Immunogenic Burkholderia pseudomallei Outer Membrane Proteins as Potential Candidate Vaccine Targets

Background Burkholderia pseudomallei is the causative agent of melioidosis, a disease of significant morbidity and mortality in both human and animals in endemic areas. There is no vaccine towards the bacterium available in the market, and the efficacy of many of the bacterium's surface and secreted proteins are currently being evaluated as vaccine candidates. Methodology/Principal Findings With the availability of the B. pseudomallei whole genome sequence, we undertook to identify genes encoding the known immunogenic outer membrane protein A (OmpA). Twelve OmpA domains were identified and ORFs containing these domains were fully annotated. Of the 12 ORFs, two of these OmpAs, Omp3 and Omp7, were successfully cloned, expressed as soluble protein and purified. Both proteins were recognised by antibodies in melioidosis patients' sera by Western blot analysis. Purified soluble fractions of Omp3 and Omp7 were assessed for their ability to protect BALB/c mice against B. pseudomallei infection. Mice were immunised with either Omp3 or Omp7, subsequently challenged with 1×106 colony forming units (cfu) of B. pseudomallei via the intraperitoneal route, and examined daily for 21 days post-challenge. This pilot study has demonstrated that whilst all control unimmunised mice died by day 9 post-challenge, two mice (out of 4) from both immunised groups survived beyond 21 days post-infection. Conclusions/Significance We have demonstrated that B. pseudomallei OmpA proteins are immunogenic in mice as well as melioidosis patients and should be further assessed as potential vaccine candidates against B. pseudomallei infection.


Introduction
Burkholderia pseudomallei is a gram-negative facultative anaerobic motile bacillus that is the causative agent of melioidosis, an infectious disease resulting in significant morbidity and mortality for both humans and animals in endemic regions such as Southeast Asia and northern Australia [1,2]. Melioidosis can present in many clinical forms, from acute pneumonia or septicaemia to chronic and subclinical forms and this poses a great challenge in rapid and accurate diagnosis of the disease [3]. Furthermore, therapy is complicated by antibiotic resistance in many clinical isolates, resulting in frequent relapse of patients and the mortality rate of patients with septic shock is approximately 80-95% despite treatment with ceftazideme, imipenem or meropenem [2]. Therefore, prevention, rather than cure of melioidosis, is critical.
Various vaccination strategies have been extensively explored. Recent work demonstrated significant protection in animal models following vaccination with attenuated strains of B. pseudomallei [4]. Attenuated strains for vaccination would require proof of avirulence as even ara + B. thailandensis, which is generally considered avirulent, can also cause clinical infection in humans [5], an important implication for vaccine research. Other approaches investigated include the use of dendritic cells as a delivery vector to generate cell-mediated immunity [6], conjugate vaccines of capsular polysaccharides and O-polysaccharide [7,8] and the DNA vaccine against fliC flagellin structural gene [9].
With the advances in whole-genome sequencing and bioinformatics, one can use the genomic information to discover novel antigens which may have been missed by conventional methods. In this study, we adopted a bioinformatics-based approach to identify potential protective antigens in B. pseudomallei by using the genome information of B. pseudomallei K96243 made available by The Wellcome Trust Sanger Institute. We selected putative outer membrane protein A (OmpA) as OmpAs are often involved in bacterial virulence and immunity, have good immunogenic properties and are therefore, important vaccine candidates [10,11]. Indeed, OmpAs from Porphyromonas gingivalis, a periodontal pathogen implicated in chronic periodontitis has been demonstrated to confer a high level of protection in the P. gingivalis murine lesion model [12]. In this study, we sought to determine the immunogenic properties of B. pseudomallei recombinant OmpAs and their ability to protect mice from B. pseudomallei infection. Our data demonstrate that two recombinant OmpAs evaluated were immunogenic in mice and show potential as candidate vaccine targets.

