDS-PAGE analysis of the wild-type protein prepared in the presence or absence of reducing agents showed that dimerization was not due to the formation of intermolecular covalent bonds (data not shown). These results suggest that the cysteine residue might contribute to stabilization of the protein structure in a tertiary configuration suitable for dimerization rather than be involved in covalent binding between two subunits. This interpretation is consistent with the result of tryptophan fluorescence analysis, which revealed subtle structural variations that were not detected by circular dichroism analysis. The maximum emission wavelength of FlaA1 was shifted from 347 to 359 nm upon introduction of the C103M mutation (Fig. 2). This indicated that the two tryptophan residues (Trp²⁰⁷ and Trp³²²) found in FlaA1 were exposed to a more polar environment in the mutant enzyme than in the wild-type protein, probably reflecting the absence of dimerization or subtle variation in the tertiary structure of the protein.

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Far ultraviolet circular dichroism of FlaA1 and its mutants. Samples were prepared at a concentration of 0.1 mg/ml in 50 mM NaH2HPO4/Na2HPO4 buffer at pH 7. Represented spectra are the average of three scans.

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Fluorescence emission spectra of FlaA1 and its mutants. Samples were prepared at a concentration of 0.1 mg/ml in 100 mM NaH2HPO4/Na2HPO4 plus 100 mM NaCl buffer at pH 7. Each spectrum is an average of three scans. The excitation wavelength used was 295 nm.

Mutagenesis of Residues Potentially Involved in the Binding of the Nucleotide Co-factor-- FlaA1 and WbpM both exhibit the GXXGXXG signature for NAD(P)⁺-binding enzymes. The addition of exogenous cofactor was necessary to obtain full activity of WbpM. In contrast, FlaA1 was fully active in the absence of exogenous cofactor, although no internally bound cofactor could be detected spectrophotometrically. To investigate whether FlaA1 indeed binds a cofactor, the glycine residue central to the GXXGXXG motif was mutated to an alanine residue. The resulting G20A mutant could be expressed in a soluble form and purified to homogeneity (data not shown), although its solubility was slightly reduced compared with wild-type FlaA1 (Table II). This mutant was totally inactive as judged by CE analysis, and activity could not be recovered by the addition of exogenous cofactor. This suggested that the loss of activity was not simply due to a slight reduction of affinity for the cofactor but might be due to structural alterations of the nucleotide-binding site.

Circular dichroism analysis revealed differences at the secondary structure level. The CD spectrum of the G20A mutant still exhibited the main features of the wild-type enzyme, with main peaks of ellipticities at 193, 209, and 222 nm (Fig. 1). This indicated that the global secondary structure of the protein was not dramatically altered. However, the decreased intensities observed suggested that the protein structure was not entirely identical to that of the wild-type enzyme. This was confirmed by tryptophan fluorescence analysis that showed a significant 8-nm shift of the protein fluorescence spectrum toward higher wavelengths upon introduction of the G20A mutation (Fig. 2). This indicates a more polar environment of the protein's tryptophan residues. The fluorescence spectrum also exhibits a shoulder at 385 nm that was not present in the wild-type spectrum. Hence, the variations in secondary structure detected by CD in the G20A mutant also resulted in subtle modifications of the protein tertiary and/or quaternary structure as detected by fluorescence. Because the site of mutation lies within the signature for the Rossmann fold that is necessary for cofactor binding, it is likely that the loss of activity observed in this mutant reflects its inability to bind the cofactor due to slight alterations of the tertiary structure of the Rossmann fold. This result indicates that the integrity of the nucleotide-binding site is critical for catalysis and confirms the participation of a cofactor in the reaction.

