Characterization of the Burkholderia pseudomallei K96243 Capsular ...

Characterization of the Burkholderia pseudomallei K96243 Capsular Polysaccharide I Coding Region Jon Cuccui,a Timothy S. Milne,b Nicholas Harmer,c Alison J. George,b Sarah V. Harding,b Rachel E. Dean,b Andrew E. Scott,b Mitali Sarkar-Tyson,b Brendan W. Wren,a Richard W. Titball,c and Joann L. Priorb Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, London, United Kingdoma; Defence Science and Technology Laboratory, Porton Down, Salisbury, Wiltshire, United Kingdomb; and College of Life and Environmental Sciences, University of Exeter, Exeter, Devon, United Kingdomc

Burkholderia pseudomallei is the causative agent of melioidosis, a disease endemic to regions of Southeast Asia and Northern Australia. Both humans and a range of other animal species are susceptible to melioidosis, and the production of a group 3 polysaccharide capsule in B. pseudomallei is essential for virulence. B. pseudomallei capsular polysaccharide (CPS) I comprises unbranched manno-heptopyranose residues and is encoded by a 34.5-kb locus on chromosome 1. Despite the importance of this locus, the role of all of the genes within this region is unclear. We inactivated 18 of these genes and analyzed their phenotype using Western blotting and immunofluorescence staining. Furthermore, by combining this approach with bioinformatic analysis, we were able to develop a model for CPS I biosynthesis and export. We report that inactivating gmhA, wcbJ, and wcbN in B. pseudomallei K96243 retains the immunogenic integrity of the polysaccharide despite causing attenuation in the BALB/c murine infection model. Mice immunized with the B. pseudomallei K96243 mutants lacking a functional copy of either gmhA or wcbJ were afforded significant levels of protection against a wild-type B. pseudomallei K96243 challenge.

B

urkholderia pseudomallei is a Gram-negative, facultatively anaerobic, motile bacterium and the causative agent of melioidosis. Both animals and humans are susceptible to infection. Although most reports of human infection emanate from Southeastern Asia and Northern Australia (9, 11), it is believed that the global burden of melioidosis is underestimated (11). Indeed, cases of infection have been documented in the Middle East, India, Africa, and the Americas (11). A contributing reason for the underreporting of disease is the difficulty associated with laboratory diagnosis. The misidentification of B. pseudomallei during clinical investigation has been attributed to the varied acute- or chronicphase disease presentations observed following infection by a pathogen termed the “great mimicker” (6, 43). Population studies in regions where the disease is endemic have identified a number of risk factors predisposing to melioidosis, including diabetes, excessive alcohol intake, chronic renal impairment, chronic lung disease, and kava consumption (7, 8). Melioidosis is primarily associated with environmental exposure to B. pseudomallei (10). This organism is a saprophyte commonly found in rural farmland, with the soils of Thailand often cited as an environmental niche (4, 10). Subcutaneous inoculation via skin cuts or abrasions is considered to be the likely route of infection, although the bacterium can be transmitted via the gastrointestinal or pulmonary route (18, 38). In light of the transmissibility of B. pseudomallei via the aerosol route, this organism has been classified as a category B biological threat agent by the Centers for Disease Control and Prevention (31). There is currently no vaccine available to protect against melioidosis, and B. pseudomallei is resistant to antibiotics, including many penicillins, the cephalosporins, and aminoglycosides. Antibiotic treatment regimens often involve protracted courses of intravenous and oral drugs to clear the infection and ensure no future relapse (42). Therefore, a priority for B. pseudomallei research is the development or assessment of effective prophylactics and therapeutics. The generation of novel medical countermeasures against melioidosis will rely on understanding the virulence

0019-9567/12/$12.00

Infection and Immunity

p. 1209 –1221

mechanisms of B. pseudomallei. Several reports have shown that capsular polysaccharide (CPS) I of B. pseudomallei is a key virulence determinant and that loss of capsule production results in severe attenuation in animal models of disease (1, 28, 29, 37, 41). In support of these findings, CPS I (also known as type I O-polysaccharide) is also an essential virulence factor for the closely related obligate equine pathogen B. mallei (13, 28). B. pseudomallei capsule is considered a promising component for inclusion in a melioidosis vaccine, as experimental evidence indicates that CPS I is both immunogenic and protective. Parthasarathy et al. demonstrated the presence of anti-CPS antibodies in convalescent patient antisera (25), and independent studies have demonstrated that CPS I plays a role in reducing B. pseudomallei phagocytosis by host cells by preventing complement factor C3b deposition on the surface of the bacterium (29). Mice vaccinated with CPS I purified from B. pseudomallei have a prolonged time to death upon subsequent B. pseudomallei challenge (23), while intravenous administration of monoclonal antibodies generated against CPS I conferred significant levels of passive protection of mice against an intraperitoneal challenge with B. pseudomallei (19). There are at least four polysaccharide-encoding gene clusters within the two chromosomes of B. pseudomallei K96243 (17). Genes involved in sugar biosynthesis and transport of CPS I are

Received 16 August 2011 Returned for modification 8 September 2011 Accepted 18 December 2011 Published ahead of print 17 January 2012 Editor: A. Camilli Address correspondence to Richard W. Titball, [email protected]. Supplemental material for this article may be found at http://iai.asm.org/. Jon Cuccui and Timothy S. Milne contributed equally to this article. Copyright © 2012, American Society for Microbiology. All Rights Reserved. doi:10.1128/IAI.05805-11

iai.asm.org

1209

Cuccui et al.

TABLE 1 Bacterial strains and plasmids used in this study Strain or plasmid

Genotype or description

Source or reference

Escherichia coli TOP10 S17-␭pir

Chemically competent cloning strain Conjugal strain for transfer of pDM4

Invitrogen 33

B. pseudomallei K96243 ⌬manC 13E8 13H8 1B9 4A4 JCX2C5 13G10 ⌬wcbG 6G8 4E6 JCXB2 ⌬wcbL ⌬gmhA ⌬wcbM 1A6 ⌬wcbP 13G3 12F5

Wild type K96243 with unmarked in-frame manC deletion K96243 transposon mutant with disrupted wcbA K96243 transposon mutant with disrupted wcbB K96243 transposon mutant with disrupted wcbC K96243 transposon mutant with disrupted wcbD K96243 transposon mutant with disrupted wzm2 K96243 transposon mutant with disrupted wcbE K96243 with unmarked in-frame wcbG deletion K96243 transposon mutant with disrupted wcbH K96243 transposon mutant with disrupted wcbI K96243 transposon mutant with disrupted wcbJ K96243 with unmarked in-frame wcbL deletion K96243 with unmarked in-frame gmhA deletion K96243 with unmarked in-frame wcbM deletion K96243 transposon mutant with disrupted wcbN K96243 with unmarked in-frame wcbP deletion K96243 transposon mutant with disrupted wcbQ K96243 transposon mutant with disrupted wcbR

33 This study 5 5 5 5 5 5 This study 5 5 5 This study This study This study 5 This study 5 5

B. thailandensis E264

Wild type

A. Scott (Dstl)

Plasmids pCRII-Blunt-TOPO pDM4

Cloning vector; Kanr Zeor Suicide vector containing sacB counterselection gene; Ampr Cmr

Invitrogen 15

found on chromosome 1 (17). The polysaccharide product of this coding region is a high-molecular-weight unbranched polymer of -3)-2-O-acetyl-6-deoxy-␤-D-manno-heptopyranose-(1- residues (20, 28). A previous signature-tagged mutagenesis screening of BALB/c mice identified transposon insertions in 12 open reading frames (ORFs) within this cluster of genes, repeatedly demonstrating the importance of this capsule in pathogenesis (5). Therefore, the aim of this study was to investigate the function of individual ORFs within this region and their role in polysaccharide biosynthesis. In this work, we investigated the phenotype of B. pseudomallei K96243 mutants containing individual disruptions in 18 of the 25 genes proposed to be responsible for CPS I production. Our analysis of the resulting phenotypes, combined with bioinformatic approaches and animal infection studies, provides new insights into the mechanism of capsule biosynthesis in B. pseudomallei. MATERIALS AND METHODS Bacterial strains and culture conditions. The bacterial strains and plasmids used in this study are described in Table 1. Broth-grown cultures were incubated with aeration for 18 h. Escherichia coli strains maintaining the pCRII-Blunt-TOPO plasmid were grown in Luria-Bertani (LB) broth or on LB agar supplemented with 50 ␮g kanamycin ml⫺1 (Sigma-Aldrich) at 37°C. E. coli strains maintaining the pDM4 plasmid (22) were grown in LB broth or on LB agar supplemented with 30 ␮g chloramphenicol ml⫺1 (Sigma-Aldrich) at 37°C. B. pseudomallei K96243 strains and B. thailandensis E264 were routinely grown in LB broth or on LB agar at 37°C unless stated otherwise. To confirm the presence of the miniTn5Km2 trans-

