Identification of a Wzy Polymerase Required for Group IV Capsular Polysaccharide and Lipopolysaccharide Biosynthesis in Vibrio vulnificus

Bacterial strains and media.E. coli S17.1λpir (44) was used as the donor strain for conjugation of the pNKTXI-SceI plasmid into V. vulnificus. A rifampin (Rf)-resistant derivative of the encapsulated V. vulnificus clinical strain 27562 obtained from the Collection de l'Institut Pasteur was used as the recipient. Both strains were grown on LB media (Sigma). Antibiotics were used at the following concentrations as indicated: ampicillin (Ap), 100 μg/ml; Rf, 100 μg/ml; kanamycin (Km), 25 μg/ml for E. coli and 160 μg/ml for V. vulnificus. IPTG (isopropyl-β-d-thiogalactopyranoside) was added to a final concentration of 1 mM.

PCR, sequencing, and phylogenetic analysis.PCR was performed with 50-μl volumes, using Platinum Pfx DNA polymerase (Invitrogen) according to the manufacturer's instructions. PCR products were TA cloned into the pCR2.1 vector (Invitrogen) and sequenced at The Centre for Applied Genomics (TCAG) at the Hospital for Sick Children (Toronto, Canada). Alignments were generated with the CLUSTALX (55) software program (version 1.8), and dendrograms were compiled by using the neighbor-joining method (computed from 1,000 independent trials) of CLUSTALW. Phylogenetic trees were generated with TREEVIEW (40).

Construction of pNKTXISce-I.The plasmid pNKBOR (49) contains a mini-Tn10 transposon (34) but lacks an origin of transfer necessary for conjugation to V. vulnificus 27562. Thus, the pNKTXISce-I fusion vector was created through ligation of EcoRI-digested pNKBOR to MfeI-digested pTXISce-IA (Fig. 1). We created the pTXISce-IA vector by introducing an I-SceI site into the multiple cloning site of pTX1A (38).

• Open in new tab
• Download powerpoint

FIG. 1.

pNKTXI-SceI mini-Tn10 delivery vector. The pNKTXI-SceI vector was obtained by ligating EcoRI-digested pNKBOR to MfeI-digested pTXISce1A. The vector contains an Ap resistance marker (Ap^r), the Tn10 transposase, an oriT_RP4 origin of transfer, a trfA-dependent origin of replication (oriV_trfA) outside of the minitransposon region, and a π-dependent R6Kγori_pir origin of replication within the minitransposon. The minitransposon also carries a Km resistance marker (Km^r). IR, inverted repeats recognized by Tn10.

Conjugations.The pNKTXI-SceI plasmid was transformed into the E. coli mobilizing strain S17.1λpir and moved to V. vulnificus as follows. Donor and recipient cultures were grown overnight in LB and LB-Rf, respectively. The next day, bacteria were diluted 100 times in fresh media and grown to an optical density at 600 nm (OD₆₀₀) of 0.7. One milliliter of each was then spun down and washed in LB, followed by incubation at 37°C for 1 h. Cultures were then pelleted, combined, and conjugated overnight without antibiotic selection on LB plates. The following day, cells were resuspended in 5 ml of LB. One milliliter of the resuspension was centrifuged at 10,000 rpm for 2 min, and the pellet was resuspended in 100 μl of LB. Exconjugants with transposon insertions were selected on LB-Rf-Km plates.

Southern analysis.Southern analysis was performed to verify single insertions of the Tn10 transposon into the genome of V. vulnificus, using the Km resistance marker from the pNKTXI-SceI plasmid as a probe. Genomic DNA was extracted from 21 mutants by using the DNAzol reagent, following the protocol provided by the manufacturer (Invitrogen), and digested with EcoRI. Following electrophoresis on a 1% agarose gel, the DNA was transferred to an N+ nylon membrane (Amersham). The membranes were probed using a Gene Images random primer labeling module (Amersham), following the manufacturer's protocol. Prehybridization, hybridization, and subsequent washes were performed at 55°C. Signal was detected using BioMax MS film (Kodak).

