Introduction

Burkholderia pseudomallei is a Gram-negative facultative intracellular bacillus that causes the disease melioidosis. Cases of human melioidosis have predominantly been reported in Southeast Asia and northern Australia, but more recent reports indicate that the disease is largely underestimated and affects both humans and animals worldwide1,2,3. Modeling estimates suggest that B. pseudomallei is likely endemic in more than 79 countries and causes 165,000 human melioidosis cases per year, from which 89,000 people die4. An outbreak across five states in the United States, which was ultimately traced to a contaminated, imported air spray, led to two human fatalities3. More recently, three cases of melioidosis occurred on the U.S. gulf coast, and B. pseudomallei was isolated from the soil, demonstrating for the first time its endemicity in North America2.

Routes of infection with B. pseudomallei include inoculation, inhalation, and ingestion, which can ultimately result in life-threatening bacteremia, sepsis, and/or pneumonia. B. pseudomallei is inherently resistant to many broad-spectrum antibiotics, and antibiotic therapy is often complicated and always prolonged. Furthermore, B. pseudomallei can establish latency and activation from latency has been documented decades after the initial infection5. B. pseudomallei is listed as a biological threat with high potential for misuse and is one of five bacterial species (B. mallei, Yersinia pestis, Bacillus anthracis, and Francisella tularensis) classified and regulated as a Tier 1 select agent under the Centers for Disease Control Federal Select Agent Program. An effective prophylactic vaccine could help protect military personnel and civilians in the event of a biological attack and reduce the global public health burden caused by B. pseudomallei.

Currently, there is no licensed vaccine for B. pseudomallei, although numerous vaccine platforms have been evaluated in pre-clinical studies. Examination of B. pseudomallei vaccines has been largely restricted to rodent models with no efficacy studies reported in non-human primates (NHP) to date. Evaluation in large animals models that mirror human disease is deemed an important step in advancing any candidate vaccine to clinical trials or licensure through the FDA Animal Rule6,7. We previously demonstrated that a vaccine composed of multivalent outer membrane vesicles (OMV) derived from B. pseudomallei are non-toxic and highly immunogenic in multiple models and afford significant protection against pneumonic and septicemic melioidosis in rodent models8,9,10. OMVs are non-infectious nanoparticles that are naturally released from the Gram-negative bacterial cell surface11. OMVs incorporate multiple protective surface antigens, including proteins, lipids, and polysaccharides, which retain their native orientation and structure and are utilized in the licensed vaccine to prevent lethal infection from Neisseria meningitidis serogroup B12. The B. pseudomallei-derived OMV vaccine confers significant cross-protection against B. mallei in rodent and NHP models of glanders, indicating that protective immune responses to conserved OMV antigens recognize both pathogenic species13. In order to maximize cross-protection against highly heterogeneous B. pseudomallei strains, we further enriched the OMV vaccine with proteins that are widely conserved across clinical isolates. This second-generation OMV vaccine, referred to as M9-OMV, provided equivalent protection to a live attenuated vaccine against inhalational melioidosis in mice10.

In this study, we utilized a cohort of twelve rhesus macaques (Macaca mulatta) to evaluate safety, protective efficacy and serological responses to vaccination using the improved second-generation OMV vaccine. To our knowledge, this is the first report of vaccine efficacy against B. pseudomallei in NHPs. We show that OMV immunization prevents lung pathology in macaques and that animals mount systemic and mucosal antibody responses to OMV surface protein and polysaccharide antigens. Using two independent seroreactivity assays, a Burkholderia immunoproteome array (BIPA) and a customized multiplex assay (BurkPx), we identified several highly conserved surface proteins and polysaccharides that were consistently recognized by OMV vaccinated macaques. We show that sera from vaccinated macaques promote opsonphagocytosis of B. pseudomallei by macrophages in vitro. We further show that humoral immune responses to three key OMV antigens are significantly associated with survival in human melioidosis patients. Collectively, these results lay the groundwork for the advancement of the OMV vaccine to human clinical trials.

Results

OMV vaccination prevents lung injury in a macaque model of pneumonic melioidosis

The compelling protective efficacy of the OMV vaccine in rodents warranted evaluation in a more advanced animal model. Rhesus macaques and other non-human primates are naturally susceptible to B. pseudomallei in the wild and develop disease that closely parallels human melioidosis14,15,16,17. Here, we evaluated the protective efficacy of the OMV vaccine in rhesus macaques against aerosol challenge with B. pseudomallei. Macaques assigned to study were pre-screened at baseline for antibodies to capsular polysaccharide (CPS-1), an antigen unique to pathogenic Burkholderia species, using a latex agglutination assay18 (kindly provided by Paul Brett and Mary Burtnick) to confirm seronegativity to Burkholderia pseudomallei. Macaques were immunized subcutaneously (SC) with OMV vaccine (n = 6) or saline vehicle control (n = 6) twice, 4 weeks apart (Fig. 1a). Animals were closely monitored by veterinary staff for changes in behavior, appetite, or injection site reactogenicity with no concerns noted. Four weeks later, macaques were exposed to aerosolized B. pseudomallei strain K96243 using a head-only exposure chamber. Prior to challenge, one animal in the OMV-immunized group (KE90) developed a surgical site infection from the biotelemetry implant that warranted antibiotic treatment and was therefore removed from the protection study, yielding five remaining animals in the OMV vaccine group. The mean inhaled dose was 4.35 × 103 CFU with no significant difference between groups of macaques immunized with sham vs OMV vaccine (p = 0.92, Supplemental Fig. 1).