Based on previous work, which indicated that a conserved histidine was essential for cofactor binding in all C4 epimerase (52), and based on the high level of sequence conservation between C4 epimerases and FlaA1, the potential role of the conserved His⁸⁶ in cofactor binding was investigated in FlaA1. This histidine was mutated into a phenylalanine or an alanine residue in FlaA1 (Table II). The H86F mutant could not be expressed in a soluble form under any of the experimental conditions tested, and its analysis was not pursued any further. Another mutant, H86A, was soluble and was purified by metal chelation affinity chromatography. Spectroscopic analysis (Figs. 1 and 2) revealed structural modifications reminiscent of those observed for the G20A mutant. However, H86A still exhibited significant catalytic activity (Fig. 3), suggesting that the structural alterations were limited. The levels of substrate conversion were lower than those obtained with wild-type enzyme at equilibrium (Fig. 4). This lower catalytic efficiency was due to a reduced affinity of the enzyme for its substrate, as shown by the 3-fold increase in the K_m of the mutant compared with that of wild-type enzyme (Table III). The catalytic constant k_cat was only slightly lower than that of wild-type enzyme. This indicates that the histidine residue is not directly involved in the catalytic process but is involved in substrate binding. The lower affinity for the substrate could reflect slight structural alterations of the substrate-binding pocket. Despite the observed role of His⁸⁶ in substrate binding, its mutation to an alanine residue did not alter the substrate and product specificity of the enzyme (data not shown). Consistent with the kinetic analysis that shows no major catalytic role of the histidine residue, the pH dependence of the H86A mutant remained very similar to that of the wild-type enzyme (data not shown).

View larger version (29K): [in this window] [in a new window]   Fig. 3.   Capillary electrophoresis analysis of the reaction products obtained after catalysis of UDP-GlcNAc by purified FlaA1, WbpM, and their mutants. All reactions were performed for 3 h with 1.9 µg of enzyme and 0.75 mM UDP-GlcNAc as the substrate. FlaA1 panel, reaction products of FlaA1 and its mutants. Reactions were performed at pH 7 and 37 °C. WbpM panel, reaction products of WbpM and its mutants. Reactions were performed at pH 10 and 30 °C with 0.14 mM NAD+. Peak 1, UDP-GlcNAc; peak 2, 4-keto, 6-deoxy intermediate; peak 3, UDP-Qui2NAc. a.u., arbitrary units.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Time course of UDP-GlcNAc catalysis by purified FlaA1, WbpM, and their mutants. Reactions were performed with 0.75 mM UDP-GlcNAc and 1.9 µg of enzyme. FlaA1 panel, reactions performed at 37 °C, pH 7, unless otherwise stated. V266E (), FlaA1 (), Y141M, pH 10 (), H86A (), and Y141M () are shown. WbpM panel, reactions performed at 30 °C, pH 10, with 0.14 mM NAD+. Shown are WbpM (), M438Y (), and E589V ().

To determine whether this histidine residue might be involved in cofactor binding, the effect of the mutation on cofactor binding and affinity were investigated. The addition of NAD⁺ had no effect on the catalytic efficiency at equilibrium or in initial rate conditions, as previously observed for wild-type enzyme. Since wild-type FlaA1 has been shown previously to use exogenous NADH, the K_m and k_cat of the H86A mutant for the cofactor NADH were determined. Both parameters were comparable with those observed for wild-type enzyme (Table IV), indicating that, contrary to what was predicted, this residue was not involved in interactions with the nucleotide cofactor. This also indicated that the observed structural alterations did not involve the cofactor-binding site. As observed in the wild-type enzyme, the low K_m of the mutant suggested a high affinity for the cofactor. Also, the fact that the k_cat for NADH was 60 times lower than the k_cat for the substrate indicated that the consumption of the reduced cofactor was most likely involved in the second step of the reaction (i.e. the reduction step), as demonstrated for the GDP-mannose dehydratase (5).

Site-directed Mutagenesis of the Catalytic Triad in FlaA1 and WbpM-- FlaA1 and the short homologues exhibit the typical SYK catalytic triad found in most members of the SDR family (Ser¹³⁰-Tyr¹⁴¹-Lys¹⁴⁵ in FlaA1). On the contrary, WbpM and other large homologues possess an altered SMK catalytic triad (Ser⁴²⁷-Met⁴³⁸-Lys⁴⁴² in WbpM) that, according to the accepted mechanism of C6 dehydratation, should not allow enzymatic activity. To investigate the role of the central tyrosine and methionine, respectively, in the catalytic mechanism of these two subfamilies of C6 dehydratases, several single mutants were constructed (Table II). Despite conservation of the aromatic character of the mutated residue, the FlaA1 Y141F mutant was totally inactive, indicating an essential role of the hydroxyl side group of the tyrosine residue for catalysis. Intriguingly, the FlaA1 Y141M mutant was moderately but significantly active (Figs. 3 and 4) and retained ~9% of the wild-type catalytic efficiency under the same experimental conditions. Kinetic analysis indicated that the Y141M mutation resulted in only a 4.5-fold reduction in affinity for the substrate (K_m = 565 µM) and in a similar reduction of the catalytic constant (Table III).

