![]()
|
|
||||||||
J. Biol. Chem., Vol. 281, Issue 23, 15653-15661, June 9, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
12
1
3
4
3
From the
School of Biosciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, United Kingdom and
Institute for Biotechnology 1, Research Centre Juelich, D-52425 Juelich, Germany
Received for publication, January 3, 2006 , and in revised form, March 31, 2006.
| ABSTRACT |
|---|
|
|
|---|
-D-arabinofuranosyl-1-monophosphoryldecaprenol to the galactan domain of the cell wall, thus "priming" the galactan for further elaboration by the arabinofuranosyltransferases. Because aftA is an essential gene in M. tuberculosis, we deleted its orthologue in Corynebacterium glutamicum to produce a slow growing but viable mutant. Analysis of its cell wall revealed the complete absence of arabinose resulting in a truncated cell wall structure possessing only a galactan core with a concomitant loss of cell wall-bound mycolates. Complementation of the mutant was fully restored to the wild type phenotype by Cg-aftA. In addition, by developing an in vitro assay using recombinant Escherichia coli expressing Mt-aftA and use of cell wall galactan as an acceptor, we demonstrated the transfer of arabinose from
-D-arabinofuranosyl-1-monophosphoryldecaprenol to galactan, and unlike the Mt-Emb proteins, Mt-AftA was not inhibited by ethambutol. This newly discovered glycosyltransferase represents an attractive drug target for further exploitation by chemotherapeutic intervention. | INTRODUCTION |
|---|
|
|
|---|
1
5-,
1
3-, and
1
2-glycosyl linkages, results from the sequential addition of arabinofuranose (Araf) residues from the sugar donor
-D-arabinofuranosyl-1-monophosphoryldecaprenol (DPA) (79), by a set of unique arabinofuranosyltransferases termed the Emb proteins, of which three paralogues exist in Mycobacterium avium (10) and M. tuberculosis (11).
The anti-tuberculosis drug ethambutol (EMB) specifically inhibits AG biosynthesis (12), and the molecular target of EMB occupies the embCAB locus in M. tuberculosis (11). Upon individual disruption of embC, embA, and embB in Mycobacterium smegmatis, the resultant mutants are viable (13, 14). However, the crucial terminal Ara6 motif, which is the structural motif for mycolylation in AG (5), is altered in both the Ms-embA and Ms-embB mutants (13). These results suggest that EmbA and EmbB are involved in the formation of the terminal Ara6 motif in AG and also presumably compensated for each other in the respective Ms-embA and Ms-embB mutants, whereas Ms-embC is probably involved in the formation of the arabinan domains of lipoarabinomannan (LAM) (14). This is in agreement with the initial studies of the Ms-embC mutant (14) and recent findings that indicate when point mutations were re-introduced into the Ms-embC mutant on a multicopy plasmid expressing the mutant allele, a truncated LAM was synthesized, which retained the basic glycosyl linkage profile of LAM (15). However, attempts to obtain deletion mutants of embA and embB in M. tuberculosis or embAB in M. smegmatis have proved unsuccessful, 6 presumably because of the essentiality of the cell wall mAGP in these bacteria (1619). In contrast, C. glutamicum has proven useful in the study of orthologous M. tuberculosis genes essential for cell viability. For instance, Cg-pks has been shown to be the key Claisen condensation enzyme involved in mycolic acid biosynthesis through the construction of a deletion mutant of C. glutamicum and its complementation with the Mt-pks13 orthologue (19, 20).
In a recent study (6), deletion of the single Cg-emb orthologue in C. glutamicum resulted in a slow growing yet viable mutant that synthesized a novel truncated AG structure possessing a galactan core and only t-Araf residues. Moreover, partial acid hydrolysis and matrix-assisted laser desorption ionization time-of-flight analysis identified the precise location of the three singular t-Araf residues attached to the galactan core on the 8th, 10th, and 12th galactofuranose (Galf) residue, thus representing the anchor points of the intact arabinan domains of AG (6). Treatment of C. glutamicum with EMB resulted in an identical phenotype comparable with the C. glutamicum
emb mutant (6). Thus, in contrast to the disruption of Ms-embA and Ms-embB, deletion of Cg-emb leads to an almost entire absence of arabinose in the cell wall, apart from specific t-Araf residues that are directly linked to the galactan backbone. This suggested the presence of a novel enzyme responsible for "priming" the galactan domain for further elaboration by the Emb proteins, resulting in the final maturation of the native AG polysaccharide. It is the aim of this study to identify this novel arabinofuranosyl-transferase, which catalyzes the addition of the first key Araf residues to the galactan domain of AG.