Identification of putative B. pseudomallei OmpA genes
We performed a BLASTP search using the conserved OmpA domain sequence (PF00691) as the query sequence and identified 14 open reading frames (ORFs) coding for proteins containing the OmpA conserved domain in the B. pseudomallei K96243 reference genome sequence. Of these 14 proteins, 13 hits had an E-value of 1e 24 or better, however, ORF 13 was not included for further studies as it was annotated as a putative cytochrome C oxidase (Table 1). Subsequently, 12 ORFs coding for OmpA-like proteins were selected for further analysis. Conservation of the OmpA domain sequence at the C-terminal of all 12 B. pseudomallei OmpA sequences as well as in 5 experimentally verified OmpAs [Klebsiella pneumoniae OmpA (ABR76422), Escherichia coli OmpA (NP_415477.1), P. gingivalis OmpA PG33 (AF175715), P. gingivalis OmpA PG32 (AF175714) and Neisseria meningitidis OmpA RmpM (YP_001599860)] is shown in Fig. 1. Residues previously proposed to be involved in N. meningitidis RmpM direct (D45, Y53, R67, and R140) and indirect (F2, G42, G49, N54, L63, G102) interactions with peptidoglycans are present in B. pseudomallei OmpA sequences and are denoted as asterisks and dots in Fig. 1. In addition, the multiple sequence alignment demonstrated that Omp3 clustered together with E. coli OmpA and K. pneumoniae OmpA while Omp7 clustered closely with P. gingivalis OmpAs. N-terminal sequences, on the other hand, were highly heterogeneous. Analysis of the global sequence similarity and calculated percentage identity demonstrated that both Omp3 and Omp7 revealed considerable similarity to reference proteins (Fig. 2). Omp3 was 40% identical to both E. coli and K. pneumoniae OmpA proteins while Omp7 was 21-23% identical to P. gingivalis OmpA proteins.