As reported earlier, FlaA1 and WbpM exhibited drastically different variations of their catalytic efficiency with pH: a bell-shaped curve with an optimum at pH 7.0-7.5 for FlaA1 and an exponential increase of activity for pH >8.5 for WbpM (1, 2) (Fig. 5). To examine the possibility that the substitution of Tyr for Met in the catalytic triad could be responsible for these differences, the pH dependence of the Y141M mutant was investigated. Fig. 5 shows that the Y141M mutant that harbored a SMK catalytic triad exhibited an exponential pH dependence similar to that observed in wild-type WbpM. This suggested that the tyrosine residue plays an essential role in governing the global pH dependence in this protein background. When the kinetic constant was determined under the optimal pH conditions of the Y141M mutant, the K_m for the substrate was found to be 251 µM, closer to the wild-type level than the value determined at pH 7, and its catalytic constant was only two-thirds that of the wild-type enzyme. Overall, in these optimal experimental conditions, the mutant retained a remarkable 42% of catalytic efficiency as compared with the wild type.

View larger version (14K): [in this window] [in a new window]   Fig. 5.   pH dependence of UDP-GlcNAc catalysis by FlaA1 and WbpM. Reactions were performed using sodium acetate (20 mM) for pH 5-6.5 and bis-Tris propane for pH 6.5-10 at 30 °C (WbpM) or 37 °C (FlaA1 and Y141M) with 1.9 µg of enzyme.

To investigate whether the drastic change of optimal pH observed in the FlaA1 Y141M mutant was due to a perturbation of the secondary structure by the introduced mutation, far UV CD spectra were recorded at different pH for both wild type and Y141M mutant. The CD spectra of the wild-type and mutant enzyme were nearly identical at pH 5, and no significant change of the secondary structure was observed in either protein from pH 5 to 9 (Fig. 6). This ruled out the possibility that important structural variations occurred due to the basic pH or to the introduction of the Y141M mutation.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Analysis of FlaA1 and Y141M at various pH values using far ultraviolet circular dichroism. Samples were prepared at a concentration of 0.1 mg/ml in 50 mM NaH2HPO4/Na2HPO4 buffer at pH 5 or 9.

The reverse experiment was performed to introduce a SYK catalytic triad in WbpM. CE analysis revealed that the resulting mutant M438Y was active (Fig. 3). Time course experiments indicated that the mutant was much less active than the wild-type enzyme, although 100% substrate conversion could still readily be obtained using slightly higher enzyme concentrations (Fig. 4). This suggested a lower rate of catalysis rather than a defect in substrate binding. This was confirmed by the determination of its kinetic parameters that showed a 3-fold reduction of the k_cat, whereas the K_m was half that of wild type (Table III). Although this mutant possessed a typical SYK catalytic triad, it still exhibited the unusual exponential pH dependence observed in wild-type WbpM (data not shown).

The substrate and product specificities of these two catalytic mutants (Y141M in FlaA1 and M438Y in WbpM) were not changed.

Role of Conserved Charged Residues Present Only in the Subfamily of Short Homologues-- FlaA1 and its short homologues exhibit a conserved aspartate residue (Asp⁷⁰) at a position usually occupied by an asparagine residue in large C6 dehydratases and in C4 epimerases. Similarly, the short homologues exhibit a conserved DK motif (Asp¹⁴⁹-Lys¹⁵⁰) at a position usually occupied by Asn-Glu in the large homologues. To examine the role of these charged residues for enzymatic activity, the mutants D70A, D70N, D149K/K150D and DK149/150AA were constructed (Table II). The D70N mutant protein was expressed in the form of insoluble inclusion bodies under all conditions tested. The other mutants could be expressed as soluble proteins, but the solubility of the purified proteins was lower than for the wild-type FlaA1. They were all totally inactive as judged by CE analysis (Table II). This points to the essential role of Asp⁷⁰ and Asp¹⁴⁹-Lys¹⁵⁰ in the enzymatic activity of FlaA1.