1210

iai.asm.org

poson, B. pseudomallei K96243 was grown in LB broth or on plates supplemented with 400 ␮g kanamycin ml⫺1 at 37°C. Bioinformatic analysis of function. Amino acid sequences from each protein within the B. pseudomallei K96243 CPS I coding locus were compared to the National Center for Biotechnology Information (NCBI) nonredundant protein database using the protein-to-protein BLAST algorithm BLASTp. Generation of B. pseudomallei K96243 mutants. Transposon mutants of B. pseudomallei K96243 were produced as described previously (5). For the generation of unmarked in-frame B. pseudomallei capsule mutants, DNA manipulations were performed as described by Sambrook et al. (32). Genomic DNA preparations were isolated using the Puregene DNA Isolation kit (Gentra Systems), and plasmid DNA was isolated from E. coli strains using the QIAprep Spin Miniprep and Plasmid Midi kits (Qiagen), both according to the manufacturer’s instructions. PCR products and digested plasmid DNA were gel purified using the QIAEX II Agarose Gel Extraction kit (Qiagen). All primers used for the construction and screening of unmarked in-frame B. pseudomallei mutants, detailed in Table 2, were designed using Clone Manager software (Scientific and Educational Software). For each gene targeted for in-frame deletion, flanking regions of approximately 500 bp were PCR amplified using the left flank forward/reverse (LFF/LFR) and right flank forward/reverse (RFF/ RFR) primer pairs under the following conditions: denaturation at 95°C for 2 min, followed by 30 cycles of denaturation at 95°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 2 min, followed by a final extension at 72°C for 2 min. Amplicons were gel purified and cloned into separate pCRII-Blunt-TOPO vectors (Invitrogen) according to the manufacturer’s instructions. The resultant constructs were digested using the restriction enzyme corresponding to the engineered restriction site on the LFR/RFF primer pair, and the linearized constructs were ligated together.


B. pseudomallei K96243 CPS I Coding Region

TABLE 2 Oligonucleotide primers used for the construction and screening of unmarked in-frame B. pseudomallei capsule mutants Oligonucleotide

Sequence (5= to 3=)a

manC LFF manC LFR manC RFF manC RFR manC screenF manC screenR wcbG LFF wcbG LFR wcbG RFF wcbG RFR wcbG screenF wcbG screenR wcbL LFF wcbL LFR wcbL RFF wcbL RFR wcbL screenF wcbL screenR gmhA LFF gmhA LFR gmhA RFF gmhA RFR gmhA screenF gmhA screenR wcbM LFF wcbM LFR wcbM RFF wcbM RFR wcbM screenF wcbM screenR wcbP LFF wcbP LFR wcbP RFF wcbP RFR wcbP screenF wcbP screenR pDM4 CAMscreenF pDM4 CAMscreenR pDM4 sacBscreenF pDM4 sacBscreenR

TCTAGACTCTGGCAATGCGGAAAGTA AGATCTCATGATGTGTGGTCCTGGAG AGATCTTCGACCTGACGCGACAGTCC TCTAGAGCGCAGGTTGCGGTCGTTCC GCTGGATTGCAAGCTGTCTG CAAAGAGCGGCGAGCGAATC TCTAGACTTGTGGAGTGCGCCTTTGA AGATCTCATTTCTTTTACCTAAACGC AGATCTTTGATGGACGCACTGCAATG TCTAGACTCGAACATCGGAAACGAGG AATTCAATCGCCGCTTCGAC GCGGTTTCAACTGGTTCTCG TCTAGAGTGATGACCACGCGGATGTT GCATGCTGCGCGAATGATTGTCGGAT GCATGCACGCAGGCATGGAGAATCGC TCTAGACTCGCGCATTTCTCCGCCGC CCTGATCGGCCGGTACTATG GCCCGCAAACGATATGTCCG TCTAGATCGGCGACGTTCGGGGGCTT CGTACGGGCCTCGGCGATGCTGTTCG CGTACGTTCGGGAAGCAGTGAAATGA TCTAGACGGTCCCGAAACGCCCTTTT GCAGCTGGCTTATCGGATCG TCTCGCCGGCAAGAATGTCC TCTAGAGCTTGCGACGGTTCGGAAAG AGATCTGGCCAAGATGATCGCTTCTC AGATCTGGAATCTGCAAGTGAAGAAT TCTAGACGCCGTCGCCACACGCGATG CGCGAATTGACGTACATCAC CCTAGAACGATCCGGCATCC GCATGCGGCATTTGCGACGCGATAGC AGATCTGGTACTGTCGCTCACGACAC AGATCTCGCGAGGAGCGATAGCGTGG GCATGCTCGAGTGTGAGCGGTAAACG CTGCAGCTCAATTCCGACTC GACCGCAAAGAACCCGTGAC ATCCCAATGGCATCGTAAAG TAAGCATTCTGCCGACATGG CGGCTACCACATCGTCTTTG GCAATCAGCGGTTTCATCAC

a

Engineered restriction sites are underlined.

The ligation mixture was used as a template in a PCR using gene-specific LFF/RFR primer pairs, and the amplified product was cloned into pCRIIBlunt-TOPO. The construct was digested using the restriction enzyme corresponding to the engineered restriction site on the LFF/RFR primer pair, and the desired DNA fragment was cloned separately into the suicide vector pDM4. Each resultant plasmid per targeted gene was individually electroporated into E. coli S17-␭pir and transferred via conjugation into B. pseudomallei as described previously (33). B. pseudomallei K96243/pDM4 transconjugants were selected on chloramphenicol-supplemented LB agar plates following incubation at 37°C, and merodiploids arising from chromosomal integration of the suicide plasmid were resolved by plating on LB agar (without NaCl) supplemented with 10% (wt/vol) sucrose (22). LB-sucrose plates were incubated at 24°C for a maximum of 5 days. Confirmation of mutants. Transposon insertion sites were identified by arbitrary PCR and confirmed by gene-specific PCR to show an increase in the ORF size due to insertion of the miniTn5Km2 element (5). To rule out the possibility of a second site transposon insertion in the Tn::wcbJ and Tn::wcbN mutants, a Southern blot assay was carried out by screening 3 ␮g of SacI-digested genomic DNA with a probe consisting of a 500-bp