Cloning, protein expression and purification of OmpA
Primers were designed to amplify the 12 OmpA-coding genes from the local clinical isolate, B. pseudomallei D286. The size of the predicted ORFs ranged from 513-1677 bp. The amplicons were successfully cloned into the pET101/D-TOPO expression vector for recombinant protein production. The full-length sequences of the inserts were verified by DNA sequencing (data not shown).
Expression of 12 B. pseudomallei recombinant outer membrane proteins was carried out and various parameters such as induction time, isopropyl b-D-1-thiogalactopyranoside (IPTG) concentration and growth temperature, were optimized. Eight recombinant proteins were successfully expressed, of which 6 were soluble and 2 were insoluble proteins. We attempted to re-solubilise both insoluble recombinant proteins using high concentrations of the denaturants urea or guanidinium hydrochloride followed by dialysis. However, neither strategy resulted in sufficient amounts of soluble protein for subsequent analysis; thus, further studies on the OmpA proteins were limited to 6 soluble recombinant proteins. All 6 soluble fusion proteins were successfully detected with anti-His tag antibody by Western blot analysis (Fig. 3A). The proteins range in size from 22.5 kDa to 65 kDa corresponding to the predicted ORF. The recombinant proteins were reacted with rabbit antiserum raised against B. pseudomallei whole cells. All 6 recombinant proteins reacted with rabbit antiserum, but the degree of reactivity varied with the strongest reactivity observed with Omp3 and Omp7 (Fig. 3B). Subsequently, all 6 recombinant proteins were also reacted with melioidosis patients' sera. We pooled sera from five patients with different disease presentations in order to detect a spectrum of reactive antigens. Pooling takes advantage of the different immune responses presented by the hosts to the pathogen throughout the course of infection. Among the 6 recombinant proteins tested, Omp10 did not show any reactivity while Omp6 reacted relatively weakly. Both Omp3 and Omp7 demonstrated strong reactivity towards pooled patients' sera ( Fig. 3C). None of the 6 recombinant proteins were recognised by antibodies from Legionella patients' sera (B. pseudomallei-negative) (data not shown). A distinct band with slightly lower molecular weight was also detected for recombinant Omp3 and Omp6. It is presently unknown if the multiple bands correspond to different isoforms of Omp3 and Omp6 or was a result of proteolytic breakdown. Antibody response and potential protective ability of recombinant OmpA Based on their strong reactivity towards anti-sera from B. pseudomallei infected humans and animals in addition to higher sequence similarity to known immunogenic OmpAs from other gram negative bacteria, Omp3 (Genbank accession number FJ746559) and Omp7 (Genbank accession number FJ746560) were selected to induce an antibody response and establish the ability of these two recombinant proteins to protect against B. pseudomallei infection in mice. To assess the immunogenicity of the recombinant Omp3 and Omp7, sera were collected and pooled from immunised mice prior to B. pseudomallei challenge, and tested for reactivity with purified OmpA by ELISA. The specific antibodies against recombinant Omp3 and Omp7 were observed in immunised mice with several dilutions screened, but, not in the pre-immunised sera (P,0.05) (Fig. 4). The absorbance value for anti-Omp7 antibodies was higher at all dilutions tested compared to anti-Omp3 antibodies, especially at higher dilutions.
To identify the type of immune response induced, we determined specific IgG subclasses (IgG 1 , IgG 2a , IgG 2b ) and IgM antibody titres by ELISA. The results indicated significantly higher titres (P,0.05) of IgG 1 , IgG 2a , IgG 2b and IgM antibodies were produced in mice immunised with recombinant Omp3 and Omp7 when compared to antibody titres in the pre-immunised sera from the same mice. Immunisation with Omp3 induced higher antibody titres in all the antibody isotypes when compared to Omp7 immunised mice but the difference was not significant. The predominant Ig subclass elicited by Omp3 and Omp7 immunisation was IgG 2a followed by IgG 2b and to a lesser extent, IgG 1 . The ratio of IgG 2a /IgG 1 was 1.2 in Omp3 immunised mice and 1.3 in Omp7 immunised mice, respectively (Fig. 5).
From our previous work, the LD 50 for the local clinical isolate used in this study, D286, was calculated to be 10 5 cfu following infection of BALB/c mice via an intraperitoneal route. Similar LD 50 values were previously reported for other B. pseudomallei clinical isolates [13] and we settled on a challenge dose of 106LD 50 as recommended by Ulett et al. [14]. The challenge dose of 106LD 50 is equivalent to LD 100 , whereby the dose should cause 100% death of the infected animals by the end of the observation period. Thus, any reduction in the mortality figures as a result of the immunisation would indicate the potential protective efficacy of these recombinant proteins.
For the control group of mice (n = 3) immunised with E. coli cells in adjuvant and subsequently challenged with B. pseudomallei, one mouse succumbed at 1-day post challenge, the second on day 6 and the last mouse died on day 9 with a median survival of 6 days ( Fig. 6). For mice immunised with Omp3, one mouse (out of 4) died on day 2 and the second on day 4 (2/4) with a median survival of 13 days, while for mice immunised with Omp7, one mouse died on day 2 and the second on day 15 (2/4) with a median survival of 18 days. The remaining two mice from each group immunised with either Omp3 or Omp7 survived beyond 21 days post-challenge (Fig. 6). Immunisation with both recombinant proteins resulted in the prolonged median survival (7-12 days) in immunised animals compared to the control group.
In this pilot experiment, both recombinant Omp3 and Omp7 demonstrated protection in mice, but the degree of protection provided by immunisation using a single protein was incomplete (50%). The challenged survivors displayed significant splenomegaly and multiple abscess formation, demonstrating incomplete bacterial clearance from the system (data not shown).