The low solubility of the mutants did not allow the analysis of their secondary structure by CD. Sufficient amounts of the D70A mutant protein could be obtained to perform fluorescence analysis. Its fluorescence spectrum indicated that a subtle variation in the protein structure had occurred, since a slight decrease of the fluorescence signal was observed compared with wild-type protein (Fig. 2). However, these variations were limited, since no shift in the maximum emission wavelength was detected. This indicated an important catalytic rather than structural role of this residue.

Role of a Conserved Glutamate Residue Found Only in the Large Homologue Subfamily-- All of the large homologues exhibit a conserved glutamate residue close to their C terminus, including WbpM (Glu⁵⁸⁹). In FlaA1 and its short homologues, alignments using the LALIGN software (available on the World Wide Web at www.expasy.ch) showed that this residue is replaced by a small hydrophobic residue (Val²⁶⁶ in FlaA1), whereas in most C4 epimerases, an aromatic residue (Tyr or His) occupies this position. In the latter enzymes, the aromatic residue was shown to interact with the 6'- and 2'-hydroxyl of the sugar moiety. It has been recently proposed that substitution of this residue by a cysteine in the human UDP-Glc C4 epimerase allows the enzyme to accommodate an N-acetylated substrate (53), whereas the E. coli enzyme cannot. Consequently, the possibility that the glutamate residue found in the large homologues would also interact with the substrate and play a role in catalysis and/or substrate specificity was investigated. For this purpose, a WbpM E589V mutant was constructed (Table II). Although E589V was still significantly active (Fig. 3), time course experiments suggested that its overall catalytic efficiency was lower than that of wild-type enzyme (Fig. 4). This was confirmed by kinetic analysis that showed that its catalytic constant was 9-fold lower than that of wild type (Table III), indicating a critical role of Glu⁵⁸⁹ for catalysis. Surprisingly, this mutant had a higher affinity for its substrate as shown by its 2.5-fold lower K_m value. These observations suggest that Glu⁵⁸⁹ might be part of the substrate-binding pocket. However, contrary to what was expected, it prevents optimal substrate binding in WbpM.

The "reverse" FlaA1 V266E mutant was constructed to investigate the effect of the introduction of a glutamate residue at an equivalent position in the short homologues (Table II). This mutant exhibited significantly higher catalytic efficiency than wild-type FlaA1 (Figs. 3 and 4), mostly due to a 40% increase in its k_cat, whereas its affinity for the substrate was unchanged (Table III).

The data reported above clearly indicate a role of this charged residue in catalysis; thus, the effect of its introduction or removal on the pH dependence of the enzymes was investigated. The V266E FlaA1 mutant exhibited the typical bell-shaped pH response curve observed for wild-type FlaA1, with only a slight shift of +0.5 pH unit. Similarly, the pH dependence was not affected by the E589V mutation in WbpM (data not shown).

Despite the potential role of the Glu⁵⁸⁹ and Val²⁶⁶ residues in substrate binding in WbpM and FlaA1, respectively, the substrate specificity of the enzymes was not affected by the mutation. No activity was detected using closely related substrates (UDP-Glc, UDP-Gal, or UDP-GalNAc) or with substrates of other known C6 dehydratases (GDP-mannose or dTDP-glucose). Likewise, the stereo-specificity of the reduction of the 4-keto, 6-deoxy intermediate was totally preserved in both mutants, so that only UDP-Qui2NAc was produced, and none of its C4 epimer UDP-FucNAc was detected (data not shown).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Due to their intrinsic differences from other members of the SDR family, FlaA1 and WbpM represent ideal model systems to investigate the molecular basis for (i) function specificity (dehydratase versus epimerase), (ii) substrate specificity (UDP-GlcNAc versus other sugar nucleotides), (iii) stereospecificity of product formation (UDP-Qui2NAc versus the C4 epimer UDP-FucNAc), and (iv) mechanism of action (altered catalytic triad, role of cofactor). To understand the molecular basis for catalysis outside the SYK dogma, several mutants were constructed by site-directed mutagenesis in this study (Table II). Residues potentially involved in cofactor binding and protein oligomerization were also targeted. The residues targeted for mutagenesis were chosen based on the crystallography data available for C4 epimerases (2, 13, 14) and related enzymes (39, 44, 47, 53-55).