March 2012 Volume 80 Number 3

fragment of the kanamycin resistance cassette amplified from the miniTn5Km2 transposon. Generation of a single band combined with gene-specific PCR and arbitrary PCR confirmed that only a single transposon insertion had occurred in each mutant and in the reported ORFs (data not shown). To identify individual B. pseudomallei capsule mutants with an unmarked in-frame deletion in gmhA, manC, wcbG, wcbL, wcbM, and wcbP, sucrose-resistant colonies were screened by PCR. The presence of a mutant copy of the targeted gene for deletion and absence of the chloramphenicol resistance cassette and sacB gene were investigated using the screening primer pairs detailed in Table 2. The region of deletion for each mutant was amplified, cloned into pCRII-Blunt-TOPO, and sequenced by Beckman Coulter Genomics to confirm the in-frame deletion. Southern blotting was performed to demonstrate the generation of B. pseudomallei K96243 strains with an unmarked in-frame deletion. Restriction enzyme digests of 2 to 3 ␮g genomic DNA from B. pseudomallei K96243 wild type and mutant for each targeted ORF were separated by electrophoresis on a 1.0% agarose gel and transferred to a nylon membrane. An amplicon generated using the ORF-specific primer pair LFF/LFR was labeled with DIG-11-dUTP during PCR amplification and used as a probe to hybridize to the membrane. DNA fragments to which the probe hybridized were detected in a chemiluminescence assay (CSPD substrate, 30 min exposure, X-ray film). Heat inactivation of B. pseudomallei. B. pseudomallei K96243 strains were grown in 10-ml LB broth aliquots overnight with aeration. Cultures were harvested by centrifugation at 6,000 ⫻ g for 15 min and washed in 5 ml phosphate-buffered saline (PBS; CaCl2 and MgCl2 free; Gibco). After harvesting, bacterial pellets were resuspended in 3 ml PBS and enumerated by culture. Suspensions were heat inactivated by incubation in a water bath maintained at 80°C for 4 h. Western blotting. Heat-killed B. thailandensis E264 and B. pseudomallei K96243 wild-type and capsule mutant preparations were separated by gel electrophoresis based on the method described by Laemmli (21). Briefly, samples were diluted in sample buffer (6 M urea, 100 mM Tris-HCl [pH 6.8], 4.5% [wt/vol] SDS, 5% [vol/vol] ␤-mercaptoethanol, 0.2% [wt/vol] bromophenol blue) and heated at 100°C for 10 min to ensure protein denaturation. Denatured samples and MagicMark XP Western Protein Standard (Invitrogen) were loaded onto 10% Trisglycine Novex gels (Invitrogen) in a Novex X-Cell gel tank (Invitrogen) and separated according to molecular weight for 120 min at 150 V. The quantity of cells for each isolate was standardized prior to gel loading to ensure that capsule expression levels were comparable across all of the immunoblots. Immunoblotting was performed using a Trans-Blot semidry transfer cell (Bio-Rad) to transfer separated proteins onto Hybond-C Extra Nitrocellulose membrane (Amersham), and membranes were blocked overnight in 10% BLOTTO (10% [wt/vol] skim milk powder and 0.1% [vol/vol] Tween 20 in PBS). Membranes were probed with the IgG 2b isotype anti-CPS I monoclonal antibody (MAb) 4VIH12, which was previously produced by immunizing BALB/c mice via the intraperitoneal route with heat-killed B. pseudomallei, collecting the splenocytes, mixing them with myeloma cells at an approximate ratio of 1 myeloma cell to 5 splenocyte cells, and identifying 4VIH12 from a hybridoma cell line secreting antibody reactive to CPS I (14). Membranes were probed with MAb 4VIH12 (10 ␮g ml⫺1) in 5% BLOTTO (5% [wt/vol] skim milk powder and 0.1% [vol/vol] Tween 20 in PBS) on a rocking platform at room temperature for 30 min (1). After being washed twice for 10 min in wash solution (0.1% [vol/vol] Tween 20 in PBS) on a rocking platform at room temperature, membranes were probed with secondary anti-mouse IgG whole molecule peroxidase conjugate (Sigma) in 5% BLOTTO at a 1:10,000 concentration on a rocking platform at room temperature for 30 min. Blots were washed once in wash solution for 10 min and then washed again with PBS for the same duration. All washes were performed with agitation at room temperature. Detection was performed using Plus Western Blotting Detection Reagents (GE Healthcare) according to the manufacturer’s instructions. Membranes were exposed to ECL Hyperfilm

iai.asm.org 1211

Cuccui et al.

(GE Healthcare) in a darkroom, and blots were developed using XOgraph ready-to-use developer and fixer chemicals (Champion Photochemistry, Essex, United Kingdom). Immunofluorescence studies. The effects of gene deletion on the expression of capsule polysaccharide in B. pseudomallei was tested by immunofluorescent staining using MAb 4VIH12 as described previously (1). Approximately 1 ⫻ 109 CFU of each B. pseudomallei K96243 capsule mutant were fixed on separate glass coverslips by air drying. After being washed in PBS, coverslips were incubated in 5% (vol/vol) fetal calf serum (FCS) in PBS for 2 h. After three 2-min washes in PBS, coverslips were incubated with MAb 4VIH12 (1 ␮l/ml in 5% [vol/vol] FCS/PBS) for 1 h at 37°C and then, following three washes in PBS, incubated with anti-mouse IgG (whole molecule) F(ab=)2 fragment-fluorescein isothiocyanate (FITC) conjugate (Sigma-Aldrich) for 1 h at 37°C. Coverslips were washed three times for 2 min in PBS and mounted on glass slides with 92% (vol/vol) aqueous glycerol. The level of fluorescence for each bacterial isolate was observed under phase-contrast and fluorescence microscopy using an Olympus FluoView laser scanning microscope. Animal studies involving B. pseudomallei K96243⌬gmhA, B. pseudomallei K96243Tn::wcbJ, and B. pseudomallei K96243Tn::wcbN capsule mutants. Female BALB/c mice (Charles River) approximately 6 to 8 weeks old were caged together with free access to food and water and subjected to a 12-h light/12-h dark cycle. For challenge with viable B. pseudomallei, the animals were handled under biosafety level 3 containment within a half-suit isolator compliant with British standard BS5726. All investigations involving animals were carried out according to the requirements of the Animal (Scientific Procedures) Act 1986. Humane endpoints were strictly observed, and animals deemed incapable of survival were humanely killed by cervical dislocation. For the attenuation study, groups of six BALB/c mice were challenged via the intranasal (i.n.) route with 1 ⫻ 103 CFU of B. pseudomallei K96243, B. pseudomallei K96243⌬gmhA, B. pseudomallei K96243Tn::wcbJ, or B. pseudomallei K96243Tn::wcbN and the infection was monitored for 5 weeks. For the protection study, six female naive BALB/c mice and survivors from the attenuation study, previously challenged with one of the B. pseudomallei K96243 capsule mutants, were challenged via the i.n. route with 1 ⫻ 103 CFU of B. pseudomallei K96243 and the infection was monitored for 5 weeks. At the end of the experiment, surviving mice were culled and the lungs, liver, and spleen were removed from each animal for processing and bacterial enumeration. Significance of attenuation and protection was determined by the Mantel-Haenszel log rank method using GraphPad Prism version 4.0.

RESULTS

Bioinformatic analysis of the B. pseudomallei K96243 CPS I coding region. The amino acid sequence encoded by each ORF within the B. pseudomallei K96243 CPS I coding locus was compared against the NCBI nonredundant protein database. The similarity and identity of resulting BLAST searches are reported in Table 3. This analysis indicates that sugar biosynthesis genes are located in a cluster flanked by wcbA and wcbO. This arrangement is typical of group 2 and group 3 CPS clusters (39), characterized by a central sugar biosynthesis cassette surrounded by genes encoding polysaccharide transport proteins. Sugar biosynthesis. A BLASTp search of the NCBI database using the amino acid sequences of GmhA, WcbL, WcbN, and WcbM identified orthologues of enzymes within Y. pseudotuberculosis H892/87 necessary for the biosynthesis of GDP-D-glycero-␣-Dmanno-heptopyranose. GmhA in B. pseudomallei is a metalloenzyme that catalyzes the interconversion of D-sedoheptulose-7-phosphate to D-glycero-␣-D-manno-heptopyranose 7-phosphate (16). WcbL was found to have similarity to HddA from Y. pseudotuberculosis. HddA catalyzes the phosphorylation of D-␣,␤-D-heptose-7-P to D-␣,␤-Dheptose-1,7,PP using ATP (24). WcbN is an orthologue of GmhB and

1212

iai.asm.org

in both B. pseudomallei and Y. pseudotuberculosis is thought to catalyze the dephosphorylation of D-␣-D-heptose-1,7,PP to generate D-␣D-heptose-1-P. WcbM is annotated as a putative D-glycero-D-mannoheptose 1-phosphate guanosyltransferase and is likely to function like HddC in Y. pseudotuberculosis and synthesize GDP-D-glycero-␣-Dmanno-heptopyranose. The next step in B. pseudomallei CPS I sugar biosynthesis is predicted to be performed by WcbK. WcbK is a protein with high amino acid similarity to Y. pseudotuberculosis DmhA, which converts GDP-D-glycero-␣-D-manno-heptopyranose to GDP-4-keto6-deoxy-D-manno-heptose. Finally, WcbJ is an ortholgoue of Y. pseudotuberculosis DmhB and is likely to be the B. pseudomallei reductase that alters GDP-4-keto-6-deoxy-D-manno-heptose to GDP-6-deoxy-D-manno-heptose. A model of sugar biosynthesis in B. pseudomallei K96243 is shown in Fig. 1 (16). Glycosyltransferases. Glycosyltransferases are enzymes that are responsible for the transfer of sugar residues from nucleotide precursors to the growing carbohydrate chain (40). BLASTp searches indicate that there are three enzymes within the cluster likely to be glycosyltransferases; WcbB, WcbE, and WcbH. Transporters. In E. coli, the kps gene cluster is involved in the export of group 3 CPS (30). KpsC and KpsS are proteins associated with the inner face of the cytoplasmic membrane involved in the attachment of a phospholipid moiety to the polysaccharide prior to translocation across the inner membrane by KpsM and KpsT (30). Our BLASTp searches identified a number of proteins coded for within the CPS I coding region that are orthologues of the Kps cluster found in E. coli. WcbA is an orthologue of KpsC, while WcbO is similar to KpsS and is therefore likely to play a similar role. Wzm2 is an orthologue of KpsM in E. coli, an ABC transporter that is part of an inner membrane polysaccharide export system (26). WcbC and WcbD have orthologues in E. coli KpsD and KpsE, respectively, and we propose a likely involvement of WcbC/WcbD in the movement of B. pseudomallei CPS I to the outer membrane. Lipid biosynthesis. A lipid anchor for B. pseudomallei CPS I has yet to be described. Bioinformatic analysis of ORFs within the cluster found five genes encoding proteins (WcbP/Q/R/S/T) whose function is putatively involved in the synthesis of a lipid(s). Most strikingly, WcbR is a putative type I polyketide synthase/ fatty acid synthase. Multiple domains can be identified within this protein, including an acyltransferase domain, a ␤-ketoacyl-acyl carrier protein synthase domain responsible for the elongation steps in fatty acid biosynthesis, a putative methyltransferase domain, a keto-reductase domain, and an enoyl-reductase domain. There is no apparent dehydratase domain, which would normally be expected in a fatty acid synthase. However, this function could be supplied by WcbP, a member of the short-chain dehydrogenase/reductase family, whose functions include dehydratase activities on sugar molecules. Within the CPS I cluster, wcbQ is the ORF adjacent to wcbR and is annotated as coding for a putative sulfatase. However, homologous genes are found in close association on the chromosome with wcbR homologues in other proteobacteria (e.g., Ralstonia sp. strain 5_7_47FAA, Geobacter lovleyi, Pseudomonas putida F1), suggesting that this protein is involved in the production of a fatty molecule. WcbS is an orthologue of LpxC in Pantoea ananatis. This enzyme may be a UDP-3-O-(R-3-hydroxymyristoyl)-Nacetylglucosamine deacetylase involved in the deacetylation of UDP-(3-O-acyl)-N-acetylglucosamine to UDP-3-O-(3-hydroxytetradecanoyl)-glucosamine in the second step of lipid A biosynthesis. Without knowledge of the nature of the lipid carrier to which the