| EXPERIMENTAL PROCEDURES |
|---|
|
|
|---|
MCR were grown in Luria-Bertani broth (LB; Difco) at 30 and 37 °C, respectively. The mutants generated in this study were grown on complex medium BHIS (21). Kanamycin and ampicillin were used at a concentration of 50 µg/ml. The minimal medium CGXII was used for C. glutamicum (21). Samples for lipid analyses were prepared by harvesting cells at an absorbance of 1015, followed by a saline wash and freeze drying. Cultivation of C. glutamicum
aftA for lipid and cell wall analysis required two pre-cultures. First, a 5-ml BHIS culture was grown for 8 h, which was then used to inoculate a 50-ml BHIS culture for 15 h. This was then used to inoculate a 100-ml BHIS culture to an absorbance of 1, which was harvested after reaching absorbance of 3.
Construction of Plasmids and StrainsThe vectors made were pET23b-Mt-aftA (Rv3792), pEKEx2Cg-aftA (NCgl0185), pEKEx2Mt-aftA, and pK19mobsacB
aftA, with the gene number of the M. tuberculosis and C. glutamicum aftA orthologue added in parentheses. To construct the E. coli expression vector pET23b-Mt-aftA, the primer pair 5'-GATCGATCCATATGCCGAGCAGACGCAAAAGCCCCCAATTC-3' and 5'-GATCGATCAAGCTTCGCGCTCTCCTGCGGCTTGCGGATGGC-3' was used with the restriction sites NdeI and HindIII underlined, with M. tuberculosis H37Rv chromosomal DNA as a template. The purified PCR fragment was ligated with accordingly digested pET23b (Novagen). To overexpress C. glutamicum aftA, the primer pair 5'-TCCCCCGGGAAGGAGATATAGATATGATTAACACCTCTGAAGATGAAG-3' and 5'-TCCCCCGGGTTACTCATTGTGCGTTACCACCAC-3' was used to amplify C. glutamicum aftA, which was ligated with SmaI-cleaved pEKEx2 to generate pEKEx2Cg-aftA. Similarly, primer pairs 5'-CAGGATCCAAGGAGATATAGATATGCCGAGCAGACGCAAAAG-3' and 5'-CAGGATCCCCATCCGCGCTCTCCTGCGGCTTGC-3'were used to clone M. tuberculosis aftA into the BamHI site of pEKEx2. To construct the deletion vector pK19mobsacB
aftA, crossover PCR was applied with primer pairs AB (A, 5'-CGTGGATCCGGTGCC-3'; B, 5'-CCCATCCACTAAACTTAAACATTCAGAGGTGTTAATCAT-3') and CD (C, 5'-TGTTTAAGTTTAGTGGATGGGGTGGGACCTTTCGTGGTGGTAACG-3'; D, 5'-GGCGTCCGTACTGTCCAG-3') and C. glutamicum genomic DNA as template. Both PCR products were used in a second PCR with primer pairs AD to generate a fragment consisting of sequences adjacent to Cg-aftA, which was blunt end-ligated with SmaI-cleaved pK19mobsacB. All plasmids were finally confirmed by sequencing. The chromosomal deletion of Cg-aftA was performed as described using two rounds of positive selection (22), and its successful deletion was verified by use of different primer pairs. Plasmid pET23b-Mt-aftA was used to transform chemically competent cells of E. coli C43 (DE3) to ampicillin resistance (100 µg/ml), and pEKEx2Cg-aftA was introduced into C. glutamicum
aftA by electroporation with selection to kanamycin resistance (25 µg/ml).