Fig. 1: OMV-vaccination prevents severe disease in rhesus macaques.
Fig. 1: OMV-vaccination prevents severe disease in rhesus macaques.The alternative text for this image may have been generated using AI.
Full size image

a NHPs were given two doses of OMV vaccine or sham 4 weeks apart, then challenged with B. pseudomallei. b Seventy-five percent (3/4) of NHPs immunized with OMV (blue triangles) vaccine survived to the day 21 study endpoint, whereas only 33% (2/6) of sham-immunized NHPs (black circles) survived, p = 0.24 as determined by the Log Rank Mantel-Cox test. One animal in the OMV group (KM49) was removed from the study and right censored at day 14 due to an unrelated event. c Febrile response of OMV- (animal ID KP07, top) and sham-immunized (KV38, bottom) macaques post-challenge. Fever (red boxes) or hypothermia (blue boxes) were defined as a 1.5-degree in core temperature increase or decrease, respectively, from the normal diurnal pattern of the macaque (green boxes). Each animal served as its own control using data collected prior to challenge (black boxes, error bars represent the range of temperatures collected by the telemetry device during the one minute sampling time interval), with post-exposure biotelemetry data plotted as an overlay, with a single temperature measurement taken at each time point for each animal. One animal is shown per group. d Gross pathology images performed at necropsy demonstrating normal appearing lungs for OMV vaccinated survivors (4 total) versus multifocal pneumonia for sham-immunized survivors (2 total). All animals that succumbed (1 OMV, 4 sham) early to infection displayed necrotizing, hemorrhagic pneumonia. e Representative images of lung histopathology demonstrating absence of alveolar inflammation (KP07), organized areas of alveolar congestion (KV38), granuloma formation (KJ33), and significant alveolar inflammation with granuloma formation (KP86). Scale bar = 1000 μm. f Histopathologic scoring performed on lungs at necropsy. Closed blue triangles -OMV-immunized, survived; open blue triangles – OMV-immunized, died, closed black circles -sham-immunized, survived, open black circles – sham-immunized, died. Four slides were analyzed per lung and scored 0–5 (OMV-immunized n = 5, sham-immunized n = 6). Non-parametric distribution was assumed, and statistical analysis was determined using the one-tailed Mann-Whitney test. g Representative images of histopathology performed on OMV-immunized animal KP07 demonstrating iBALT formation. Left, scale bar = 1000 μm, right, scale bar = 200 μm.

Following the challenge, 4 out of 6 sham-immunized and 1 out of 5 OMV-immunized animals demonstrated a rapid onset of illness and terminal respiratory distress within 3 to 4 days and were humanely euthanized and necropsied. All remaining animals (2 sham; 4 OMV) survived to the day 21 study endpoint with the exception of animal KM49 (OMV), who was humanely euthanized on day 14 due to an untreatable severe ocular infection with S. aureus. (Fig. 1b). Body temperature measurements via biotelemetry in animals that survived demonstrated a febrile response to bacterial challenge that resolved within 72 h in the OMV vaccinated animals, whereas the sham vaccinated animals maintained an elevated body temperature throughout the study observation period (Fig. 1c). Sham vaccinated animals that survived also displayed considerable anorexia and weight loss, while OMV vaccinated survivors (with the exception of KM49) maintained their appetite and regained much of their early weight loss by the study endpoint (Supplementary Table 1).

Necropsy was performed upon euthanasia of unprotected animals and at the study endpoint for survivors. Pulmonary lesions were the most common pathological finding with severe necrohemorrhagic pneumonia evident in unprotected animals (Fig. 1d). In survivors, lungs from OMV vaccinated animals (including KM49) displayed no overt gross lung pathology, while pyogranulomatous-like lesions were evident in sham vaccinated survivors (Fig. 1d). Histopathological analysis and scoring demonstrated OMV-immunized animals had reduced numbers of pyogranulomas and significantly reduced alveolar inflammation and consolidation compared to sham immunized animals (Fig. 1e, f). Moreover, 4 out of 5 OMV-immunized animals demonstrated areas of lymphoid hyperplasia that resembled inducible bronchoalveolar lymphoid (iBALT) tissue (Fig. 1g), which were absent in sham-immunized animals.