Discussion
B. pseudomallei is an intracellular pathogen and is known to survive and multiply within both phagocytic and non-phagocytic host cells and may be able to spread directly from cell to cell [15]. Thus, the nature of this bacterium highlights the need to generate cell-mediated immunity to combat infection. In an attempt to identify proteins capable of inducing a cell-mediated immune response, we selected putative B. pseudomallei proteins homologous to immunogenic OmpA to demonstrate proof-of-concept. We adopted a bioinformatics-based approach, and by utilising the reference genome sequence of B. pseudomallei K96243, we identified 12 putative OmpA genes in the genome. These genes contain the conserved OmpA domain at the C-termini (Fig. 1) while their Nterminal sequences share little similarity, another characteristic of OmpAs as they are generally heterogeneous at the N-terminal [16]. The C-terminal domain contains conserved residues previously proposed to be important for protein function and interaction with peptidoglycans in N. meningitidis [17]. Some residues may interact directly with peptidoglycans while others may have an indirect interact with peptidoglycans by stabilising residues involved in the direct interaction. All identified residues were well conserved in B. pseudomallei OmpAs, suggesting that they may play similar roles in B. pseudomallei OmpA interaction with peptidoglycans for maintenance of membrane and cell morphology.
OmpAs are known to be involved as virulence factors in bacterial pathogenesis and are capable of inducing both humoral and cytotoxic responses [11,18]. OmpAs were previously proposed in the design of vaccines for bacterial infections caused by Gram negative bacteria such as Vibrio cholerae [10], K. pneumoniae [19] and P. gingivalis [12] and promising data from animal models have been reported. Thus, we expressed 6 soluble recombinant B. pseudomallei OmpAs to assess their immunogenicity and evaluate the potential protective efficacy in an animal model of melioidosis. The observed presence of antibodies towards this class of membrane protein in both the melioidosis animal model as well as human patients implicates a functional role for OmpA during active B. pseudomallei infection. Indeed, evidence indicates that OmpA is a determinant of bacterial virulence and that it is a major target of the mammalian host cell defence [20,21]. The differences in immunoreactivity demonstrated by B. pseudomallei OmpAs may somewhat correlate to the sequence similarity of these OmpAs to experimentally verified immunogenic OmpAs. The strongly immunoreactive Omp3 and Omp7 were clustered together with known immunogenic OmpAs (from E. coli, K. pneumonia and P. gingivalis) whereas weakly reactive proteins (Omp6 and Omp10) were not as closely related. The percentage of sequence identity observed between Omp3/Omp7 and the reference proteins was moderate and predominantly at the C-terminal domain, however, they showed considerable global  sequence similarity, and thus we speculate that they may be similar at the structural level when they are exposed to the host immune system. Furthermore, conservation of the C-terminal domain sequence is important since the C-terminal domain of E. coli OmpA is known to contain the immunodominant epitope and is efficiently recognised by antibodies formed during the course of infection [22]. Similar immunodominant epitope(s) may be better conserved in Omp3 and Omp7 than in the other B. pseudomallei OmpAs and/or exposed more efficiently, leading to better recognition by antibodies from infected individuals.
Vaccine development for melioidosis is made difficult because B. pseudomallei is an intracellular pathogen that can successfully evade the host immune system for an extended period of time. Effective vaccination against this pathogen is therefore required to elicit a cytokine-mediated cellular response so as to combat potential disease recurrence [23]. Various approaches have been undertaken towards the development of effective vaccines against B. pseudomallei infection in animal models. Those investigated include heat-killed or live attenuated mutants of B. pseudomallei, polysaccharides and a DNA-vaccine coding for the fliC flagellin structural gene [4,[7][8][9]. These studies demonstrated a significant protection against B. pseudomallei infection in immunised animal models, however, the level of protection varied depending on the different vaccine strategies. For example, immunisation with attenuated mutants of B. pseudomallei resulted in 60% survival rate at 35 days post challenge in immunised mice [4] while DNAvaccination protected 85% of vaccinated mice up to 7 days post infection [9] but the observation period was limited. Active immunisation with B. pseudomallei lipopolysaccharide resulted in 50% survival at day 35 post challenge in immunised mice [8]. Thus, data from the present pilot study is generally comparable to these previously reported efficacies whereby immunisation with recombinant Omp3 or Omp7 could be seen to be as effective as immunisation with lipopolysaccharide in murine models of melioidosis. Furthermore, our assessment of polyclonal antibody isotypes in the sera from Omp3-and Omp7-immunised mice indicated a predisposition towards a T h 1-driven immune response as demonstrated by an IgG 2a /IgG 1 ratio of .1 [24]. Both recombinant proteins demonstrated similar potential protective efficacy, however, Omp7 appeared to be more efficient in inducing a higher titre of anti-Omp7 antibodies and higher IgG 2a /IgG 1 ratio. This may have contributed to the improvement of the median survival in Omp7-immunised mice when compared to Omp3-immunised mice. However, challenged survivors were unable to clear the infection and a chronic infectious status was evidently present, whereby mice eventually succumbed to the infection. Such findings are consistent with previously reported studies that immunisation strategies alone are not sufficient to protect animals from infection. Furthermore, recombinant antigens as vaccination candidates typically induce strong humoral antibodies but may be hampered in their ability to induce the cellular-mediated immune response [25]. Eradication of B. pseudomallei has been proposed to be dependent on the induction of a T h 1 type immune response at the onset of melioidosis [6]. This is because in the presence of protective humoral antibodies, B. pseudomallei is still able to invade epithelial or phagocytic cells, resulting in a long term intracellular latent infection in the infected individuals before relapse occurs many years later [26,27]. The use of vaccine delivery vectors such as dendritic cells (DCs) has been proposed to be effective in inducing significant levels of protective immune response against heterologous strains of B. pseudomallei by eliciting strong cell-mediated immune responses and lower but adequate levels of antibody responses [28]. Although the use of cultured DCs as a human vaccine candidate is simply not viable, a vaccine strategy that actively targets DCs is feasible by means of a formulation carrying recombinant B. pseudomallei antigens and Toll-like receptor (TLR) ligands. B. pseudomallei binding to its known receptor, TLR2, would activate the DCs, initiating the process of antigen uptake, processing and presentation required for the generation of protective immune responses. Recombinant OmpAs from this study could be good candidate antigens for coupling to DCs.
In summary, this study has demonstrated that by using publicly available bacterial genome sequence information, we successfully identified two immunogenic antigens, Omp3 and Omp7. Although more extensive studies are required, data from this pilot experiment have demonstrated the promising protective efficacy of B. pseudomallei OmpAs against an animal model of melioidosis.