|
Isolation of the mAGP ComplexThe thawed cells were resuspended in phosphate-buffered saline containing 2% Triton X-100 (pH 7.2), disrupted by sonication, and centrifuged at 27,000 x g (4, 6, 23). The pelleted material was extracted three times with 2% SDS in phosphate-buffered saline at 95 °C for 1 h to remove associated proteins, successively washed with water, 80% (v/v) acetone in water, and acetone, and finally lyophilized to yield a highly purified cell wall preparation (4, 6, 23).
|
|
DPA and Cg-Emb Biosynthetic Activity within Membrane Preparations of C. glutamicum and C. glutamicum
aftAMembranes from C. glutamicum and C. glutamicum
aftA were prepared as described previously to determine DPA biosyntheticactivity (24, 25). Membrane protein (1 mg) was added to 5-phospho[14C]ribofuranose pyrophosphate (2x 106 cpm), 50 µg of decaprenol monophosphate, 60 µM ATP, 0.5 mM NADP in 50 mM MOPS (pH 7.9), 5 mM
-mercaptoethanol, and 10 mM MgCl2 (buffer A) to a final volume of 160 µl. The reaction mixture was incubated for 1 h at 37°Cand stopped by the addition of 3 ml of CHCl3/CH3OH (2:1, v/v). Radiolabeled lipid linked sugars were extracted, as described previously, prior to scintillation counting and subjected to TLC using silica gel plates (5735 Silica Gel 60F254, Merck) in CHCl3/CH3OH/H2O/NH4OH (65:25:3.6:0.5, v/v) with reaction products visualized by autoradiography (24, 25). Analysis of Cg-Emb activity was determined using the synthetic
-D-Araf-(1
5)-
-D-Araf-O-C10:1 acceptor in a cell-free assay as described previously (8).
Expression and Analysis of Mt-aftA Gene ProductFor expression studies, E. coli (C43) was transformed with pET23b-Mt-aftA and cultured in Terrific Broth at 37 °C until an absorbance of 0.5, followed by the addition of 1 mM isopropyl 1-thio-
-D-galactopyranoside and further incubation for 12 h at 16 °C. Cells were harvested by centrifugation at 5000 rpm, and the resulting pellet was resuspended in buffer A. Resuspended cells were sonicated and centrifuged at 23,000 x g for 20 min at 4 °C, and the resulting supernatant was recentrifuged at 100,000 x g for 90 min at 4 °C to isolate cell membranes that were collected and concentrated to a protein concentration of 1520 mg/ml.
|
0.0150.15 mM galactan acceptor (6, 27, 28)) from C. glutamicum
aftA in buffer A to a final volume of 300 µl. The reaction mixture was incubated for 1 h at 37 °C, and the pellet was recovered following centrifugation at 18,000 x g. The pellet was washed twice with 1 ml of buffer A containing 1% IgePal CA-630 and further washed with 1 ml of CHCl3/CH3OH (2:1, v/v) until the radioactivity was recorded as background in the final supernatant washes (less than 30 cpm). The pellet was finally resuspended in buffer A and transferred to a scintillation vial, and 10 ml of EcoScint was added, and the amount of [14C]arabinose incorporation into cell wall galactan was measured by scintillation counting.
Analysis of Mt-AftA Assay ProductThe basic assay was repeated five times as described above, but using nonradiolabeled DPA (200 µg, 0.75 mM) (9, 26), and 1 mg of cell wall galactan polymer prepared from C. glutamicum
aftA. Following an initial incubation at 37 °C for 1 h, the assay was replenished with fresh membranes containing Mt-AftA (1 mg) and re-incubated for 1 h at 37°C with the entire process repeated three times with the addition of fresh membranes. The reaction mixture was processed as described above using two 1% IgePal CA-630 detergent washes in buffer A (1 ml) and five CHCl3/CH3OH (1 ml, 2:1, v/v) extractions followed by centrifugation at 18,000 x g. The recovered pellets were pooled, per-O-methylated, hydrolyzed using 2 M trifluoroacetic acid, reduced with NaB2H4, per-O-acetylated, and analyzed by GC/MS as described previously (6).