TABLE 3 Summary of bioinformatic, Western blotting, and immunofluorescence analysesa Category and gene Glycosyltransferases wcbB wcbE wcbH

Transporters BPSL2806a wcbA wcbC wcbD wcbO wzm2 wzt2 Sugar biosynthesis gmhA manC wcbJ wcbK wcbL wcbM wcbN

Lipid biosynthesis wcbP wcbQ wcbR wcbS wcbT

Unknown BPSL2786 wcbF wcbG wcbI a

Western blot assay

Immunofluorescence

Negative

Negative

Negative

Negative

Negative

Negative

Pseudogene; CPS export/inner membrane Polysaccharide transfer; 32% identity and 46% amino acid similarity to E. coli KpsC Outer membrane transporter; 40% identity and 61% amino acid similarity to Haemophilus influenzae BexD CPS inner membrane transporter; 44% identity and 68% amino acid similarity to H. influenzae BexC Polysaccharide transfer; 38% identity and 53% similarity to E. coli KpsS ABC-2-type inner membrane transporter; 43% identity and 66% amino acid similarity to H. influenzae BexB ABC CPS transporter; 48% identity and 66% amino acid similarity to E. coli kpsT

Not determined Weakly positive Positive

Not determined Weakly positive Weakly positive

Positive

Weakly positive

Not determined Positive

Not determined Negative

Not determined

Not determined

Sedoheptulose-7-phosphate isomerase; 59% identity and 74% amino acid similarity to Y. pseudotuberculosis H892/87 GmhA Mannose metabolism; 50% identity and 69% amino acid similarity to E. coli ManC Reductase; 36% identity and 55% amino acid similarity to Y. pseudotuberculosis H892/87 DmhB GDP sugar epimerase/dehydratase; 50% identity and 66% amino acid similarity to Y. pseduotuberculosis H892/87 DmhA GmhP kinase; 49% identity and 69% amino acid similarity to Y. pseudotuberculosis H892/87 HddA D[R]-glycero-D[R]-manno-Heptose 1-phosphate guanosyltransferase; 39% identity and 54% amino acid similar to Y. pseudotuberculosis H892/87 HddC D[R]-glycero-D[R]-manno-Heptose-bisphosphate-phosphatase; 33% identity and 54% amino acid similarity to Y. pseudotuberculosis H892/87 GmhB

Positive

Positive

Weakly positive Weakly positive

Negative Positive

Not determined

Not determined

Negative

Negative

Negative

Negative

Weakly positive

Positive

Short-chain dehydrogenase/reductase;39% identity and 59% amino acid similarity to Dictyostelium discoideum DDB_G0289059 Sulfatase membrane protein; 52% identity and 64% amino acid similarity to Ralstonia pickettii Rpic_1147 CPS biosynthesis, fatty acid synthase; 37% identity and 53% amino acid similarity to Sinorhizobium fredii RkpA Lipid biosynthesis; 34% identity and 52% amino acid similarity to Pantoea ananatis LpxC; UDP-3-O-[3-hydroxymyristoyl] N-acetylglucosamine deacetylase 2-Amino-3-ketobutyrate CoA ligase; 36% identity and 57% amino acid similarity to E. coli O157:H7 Kbl

Weakly positive

Negative

Weakly positive

Negative

Weakly positive

Negative

Not determined

Not determined

Not determined

Not determined

Acetyltransferase; 52% identity and 63% amino acid similarity to Ralstonia pickettii GCN5-related acetyltransferase CPS biosynthesis/sulfotransferase family protein; 26% identity and 46% amino acid similarity to Desulfovibrio sp. sulfotransferase No conserved domains/CPS biosynthesis No conserved identity domains/CPS biosynthesis

Not determined

Not determined

Not determined

Not determined

Negative Negative

Negative Negative

Putative function Glycosyltransferase; 64% identity and 76% amino acid similarity to Ralstonia pickettii group 1 glycosyltransferase Glycosyltransferase/mannosyltransferase; 28% identity and 43% amino acid similarity to Aquifex aeolicus mannosyltransferase B Glycosyltransferase group 1; 26% identity and 45% amino acid similarity to Clostridium lentocellum transferase

Pairwise BLAST analysis was performed. Levels of CPS expression were determined by reactivity of B. pseudomallei K96243 capsule mutants with anti-CPS MAb 4VIH12.

polysaccharide is attached, an exact role for this enzyme in lipid biosynthesis remains unclear. Generation of B. pseudomallei K96243 CPS I mutants using transposon and targeted in-frame deletion mutagenesis. For this study, we disrupted 18 of the 25 ORFs predicted to be involved in B.


pseudomallei K96243 CPS I biosynthesis and export (Fig. 2). Twelve of the mutants successfully generated were isolated during a previously reported mutagenesis study using miniTn5Km2 (5). This Tn5 derivative does not contain a transcriptional terminator, and therefore, genes downstream of the transposon may be tran-

iai.asm.org 1213

Cuccui et al.

FIG 1 Biosynthesis of GDP-6-deoxy-␣-D-manno-heptose from D-sedoheptulose 7-P.

scribed from a Tn5 promoter (12). Our results also seem to confirm this feature. For example, while the B. pseudomallei K96243Tn::wzm2 mutant indicates the presence of CPS I by Western blot analysis, the function of a glycosyltransferase-encoding gene (wcbE) immediately downstream is not affected (see later). However, when wcbE is interrupted in B. pseudomallei K96243Tn:: wcbE, a complete loss of CPS I is observed. In order to demonstrate transcriptional activity in ORFs upstream and downstream of the insertion site, RNA was extracted

from all 12 transposon mutants. First, transposon insertions were confirmed in the individual mutants by using genomic DNA as template for a PCR to demonstrate the generation of an increased amplicon corresponding to the size of the transposon (1.5 kb). The RNA was then DNase treated and subjected to reverse transcription in order to generate cDNA. The cDNA was used as a template to amplify regions specific for ORFs downstream and upstream of each insertion. As a control, DNase-treated RNA only was also used as a template. The results demonstrated transcrip-

FIG 2 Genetic organization of the CPS I coding locus of B. pseudomallei K96243. Red arrows indicate in-frame deletions, black arrows indicate orientation of miniTn5km2 insertions. A cassette of genes orthologous to Y. pseudotuberculosis H897/87 involved in the biosynthesis of GDP-6-deoxy-D-manno-heptose is indicated by broken lines.