OMV vaccination induces mucosal and systemic antigen-specific antibodies in macaques

Prior to challenge, sera, bronchoalveolar (BAL) fluid, and PBMCs were collected in order to confirm that animals had mounted antigen-specific antibodies and T cells in response to OMV vaccination. As expected, sham-immunized macaques displayed very low levels of sera IgG reactivity to B. pseudomallei or OMV at baseline and after immunizations (Fig. 2a, b). After two doses of OMV vaccine, 5 of 6 immunized macaques demonstrated significant increases in sera IgG specific for B. pseudomallei (Fig. 2c) and OMV (Fig. 2d) as compared to baseline values. One animal, KT39, demonstrated high levels of cross-reactive IgG at baseline, which appeared to limit its response to OMV vaccination (Fig. 2d) and could have contributed to its survival. The IgG response in the BAL fluid largely mirrored the systemic response (Supplemental Fig. 2), likely reflecting transudation of sera IgG as the vaccine was given SC. Sham-immunized animals possessed little to no BAL IgG specific to B. pseudomallei or OMV, whereas 5 of 6 OMV-immunized animals displayed significant increases in antigen-specific BAL IgG compared to baseline IgG levels (Supplemental Fig. 2).

Fig. 2: OMV-immunization elicits sera IgG specific for B. pseudomallei and OMV vaccine.
Fig. 2: OMV-immunization elicits sera IgG specific for B. pseudomallei and OMV vaccine.The alternative text for this image may have been generated using AI.
Full size image

a OMV- and (b) B. pseudomallei-specific sera IgG were measured in sham-immunized and OMV-immunized (boxed) rhesus macaques by ELISA. Antibodies were measured prior to immunization (baseline), four weeks after the first dose (prime), and two weeks after the second dose (boost). Convalescent sera obtained from B. pseudomallei experimentally-infected macaques was used as a positive control. Reciprocal titers are presented for each dilution. Pairwise comparison of baseline versus boost sample curves of OD450 value vs dilution number was performed by the two-tailed modified chi-squared test. ns = not significant, **** = p < 0.0001.

To examine antigen-specific cellular immunity, we performed an antigen restimulation assay on PBMCs obtained from the blood of macaques two weeks after the final immunization. There was no detectable difference in peripheral IFN-g producing CD4+ T cells between sham- and OMV-immunized animals (Supplemental Fig. 3). Interestingly, PBMCs from the single OMV-immunized animal (KJ33) that was not protected exhibited little to no response to the mitogen PMA in the restimulation assay. Examination of the animal’s clinical record revealed no sign of underlying illness or co-morbidity that could account for this observation. Despite these results, increases in antigen-specific IgG in immunized NHPs would not have been achieved without CD4 T cell help. Therefore, to demonstrate that OMV vaccination can drive cellular immunity, we turned to a human surrogate system termed the Modular Immune In vitro Construct (MIMIC). The MIMIC System is a highly automated, laboratory-based methodology that utilizes individual human donor cells to replicate the human immune response ex vivo19. As shown in Supplemental Fig. 4a, b, the OMV vaccine, as well as a comparator OMV vaccine (Bexsero), elicited vaccine-specific IFN-γ, TNF-α, and IL-2 secreting CD4 and CD8 T cells.

Identification of OMV antigens associated with vaccine response

Correlates of vaccine protection are a key component of vaccine development and licensure. OMVs contain hundreds to thousands of proteins that could potentially contribute to their protective efficacy. We therefore wished to identify OMV proteins that were recognized by NHP IgG generated in response to vaccination. To accomplish this, we first utilized an unbiased approach by testing NHP sera on a previously described Burkholderia immunoproteome microarray (BIPA) that incorporates approximately 3000 open reading frames (ORFs) from B. pseudomallei20. The top 10 seroreactive antigens were determined by fold change in MFI between pre-and post-immune sera (Table 1 and Supplementary Fig. S5a, b). Some of the most reactive proteins included BPSL2522 (OmpA), BPSL2705 (lipoprotein), and BPSL0280 (FlgK). These three proteins were previously identified in the OMV proteome by LC-MS10 and were incorporated into a customized, multiplex bead assay called BurkPx which contains a subset of purified B. pseudomallei proteins and carbohydrates from various known or predicted sub-cellular locations (Table 2)21. Thus, to independently confirm some of the BIPA array results, NHP sera and BAL IgG and IgA reactivity were measured using the BurkPx assay. As indicated in the vaccination fold change heat map (Fig. 3a), there was a notable clustering of sera IgG and sera and BAL IgA reactive to predicted or known outer membrane (OM) or extracellular (E) proteins and polysaccharides in OMV vaccinated animals, aligning with expectations that the OMV vaccine elicits antibody responses to bacterial surface antigens. In particular, mean IgG and IgA responses to the lipopolysaccharides (LPS) serotype A and the capsular polysaccharide-1 (CPS-1), and proteins OmpA, lipoprotein BPSL2705, and FlgK increased greater than two-fold consistently across vaccinated animals when comparing baseline to post-vaccination sera and BAL samples (Fig. 3b–e). For LPS-A, OmpA, and FlgK, the increase in sera and BAL IgG was statistically different from paired baseline samples in OMV vaccinated animals (Fig. 3b, c). BAL IgA specific for LPS-A and CPS was also significantly increased after OMV vaccination (Fig. 3e).