FlaA1 was shown to exist as a dimer, but the role of dimerization for activity was not determined, and the specific residues involved in the process were not identified. In the short homologues, two conserved cysteine residues are present in -helix 4 that is predicted to be located at the dimerization interface between two subunits: Cys¹⁰³ and Cys¹¹⁸ in FlaA1. Hence, the contribution of two conserved cysteines to dimer formation and to protein structure stabilization was investigated by site-directed mutagenesis. The only soluble cysteine mutant that could be obtained, C103M, was inactive and had lost its ability to dimerize, whereas its secondary structure was intact. These data suggest that the conserved cysteine residue might play a role in the dimerization process via stabilization of an intramolecular configuration that would be suitable for dimerization. This is consistent with the protein fluorescence analysis, which revealed that the C103M mutation increased the polarity of the local environment of the two tryptophan residues present in the protein (Trp²⁰⁷ and Trp³²²), despite their remote position from Cys¹⁰³ at the primary structure level. This change of local environment could arise from variations in the tertiary and/or quaternary structures resulting from two nonexclusive phenomena. First, the absence of intramolecular disulfide bonds due to the C103M mutation could result in a less compact structure with more exposure of core residues to the aqueous buffer. Second, because the C103M protein can no longer dimerize, the tryptophan residues that are usually buried at the protein/protein interface in the wild-type enzyme might be exposed to the aqueous buffer in the mutant protein. Solving the protein structure by x-ray crystallography will determine which interpretation is correct.

A mutation was introduced in the glycine central to the GXXGXXG signature for nucleotide-binding proteins of FlaA1 (G20A) to determine whether a cofactor was involved in the catalytic process. In agreement with results obtained with other dehydratases (3, 9, 12, 46), the data concerning the G20A mutant suggest that an enzyme-bound cofactor is involved in the catalytic process. The loss of activity observed for this mutant seems to be due to variations at the secondary and tertiary structure level. This is consistent with improper or altered folding of the Rossmann fold that would prevent cofactor binding. The limited secondary structure perturbation revealed by CD analysis was unexpected, since in the case of other enzymes that internally recycle a tightly bound cofactor, the removal or absence of co-factor results in irreversible denaturation of the protein (34). Only in the human GalE can the NAD⁺ be readily removed without protein denaturation (39). Hence, in the case of FlaA1, the mutation did not appear to prevent the formation of the Rossmann fold, which would have been reflected by a drastic change of the CD and fluorescence spectra, but rather might have affected its size and flexibility (44). To investigate the potential correlation between cofactor binding and dimerization, attempts were made to determine the dimerization status of this mutant. However, its tendency to aggregate over time prevented such analysis.

Previous chemical mutagenesis studies on a C4 epimerase implied an interaction between a histidine residue and the cofactor, but the specific residue involved was not identified (52). Sequence alignments showed that His⁸⁶ from FlaA1 is conserved in all C4 epimerases (33). This residue is also conserved in all known C6 dehydratases whether or not they belong to the FlaA1/WbpM subclass of dehydratases/reductases. The possibility that His⁸⁶ would correspond to the histidine involved in cofactor binding alluded to above was investigated. However, our results showed that this histidine was involved in substrate binding rather than in cofactor binding. This was unexpected, since previous crystallography data did not reveal any proximity of this residue with the substrate. This suggests that the participation in cofactor binding could be indirect, especially since the H86A did not alter the substrate specificity of the enzyme.