1214

iai.asm.org



tional activity in every upstream and downstream gene of every transposon insertion site (see Fig. S1 in the supplemental material). Despite these tests, it is important to note that polar effects due to transposon insertions cannot be completely ruled out and remain a possibility. A further six B. pseudomallei K96243 unmarked in-frame deletion mutants were generated for the manC (BPSL2810), wcbG (BPSL2801), wcbL (BPSL2796), gmhA (BPSL2795), wcbM (BPSL2794), and wcbP (BPSL2791) genes. Validation of these inframe deletion mutants was obtained by Southern blotting and sequencing of the ORF after mutagenesis to confirm each deletion genotype (data not shown). Western blot assay characterization of B. pseudomallei K96243 CPS I mutants. The relative levels of CPS I produced by each mutant were compared to wild-type B. pseudomallei K96243 using Western blotting with a monoclonal antibody reactive against CPS I (Fig. 3 and Table 3). The production of CPS I by wild-type B. pseudomallei K96243 grown in LB broth was demonstrated by the presence of a high-molecular-weight ladder reactive with MAb 4VIH12. This ladder was not observed in extracts from capsule-negative B. thailandensis E264. Extracts from B. pseudomallei K96243 mutants lacking wcbB, wcbE, wcbG, wcbH, wcbI, wcbL or wcbM did not react with MAb 4VIH12. These results suggested that inactivating each of these ORFs abolishes the ability of B. pseudomallei K96243 to produce the CPS I epitope recognized by this antibody. B. pseudomallei K96243 mutants individually lacking a functional manC, wcbA, wcbJ, wcbN, wcbP, wcbQ, or wcbR gene showed a reduced reactivity with MAb 4VIH12. The levels of CPS production by B. pseudomallei K96243Tn::wcbQ and B. pseudomallei K96243Tn::wcbR seemed directly comparable to one another. The remaining mutants, lacking a functional gmhA, wcbC, wcbD, or wzm2 gene, all retained the ability to generate CPS I at levels equivalent to that of the wild type. In summary, Western blotting indicated that a subset of ORFs within the B. pseudomallei K96243 CPS I coding region are essential for the production of the CPS epitope recognized by MAb 4VIH12. However, the inactivation of other ORFs reduced, but did not completely abolish, reactivity with MAb 4VIH12. The results also indicated that some ORFs are not essential for the production of the epitope recognized by MAb 4VIH12. Immunofluorescence characterization of B. pseudomallei K96243 CPS I mutants. External presentation of antibodyreactive CPS I on the surface of each B. pseudomallei K96243 CPS I mutant and the wild type was determined using immunofluorescence (Fig. 4 and Table 3). B. pseudomallei K96243 mutants lacking a functional copy of wcbB, wcbE, wcbG, wcbH, wcbI, wcbL, or wcbM failed to react with MAb 4VIH12. These findings were in agreement with the Western blotting results obtained using MAb 4VIH12. Although B. pseudomallei K96243⌬manC, B. pseudomallei K96243⌬wcbP, B. pseudomallei K96243Tn::wcbQ, B. pseudomallei K96243Tn::wcbR, or B. pseudomallei K96243Tn::wzm2 reacted with MAb 4VIH12 in Western blot assays, these mutants did not react with the same antibody when used for whole-cell immunofluorescence. This finding suggests that these mutants were unable to export 4VIH12-reactive CPS I onto the surface of the bacterium. Compared to the wild type, B. pseudomallei K96243Tn::wcbC or B. pseudomallei K96243Tn::wcbD showed reduced reactivity with MAb 4VIH12 by immunofluorescence. This finding con-


trasts with the result of Western blotting, which indicated that these mutants were able to produce high-molecular-weight CPS I at levels equivalent to that of the wild type. Although the B. pseudomallei K96243Tn::wcbA mutant showed a reduced immunofluorescence signal, this result was in agreement with the corresponding Western blot analysis. Perhaps the most striking observation from immunofluorescence studies was that B. pseudomallei K96243 CPS I mutants lacking a functional gmhA, wcbJ, or wcbN gene appeared to retain the immunogenic epitope recognized by MAb 4VIH12. Any interruption within the sugar biosynthesis pathway was expected to lead to a complete loss of CPS I generation. However, the ability of the B. pseudomallei K96243⌬gmhA mutant to express CPS I at levels comparable to that of the wild type was in direct agreement with the corresponding Western blot assay result. In the case of the B. pseudomallei K96243Tn::wcbJ and B. pseudomallei K96243Tn:: wcbN transposon derivatives, Western blot analysis indicated a reduced expression of CPS I, relative to either the B. pseudomallei K96243 wild type or the B. pseudomallei K96243⌬gmhA mutant. In summary, the use of immunofluorescence to characterize capsule expression by the panel of B. pseudomallei K96243 CPS I mutants indicated that the majority of genes are essential for a B. pseudomallei K96243 phenotype expressing CPS I, reactive with MAb 4VIH12, on the bacterial cell surface. Although differences were observed between the Western blot assay and immunofluorescence data, these findings indicate that subsets of ORFs within the B. pseudomallei K96243 CPS I coding region can be further classified, according to the reactivity of the corresponding B. pseudomallei K96243 mutant against MAb 4VIH12, as follows: essential for CPS I production (wcbB, wcbE, wcbG, wcbH, wcbI, wcbL, and wcbM); if deleted, result in reduced but not abolished CPS I production and/or expression (manC, wcbA, wcbJ, wcbN, wcbP, wcbQ, and wcbR); or not essential for CPS I production/expression (gmhA). We can also highlight the presence of ORFs required for export of the polysaccharide to the bacterial outer membrane (wcbC, wcbD, and wzm2). Assessment of attenuation of B. pseudomallei K96243⌬ gmhA, B. pseudomallei K96243Tn::wcbJ, or B. pseudomallei K96243Tn::wcbN in BALB/c mice. BALB/c mice were infected via the i.n. route with B. pseudomallei K92643, B. pseudomallei K96243⌬gmhA, B. pseudomallei K96243Tn::wcbJ, or B. pseudomallei K96243Tn::wcbN to investigate whether these B. pseudomallei CPS I mutants retained virulence. The exact challenge doses were as follows: the B. pseudomallei K96243 wild type, 1,006 CFU (approximately 5 minimum lethal doses [MLD]); B. pseudomallei K96243⌬gmhA, 1,060 CFU; B. pseudomallei K96243Tn:: wcbJ, 896 CFU; B. pseudomallei K96243Tn::wcbN, 929 CFU. The MLD for wild-type B. pseudomallei K96243 has previously been calculated to be 195 CFU ml⫺1 by the i.n. route in BALB/c mice (S. V. Harding, personal communication). All mice challenged with the B. pseudomallei K96243⌬gmhA, B. pseudomallei K96243Tn:: wcbJ, or B. pseudomallei K96243Tn::wcbN mutant survived (P ⬍ 0.01 for all mutants, data not shown). Assessment of protection afforded by B. pseudomallei K96243⌬gmhA, B. pseudomallei K96243Tn::wcbJ, or B. pseudomallei K96243Tn::wcbN in BALB/c mice. To investigate whether mice that survived a challenge with the B. pseudomallei K96243⌬gmhA, B. pseudomallei K96243Tn::wcbJ, or B. pseudomallei K96243Tn::wcbN capsule mutant were protected against a subsequent challenge, these mice were challenged via the i.n.

iai.asm.org 1215

Cuccui et al.

FIG 3 Western blot assays investigating the expression of CPS I by wild-type (WT) B. pseudomallei K96243 and capsule mutants. Loading of B. pseudomallei K96243 strains was standardized across all of the gels, and blots were probed with anti-CPS MAb 4VIH12 as described in Materials and Methods. The values to the left of each panel are molecular sizes in kilodaltons. BT, B. thailandensis cell lysate.

1216

iai.asm.org



FIG 4 Immunofluorescence images investigating the expression of CPS I by wild-type (WT) B. pseudomallei K96243 and capsule mutants. Heat-killed B. pseudomallei K96243 strains were labeled with anti-CPS MAb 4VIH12 and anti-mouse IgG-FITC conjugate as described in Materials and Methods. Each panel shows both immunofluorescence (left half) and phase-contrast (right half) images.

route with 1,000 CFU of wild-type B. pseudomallei K96243. All control mice succumbed to the infection by day 11 (Fig. 5), and the median survival time of the animals in this group was 3 days. Mice previously immunized via the i.n. route with B. pseudomallei K96243Tn::wcbN died by day 31 (median survival time, 11 days). However, despite an increased time to death, relative to that of the control group, this was not statistically significant (P ⫽ 0.0759). Mice previously immunized via the i.n. route with B. pseudomallei K96243⌬gmhA died by day 34 following a wild-type B. pseudomallei K96243 challenge (median survival time, 13.5 days), which was statistically significant (P ⫽ 0.006). Mice previously immunized via the i.n. route with B. pseudomallei K96243Tn::wcbJ and challenged with wild-type B. pseudomallei K96243 showed 33% survival (median survival time, 29.5 days), with 2/6 mice remaining alive at day 35. This level of protection was statistically significant (P ⫽ 0.0007). At the end of the experiment, B. pseudomallei was recovered from the lungs, livers, and spleens of survivors, indicating that sterile immunity was not achieved.