Ethics statement
All animal experiments were performed in accordance with the animal ethics guideline compiled by Universiti Kebangsaan Malaysia Animal Ethics Committee (UKMAEC). Animals were maintained under specific-pathogen-free conditions and on a 12-h  light/dark cycle. New Zealand White rabbits and BALB/c mice were obtained from the Universiti Kebangsaan Malaysia Animal House Facility. The rabbits were housed in individual cages and had free access to food and water during the study. The mice were provided with pelleted protein-enriched diet and water ad libitum.

In silico genome analysis and multiple sequence alignment
The conserved OmpA domain sequence PF00691 (Table 1) was used as a query sequence to perform BLASTP [29] analysis against the annotated genome sequence of B. pseudomallei clinical strain K96243 deposited at http://www.sanger.ac.uk/cgi-bin/ blast/submitblast/b_pseudomallei. The sequences containing the conserved OmpA domain were selected using a cut-off threshold of 1e 24 . Subsequently, full ORFs of these genes were determined using Artemis [30]. Multiple sequence alignment of B. pseudomallei OmpA sequences together with the 5 experimentally verified OmpAs from other gram negative bacteria [K. pneumoniae OmpA (ABR76422), E. coli OmpA (NP_415477.1), P. gingivalis OmpA PG33 (AF175715), P. gingivalis OmpA PG32 (AF175714) and N. meningitidis OmpA RmpM (YP_001599860)] was performed using ClustalW [31]. The percentage identity between the predicted amino acid sequences of B. pseudomallei OmpAs and immunogenic OmpAs from other bacteria was calculated using Jalview [32].