Fig. 3: Antibodies to key OMV surface proteins and polysaccharides represent potential immune correlates of vaccine protection.
Fig. 3: Antibodies to key OMV surface proteins and polysaccharides represent potential immune correlates of vaccine protection.The alternative text for this image may have been generated using AI.
Full size image

a Antigens recognized following OMV immunization were determined by the BurkPx assay, which measures IgG and IgA responses to 19 different B. pseudomallei antigens (C – cytoplasmic; P – periplasmic; OM – outer membrane; E – extracellular/secreted). Color gradient represents fold-change from baseline for each individual animal. OMV-immunized animals that survived challenge are indicated by +; OMV-immunized animal that succumbed is indicated by -; OMV-immunized animal removed from challenge study is indicated by ND (not determined). IgG antibodies to OmpA (BPSL2522), FlgK (BPSL0280) and lipoprotein (BPSL2705), and the lipo- and capsular polysaccharides, LPS serotype A and CPS-1, are shown for (b) serum and (c) BAL of NHPs, depicted as IgG MFI. IgA antibodies are shown in (d) serum and (e) BAL. f Sera from OMV-vaccinated macaques promote bacterial opsonophagocytosis by macrophages. Phagocytic index (PI) is calculated as the percentage of macrophages that have phagocytosed multiplied by the MFI. For (b–f) OMV-immunized n = 6 and sham-immunized n = 6. Individual OMV-immunized NHPs are provided as triangles (blue/open for succumbed; blue/filled for survived; purple/filled for not determined). Line denotes the median. Significant differences were determined between baseline and post-vaccination antibody MFI or PI using the two-tailed Wilcoxon signed-rank test.

Table 1 Top 10 antigens recognized by NHP immune sera in the Burkholderia Immuno-Proteome Array (BIPA)
Full size table
Table 2 BurkPx expressed and purified antigens used to evaluate NHP antibody responses to OMV vaccine
Full size table

To determine if antigen-specific antibodies in vaccinated macaques possessed anti-bacterial function, we performed an in vitro opsonophagocytosis assay using macaque macrophages. As shown in Fig. 3f, sera from OMV-vaccinated animals increased bacterial opsonophagocytosis by more than 1-log compared to paired pre-immune sera collected at baseline. Collectively, these data indicate that OMV vaccination of NHPs is associated with the production of system and mucosal antibodies to, OmpA, FlgK, LPS-A, and to some extent, lipoprotein BPSL2705 and CPS-1 and that immune sera promotes anti-bacterial effector responses.

Sera IgG responses to key OMV antigens are associated with survival in human melioidosis patients

Having identified several key surface antigens associated with OMV vaccine responses in NHPs, we next assessed whether human survivors of melioidosis mounted IgG responses against these antigens. As shown in Fig. 4, sera IgG to LPS-A, CPS, and OmpA were significantly increased in patients that survived versus those that did not, indicating these are important antibody targets during natural B. pseudomallei infection.

Fig. 4: Human survivors of melioidosis possess higher levels of sera IgG to LPS, CPS, and OmpA.
Fig. 4: Human survivors of melioidosis possess higher levels of sera IgG to LPS, CPS, and OmpA.The alternative text for this image may have been generated using AI.
Full size image

Antigen-specific IgG was determined using sera from culture-confirmed human melioidosis patients. Individuals who survived (n = 84, open circles) or died (n = 8, filled circles) of melioidosis were stratified, and samples were selected based on the succumbed time range after admission (0–134 days). An individual sample was selected from the survivors in the same time range. Reactive IgG to OmpA (BPSL2522), and the capsular – and lipo- polysaccharides, CPS-1 and LPS serotype A, were determined using the BurkPx assay. Significant differences were determined between the groups using a two-tailed Mann-Whitney test.

Discussion

The increasing global public health burden of melioidosis in both humans and animals underscores the urgent need for a vaccine against B. pseudomallei. Vaccine development against this complex facultative intracellular bacterium has been challenging due to its multiple virulence factors and strain heterogeneity. A number of promising vaccine candidates have demonstrated protection in rodent models of melioidosis22,23 with several excellent reviews on the subject24,25,26. However, up until now, none has been evaluated for protection against melioidosis in an NHP model. While mice are excellent models for down-selection of vaccine candidates, they exhibit tissue and structural differences from the human respiratory tract that limit their translational value. Here, we provide compelling evidence of OMV vaccine efficacy against inhalational melioidosis in rhesus macaques using survivability, clinical signs, and lung pathology assessments. Absence of protection in macaques manifested as severe, necrotizing pneumonia within 3-4 days, attesting to the virulent nature of B. psuedomallei infection following inhalation. It is quite remarkable that OMV-vaccinated animals exhibited few clinical signs and highly preserved lung architecture upon necropsy. It is unclear why one OMV-immunized macaque (KJ33) was not protected, however, it is notable that PBMCs from this animal responded poorly to stimulation with PMA mitogen. Moreover, the BAL IgG response for KJ33 was consistently below the mean for all antigens quantitatively examined in the BurkPx assay. This absence of breadth and/or magnitude of mucosal IgG may have hampered protection in the lung. Surprisingly, two of six sham-immunized animals survived the challenge despite clinical signs of persistent infection, including fever and weight loss. In addition, sham animals displayed overt lung pathology at necropsy. Similar to the human population, the outbred macaques used in our study are genetically diverse, and some variability is expected. The macaques are also exposed to soil bacteria in their outdoor habitats. Although animals were screened prior to study enrollment for antibodies to B. pseudomallei, it is possible that other pre-existing, cross-reactive immune responses (e.g., T cells) to highly conserved proteins of closely related Gram-negative species could have contributed to the survival of sham-immunized animals.