FIG 5 Survival of BALB/c mice previously immunized with B. pseudomallei K96243⌬gmhA, B. pseudomallei K96243Tn::wcbN, or B. pseudomallei K96243Tn::wcbJ and challenged i.n. with 1 ⫻ 103 CFU of B. pseudomallei K96243. Symbols: unvaccinated mice, ; B. pseudomallei K96243Tn::wcbNimmunized mice, ; B. pseudomallei K96243⌬gmhA-immunized mice, x; B. pseudomallei K96243Tn::wcbJ-immunized mice, ●. Mice were challenged via the i.n. route as described in Materials and Methods. For B. pseudomallei K96243Tn::wcbN, P ⬎ 0.05. For B. pseudomallei K96243⌬gmhA and B. pseudomallei K96243Tn::wcbJ, P ⬍ 0.01.

iai.asm.org 1217

Cuccui et al.

DISCUSSION

B. pseudomallei requires CPS I for full virulence, as the production of this polysaccharide contributes to the survival of B. pseudomallei in vivo by preventing opsonization and phagocytosis in the infected host (29). At a size of 34.5 kb, the CPS I cluster represents the largest known bacterial locus involved in the biosynthesis and transport of a monomeric repeating sugar unit. Bioinformatic comparisons of the amino acid sequences of proteins within the B. pseudomallei K96243 CPS coding region with the NCBI database identified orthologues of six proteins involved in 6-deoxy-Dmanno-heptose biosynthesis in Y. pseudotuberculosis H892/87 O antigen, with gene order maintained in both organisms (Fig. 2) (24). In addition to this glycan, loci found in various Y. pseudotuberculosis strains include ORFs responsible for the synthesis of the rare sugars abequose, paratose, and tyvelose (present in IIA, IA, and IVB, respectively), and despite the need to generate a more complex polysaccharide, none of the clusters are more than 20 kb in size or carry more than 18 ORFs (24). Even when the search is expanded to include Gram positives and compared to all of the known CPS coding loci from Streptococcus pneumoniae, none of the sequenced serotypes (over 90) match the size of the B. pseudomallei CPS I coding region (3). Although previous studies have identified the coding cluster responsible for the biosynthesis and transport of the -3)-2-Oacetyl-6-deoxy-␤-D-manno-heptopyranose-(1- repeating sugar unit (13, 28), there is currently no proposed model for the biosynthesis and transport of CPS I by B. pseudomallei. Our study represents the most comprehensive investigation of the role of individual ORFs within the CPS I coding region of B. pseudomallei to date. The data presented here suggest that the majority of these genes are essential for CPS I production/expression. Our investigations have allowed conclusions to be drawn regarding CPS I sugar biosynthesis. Despite evidence for the alteration of sedoheptulose-7-phosphate in order to generate GDP-6deoxy-D-manno-heptose, we have also identified another ORF that seems to be important for CPS biosynthesis. Deletion of manC resulted in a reduction of CPS I production by B. pseudomallei K96243. This result is in agreement with a previously published report by DeShazer et al. (13). This ORF is thought to represent one end of the CPS I coding region (13) and is predicted by BLASTp searches to be involved in the generation of GDPmannose. The encoded enzyme is predicted to be a GDP-mannose pyrophosphorylase with 50% amino acid identity and 69% similarity to ManC in E. coli. It is feasible that mannose serves as the “priming” sugar onto which the polysaccharide is built and that ManC functions to generate a supply of the GDP-mannose precursor in situ, which then acts as the first sugar onto which the nascent capsule polysaccharide chain is subsequently assembled. The mannose sugar has not been reported to be part of the B. pseudomallei K96243 CPS. This may be due to the harsh acid treatment used to purify CPS I from the organism (27). It is likely that only the distal regions of the capsule are purified in any great amount; therefore, the first mannose residue is missing. A lack of chain length determinant or polymerase found within the B. pseudomallei CPS I coding region is consistent with other known bacterial group 3 CPS coding clusters. Polymerization instead is likely to proceed via the action of glycosyltransferases. The findings that transposon insertions into wcbB, wcbE, or wcbH caused a complete loss of CPS I is in agreement with observations

1218

iai.asm.org

from independent laboratories that genetic modification of B. pseudomallei strains containing a deletion in one of the CPS I glycosyltransferases results in an acapsular phenotype (1, 5, 28). We propose that one of the glycosyltransferases attaches an initial mannose sugar to a lipid precursor, possibly undecaprenyl phosphate. Following a second transferase attaching the -3)-2-Oacetyl-6-deoxy-␤-D-manno-heptopyranose sugar found in B. pseudomallei CPS I to the mannose, the final transferase polymerizes the nascent CPS I chain by adding further heptopyranose residues. A final modification of the sugar unit during CPS I biosynthesis is the acetylation of the -OH group at the C2 position to form 2-O-acetyl-6-deoxy-␤-D-manno-heptopyranose. BLAST analysis of WcbT indicates that it is a 2-amino-3-ketobutyrate coenzyme A (CoA) ligase (KBL)-like protein (E value, 2e-101). KBL catalyzes the second reaction step of the metabolic degradation pathway for threonine, converting 2-amino-3-ketobutyrate to glycine and acetyl-CoA. WcbT therefore may be responsible for generating the acetyl-CoA substrate required to acetylate 6-deoxy-␤-D-mannoheptopyranose or act as a substrate for fatty acid synthesis by WcbR. Our analysis failed to identify a gene coding for the acetyltransferase required to carry out acetylation of the -OH group, at the C2 position of the polysaccharide, in any of the 25 ORFs present within the CPS coding region. A BLASTp search using the amino acid sequence of BPSL2786 (the coding ORF located immediately downstream of wcbT) indicates a possible candidate for this acetyltransferase. Using the bioinformatic and immunological investigation described in this study, we propose a model describing B. pseudomallei CPS I sugar biosynthesis and transport through the inner and outer bacterial cell membranes (Fig. 6). Although it is likely that the CPS biosynthesis and transportation machinery works in a coordinated manner, it is possible that, initially, the repeating polysaccharide is attached to a phospholipid moiety via the action of WcbA/WcbO. Subsequent translocation of CPS I across the inner membrane is suggested to occur via the action of Wzt2/ Wzm2, followed by transportation across the periplasm and presentation on the outer surface of the bacterium. Although we did not generate a B. pseudomallei K96243 mutant lacking a functional wcbO gene, we propose that WcbO may work in concert with WcbA in a way similar to that in which KpsC and KpsS function in E. coli. Wzt2 and Wzm2 are likely to form an ABC transporter system involved in the movement of CPS I across the inner membrane and into the periplasm. Insertion of a transposon into wzm2 resulted in the detection of capsule by Western blot analysis but not by immunofluorescence. As samples tested by Western blotting were prepared by boiling bacteria, this process may have lysed the organism and exposed CPS potentially accumulating in the cytoplasm, thus explaining the observed discrepancy between the immunofluorescence and Western blot assay data. Although a B. pseudomallei K96243 mutant lacking a functional Wzt2 mutant was not generated, it is likely that Wzt2 is the ATP binding component of the ABC transporter. The final stage of the transport of B. pseudomallei CPS I is likely to require transportation across the periplasm and presentation on the outer surface of the bacterium. In E. coli, mutations in kspD or kpsE result in the localization of the polysaccharide to the periplasm, suggesting a role for transport across the periplasmic space (30). A similar role for the B. pseudomallei orthologues wcbC and wcbD is therefore postulated. In this study, we observed that B. pseudomallei K96243⌬gmhA,



FIG 6 Proposed model of B. pseudomallei CPS I biosynthesis. WcbA and WcbO coordinate the glycosyltransferases to initiate and synthesize the CPS chain on a lipid anchor. GmhA, WcbL, WcbN, WcbM, WcbK, and WcbJ synthesize the glycan that comprises CPS I, which is acetylated by an undetermined enzyme (possibly BPSL2786, WcbI, WcbF, or WcbG), assembled on a priming sugar encoded by ManC, and polymerized by the action of the glycosyltransferases WcbB, WcbE, and WcbH. The assembled polysaccharide is moved from the cytoplasm into the periplasm via a complex composed of Wzm2 and Wzt2. The assembled polysaccharide is presented on the bacterial cell surface via the action of WcbD and WcbC. C, cytoplasm; IM, inner membrane; P, periplasm; OM, outer membrane.