| RESULTS |
|---|
|
|
|---|
emb possessed no known arabinofuranosyltransferase activity yet contained single t-Araf units attached to the galactan core (6), we analyzed a 14-kb chromosomal region of M. tuberculosis encompassing embC, embA, and embB in detail and compared it with that of other Corynebacterianeae (Fig. 1A). This region includes the recently discovered decaprenylphosphoryl-5-phosphoribose (DPPR) epimerizing enzymes encoded by Rv3790 and Rv3791, which eventually provide Araf units from the sugar donor DPA (8, 24, 25, 29). These genes are followed by Rv3792 which is adjacent to embC. This particular region consisting of four genes is syntenic with regions in Mycobacterium bovis, M. avium subsp. paratuberculosis, and M. leprae as well as with those of C. efficiens, C. diphtheriae, and C. glutamicum. In addition, Nocardia farcinica also shows ample synteny to the M. tuberculosis gene locus, as well as Corynebacterium jeikeium, indicating a fundamental function of Rv3792. Based on the results described below, the Rv3792 gene and its orthologues was designated aftA (acronym for arabinofuranosyltransferase A). M. tuberculosis aftA is predicted to encode a membrane protein of 643 amino acid residues. The predicted topology of the N-terminal region (residues 1459) contains several hydrophobic segments based on dense alignment surface analysis (30), which probably form 11 trans-membrane-spanning helixes, whereas the C-terminal region (residues 460643) is predicted to be directed toward the periplasm. This domain organization and localization somewhat resemble that of EmbC (Fig. 1B). Nevertheless, the AftA proteins show no significant sequence similarity to the Emb proteins, and unlike the Emb proteins, they are not included in the CAZy data base of glycosyltransferases (31). However, the similarity of the AftA proteins among each other is very high over their entire sequence. Even for the most distant pairs, M. tuberculosis and C. diphtheriae, there is still 35.1% identity spanning 555 amino acid residues. The three regions of maximal similarity extend from 111 to 191 (I), 474 to 498 (II), and 516 to 551 (III) (amino acid residues of M. tuberculosis AftA). These regions contain several strictly conserved acidic and polar side chains as exemplified for part of region I of AftA (Fig. 1C), which are also known to play roles of general base and nucleophilic residues in glycosyl hydrolysis (31). Interestingly, region I (marked by a star in Fig. 1B) contains a conserved stretch of amino acids similar to that of the GT-C motif in EmbC, which is located in a periplasmic loop between membrane-spanning domains 3 and 4 (15). Taken together, the structural features of AftA as well as the localization of aftA within the gene cluster involved in arabinan biosynthesis suggest that AftA represents a putative glycosyltransferase involved in arabinan polymerization.
|
aftADespite the fact that Rv3792 is essential for M. tuberculosis (32), we attempted to delete its orthologue in C. glutamicum. The nonreplicative plasmid pK19mobsacB
aftA was constructed carrying sequences adjacent to Cg-aftA. The vector was introduced into C. glutamicum, and in several electroporation assays kanamycin-resistant clones were obtained, indicating integration of pK19mobsacB
aftA into the genome by homologous recombination (Fig. 2A). The sacB gene enables positive selection of a second homologous recombination event that can result either in the original wild type genomic organization or in clones deleted of aftA (20). More than 200 clones were obtained after 24 days and analyzed by PCR, but all were wild type, illustrating a strong disadvantage of aftA deletion. We continued with this analysis to eventually obtain in three independent approaches three rough textured colonies appearing after about 15 days. These were shown by PCR to have aftA deleted, whereas inspection of the adjacent C. glutamicum open reading frames resulted in identical PCR products to that of controls derived from the wild type (see Fig. 2A).
Growth of C. glutamicum
aftA in liquid brain-heart-infusion medium is shown in Fig. 2B. The mutant was almost unable to grow on this medium, whereas the presence of plasmid-encoded Cg-aftA (pEKEx2Cg-aftA) almost fully restored growth. Also, upon complementation of the mutant with Mt-aftA (pEKEx2Mt-aftA), growth restoration was obtained, albeit somewhat reduced, which might be due to the biased codon usage of M. tuberculosis. Transformation of C. glutamicum
aftA with pEKEx2emb (6) did not restore the growth defect, showing that overexpression of Cg-emb is unable to substitute Cg-aftA, thus confirming the unique specificity of both AftA and Emb. The identical strains were also grown on the same medium as before (brain-heart-infusion) but osmotically stabilized with 0.5 M sorbitol (Fig. 2C). Surprisingly, under these conditions, substantial growth was possible for C. glutamicum
aftA, which is presumably indicative of sorbitol stabilizing the cell wall mutant. On this medium the growth rate obtained for the wild type was 0.66 h1 with a final absorbance of 17 and that for the deletion mutant was 0.31 h1 with a final absorbance of 15.