Determination of transposon insertion sites.Translucent V. vulnificus colonies were grown overnight on LB-Km plates. Bacteria were scraped from the plate, and genomic DNA was extracted using the DNAzol reagent. The DNA was cut using EcoRI, self-ligated, and transformed into DH5αλpir. Transformants were selected on LB-Km plates. Plasmids were purified using a Sigma GenElute plasmid mini prep kit and sequenced using the primer pirseq (ACACTTAACGGCTGACATGG), designed to obtain sequence going outwards from the R6K origin of replication within the transposon region on the pNKTXI-SceI plasmid. Complete plasmid sequence was obtained by primer walking, and open reading frames (ORFs) were identified using ORFfinder. Putative functions for the disrupted genes were assigned by BLAST analysis (http://www.ncbi.nlm.nih.gov/BLAST/). Transmembrane helix predictions for the Wzy sequences were obtained using the TMpred online program (http://www.ch.embnet.org/software/TMPRED_form.html).

Construction of pTRC99AoriT_RP4 expression vector.A NotI site was introduced into the backbone of the expression vector pTRC99A (20) by inverse PCR using the primers pTRC99aNot3 (AAAAGCGGCCGCTTCTCCTTACGCATCTGTGC) and pTRC99aNot4 (AAAAGCGGCCGCAATACCGCATCAGGCGCTC) (underlining indicates the restriction enzyme site). The RP4 origin of transfer was amplified with primers oriT1Not (CTGCGGCCGCCTTTTCCGCTGCATAACCCTG) and oriT2Not (CTGCGGCCGCCGGCCAGCCTCGCAGAGC), using the plasmid pSW23oriT (21) as a template. Both PCR products were cut with NotI, ligated together, and transformed into DH5α to create pTRC99AoriT.

Complementation of the wzy::Tn10 mutant.The sequence of the putative wzy gene was amplified from genomic DNA of the 27562 parent strain with primers WzyBamHI (CGGGATCCTCATAAGGGTATGCGTTTGTTA) and WzyFBspHI (GGGTTTTCATGATTACATACAACGTTAGATTAA). The wzy gene was cloned into the NcoI and BamHI sites of pTRC99A:oriT. The construct was conjugated to the V. vulnificus wzy::Tn10 mutant as described above, and gene expression was induced with IPTG.

Antiserum production.A 2-liter flask containing 200 ml of LB was inoculated with 1 ml of seed culture suspension (ca. 1 × 10⁹ CFU) of strain 27562, and the bacteria were grown at 30°C for 7 to 8 h with shaking. The bacteria were collected by centrifugation, washed twice with phosphate-buffered saline (PBS; NaCl [130 mM], Na₂HPO₄ [5 mM], KH₂PO₄ [1.5 mM]), and resuspended in 50 ml of PBS. The number of CFU per ml of suspension (usually ca. 5 ×10¹⁰ CFU/ml) was determined by plate counts. The bacteria in 10-ml portions of the suspension were killed by exposure to 0.8% (vol/vol) formalin for 1 day at ca. 25°C and then at 4°C for another 24 h. The bacteria were washed three times in PBS, and the vaccines were stored at 4°C. Complete inactivation was confirmed by testing the growth of suspensions on LB plates at 37°C for 3 days. One-milliliter portions were diluted with sterile PBS to ca. 5 × 10⁹ CFU/ml. Adult New Zealand White rabbits were immunized intravenously with 1 ml of formalin-killed bacteria without adjuvant. Five injections were given on days 1 (0.25 ml), 2 (0.5 ml), 3 (1.0 ml), 4 (1.0 ml), and 11 (1.0 ml). One week after the last immunization, 10 to 20 ml of blood was collected from the marginal ear vein of each rabbit and tested for the presence of antibodies by using the slide agglutination test described below. Following a positive test result, two more boosts of 1.0 ml at 1-week intervals were administered. One week after the second boost, animals were exsanguinated to collect the antisera. The blood was allowed to clot, and the sera were separated and stored at −30°C.