The required elements for protection against B. pseudomallei in humans are not fully resolved, yet there is evidence that both humoral and cellular immunity are crucial for protection against melioidosis20,27,28,29. In animal models of acute melioidosis, vaccine-mediated antibody responses are often sufficient for protection, while cellular immunity is likely critical in more chronic stages of disease24. In our study, sera from OMV-vaccinated macaques promoted increased opsonphagocytosis of B. pseudomallei by macrophages, providing evidence of humoral protective responses. The multivalent nature of the OMV vaccine appears advantageous because it directs antibodies to multiple surface antigens that are highly conserved among heterologous clinical isolates, including LPS-A, OmpA, and FlgK. The OMV vaccine is derived from a strain that produces LPS-A that is highly conserved in a wide variety of clinical isolates30. The observed systemic and mucosal antibodies directed against LPS-A indicate this polysaccharidemay contribute to the protective capacity of the OMV vaccine. Survival in humans has been previously correlated with the level of serum antibodies to B. pseudomallei LPS31. We further corroborated those results using a different melioidosis patient cohort in this study. Therefore, LPS-specific sera IgG, and perhaps mucosal LPS- and CPS-specific IgA, represent possible immune correlates of protection for the OMV vaccine against inhalational melioidosis. While little is known about the protective role of mucosal IgA in human melioidosis, infection and vaccine studies in mice suggest that IgA may contribute to survival32,33,34. Since B. pseudomallei can infect humans by multiple routes, it is possible that IgA plays a more critical role against mucosal infection.

Using the BIPA and BurkPx assays, we demonstrated that OMV vaccination also promotes sera IgG to a number of membrane proteins and identified OmpA and FlgK as vaccine antigens potentially important for protection. OmpA and FlgK are conserved among > 99% of 407 B. pseudomallei isolates examined10. We show for the first time that IgG directed against OmpA is associated with human survival in melioidosis. Patients naturally infected with B. pseudomallei develop antibodies against both OmpA and FlgK35. OmpA proteins are known virulence determinants capable of inducing humoral immune responses36 and have been considered in vaccine platforms against other Gram-negative bacteria, including K. pneumoniae37. Mice immunized with recombinant B. pseudomallei OmpA demonstrated 50% survival against intraperitoneal challenge with B. pseudomallei35. These data, combined with our own results, suggest that OmpA-specific IgG may also represent a correlate of immune protection for the OMV vaccine. It is certainly possible that antibody responses to OmpA, FlgK, and LPS-A enhance protection by working in an additive or synergistic fashion, which could be an important benefit of the multivalent OMV vaccine compared to monovalent vaccines. Several other antigens of interest were detected by protein microarray that were not available for study in the BurkPx assay. Future incorporation of these additional antigens into the BurkPx will enable us to better understand their contribution to the overall OMV-specific antibody response.

Human survivors of melioidosis display elevated IFN-γ producing T-cells compared to humans who do not survive, suggesting that cellular immunity is also critical for protection28. One limitation of the current study was our inability to detect differences in antigen-specific T cell responses using PBMCs from immunized macaques. It is possible that the majority of antigen-specific T cells were sequestered in lymphoid tissues after immunization and were exceptionally rare in the blood prior to infectious challenge38. Assessment of OMV vaccine-mediated cellular immunity in macaques, particularly focusing on lung T cells, cytokines, and tissue resident memory responses, should be investigated by integration into future studies within the context of inhalational melioidosis. Nonetheless, we previously demonstrated vaccine-specific cellular immune responses in OMV-immunized mice10. To bridge this gap, we utilized the human MIMIC system and demonstrated for the first time the ability of the OMV vaccine to drive human antigen-specific Th1 CD4 and CD8 T cell responses, albeit ex vivo. Taken together, these studies provide evidence that cellular immunity following OMV vaccination is achievable.

OMV-based vaccines have been safely and successfully used in humans across the globe for decades to prevent disease caused by Neisseria meningitidis serogroup B12. Vaccine-mediated protection against meningococcemia is afforded by induction of serum bactericidal IgG against Neisseria surface proteins or polysaccharides12. Similarly, the Burkholderia-derived OMV vaccine induces IgG against multiple conserved surface proteins and polysaccharides and subsequently prevents lung disease during inhalational melioidosis. Given that pneumonia is the main presenting feature of patients with melioidosis and is commonly fatal39, prevention of acute, necrotizing pneumonia and preservation of lung function has the potential to significantly reduce mortality from B. pseudomallei. Although our study was constrained by the limited number of non-human primates, it nonetheless provides compelling proof of concept for evaluating OMV vaccines in a large animal model. The application of the rhesus macaque model for pneumonic melioidosis, combined with the translational serologic and cellular immune analyses detailed herein, is expected to significantly advance the development of the OMV vaccine—as well as other promising B. pseudomallei vaccine candidates—toward Phase 1 clinical trials in humans40.