|
aftA to its cellular composition, C. glutamicum
aftA and its complemented strain along with C. glutamicum were analyzed for arabinogalactan-esterified corynomycolic acids. C. glutamicum exhibited the known profile of corynomycolic acid methyl esters (CMAMEs) (Fig. 3, lane 1), whereas cell wall-bound corynomycolic acids were absent in C. glutamicum
aftA (Fig. 3, lane 2). The complementation of C. glutamicum
aftA with Cg-aftA (Fig. 3, lane 3) led to the restoration of cell wall-bound corynomycolic acids. These results suggest that Cg-aftA was involved in a key aspect of arabinan biosynthesis, whereby deletion perturbs tethering of corynomycolic acids to AG. Analysis of alditol acetate derivatives prepared from purified cell walls of C. glutamicum by GC revealed the sugar composition rhamnose (Rha), Ara, and Gal (Fig. 4A) (6). GC analysis of alditol acetates prepared from C. glutamicum
aftA (Fig. 4B) revealed a total loss of cell wall arabinose, which was restored upon complementation with plasmid pEKEx2Cg-aftA (Fig. 4C). GC/MS of per-O-methylated alditol acetate derivatives of C. glutamicum, C. glutamicum
aftA, and C. glutamicum
aftA pEKEx2Cg-aftA is shown in Fig. 4, AC. Apart from the presence of 2,5-Araf and t-Rhap associated with the arabinan domain of AG in C. glutamicum, other glycosidic linkages are comparable between M. tuberculosis (4, 33) and C. glutamicum (6, 27, 33) and include t-Araf, 2-Araf, 5-Araf, 3,5-Araf, 4-Rhap, t-Galf, 5-Galf, 6-Galf, and 5,6-Galf (Fig. 4A). As expected, C. glutamicum
aftA was devoid of t-Araf, t-Rhap, 2-Araf, 5-Araf, 3,5-Araf, and 2,5-Araf, whereas the galactan domain (apart from the 5,6-Galf branching residues resulting from arabinan side chains) was completely unaffected by the deletion of Cg-aftA and contained 4-Rhap, t-Galf, 5-Galf, and 6-Galf (Fig. 4B). Complementation of C. glutamicum
aftA with plasmid pEKEx2Cg-aftA restored the glycosyl linkage profile to that of C. glutamicum (Fig. 4, A and C). The previous deletion of emb in C. glutamicum and chemical analysis of the cell wall revealed a drastically truncated AG structure possessing only t-Araf residues and an unaltered galactan domain (6). Thus, the results indicate that aftA represents a putative arabinofuranosyltransferase responsible for priming the galactan domain with Araf residues for subsequent elaboration by Emb proteins.
Analysis of DPA Synthesis and Cg-Emb ActivityIn order to ensure that in C. glutamicum
aftA the biosynthesis of DPA is not reduced thereby disabling Araf delivery to the cell wall, we examined DPA and DPPR biosynthesis. Based on previous studies (24, 25) using 5-phospho[C14]ribofuranose pyrophosphate and decaprenol phosphate to monitor DPA formation, membranes were prepared from C. glutamicum and C. glutamicum
aftA. Both preparations afforded DPA synthesis (Fig. 5A), demonstrating that there was no reduced ability of C. glutamicum
aftA to synthesize DPA. In fact, an accumulation of DPA and DPPR was evident in C. glutamicum
aftA (
75 and 80%, respectively) as compared with C. glutamicum. Using endogenous acceptor and membrane preparations of either C. glutamicum or C. glutamicum
aftA, we observed no apparent Emb transferase activity with C. glutamicum
aftA compared with a sustained level of activity with C. glutamicum (data not shown), thus suggesting that the endogenous acceptor from C. glutamicum
aftA requires priming with Araf residues. In contrast, use of the synthetic acceptor
-D-Araf-(1
5)-
-D-Araf-O-C10:1 (8, 9) resulted in a significant transfer of [14C]arabinose from decaprenol phospho[14C]arabinose (8) affording an organic soluble trisaccharide product with membrane preparations from both strains (Fig. 5B). Thus, Emb is fully functional in C. glutamicum
aftA, as demonstrated above and by the earlier genetic experiments (Fig. 2, B and C). Furthermore, these results follow our hypothesis that the galactan domain of AG requires priming by the addition of a single Araf unit to the C-5 OH of a
(1
6)-linked Galf sugar for recognition and further extension by the Emb proteins.