Slide agglutination tests.Overnight cultures of V. vulnificus wild-type, wzy::Tn10, and wzy::Tn10C cells were diluted to an OD₆₀₀ of 1. Twenty-microliter aliquots of each were applied in duplicate onto microscope slides. One aliquot was mixed with 20 μl of rabbit antisera raised against formalin-killed whole cells of the V. vulnificus parent strain. The other aliquot was mixed with sera obtained from the baseline bleed and served as the negative control. Agglutination results were scored after incubation of the mixtures at room temperature for 5 min. A distinct and immediately occurring agglutination was registered as positive, while weak or no agglutination after 5 min was considered a negative test result.

CPS extraction and visualization.The CPS extraction procedure was essentially as described by Gunawardena et al. (29) with some modifications. Single bacterial colonies were inoculated into LB and grown for 18 h at 30°C. Cultures were diluted to an OD₆₀₀ of 0.9, and 1 ml was plated and grown on LB plates for 18 h at 30°C. The cells were collected in 15 ml of LB, pelleted, and resuspended in an equal volume of PBS. Samples were shaken for 1 h at 200 rpm at 25°C. The OD₆₀₀ was measured, and CFU counts were determined after each step to ensure that cell lysis was not occurring and that equal numbers of bacteria were used for CPS isolation. The bacteria were pelleted, and the supernatant was dialyzed against three changes of distilled water (2l for 12 h), followed by concentration through a 10-kDa Amicon Ultra-15 centrifugal filter device column. This was followed by sequential treatment of the retentate with DNase (50 μg/ml plus 1 mM MgCl₂), RNase (100 μg/ml), and Pronase E (250 μg/ml) at 2-h intervals at 37°C. The retentate was spun at 100,000 × g for 18 h at 20°C and extracted twice with phenol chloroform, and the aqueous layer was dialyzed as described above. The retentate was concentrated using a 100-kDa-molecular-mass-cutoff Centriprep unit (Amicon) spun at 500 × g for 1 h. CPS samples were run on a 40% nondenaturing polyacrylamide gel electrophoresis (PAGE) gel or a 4% stacking/12.5% separating Tris-HCl-glycine sodium dodecyl sulfate (SDS)-PAGE gel, using the procedure described by Laemmli (36). CPS was visualized by immunostaining using antisera against formalin-killed whole cells of V. vulnificus 27562 and Alcian Blue staining of acidic polysaccharides as described by Pelkonen et al. (42, 43).

LPS extraction and visualization.Whole-cell lysates were prepared by following the method of Hitchcock and Brown (32) with several modifications. Five-milliliter cultures were grown overnight in appropriate antibiotics, and the OD₆₀₀ was adjusted to 0.9. Five milliliters of each was centrifuged at 5,000 rpm for 10 min, and the cells were washed six times in PBS. The pellet was resuspended in 375 μl of HB buffer (1% SDS, 4% β-mercaptoethanol, 10% glycerol, 1 M Tris-HCl, pH 6.8, 0.002% bromophenol blue) and boiled for 30 min. Proteinase K (1.5 μl; 20 mg/ml) was then added, and the reaction mixtures were incubated at 56°C overnight. The LPS samples were boiled for 15 min and run on a 4% stacking/12.5% separating Tris-glycine SDS-PAGE gel. LPS was visualized by immunoblotting.