Methods

Bacterial strains and growth conditions

B. pseudomallei Bp82 was kindly provided by Herbert Schweizer and is a ∆purM derivative of B. pseudomallei 1026b42. Bacteria were cultured from glycerol stocks immediately prior to use, and single colonies were selected from freshly streaked Luria Broth (LB) agar plates. For the challenge experiments, overnight cultures of B. pseudomallei K96243 (BEI resources) were diluted 1:100 in fresh LB with 4% glycerol (MilliporeSigma) and incubated with shaking at 37 °C until OD600 reached 0.75. Cultures were centrifuged, and bacterial pellets were washed and re-suspended to achieve the desired concentration.

OMV Purification

OMVs were purified as previously described43 with minor modifications. B. pseudomallei strain Bp82 was freshly streaked from a glycerol stock onto LB agar and incubated for 48–72 h at 37 °C. For preparation of M9 OMVs, an individual colony was inoculated into M9CG media consisting of M9 minimal salts agar (BD Difco) supplemented with 0.4% glucose (Sigma) and 0.5% casamino acids (Amresco) [9], with 100 μg/ml adenine (Sigma) and 5 μg/ml thiamine hydrochloride (Sigma). Cultures were incubated at 37 °C for 16–18 h. The overnight cultures were diluted 1:100 into M9CG media and incubated at 37 °C for 16–18 h until late log phase (OD600 4.5–5.0). Intact bacteria were pelleted by centrifugation (6000 × g for 30 min at 4 °C) using an SLA-1500 fixed-angle rotor. Following centrifugation, the supernatant was filtered through a 0.22 μm polyethersulfone (PES) membrane (MilliporeSigma) to remove any remaining bacteria or large bacterial fragments. Absence of bacterial contamination was verified by incubating 1 mL per liter of supernatant on LB agar for 48–72 h at 37 °C. OMVs were precipitated by incubating with 1.5 M ammonium sulfate (Fisher Scientific, Pittsburgh, PA, USA) overnight and then harvested by centrifugation (11,000 × g, 45 min, 4 °C) using an SLA-1500 rotor. Crude vesicles were resuspended in 60% sucrose (MilliporeSigma) in 30 mM Tris-HCL pH 8.0, layered at the bottom of 35–60% density gradient, and subjected to ultracentrifugation (200,000 × g, 3 h, 4 °C) using a 50.2Ti rotor. Fractions of equal volume were removed from the top, then evaluated by SDS-PAGE to visualize protein profiles by Coomassie blue staining as previously described8. Fractions containing identical protein profiles were pooled and subjected to ultracentrifugation (200,000 × g, 19 h, 4 °C) to obtain highly purified vesicles. Purified vesicles were re-suspended in LPS-free water, visually confirmed by transmission electron microscopy, and quantitated by Bradford assay, as previously described43.

Immunization and challenge

Twelve male Indian rhesus macaques (Macaca mulatta) were used in this study. Six animals (KM30, KT39, KJ33, KP07, KM49, KE90) were immunized subcutaneously (SC) with 100 μg M9 OMVs. Six animals (KP86, KV38, KA94, KF36, KJ83, KM28) were immunized SC with saline vehicle only (sham) and served as controls. Immunizations were performed on days 0 and 28. Blood and bronchoalveolar lavage fluid was obtained prior to immunization (basline), four weeks after the first immunization (prime), and two weeks after the second immunization (boost) for measurement of humoral and cellular immune responses to vaccination.

Four weeks after the final immunization, five OMV- and six sham-immunized macaques were challenged with B. pseudomallei strain K96243 (BEI Resources) by small particle aerosol using a head-only configuration as we previously described44. Actual infectious doses delivered to the macaques were determined by active sampling of the aerosol within the head-only chamber during aerosol exposure and respiratory measurement by whole body plethysmography immediately prior to aerosol exposure. The product of the corresponding aerosol concentration (CFU/liter of aerosol) and the cumulative tidal volume based upon exposure duration for each animal were used to determine individual exposure dose and is presented in Supplementary Fig. S1.

Monitoring of respiratory function

The respiratory function of the macaques was measured by subjecting each animal to whole-body plethsymography for challenge dose calculation. Animals were anesthetized using tiletamine/zolazepam (Telazol, 6–8 mg/kg) and then placed in dorsal recumbency into a custom, sealed acrylic whole body chamber fitted with a 3 mm rubber dam surrounding the neck. The animal was allowed to breathe normally, and thoracic movement produced volumetric displacement of air in the sealed chamber, measured by a pneumotachograph. Digital signal from passive respiratory maneuvers were acquired and converted to wave form for further analysis using specialized software (IOX2, SciReq, Montreal, QC, Canada). Respiratory function measurements were collected continuously for three minutes for a minimum of 18 sampling points per animal per sampling event.