As mentioned above, a striking feature of WbpM and its large homologues is their capacity to perform catalysis via the formation of a 4-keto intermediate although they contain an altered catalytic triad. Hence, possessing two subfamilies of enzymes that exhibit the same substrate and function specificities but differ in their catalytic triad was a golden opportunity to investigate the mechanism involved. The observation that Y141F mutant of FlaA1 was inactive provided convincing evidence that the tyrosine of the triad plays a central role in catalysis. Also, it implied that the nucleophilic character of the side chain of the tyrosine was important for activity and not its aromatic character. This is consistent with the participation of the tyrosine in a charge-transfer complexation with the cofactor and substrate (35). Intriguingly, significant activity was obtained in the Y141M mutant that mimics the situation found in the subfamily of large homologues represented by WbpM. The 42% catalytic efficiency retained by the Y141M mutant is remarkable. This is the first report of a significantly active mutant of the tyrosine central to the catalytic triad of SDR enzymes. The activity might involve electron delocalization from the sulfur atom of the methionine side chain. The fact that the methionine residue found in the SMK triad of WbpM is actively involved in the catalytic process was demonstrated using the M438Y mutant. This mutant had a lower activity than the wild type, although it harbored the traditional SYK triad.

The introduction of the methionine in the catalytic triad in the Y141M mutant of FlaA1 was sufficient to induce a dramatic change of pH dependence of the reaction. This could reflect a shift in the local pK_a within the catalytic triad, which is expected if a residue different from tyrosine was involved in the catalysis. Indeed, in C4 epimerases, the pK_a of the tyrosine of the catalytic triad is 6.1 (e.g. four units lower than the corresponding value in aqueous solution (10.1)). This results in the stabilization of the phenolic form of tyrosine within the active site and is caused by the neighboring positive charges of the lysine residue of the triad and of the nicotinamide ring of the cofactor (35, 56). Mutation of the catalytic tyrosine has been shown to totally abolish this lowering of the local pK_a within the active site (35, 56). This is similar to what was observed for the Y141M FlaA1 mutant and for wild-type WbpM, which do not have a catalytic tyrosine residue and have optimum activity at basic pH. In addition, the important variation of the K_m of the Y141M mutant with the pH suggests that, when present at this position, the methionine might also be involved in direct substrate binding. This is consistent with the mechanism proposed by Liu et al. (35, 56) whereby the residue central to the catalytic triad could trigger direct proton abstraction from the substrate instead of only acting as a recipient of a proton that would be preabstracted via another amino acid.

The conserved catalytic efficiency observed despite the removal of the central tyrosine in FlaA1 or of the methionine in WbpM is consistent with the fact that residues other than Tyr/Met are also involved in the catalysis. The absence of pH conversion by the M438Y mutation in WbpM suggests that a strongly charged residue can override the pH effect due to the tyrosine in the WbpM background. The sequences of both subfamilies of FlaA1 and WbpM homologues were analyzed for the presence of charged residues that could potentially be involved in catalysis and explain the significant levels of activity observed in both enzymes in the absence of the wild-type catalytic triad. This analysis revealed the existence of charged residues that are conserved in one of the families but are absent in the other. Among them, the Asp⁷⁰ and Asp¹⁴⁹-Lys¹⁵⁰ residues present only in FlaA1 and in the short homologues and the Glu⁵⁸⁹ residue found only in the subfamily of large homologues that exhibit an altered catalytic triad were each found to play a critical catalytic role. In addition, the "subfamily specificity" of the novel residues essential for catalysis that were identified in this study suggests that the two subfamilies of UDP-GlcNAc C6 dehydratases/C4 epimerases have had a divergent evolution from a common ancestor. The protein sequences have evolved so that maximum efficiency is obtained using a specific catalytic triad within a determined protein background that also includes subfamily-specific features (Asp⁷⁰ and Asp¹⁴⁹-Lys¹⁵⁰ in short homologues, Glu⁵⁸⁹ in large homologues). This is consistent with the early divergence described within the rest of the SDR family (17).