LD₅₀ determinations.Strains were inoculated from single colonies into LB and were cultured overnight at 30°C. The overnight culture was seeded at a 1:100 dilution into fresh media and incubated at 30°C until the bacteria reached an OD₆₀₀ of 1.0. Bacteria were collected by centrifugation and washed once with PBS. Bacterial-cell concentrations were confirmed by plate counts. Five-week-old female CD-1 mice housed under pathogen-free conditions were pretreated with an intraperitoneal injection of iron dextran (250 μg of iron-dextran per g of body weight) 60 min before inoculation with the bacterial strains. Groups of mice (n = 5) were injected intraperitoneally with 0.1 ml of serial log dilutions (10² to 10⁸ CFU) of each bacterial strain. Injection of PBS alone served as a control. Mice were observed for up to 24 h postinfection with the recording of deaths or moribund animals. The 50% lethal dose (LD₅₀) was quantified using the simplified linear approximation method of Reed and Muench (47).

Nucleotide sequence accession number.The sequence determined for this study has been deposited in GenBank under accession number EU179506.

Transposon mutagenesis with the pNKTXISce-I mini-Tn10 delivery plasmid.To facilitate the generation of transposon mutants and the identification of the disrupted genes, we created the mini-Tn10 delivery plasmid pNKTXISce-I (Fig. 1). This vector was generated by fusing the plasmid backbones of pNKBOR and pTXISce-IA. pNKTXISce-I carries a Km^r marker and a π-dependent origin of replication in its mini-Tn10 transposon, while a TrfA-dependent origin of replication is present in the vector backbone. Thus, this plasmid can be maintained and its copy number controlled in cells expressing either the π (35, 39) or TrfA (26, 27) protein. The presence of an origin of replication in the mini-Tn10 transposon also allows for the isolation of the transposon insertion site and flanking DNA as Km-resistant clones in E. coli strains expressing the π protein. It also carries an RP4 origin of conjugative transfer for conjugation of the plasmid to a wide range of bacteria and an I-SceI restriction enzyme site for plasmid verification and cloning of additional modules. The plasmid is fully defined, as the complete sequences of both parent plasmids are known, and the Tn10 insertions are stable.

Identification of a capsule biosynthesis locus in V. vulnificus 27562.Transposon mutagenesis was conducted as described in Materials and Methods. Acapsular mutants were selected from the conjugation experiments, and Southern analysis was used to verify that each mutant contained a single transposon insertion in its chromosome. The transposon insertion points were identified by sequencing and BLAST analysis. The disrupted gene in one mutant (Fig. 2) was identified as a putative polysaccharide polymerase (wzy). Subsequently, 9.3 kb of the region surrounding the insertion site was cloned, sequenced, and annotated (Fig. 3). Potential functions were assigned based on homology to previously described genes (Table 1) . The ORFs appeared to encode gene products common to CPS/O-antigen biosynthesis clusters, including four glycosyltransferases, a putative wzy polymerase, and a wzx flippase homologue. Of the eight ORFs identified, seven appeared to be transcribed in the same orientation. This organization is reminiscent of the operon-like structure common to CPS biosynthesis clusters in a variety of encapsulated bacterial species. The lone exception encoded a homologue of WecA, the initiating glycosyltransferase that in part defines group IV CPS biosynthesis loci (60).

• Open in new tab
• Download powerpoint

FIG. 2.

Restoration of CPS production in the wzy::Tn10 mutant by the wild-type wzy gene. (A) Wild-type strain 27562; (B) wzy::Tn10 mutant; (C) wzy::Tn10 mutant complemented with the wild-type wzy gene.

• Open in new tab
• Download powerpoint

FIG. 3.

V. vulnificus 27562 capsule polysaccharide biosynthesis cluster. Open arrows represent the locations and directions of transcription of the respective ORFs. Horizontally striped, spotted, black, and white arrows indicate nucleotide sugar biosynthesis, transferase, insertion sequence, and processing genes, respectively. The closed triangle denotes the transposon insertion site. Gene characteristics are summarized in Table 1.