Biotelemetry measurements

Remote biotelemetry was used to monitor 6 of 12 macaques and began two weeks prior to infection to establish baseline parameters. Mean Analysis was performed over 30–60 min intervals every hour and control temperatures were averaged for 24 h periods, with each animal serving as its own control. A hyperthermic response (fever) was defined as a 1.5-degree in core temperature increase and departure from normal diurnal variation in NHP body temperature. A hypothermic response was defined as a 1.5-degree in core temperature decrease and departure from normal diurnal variation in NHP body temperature.

Necropsy, gross pathology, histopathology

For euthanasia, animals were first anesthetized (Telazol, 6–8 mg/kg) and provided an opioid analgesic (Buprenex, 0.01 mg/kg), followed by an overdose of pentobarbital. Necropsy and gross examination were performed by veterinary pathologists who remained blinded to the treatment groups, and tissues were collected for subsequent analysis. Formalin-fixed, paraffin-embedded tissues were sectioned at 6 μm and mounted onto positively charged glass slides. Sections were baked for 1 h at 60 °C, deparaffinized in xylene, and then rehydrated in graded concentrations of ethanol. Slides were stained with hematoxylin and eosin to allow histologic examination of the tissues. For histopathology scoring, a minimum of 4 slides were taken from affected areas and analyzed by a blinded clinical pathologist. Granulomas and iBALT like structures were quantified for each slide, and the total number were averaged for each animal. For inflammation, slides were scored on a 0-5 basis according to the following criteria: 0= no parenchymal inflammation; 1 = mild, non-organized inflammation, less than 15 % of lung with no discernable alveolar congestion; 2 ≤ 25% of lung inflamed with obvious organized areas of alveolar congestion; 3 ≤ 50% of alveolar space inflamed and congested; 4 ≤ 75% of alveolar space inflamed and congested; 5 = no discernable unaffected lung. For inflammation scoring analysis, a non-parametric distribution was assumed, and significance was determined using the Mann-Whitney test (one-tailed).

Assessment of immune responses to vaccination

Blood and bronchoalveolar lavage (BAL) fluid were collected from rhesus macaques 2 weeks prior to the first immunization (baseline), four weeks after the first immunization (prime, blood only), and two weeks after the final immunization (boost) to evaluate antigen-specific antibody responses. High-binding microtiter plates (Greiner Bio-One, Monroe, NC, USA) were coated with M9 OMVs or heat-killed Burkholderia pseudomallei K96243 at a concentration of 500 ng/well in coating buffer and incubated overnight at 4 °C overnight. All subsequent incubations were done at room temperature for two hours on an orbital shaker at 300 rpm. Coated plates were washed three times with wash buffer (1 × PBS with 0.1% Tween 20). Washed plates were incubated with blocking buffer (5% skim milk in wash buffer) for 2 h. Plates were then washed three times. Washed plates were incubated with NHP serum serially diluted in blocking buffer from 1:4000 to 1:256,000. Plates were washed three times and then incubated for 1 h with goat anti-monkey IgG secondary antibody (Fitzgerald, Acton, MA, USA) diluted 1:500 in blocking buffer. Plates were washed three times and developed using 3,3′,5,5′-Tetramethylbenzidine (SeraCare, Milford, MA, USA). Color development was stopped using 1.0 M H2SO4. Plates were immediately read at 450 nm. Results for each animal were plotted as reciprocal titers versus absorbance. For each sample, the log of ELISA optical density versus dilution number generated a smooth curve that could be fit by a second-order polynomial by non-linear regression. From 64 such fits, a consistent proportional uncertainty for ELISA data was obtained and applied to all OD values to estimate the standard deviation of each measurement. Pairwise comparison of pre immune versus post-boost sample curves of OD450 vs dilution number was done by the modified chi-squared test45.

MIMIC® analysis

Apheresis blood products were collected from healthy human donors. The collections and study protocol were reviewed and approved by Chesapeake Research Review Inc (Columbia, Maryland) under IRB 0906009. Donor plasma and peripheral blood mononuclear cells were cryopreserved and stored for an extended period of time in vapor-phase nitrogen tanks. To evaluate T cell responses, differentiated donor antigen-presenting cells (APCs) were stimulated with either untreated (no antigen-stimulated control wells) or designated Ag of interest. After priming with antigens, APCs were co-cultured with enriched autologous CD4 + T cells for 14 days. Cells from day 14 MIMIC® co-cultures were challenged by autologous cytokine-derived dendritic cells loaded with specific Ag for 4 h in the presence of Brefeldin A to measure cytokine stimulation upon Ag exposure as previously described46. After cells were harvested and stained with specific Abs, cells were acquired on an fortessa flow cytometer (BD Biosciences), and data were analyzed using FlowJo software (FlowJo, LLC).