Bacterial strain and preparation of genomic DNA
The B. pseudomallei strain D286 was previously isolated from a melioidosis patient at the Kuala Lumpur Hospital [33]. A stock culture of this strain was obtained from the Pathogen Laboratory, Faculty of Science and Technology, Universiti Kebangsaan Malaysia. Genomic DNA was extracted according to the standard protocol [34] with modification. Briefly, B. pseudomallei strain D286 was grown at 37uC in Brain Heart Infusion broth (Pronadisa, Spain) and then plated on Ashdown selective agar plates. A single colony was inoculated into fresh Brain Heart Infusion broth and grown overnight at 37uC, following which the bacterial cells were harvested and the cell pellet was dissolved in TE buffer (1 mM Tris-Cl, pH 7.5, 1 mM EDTA). The pellet was treated with Proteinase K at a final concentration of 25 mg/ml and 10% SDS (Amresco, USA) and incubated at 37uC for one hour. CTAB-NaCl (10% CTAB (Sigma, USA), 0.7 M NaCl) was added and the incubation was continued at 65uC for 20 min. Purification of genomic DNA was carried out according to the standard phenol chloroform method and precipitated using 0.6 V of isopropanol.

Amplification of B. pseudomallei OmpA genes and cloning
Primer sequences used for amplification of the twelve ORFs from the B. pseudomallei strain D286 genome were designed based on the 12 annotated ORFs and are listed in Table 2. The forward primers contained an additional CACC sequence at the 59-end to allow direct cloning into the expression vector. PCR was conducted in a final volume of 50 ml that contained 5 ml 106 Expand High Fidelity reaction buffer with 15 mM MgCl 2 (Roche, Germany), 1 ml B. pseudomallei genomic template, 1 ml deoxynucleotide mix (10 mM each of dNTP) (Promega, USA), 1 ml each of forward and reverse primers (25 pmoles each), 5 U of Expand High Fidelity enzyme mix (Roche, Germany) and 40.25 ml of MilliQ water. Amplification was performed under the following conditions; 1 cycle of 95uC for 5 min; 10 cycles of 94uC for 15 s, 55uC for 30 s, and 72uC for 1 min; 20 cycles of 94uC for 15 s, 55uC for 30 s and 72uC for 1 min with 5 s cycle elongation for each successive cycle for the final step; and 1 cycle of 72uC for 7 min. To achieve high efficiency and regulation for cloning and expression in an E. coli expression system, PCR products were cloned into pET101/D-TOPOH according to the manufacturer's recommendations (Invitrogen, USA). The gene orientation and sequence fidelity was verified by DNA sequencing.

Recombinant OmpA protein production and purification
Transformation of host BL21 Star (DE3) cells was performed according to the manufacturer's recommendations (Invitrogen, USA). The positive control used for expression was E. coli BL21 Star (DE3) transformed with pET-101-LacZ (Invitrogen, USA). The colonies were cultured overnight and used to inoculate fresh LB medium supplemented with carbenicillin (50 mg/ml) (Phytotechnology Laboratories, USA). At OD 600nm = 0.5-0.6, expression was induced with 1 mM IPTG (Promega, USA) for 4 hr at 30uC. At the end of protein induction, cells were harvested and all protein extraction procedures were performed at 4uC. The cell pellet was resuspended in l ml lysis buffer (50 mM NaH 2 PO 4 , 300 mM NaCl, 10 mM imidazole, pH 8.0) and stored at 220 uC overnight. The cells were then sonicated, centrifuged and the supernatant containing the soluble proteins was collected. His6-tagged recombinant protein purification was performed with the Ni-NTA purification system (QIAgen, USA). The column was washed with wash buffer (50 mM NaH 2 PO 4 , 300 mM NaCl, 20 mM imidazole, pH 8.0) and then protein fractions were eluted with elution buffer (50 mM NaH 2 PO 4 , 300 mM NaCl, 250 mM imidazole, pH 8.0). The eluted protein fractions were concentrated by centrifugation (Vivaspin, MW cutoff 10 kDa) (Vivascience, Germany) and protein concentration was determined using the bicinchoninic acid assay (BCA) (Sigma, USA) [35].