The complete Wzy protein was analyzed using TMpred software. The result was compared with those of Wzy proteins from Escherichia coli (GenBank accession no. AAO37717.1), Yersinia enterocolitica (GenBank accession no. AAC60768), Shigella boydii (GenBank accession no. AAW29815), Salmonella enterica (GenBank accession no. CAA43077), and Bacillus cereus (GenBank accession no. AAS44289.1). Eleven inside-to-outside transmembrane helices were suggested for the putative V. vulnificus Wzy protein. This was in agreement with the 10 or 11 transmembrane helices predicted for the Wzy proteins of E. coli, Y. enterocolitica, S. boydii, Salmonella enterica, and B. cereus. Phylogenetic comparisons with previously described bacterial Wzy and WaaL enzymes (Fig. 4) also supported the designation of a Wzy polymerase function for the disrupted gene.

• Open in new tab
• Download powerpoint

FIG. 4.

Phylogenetic relationship of bacterial Wzy and WaaL enzymes. An unrooted dendrogram based on representative Wzy and WaaL bacterial enzymes is shown. The organism abbreviations and GenBank accession numbers (in brackets) are as follows: Bce, Bacillus cereus (NP_981681); Sac, Syntrophus aciditrophicus (YP_460441); Sth, Streptococcus thermophilus LMG 18311 (YP_139549); Spn, Streptococcus pneumoniae (AAK20717); Cph, Clostridium phytofermentans ISDg (ZP_01352529); Cat, Croceibacter atlanticus HTCC2559 (ZP_00951257); Eca, Erwinia carotovora subsp. atroseptica SCRI1043 (YP_048289); Eco, Escherichia coli (AAO39700); Sbo, Shigella boydii (AAV41072); Sdy, Shigella dysenteriae (AAR97964); Lrh, Lactobacillus rhamnosus (AAW22494); Sen, Salmonella enterica (CAA43077); Pae, Pseudomonas aeruginosa PAO1 (AAG08384); Vch, Vibrio cholerae (AAL77359); Vvu, V. vulnificus; Kpn, Klebsiella pneumoniae (AAD37765); Xca, Xanthomonas campestris pv. vesicatoria strain 85-10 (YP_365325); Yen, Yersinia enterolitica (AAC60768). The two V. vulnificus Wzy polymerases are boxed, and the bootstrap values for each node are indicated.

Demonstration of a wzy-dependent CPS biosynthesis pathway in V. vulnificus.To verify that disruption of the wzy gene prevented CPS production, we conducted agglutination assays with the wild-type and mutant strains. Antisera against formalin-killed whole cells reacted strongly with the 27562 parent strain in slide agglutination tests, while no agglutination was observed with the wzy::Tn10 mutant. A positive agglutination reaction was obtained when the pTRC99aoriT expression vector carrying the wild-type wzy gene was introduced into the wzy::Tn10 mutant, while no agglutination was detected when the wzy::Tn10 mutant carried the empty pTRC99AoriT vector. CPS was extracted from the wild type, the wzy::Tn10 mutant, and the wzy::Tn10 mutant strain that was complemented with the wild-type wzy gene (wzy::Tn10C). The CPS was then separated by PAGE and visualized by Alcian Blue staining for acidic polysaccharides and immunoblotting (Fig. 5 and 6). Both detection methods revealed the presence of a high-molecular-weight polysaccharide and a prominent lower-molecular-weight product (designated the K band) in the wild-type strain, while no such material was observed in the wzy::Tn10 mutant. Complementation of the wzy::Tn10 mutant with the wild-type wzy gene restored production of the high-molecular-weight polysaccharide and the K band. These results confirmed the absence of CPS inferred from the translucent phenotype of the mutant and supported the involvement of the wzy gene in CPS production in V. vulnificus 27562.

• Open in new tab
• Download powerpoint

FIG. 5.

Alcian Blue stain of V. vulnificus CPS. CPS was extracted from wild-type (WT), wzy::Tn10 mutant (wzy polymerase knockout), and wzy::Tn10C (wzy::Tn10 complemented with pTRC99aoriT::wzy) cells and separated on a nondenaturing polyacrylamide gel. CPS was visualized by Alcian Blue staining.

• Open in new tab
• Download powerpoint

FIG. 6.