|
aftA and incubated with the E. coli membrane preparation expressing His6-tagged Mt-aftA. Following incubation, the residual decaprenol phospho[14C]arabinose substrate was removed from the cell wall galactan by detergent and several repeated extractions using CHCl3/CH3OH (2:1, v/v). The remaining insoluble cell wall core was then subjected to scintillation counting revealing an increased amount of [14C]Araf incorporation in the presence of increasing amounts of cell wall galactan acceptor (Fig. 6B). Surprisingly, a 10-fold increase in cell wall galactan acceptor resulted in only a 3-fold increase in transferase activity. This poor turnover is presumably because of the inefficiency of the assay, which utilizes insoluble cell wall galactan as an acceptor. In addition, the activity of Mt-AftA remained unaffected in the presence of 100 µg/ml EMB (Fig. 6B), a known inhibitor of Cg-Emb (6), Mt-EmbA, and Mt-EmbB (11). The conversion of the galactan acceptor to a sugar polymer containing t-Araf units was further confirmed by glycosyl linkage analysis of the enzymatic reaction product (Fig. 6C). The newly synthesized product of several scaled up nonradiolabeled reactions was recovered, per-O-methylated, and derivatized to alditol acetates, which were analyzed by GC/MS. The linkages identified included those associated with the original cell wall galactan acceptor (Fig. 4B) plus the appearance of t-Araf, and as a result the branched 5,6-Galf residues (Fig. 6C).
Thus, our results describe the first report of a novel EMB-resistant arabinofuranosyltransferase, AftA, as the key initial glycosyltransferase involved in cell wall arabinan biosynthesis in Corynebacterianeae (Fig. 7) like M. tuberculosis and C. glutamicum.
| DISCUSSION |
|---|
|
|
|---|
(1
6)-linked Galf of a presynthesized galactan core (6). This was derived from a thorough analysis of a C. glutamicum mutant deleted of its single arabinofuranosyltransferase Emb (6), which still synthesizes a linear galactan extending from the reducing Rha with single t-Araf residues attached to the 8th, 10th, and 12th Galf residue. Apparently, these specific Araf residues serve for recognition and extension by the known Emb proteins resulting in the formation of the mature arabinan chains (6, 10, 11).
The in vivo analysis of C. glutamicum
aftA, as well as the in vitro study with the AftA protein of M. tuberculosis, identifies a bona fide arabinofuranosyltransferase. In principle, the absence of Araf residues in C. glutamicum
aftA could be due to the unavailability of precursors, as we demonstrated previously with a C. glutamicum mutant devoid of the polyprenyl transferase (UbiA) activity involved in the synthesis of DPA (6). However, we established that DPA biosynthesis was maintained in C. glutamicum
aftA, as well as Emb-catalyzed arabinan biosynthesis in vitro. This, together with our previous study on C. glutamicum
emb, shows that AftA functions to link the first Araf residue to the cell wall galactan core. More importantly, the recombinant Mt-AftA transferred Araf units from DPA to a cell wall galactan core acceptor in vitro, thus further confirming the unique activity of the enzyme. Although both Emb and AftA are arabinofuranosyltransferases, the proteins cannot functionally replace each other. Thus, despite some functional relationship, both glycosyltransferases have inherent specific features as is also evident from the insensitivity of Mt-AftA and Cg-AftA toward EMB, whereas the single Cg-Emb (6, 35) and Mt-Emb proteins are sensitive toward EMB (11).