B. pseudomallei K96243Tn::wcbJ, and B. pseudomallei K96243Tn:: wcbN mutants produced a polysaccharide that reacted with MAb 4VIH12. We postulated that these mutants might have a modified form of capsule or a reduced amount of glycan coating the organism, perhaps as a result of partial complementation by other homologous genes within the chromosome. As disruption of functional CPS I is known to be attenuating, an immunization study was performed to investigate whether full virulence was retained by these mutants. However, the B. pseudomallei K96243⌬gmhA, B. pseudomallei K96243Tn::wcbJ, and B. pseudomallei K96243Tn:: wcbN mutants showed levels of attenuation comparable to those of acapsular B. pseudomallei (1). When investigating the B. pseudomallei K96243Tn::wcbJ and B. pseudomallei K96243Tn::wcbN mutants by Western blot analysis, we found what appeared to be smaller amounts of antibody-reactive glycan. This finding could be the explanation for the attenuation in the BALB/c mouse model. In contrast, the level of CPS I produced by the B. pseudomallei K96243⌬gmhA mutant is comparable to that of wildtype B. pseudomallei K96243 in Western blot assays. A possible explanation for the apparent discrepancy among Western blot analysis, immunofluorescence, and animal infection studies could be that the functions of gmhA and wcbJ/wcbN are complemented by the nutrient-rich environment provided by LB broth. However, this state may not be achieved in a nutrient-restricted environment, such as the mouse lung, thus explaining the in vivo attenuation of these mutants. To test this hypothesis, we grew the B. pseudomallei K96243⌬gmhA, B. pseudomallei K96243Tn::wcbJ, and B. pseudomallei K96243Tn::wcbN mutants in minimal M9 medium using glucose as the carbon source but observed that CPS I polysaccharide was still generated in these mutants (data not shown). The lack of B. pseudomallei phenotype alteration in broth when gmhA is knocked out may be due to partial complementation by other ORFs within the genome. A search for GmhA, a


sedoheptulose 7-phosphate isomerase, using the John Craig Venter Institute (JCVI, Rockville, MD) database and the Pathema Burkholderia gene search tool, identified 5 ORFs annotated as isomerases involved in cell envelope biosynthesis. One of these, BPSL3273, is even annotated as gmhA in the JCVI annotation. ORFs with predicted roles homologous to those of wcbJ and wcbN can also be identified. However, it is unlikely that transcriptional control of gmhA, wcbJ, and wcbN is the same as that within the CPS I coding region when bacteria are placed in a nutrient-rich environment, providing sufficient complementation to restore the CPS I phenotype, but altered in a nutrient-restricted environment (e.g., mouse lung) to prevent complementation and confer in vivo attenuation. Immunization of mice with either B. pseudomallei K96243⌬ gmhA or B. pseudomallei K96243Tn::wcbJ afforded significant levels of protection against a subsequent i.n. challenge with wild-type B. pseudomallei K96243. Previously, it has been reported that attenuated B. pseudomallei strains are capable of inducing protective immune responses in mice (2, 35). However, capsule-negative mutants have not previously been reported to induce protective immunity (1, 36). To our knowledge, this is the first report of B. pseudomallei CPS I mutants that are attenuated yet capable of stimulating significant levels of protection in an animal model of disease. Given the ability of purified CPS I to offer an increased time to death in mice (23), this polysaccharide is considered a suitable component of a subunit vaccine against melioidosis. If the B. pseudomallei K96243⌬gmhA and B. pseudomallei K96243Tn:: wcbJ strains do indeed retain the capsular epitope required to stimulate protective immunity, these findings again support the suitability of CPS I as an important Burkholderia immunogen. Although likely functions can be assigned to the majority of the ORFs within the CPS I coding cluster, there are still four genes (wcbF, wcbG, wcbI, and wcbP) for which a role in CPS I biosynthe-

iai.asm.org 1219

Cuccui et al.

sis cannot be assigned. A BLASTp search of the NCBI database using the amino acid sequence of WcbF indicated that the protein functions as a putative sulfotransferase. A sulfate modification is not apparent on the glycan that makes up CPS I, so we cannot yet assign a definitive role in CPS I biosynthesis to this protein. No prokaryotic sulfotransferase family protein has ever been experimentally characterized; therefore, assigning a function to wcbF is difficult. A BLASTp comparison of the amino acid sequence of WcbG and WcbI did not identify conserved domains with any protein in the NCBI database. However, functional copies of wcbG and wcbI are necessary for CPS I production, as in-frame deletion of each individual ORF resulted in loss of detectable CPS I. Findings from this study indicate that deleting wcbP, which encodes a putative dehydrogenase/reductase, severely reduces the amount of CPS I within and on the surface of B. pseudomallei. We propose that WcbP may also be involved in lipid biosynthesis. In most group 3 capsules, the genes for WcbA- and WcbO-like proteins act as the first and last genes in the cluster. Therefore, wcbP to wcbT may be a separate cluster of genes required for lipid formation. Within this subcluster, we hypothesize that WcbQ forms a complex with WcbR to aid lipid biosynthesis. We noticed that a transposon insertion in wcbQ abolished polysaccharide presentation on the surface of the organism and that the amount of CPS I detectable on Western blot assays seemed comparable to the amount observed in the B. pseudomallei K96243Tn::wcbR mutant. Insertion of a transposon into wcbR was found to abolish CPS I expression by immunofluorescence, although a weakly reactive product was observed by Western analysis. This large ORF codes for a putative polyketide synthase, an enzyme that can play a role in lipid metabolism. If WcbR is involved in the synthesis of the final lipid anchor onto which CPS I is attached, the weakly reactive material detected by Western blot assay may be CPS I attached to a lipid precursor, which would occur prior to involvement with the external lipid anchor. Further studies are therefore necessary to understand the roles of WcbF/WcbG/WcbI/WcbP. It is tempting to speculate that some, if not all, of the enzymes that fall into the categories of unknowns may also be involved in the biosynthesis of the final lipid anchor, the nature of which remains undefined. It should be noted that lipid biosynthesis by group 2 and group 3 CPS clusters is not well studied (39), but the colocalization of putative lipid biosynthesis genes with the rest of the cluster in B. pseudomallei provides an opportunity to investigate the nature of the endogenous acceptor. Twelve of the mutants used in this study were generated via the insertion of a miniTn5Km2 transposon lacking transcriptional terminators. Despite this, and our reverse transcription-PCR testing of every mutant, one cannot completely rule out the possibility that the insertion caused polar effects and this possibility must be considered when interpreting data from individual transposon mutants. During the preparation of this report, a study by Sim et al. identified a CPS I cluster in B. thailandensis strains E555 and CDC3015869 (34). Sequencing and Western blot assay data indicated that the cluster was responsible for generating a polysaccharide that reacted with an antibody raised against B. pseudomallei CPS I (34). Increased resistance to C3b deposition on the surface of B. thailandensis, as well as enhanced survival in RAW macrophage cells, was seen compared to sequenced acapsular strain E264 (34). However, presence of CPS I in B. thailandensis did not enhance virulence in the BALB/c mouse model (34). One explanation for this apparent discrepancy is that other factors are nec-

1220

iai.asm.org

essary for the virulence of B. thailandensis in a mammalian model. However, two CPS I-specific explanations must also be considered. First, our study suggests that any alteration to CPS I production in B. pseudomallei attenuates the organism. Despite the presence of what appeared to be an intact epitope for antibody 4VIH12 in a number of our mutants, we were unable to find any that maintained their virulence capability in BALB/c mice. It may be that B. thailandensis is coated with a reduced amount of polysaccharide compared to B. pseudomallei. Second, it is also possible that the structure of the glycan in B. thailandensis E555 and CDC3015869 differs from that of B. pseudomallei. Further studies are necessary to understand the implications of these novel findings on B. thailandensis. Conclusions. This investigation into B. pseudomallei has attempted to further clarify CPS I biosynthesis. Although the CPS I coding region is possibly the largest known CPS locus found in bacteria, by combining bioinformatic approaches, Western blotting, and immunofluorescence imaging, we have been able to generate models for sugar biosynthesis and transport. In addition, we have found that knockout mutants of three separate ORFs involved in sugar biosynthesis (gmhA, wcbJ, and wcbN), generated B. pseudomallei strains that were attenuated in the BALB/c mouse model of infection despite retaining the ability to express CPS I. In the case of gmhA and wcbJ, the generation of phenotypes showing characteristics of a live attenuated vaccine has been demonstrated. ACKNOWLEDGMENTS This work was supported by funding from the United Kingdom Ministry of Defense. We are grateful to T. Laws for his statistical analysis of the animal data and operation of the fluorescence microscope.