Immunoblot of V. vulnificus CPS. CPS was extracted from wild type (WT), wzy::Tn10 mutant (wzy polymerase knockout), and wzy::Tn10C (wzy::Tn10 complemented with pTRC99aoriT:wzy) cells and separated by denaturing SDS-PAGE. CPS was visualized using immunoblotting with antisera to WT formalin-killed whole cells.

Wzy is involved in LPS biosynthesis.To investigate whether the wzy CPS mutant was also defective in LPS production, LPS was extracted from the wild-type, the wzy::Tn10 mutant, and the wzy::Tn10C strains and separated by SDS-PAGE. Similar to previous reports (2, 5, 6), silver staining failed to detect the characteristic ladder pattern observed for smooth LPS profiles among the Enterobacteriaceae (data not shown). However, immunoblotting using antisera against the wild-type strain revealed a ladder pattern characteristic of smooth LPS. Two fast-migrating bands corresponding to the lipid A-core and the core-plus-one O-antigen structures were also observed (Fig. 7). Attempts to better resolve the fastest-migrating bands by using the Novex Tricine and 4 to 20% Bis-Tris gel systems were unsuccessful. The same prominent K band present in the CPS sample from wild-type bacteria was also evident in the LPS from this strain. Neither the ladder nor the K band was present in the LPS extract of the wzy::Tn10 mutant. Complementation with the wild-type wzy gene resulted in the reappearance of both the ladder and the K band in the LPS gels. An increase in the intensity of the banding pattern was observed following IPTG induction. In particular, the accumulation of higher-molecular-weight bands in the LPS ladder following IPTG induction supports a polymerization activity for the wzy gene. The empty vector did not result in the reappearance of the banding pattern. The results indicated that Wzy was involved in the production of both CPS and the O-antigen in V. vulnificus.

• Open in new tab
• Download powerpoint

FIG. 7.

Immunoblot of V. vulnificus LPS. LPS was extracted from wild-type (WT), wzy::Tn10 mutant (wzy polymerase knockout), and wzy::Tn10C (wzy::Tn10 complemented with pTRC99aoriT::wzy) cells. The LPS was separated on a denaturing SDS-PAGE gel and visualized by immunoblotting. Expression of the wzy gene was induced by the addition of IPTG from the media (+IPTG). The position of the K band band is shown. The black bar denotes the higher-molecular-weight O-antigen bands that accumulate upon induction of wzy.

The wzy::Tn10 mutant is attenuated in vivo.The LD₅₀ of V. vulnificus is dramatically reduced in the presence of iron, and the iron-overloaded mouse model is widely used to demonstrate the enhanced in vivo virulence of V. vulnificus strains in the presence of excess serum iron (67). The LD₅₀ values for the wild type and the wzy::Tn10 and wzy::Tn10C mutant strains were determined. The LD₅₀ of the wild-type strain (ATCC 27562) was 2.4 × 10⁴. Disruption of the wzy gene (wzy::Tn10) caused an increase in the LD₅₀ to >4.2 × 10⁷. Virulence was restored to the attenuated mutant following complementation with the wild-type gene (wzy::Tn10C; LD₅₀, 1.4 × 10⁴). Thus, the activity of the Wzy polymerase is required for virulence in V. vulnificus.

A new genetic organization for the group IV CPS locus of V. vulnificus 27562.Both group I and group IV CPS biosynthesis loci code for Wzy-dependent systems that include a Wzx flippase; however, several differences have been used to discriminate these CPS groups. In group I CPS biosynthesis, WbaP is the initiating transferase and the wza, wzb, wzc, wzx, and wzy genes reside within the same genetic locus. This feature distinguishes group I Wzy-dependent loci from O-antigen loci (60). In group IV CPS biosynthesis, the initiating transfer is catalyzed by WecA and the wza, wzb, and wzc genes are found at a chromosomal location different from that of the wzx and wzy genes, making group IV CPS loci allelic to many Wzy-dependent O-antigen loci (60). Despite this, the wza, wzb, and wzc genes still participate in group IV capsule assembly, as recently demonstrated with E. coli (41).