The 40% increase of activity observed after introduction of a glutamate residue in FlaA1 (mutant V266E) at a position equivalent to Glu⁵⁸⁹ in WbpM indicated that the presence of such an acidic residue is highly beneficial for catalytic activity. The enhanced activity observed in the V266E mutant is unequaled by any wild-type enzyme studied, suggesting that high levels of UDP-GlcNAc dehydration could be detrimental to the bacteria. This is understandable, considering that UDP-GlcNAc is a common precursor for the synthesis of several essential cellular components, including peptidoglycan. Hence, high levels of UDP-GlcNAc dehydration might deplete the cellular pool of UDP-GlcNAc and adversely affect peptidoglycan biosynthesis and cell viability. Moreover, the introduction of the glutamate residue at position Val²⁶⁶ does not influence substrate binding. This is different from what was observed in C4 epimerases in which this position is usually occupied by an aromatic or cysteine residue suspected to interact with hydroxyl groups of the sugar moiety of the substrate (33, 37-39, 53) and consequently play a role in substrate binding and specificity. This could account for the differences observed with crystallography-based data. The determination of the structures of FlaA1 and WbpM is currently under investigation by x-ray crystallography. Comparison of the three-dimensional structures of the proteins will determine whether these residues are also aligned at the tertiary structure level.

The novel residues essential for catalysis that were identified in this study are different in each enzyme subfamily. However, they are in each case negatively charged residues (aspartate and glutamate) that might contribute to abstraction of a proton from the sugar moiety and act as a catalytic base. Hence, in essence, the global mechanism for the formation of the 4-keto intermediate might be conserved between both subfamilies and between epimerases and dehydratases, although the specific amino acids involved might be different. In addition, these residues occupy very different positions in the sequence of each protein. Only the Asp¹⁴⁹-Lys¹⁵⁰ are in close proximity of the catalytic triad in FlaA1, and this location would favor the role of the aspartate as a classical base during catalysis. However, the Asp⁷⁰ residue is located in the N-terminal half of the protein, which is usually associated with nucleotide cofactor binding. In WbpM, the Glu⁵⁸⁹ residue is located in the C-terminal domain of the protein associated with substrate binding according to previous studies on C4 epimerases (33, 34). Since its presence seems to be essential for the high activity observed in wild-type WbpM, and since it is only found in the large homologues that exhibit an altered catalytic triad, it is possible that this glutamate residue acts as the proton acceptor during catalysis. This suggests that further transfer of the removed hydrogen to the acceptor cofactor might occur via slightly different mechanisms in each enzyme subfamily and probably involves a relay mechanism. The determination of the structure of both proteins and their catalytic mutants will help to understand how the transfer occurs and will suggest additional residues potentially involved. In particular, additional charged residues conserved only within the family of short homologues are also likely to participate in the process.

Although this study revealed several residues that seem directly involved in substrate binding, the substrate and stereo-specificity of the enzymes were conserved in all mutants. Notwithstanding the fundamental interest in understanding the molecular basis for these specificities, it would be interesting to be able to alter the enzyme so that it would produce UDP-FucNAc instead of UDP-Qui2NAc. UDP-FucNAc is not commercially available, and its chemical synthesis is not straightforward. However, this compound is necessary for the study of transferases that are involved in the assembly of capsule or lipopolysaccharide molecules and might be interesting targets for the development of antimicrobial agents. Hence, its enzymatic synthesis using a modified version of FlaA1 or WbpM would allow further studies of lipopolysaccharide or capsule biosynthetic pathways. Ultimately, understanding the molecular basis for substrate and function specificity is essential for the rational design of inhibitors active against this novel subfamily of enzymes. The highly conserved nature of this family of enzymes in medically relevant bacteria suggests that an inhibitor developed against one of these enzymes is likely to exhibit a broad spectrum of antimicrobial activity and a high level of biochemical specificity.

In conclusion, the approach of performing structure-function studies of two novel UDP-GlcNAc C6 dehydratases/C4 reductases is a first step toward the understanding of the catalytic properties of a new class of dehydratases that do not follow the classical SYK dogma. The data reveal that the differences between the two subfamilies of enzymes go beyond their size and cellular localization. It is apparent that fundamental differences between the two subfamilies exist at the catalytic level. The essential catalytic role of an altered catalytic triad was demonstrated, and additional residues that confer improved catalytic efficiency within a specific protein background were identified. Future x-ray crystallography data combined with the structural and biochemical data presented in this study will aid in understanding the contribution of these newly identified residues to the reaction mechanism.