REFERENCES 1. Atkins T, et al. 2002. Characterisation of an acapsular mutant of Burkholderia pseudomallei identified by signature tagged mutagenesis. J. Med. Microbiol. 51:539 –547. 2. Atkins T, et al. 2002. A mutant of Burkholderia pseudomallei, auxotrophic in the branched chain amino acid biosynthetic pathway, is attenuated and protective in a murine model of melioidosis. Infect. Immun. 70:5290 – 5294. 3. Bentley SD, et al. 2006. Genetic analysis of the capsular biosynthetic locus from all 90 pneumococcal serotypes. PLoS Genet. 2:e31. 4. Brook MD, Currie B, Desmarchelier PM. 1997. Isolation and identification of Burkholderia pseudomallei from soil using selective culture techniques and the polymerase chain reaction. J. Appl. Microbiol. 82:589 –596. 5. Cuccui J, et al. 2007. Development of signature-tagged mutagenesis in Burkholderia pseudomallei to identify genes important in survival and pathogenesis. Infect. Immun. 75:1186 –1195. 6. Currie BJ, Fisher DA, Anstey NM, Jacups SP. 2000. Melioidosis: acute and chronic disease, relapse and re-activation. Trans. R. Soc. Trop. Med. Hyg. 94:301–304. 7. Currie BJ, et al. 2000. Endemic melioidosis in tropical northern Australia: a 10-year prospective study and review of the literature. Clin. Infect. Dis. 31:981–986. 8. Currie BJ, et al. 2004. Melioidosis epidemiology and risk factors from a prospective whole-population study in northern Australia. Trop. Med. Int. Health 9:1167–1174. 9. Dance DA. 1991. Melioidosis: the tip of the iceberg? Clin. Microbiol. Rev. 4:52– 60. 10. Dance DA. 2000. Ecology of Burkholderia pseudomallei and the interactions between environmental Burkholderia spp. and human-animal hosts. Acta Trop. 74:159 –168. 11. Dance DA. 2000. Melioidosis as an emerging global problem. Acta Trop. 74:115–119. 12. de Lorenzo V, Herrero M, Jakubzik U, Timmis KN. 1990. Mini-Tn5 transposon derivatives for insertion mutagenesis, promoter probing, and



13.

14. 15. 16. 17. 18. 19.

20. 21. 22.

23. 24. 25.

26. 27.

chromosomal insertion of cloned DNA in gram-negative eubacteria. J. Bacteriol. 172:6568 – 6572. DeShazer D, Waag DM, Fritz DL, Woods DE. 2001. Identification of a Burkholderia mallei polysaccharide gene cluster by subtractive hybridization and demonstration that the encoded capsule is an essential virulence determinant. Microb. Pathog. 30:253–269. Ellis J. 2000. Antibody detection of Burkholderia pseudomallei and Burkholderia mallei. Ph.D. dissertation. Aston University, Birmingham, United Kingdom. Essex-Lopresti AE, et al. 2005. A type IV pilin, PilA, contributes to adherence of Burkholderia pseudomallei and virulence in vivo. Infect. Immun. 73:1260 –1264. Harmer NJ. 2010. The structure of sedoheptulose-7-phosphate isomerase from Burkholderia pseudomallei reveals a zinc binding site at the heart of the active site. J. Mol. Biol. 400:379 –392. Holden MTG, et al. 2004. Genomic plasticity of the causative agent of melioidosis, Burkholderia pseudomallei. Proc. Natl. Acad. Sci. U. S. A. 101: 14240 –14245. Howe C, Sampath A, Spotnitz M. 1971. The pseudomallei group: a review. J. Infect. Dis. 124:598 – 606. Jones SM, Ellis JF, Russell P, Griffin KF, Oyston PC. 2002. Passive protection against Burkholderia pseudomallei infection in mice by monoclonal antibodies against capsular polysaccharide, lipopolysaccharide or proteins. J. Med. Microbiol. 51:1055–1062. Knirel YA, et al. 1992. Structure of the polysaccharide chains of Pseudomonas pseudomallei lipopolysaccharides. Carbohydr. Res. 233:185–193. Laemmli UK. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680 – 685. Logue CA, Peak IR, Beacham IR. 2009. Facile construction of unmarked deletion mutants in Burkholderia pseudomallei using sacB counterselection in sucrose-resistant and sucrose-sensitive isolates. J. Microbiol. Methods 76:320 –323. Nelson M, et al. 2004. Evaluation of lipopolysaccharide and capsular polysaccharide as subunit vaccines against experimental melioidosis. J. Med. Microbiol. 53:1177–1182. Pacinelli E, Wang L, Reeves PR. 2002. Relationship of Yersinia pseudotuberculosis O antigens IA, IIA, and IVB: the IIA gene cluster was derived from that of IVB. Infect. Immun. 70:3271–3276. Parthasarathy N, DeShazer D, England M, Waag DM. 2006. Polysaccharide microarray technology for the detection of Burkholderia pseudomallei and Burkholderia mallei antibodies. Diagn. Microbiol. Infect. Dis. 56:329 –332. Pavelka MS, Jr, Wright LF, Silver RP. 1991. Identification of two genes, kpsM and kpsT, in region 3 of the polysialic acid gene cluster of Escherichia coli K1. J. Bacteriol. 173:4603– 4610. Perry MB, MacLean LL, Schollaardt T, Bryan LE, Ho M. 1995. Structural characterization of the lipopolysaccharide O antigens of Burkholderia pseudomallei. Infect. Immun. 63:3348 –3352.


28. Reckseidler SL, DeShazer D, Sokol PA, Woods DE. 2001. Detection of bacterial virulence genes by subtractive hybridization: identification of capsular polysaccharide of Burkholderia pseudomallei as a major virulence factor. Infect. Immun. 69:34 – 44. 29. Reckseidler-Zenteno SL, DeVinney R, Woods DE. 2005. The capsular polysaccharide of Burkholderia pseudomallei contributes to survival in serum by reducing complement factor C3b deposition. Infect. Immun. 73: 1106 –1115. 30. Roberts IS. 1996. The biochemistry and genetics of capsular polysaccharide production in bacteria. Annu. Rev. Microbiol. 50:285–315. 31. Rotz LD, Khan AS, Lillibridge SR, Ostroff SM, Hughes JM. 2002. Public health assessment of potential biological terrorism agents. Emerg. Infect. Dis. 8:225–230. 32. Sambrook J, Fritsch EF, Maniatis T. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 33. Sarkar-Tyson M, et al. 2007. Polysaccharides and virulence of Burkholderia pseudomallei. J. Med. Microbiol. 56:1005–1010. 34. Sim BM, et al. 2010. Genomic acquisition of a capsular polysaccharide virulence cluster by non-pathogenic Burkholderia isolates. Genome Biol. 11:R89. 35. Stevens MP, et al. 2004. Attenuated virulence and protective efficacy of a Burkholderia pseudomallei bsa type III secretion mutant in murine models of melioidosis. Microbiology 150:2669 –2676. 36. Ulrich RL, Amemiya K, Waag DM, Roy CJ, DeShazer D. 2005. Aerogenic vaccination with a Burkholderia mallei auxotroph protects against aerosol-initiated glanders in mice. Vaccine 23:1986 –1992. 37. Warawa JM, Long D, Rosenke R, Gardner D, Gherardini FC. 2009. Role for the Burkholderia pseudomallei capsular polysaccharide encoded by the wcb operon in acute disseminated melioidosis. Infect. Immun. 77:5252– 5261. 38. West TE, et al. 2010. Pathogenicity of high-dose enteral inoculation of Burkholderia pseudomallei to mice. Am. J. Trop. Med. Hyg. 83:1066 –1069. 39. Whitfield C. 2006. Biosynthesis and assembly of capsular polysaccharides in Escherichia coli. Annu. Rev. Biochem. 75:39 – 68. 40. Whitfield C, Keenleyside WJ. 1995. Regulation of expression of group IA capsular polysaccharides in Escherichia coli and related extracellular polysaccharides in other bacteria. J. Ind. Microbiol. 15:361–371. 41. Wikraiphat C, et al. 2009. Comparative in vivo and in vitro analyses of putative virulence factors of Burkholderia pseudomallei using lipopolysaccharide, capsule and flagellin mutants. FEMS Immunol. Med. Microbiol. 56:253–259. 42. Wuthiekanun V, Peacock SJ. 2006. Management of melioidosis. Expert Rev. Anti Infect. Ther. 4:445– 455. 43. Yee KC, Lee MK, Chua CT, Puthucheary SD. 1988. Melioidosis, the great mimicker: a report of 10 cases from Malaysia. J. Trop. Med. Hyg. 91:249 –254.

iai.asm.org 1221