Based on the location of the wza, wzb, and wzc genes, a group I capsule was proposed for several V. vulnificus strains (12, 66). BLAST analysis indicated that neither V. vulnificus strain CMCP6 nor YJ016 contained a wecA homologue, supporting their group I classification (12, 66). Homologues of wecA, wzy, and wzx were identified in a CPS cluster in V. vulnificus 1003 (O) (51). Based on the presence of wecA, a group IV capsule was proposed for this strain; however, the chromosomal location of these genes relative to wza, wzb, and wzc was not determined. Furthermore, functional assignments for the ORFs were similarity based; none were shown to catalyze the proposed reaction, and none of the disrupted genes were genetically complemented with their functional counterparts to demonstrate their specific role in CPS production. Hence, a CPS grouping for the capsule could not be assigned with certainty. The CPS of strain 27562 is a serine-linked polysaccharide composed of d-GlcNAc, MurNAc, d-GalA, and l-Rha (29). The 9.3-kb CPS loci identified here contained wecA, wzx, and wzy genes, and we have demonstrated that Wzy is a functional polysaccharide polymerase. Southern analysis and sequencing have indicated that the wza, wzb, and wzc genes reside within the same genetic locus as the wecA, wzx, and wzy genes that we identified in strain 27562 (A. Nakhamchik, C. Wilde, and D. A. Rowe-Magnus, unpublished results). Hence, this strain appears to have a group IV Wzy-dependent CPS with a locus organization that mirrors that of group I CPS loci.

A comparison of the CPS biosynthesis regions of strains 27562 and 1003 (O) revealed that the two loci were similar but not identical in organization (Fig. 8). In 1003 (O), the wcvF rhamnosyltransferase is 5′ of wzx, while it is 3′ of wzx in 27562. IS1348, wecA, wcvB, and wcvA occur in the same order in the two strains, but strain 1003 (O) also contains INSE, IS3Tn, and a nonfunctional wcvA duplication. The presence of wecA and a rhamnosyltransferase gene, wcvF, in strain 27562 accounts for two of the transfer activities required for the assembly of its CPS. Transfer of the d-GalA and MurNAc moieties could be catalyzed by the wcvD and wcvE transferases to complete the structure.

• Open in new tab
• Download powerpoint

FIG. 8.

Genomic comparison of the group IV capsule loci of V. vulnificus strains 27562 and 1003 (O) and V. cholerae NRT36S. Open arrows represent the locations and directions of transcription of the respective ORFs. Horizontally striped, spotted, black, and white arrows indicate nucleotide sugar biosynthesis, transferase, insertion sequence, and processing genes, respectively. Dotted lines highlight the similar gene arrangements of the strains. The gene designations for strains 1003 (O) and NRT36S are as described in references 51 and 14.

Recent acquisition of the CPS biosynthesis cluster.The new arrangement that we observed for the group IV CPS locus of strain 27562 may be linked to our finding that the CPS region appears to have been recently acquired by horizontal transfer. While the average G+C content for V. vulnificus strains CMCP6 and YJ016 is 46.7% (13), the G+C contents of the genes in the CPS cluster identified here dropped from 41.7% for the wcvB gene to 30.8% for the wzx gene (Fig. 2). The five genes with the lowest G+C contents (wzx, wcvD, wzy, wcvE, and wcvF) were adjacent to IS1358, appear to be organized as an operon, and showed the highest similarity to genes from gram-positive or archaeal organisms (Table 1). The horizontal acquisition of the CPS biosynthesis clusters can clearly contribute to the tremendous structural diversity observed for the CPS of V. vulnificus clinical and environmental isolates and may play an important role in the evolution of new strains, similar to what has been observed for Vibrio cholerae O139. The epidemic Vibrio cholerae O139 serovar emerged from the pandemic biotype El Tor O1 serovar through the replacement of a 22-kb region by a 40-kb O139-specific DNA fragment (7) that contained the IS1358 insertion sequence. IS1358 was shown to be active for transposition (25) and is also associated with LPS loci in serovars of V. cholerae (52, 53), V. vulnificus (51), and Vibrio anguillarum (33, 53), making it possible to envision IS1358-mediated genetic exchange between Vibrio species that leads to the genesis of new capsule regions. It is tempting to speculate that IS1358 participated in the acquisition of the CPS loci in strain 27562; however, low-G+C-content polysaccharide biosynthesis genes are found on both sides of IS1358 and only a single copy of IS1358 is present on the chromosome (A. Nakhamchik and D. Rowe-Magnus, unpublished results). At present, the role, if any, of IS1358 in the evolution of CPS clusters in the Vibrionaceae remains unclear.