The discovery of AftA sheds new light on the key arabinofuranosyltransferases that build the arabinan domain, which is typical for Corynebacterianeae. An elementary structure of this sugar polymer is apparent in C. glutamicum, and this bacterium has proven useful for a number of studies on mAGP biosynthesis (6, 19, 20). It represents the archetype of the Corynebacterianeae and has a low frequency of structural alterations as manifested, for instance in a low number of gene duplications (36). Corynebacterium species have only one emb gene, whereas Myco-bacterium species have up to three. Nevertheless, the glycosidic linkage analysis of C. glutamicum AG shows that 2-Araf, 5-Araf, and 3,5-Araf linkages are present, which are analogous to those found in the AG of M. tuberculosis (4). Furthermore, in C. diphtheriae and C. glutamicum 2,5-Araf linkages are also evident (6, 37). This suggests the possibility that one single Emb glycosyltransferase enables the formation of different linkage types, a feature reminiscent of the bi-functional galactofurano-syltransferase (GlfT) of M. tuberculosis that produces alternating 5-Galf and 6-Galf linkages within the galactan core of AG (38, 39). The high syntenies of the M. tuberculosis Rv3790, Rv3791, aftA, and embC to the maps of all other Mycobacterium and Corynebacterium species are in agreement with this view (Fig. 1A). Also in N. farcinica this general organization is retained with an additional membrane protein-encoding gene. This largely retained organization, as well as the separation of the paralogous embAB genes in M. leprae and M. avium subsp. paratuber-culosis is indicative of an ancient core function of Rv3790, Rv3791, aftA, and embC within the Corynebacterianeae involved in the synthesis of Araf donors and their use to assemble a basic periplasmic arabinan domain that serves as a scaffold to tie mycolic acids. The identification of new cell wall biosynthetic drug targets is of great importance, especially with the emergence of multidrug-resistant tuberculosis. This newly discovered DPA-dependent arabinofuranosyltransferase represents a promising candidate for further exploitation as a potential drug target to disrupt the essential mycolyl-arabinogalactan-peptidoglycan complex in mycobacterial species, such as M. tuberculosis.
| FOOTNOTES |
|---|
1 Both authors contributed equally to this work. ![]()
2 Biotechnology and Biological Sciences Research Council Quota Student. ![]()
3 Supported by the Fonds der Chemischen Industrie. ![]()
4 Supported as a Lister Institute-Jenner Research Fellow and the Medical Research Council (UK). To whom correspondence should be addressed. Tel.: 121-415-8125; Fax: 121-414-5925; E-mail: g.besra{at}bham.ac.uk.
5 The abbreviations used are: AG, arabinogalactan; Ara, arabinose; CMAME, corynomycolic acid methyl ester; DPA, decaprenol phosphoarabinose; DPPR, decaprenylphosphoryl-5-phosphoribose; EMB, ethambutol; GC, gas chromatography; GC/MS, gas chromatography/mass spectrometry; mAGP, mycolyl-arabinogalactan-peptidoglycan; Rha, rhamnose; Araf, arabinofuranose; MOPS, 4-morpholinepropanesulfonic acid. ![]()
6 G. S. Besra, unpublished results. ![]()
| ACKNOWLEDGMENTS |
|---|
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
X. Meniche, C. de Sousa-d'Auria, B. Van-der-Rest, S. Bhamidi, E. Huc, H. Huang, D. De Paepe, M. Tropis, M. McNeil, M. Daffe, et al. Partial redundancy in the synthesis of the D-arabinose incorporated in the cell wall arabinan of Corynebacterineae Microbiology, August 1, 2008; 154(8): 2315 - 2326. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Shi, R. Zhou, Z. Liu, T. L. Lowary, P. H. Seeberger, B. L. Stocker, D. C. Crick, K.-H. Khoo, and D. Chatterjee Transfer of the First Arabinofuranose Residue to Galactan Is Essential for Mycobacterium smegmatis Viability J. Bacteriol., August 1, 2008; 190(15): 5248 - 5255. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Goude, A. G. Amin, D. Chatterjee, and T. Parish The Critical Role of embC in Mycobacterium tuberculosis J. Bacteriol., June 15, 2008; 190(12): 4335 - 4341. [Abstract] [Full Text] [PDF] |
||||