	ACKNOWLEDGEMENTS

We thank R. Merrill and R. Yada (University of Guelph) for assistance in performing protein fluorescence and circular dichroism analysis, respectively.

	FOOTNOTES

^* This work was supported by operating grants from the Canadian Institute for Health Research (CIHR) (Grant MOP-14687) (to J. S. L.) and from the Canadian Bacterial Diseases Network, a consortium of the Federal Networks of Centres of Excellence program. This work was also supported by the Collaborative Health Research Projects program of the Natural Sciences and Engineering Research Council Grant CHRPJ-251007 (to J. S. L. and C. C.) and by CIHR Grant MMA-41558 (to J. S. L.) for the purchase of the capillary electrophoresis instrument.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ Recipient of a postdoctoral fellowship from the Canadian Cystic Fibrosis Foundation (CCFF).

A Zellers Senior Scientist and a recipient of a Marsha Morton Scholarship from CCFF. To whom correspondence should be addressed. Tel.: 519-824-4120 (ext. 3823); Fax: 519-837-1802; E-mail: jlam@ uoguelph.ca.

Published, JBC Papers in Press, May 9, 2002, DOI 10.1074/jbc.M202882200

	ABBREVIATIONS

The abbreviations used are: Qui2NAc, 2-acetamido-2,6-dideoxy-D-glucose (also known as N-acetylquinovosamine); SDR, short chain dehydrogenase/reductase; CE, capillary electrophoresis; bis-Tris, 2- [bis(2-hydroxyethyl)amino]-2-(hydroxymethyl)propane-1,3-diol.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				M. J. Coyne, M. Chatzidaki-Livanis, L. C. Paoletti, and L. E. Comstock Role of glycan synthesis in colonization of the mammalian gut by the bacterial symbiont Bacteroides fragilis PNAS, September 2, 2008; 105(35): 13099 - 13104. [Abstract] [Full Text] [PDF]

				Y. A. Knirel, O. V. Bystrova, N. A. Kocharova, U. Zahringer, and G. B. Pier Review: Conserved and variable structural features in the lipopolysaccharide of Pseudomonas aeruginosa Innate Immunity, December 1, 2006; 12(6): 324 - 336. [Abstract] [PDF]

				N. Ishiyama, C. Creuzenet, W. L. Miller, M. Demendi, E. M. Anderson, G. Harauz, J. S. Lam, and A. M. Berghuis Structural Studies of FlaA1 from Helicobacter pylori Reveal the Mechanism for Inverting 4,6-Dehydratase Activity J. Biol. Chem., August 25, 2006; 281(34): 24489 - 24495. [Abstract] [Full Text] [PDF]

				E. F. Mulrooney, K. K. H. Poon, D. J. McNally, J.-R. Brisson, and J. S. Lam Biosynthesis of UDP-N-acetyl-L-fucosamine, a Precursor to the Biosynthesis of Lipopolysaccharide in Pseudomonas aeruginosa Serotype O11 J. Biol. Chem., May 20, 2005; 280(20): 19535 - 19542. [Abstract] [Full Text] [PDF]

				A. Merkx-Jacques, R. K. Obhi, G. Bethune, and C. Creuzenet The Helicobacter pylori flaA1 and wbpB Genes Control Lipopolysaccharide and Flagellum Synthesis and Function J. Bacteriol., April 15, 2004; 186(8): 2253 - 2265. [Abstract] [Full Text] [PDF]

				L. S. Forsberg, K. D. Noel, J. Box, and R. W. Carlson Genetic Locus and Structural Characterization of the Biochemical Defect in the O-Antigenic Polysaccharide of the Symbiotically Deficient Rhizobium etli Mutant, CE166: REPLACEMENT OF N-ACETYLQUINOVOSAMINE WITH ITS HEXOSYL-4-ULOSE PRECURSOR J. Biol. Chem., December 19, 2003; 278(51): 51347 - 51359. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 277/30/26769    most recent M202882200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Creuzenet, C. Articles by Lam, J. S. Search for Related Content PubMed PubMed Citation Articles by Creuzenet, C. Articles by Lam, J. S. Social Bookmarking                      What's this?