Does V. vulnificus harbor a hybrid CPS/LPS biosynthetic cluster?Wzy polymerases participate in O-antigen and CPS (K-antigen) production in group I and group IV capsules (3, 4, 23, 61, 63). Hence, it is possible for the same Wzy polymerase to be involved in the production of both the O and K antigens in the same strain. E. coli O9a:K30 has the same antigen, K30, present in two forms: as a smaller polymer attached to lipid A (K_LPS) and as a capsular antigen. A Wzy polymerase mutation in this strain results in the loss of both the K30_LPS and the capsular K30 antigen (23). Campbell et al. isolated Sinorhizobium meliloti mutants that were defective in the production of K antigen; these mutants also had an altered LPS profile (11). The LPS and CPS of V. cholerae O139 are structurally identical (19, 58), and the functions for the biosynthesis of the O antigen and the CPS were shown to be dependent upon shared genes that were carried on the same locus (17, 18). Interestingly, the overall organization (Fig. 8) and effect of transposon insertions on CPS and LPS biosynthesis that we observed in strain 27562 paralleled those reported for V. cholerae NRT36S (14). In NRT36S, the CPS and LPS biosynthesis genes share the same genetic locus. It was proposed that such an arrangement could play a key role in the evolution of new V. cholerae strains by providing a mechanism for the simultaneous emergence of new K and O antigens in a single strain. This notion is supported by the isolation and characterization of genetically similar strains of V. parahaemolyticus that express entirely different O and K antigens (15). Our results clearly demonstrate a link between CPS and LPS production in V. vulnificus 27562. Thus, it is conceivable that enzymes common to these two biosynthetic pathways are encoded by the same locus in this strain. We are examining the effects of targeted, nonpolar knockouts of other genes in this region on CPS and O-antigen production, and we are determining the structure of its O antigen.

Does V. vulnificus produce a K_LPS?The persistence of the K band in both the LPS and CPS preparations was unexpected. The K band remained in the LPS preparations despite the performance of six PBS washes. Hence, it is tightly linked to the cell surface rather than loosely associated with it. This suggested that either the K band represents an O antigen of a preferred chain length whose size is governed by an as yet unidentified chain length determinant (23, 69) or strain 27562 produces an additional polysaccharide (K_LPS) that is linked to lipid A in a manner similar to that for the K_LPS in E. coli serotype K40 (3, 60). The presence of the K band in CPS preparations can be best explained by the production of a K_LPS. The OD₆₀₀ and CFU counts were monitored before and after the PBS wash during preparation of the CPS. Both the OD₆₀₀ and CFU/ml values increased, indicating that the bacteria remained intact and viable. Thus, the release of LPS due to cell lysis was not the source of the K band in the CPS preparation. Furthermore, a 100-kDa-cutoff Amicon Centriprep column was used to minimize LPS cross-contamination of the CPS. We believe that the K band produced by V. vulnificus 27562 is actually a K_LPS and its presence in the CPS preparations is a result of unlinked K_LPS associating with the higher-molecular-weight CPS during purification.