Human sera samples
Culture-confirmed melioidosis sera of five adult patients from Malaysia presenting with acute and chronic disease manifestations were pooled and used. The serum samples from two patients presenting with Legionella infection but confirmed B. pseudomalleinegative were pooled and used as the control. A series of dilutions was performed on the serum samples to determine the optimal titres for Western analysis. All the serum samples were generously provided by the Institute for Medical Research, Malaysia.
Rabbit anti-B. pseudomallei serum Rabbit anti-serum was raised against heat killed (80uC, 1 h) whole B. pseudomallei cells. Heat-treated cells were injected into two male New Zealand White rabbits. The first rabbit received 5610 5 cfu in Freund's incomplete adjuvant (FIA) while the second rabbit received 2.5610 5 cfu in FIA, respectively, for 2-week intervals over 2 doses. Rabbits were bled 2 weeks after the second dose and sera samples were collected. ELISA was performed according to the protocol described elsewhere [38] using heatkilled whole B. pseudomallei cells to test the sera reactivity and calculate the optimal titre for immunoblotting.

Antibody analysis and mouse protection study
BALB/c mice (n = 4, 6 weeks, female) were immunised with three doses of 50 mg purified recombinant Omp3 or Omp7, via the intraperitoneal route (i.p.) over 9 weeks. The first dose was administered in Freund's complete adjuvant (FCA) while the latter two doses were in FIA, three weeks apart. For antibody analysis, the sera samples were collected (n = 4) prior to immunisation one week after the second booster and pooled. The antibody titre of total IgG was analysed by ELISA as previously described [38]. Briefly, the 96-well microplates were coated with 0.5 mg per well of either purified recombinant Omp3 or Omp7, or crude E. coli protein in coating buffer (0.1 M sodium bicarbonate, p.H 8.6) and incubated overnight at 4uC. The plates were blocked, washed, then incubated with serial dilutions of sera samples and incubated at 37uC for 1 h. The diluted Rabbit anti-mouse IgG conjugated with peroxidase (1:4000, Pierce, USA) was added and incubation continued for 1 h at 37uC. At the end of the incubation, the plates were washed and 1-step TM Ultra TMB-ELISA (Pierce, USA) was added as substrate. The reaction solutions were read at 450 nm using a microplate reader.
The LD 50 of the B. pseudomallei local isolate D286 via an i.p. route of infection was previously calculated at approximately 1610 5 cfu in a BALB/c mouse model (data not shown). Thus, 1610 6 cfu (106LD 50 ) were used in this study as the lethal dose for infection. Mice were challenged intraperitoneally 2 weeks after the second booster with approximately 1610 6 cfu B. pseudomallei D286. Mice were examined daily for 21 days post-challenge. Control mice (n = 3) received E. coli BL21 (DE3) cells in FIA followed by challenge.

Statistical analysis
Statistical analysis on anti-OmpA mice sera reactivity was performed using the Mann Whitney test within the SPSS (version 14.0) software package. Data are expressed as the mean6SEM. Survival (%) was calculated using the Kaplan-Meier Cumulative Survival Plot for Time (nonparametric survival analysis) and survival curves were compared using Log Rank tests (GraphPad Prism) where P values of ,0.05 were taken to be significant. All experiments were replicated in a comparable fashion and found to produce findings similar to the data presented. The median survival is defined as the time at which half the subjects have died.