BIPA printing and probing

Protein microarray chips consisting of 1205 B. pseudomallei antigens were fabricated as described previously20. Briefly, custom PCR primers comprising 20 bp of a gene-specific sequence and 33 bp of “adapter” sequences were used with B. pseudomallei, K96243 genomic DNA as a template. The adapter sequences, which become incorporated into the termini flanking the amplified gene, are homologous to the cloning site of the linearized T7 expression vector pXT7 and allow PCR products to be cloned by in vivo homologous recombination in competent DH5a cells. The resulting fusion protein incorporates a 5’ polyhistidine epitope, an ATG translation start codon, and a 3’ hemagglutinin epitope and T7 terminator. Sequence-confirmed plasmids are expressed in 5 h in vitro transcription-translation reactions (RTS 100 kits from Roche) according to the manufacturer’s instructions. Protein expression is monitored either by dot blot or microarray using monoclonal anti-polyhistidine (clone His-1, Sigma) and anti-hemagglutinin (clone 3F10, Roche). Microarrays are printed onto nitrocellulose-coated glass FAST slides (Whatman) using an Omni Grid 100 microarray printer (Genomic Solutions).

Sera used in this study was collected from OMV- and sham-vaccinated NHPs before and after vaccination. Prior to probing, the sera were diluted to 1/100 in Protein Array Blocking Buffer (GVS) containing E. coli lysate at a final concentration of 30% and incubated at room temperature for 30 minutes with constant mixing. The arrays were rehydrated in blocking buffer for 30 min and probed with the pretreated sera overnight at 4 °C with constant agitation. The slides were then washed 3 times for 5 minutes with agitation in Tris(hydroxymethyl)aminomethane (Tris) buffer containing 0.05% (vol/vol) Tween 20 and incubated in biotin-conjugated goat anti-mouse Ig (Jackson Immuno Research) diluted 1/400 in blocking buffer for 1 h. After washing the slides 3 times in 10 mM Tris (pH 8.0) and 150 mM NaCl containing 0.05% (vol/vol) Tween 20 (TTBS), bound antibodies were detected by incubation with streptavidin conjugated Quantum Dots (Invitrogen) diluted 1/250 in blocking buffer. The slides were then washed 3*5 min in TTBS for 5 min and 3 × 5 min in Tris buffer without Tween, followed by a final wash with deionized water. The slides were air dried after brief centrifugation and analyzed using an ArrayCam Imager (Grace Bio). Intensities were quantified using Arraycam software. All signal intensities were corrected for spot-specific background. Proteins were considered expressed if the carboxy-terminal HA tag signal intensity was greater than the average No DNA signal intensity plus 2.5-times the standard deviation. Protein expression efficiency was determined to be 99.2% by probing against a carboxyterminal HA tag for quality control.

Protein purification and bead conjugation for multiplex serology assays (BurkPx)

Nineteen B. pseudomallei antigens that include recombinant proteins or purified carbohydrates were utilized in the BurkPx assay (Table 2) and have been previously described21. These antigens were previously identified by mass spectrometry and/or western blot analysis of OMV preparations and represent a subset of the total antigen content of the OMV vaccine10. The cloning and purification of these proteins have been previously described21. Briefly, the histadine-tagged proteins were expressed in E. coli. We initially purified the protein using soluble and insoluble (thioredoxin solubility tag or N-lauroylsarcosine detergent solubilization) expression methods and nickel affinity chromatography (HisPrep FF 16/10, GE Life Sciences) according to the manufacturer’s protocols47. Protein concentrations were determined using the Bradford assay with bovine serum albumen (BSA) as a standard. Capsular polysaccharide (CPS) and Lipopolysaccharide type A (LPS A) were prepared as previously described48. The amount of purified polysaccharide was determined by dry weight after lyophilization.

Purified soluble and chemically solubilized proteins were conjugated to different fluorescently labeled BurkPx beads (regions) as previously described21. For proteins, the standard amine coupling reaction was used that included Sulfo-NHS (N-hydroxysulfosuccinimide) and EDC (1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride) conjugation chemistry (Luminex xMAP Cookbook). Successful coupling of proteins to beads were confirmed using the anti-6x His monoclonal antibody (clone AD1.1.10) biotin conjugate (Abcam) and streptavidin, R-Phycoerythrin Conjugate (SAPE, Life Technologies). For purified carbohydrates, we used the 4-(4,6-dimethoxy[1,3,5]triazin-2-yl)-4-methyl-morpholinium (DMTMM) as previously described49. Purified carbohydrates LPS A, LPS B, or CPS were coupled to BurkPx beads, and the coupling was confirmed using LPS or CPS specific monoclonal antibodies kindly provided by Dr. David AuCoin at the University of Nevada, Reno50,51 Goat anti-mouse IgG biotin conjugate (ThermoFisher), and SAPE. In addition to the B. pseudomallei antigens, we conjugated positive (species-specific antibody type) controls. All conjugated beads were blocked and stored in PBS-TBN buffer (0.01 phosphate, pH 7.4, 138 mM NaCl, 2.7 mM KCl, 0.1% BSA, 0.02% Tween-20, and 0.05% sodium azide). A summary of the coupled beads is shown in Table 1.

BurkPx multiplex assays

Protein or carbohydrate conjugated beads were diluted to 10,000 beads per bead region per ml in 1 x Blocker BSA solution (Fisher Scientific) after bead sonication, and 100 μl of beads were aliquoted per well. Beads were washed 2X with wash buffer (11.9 mM Phophate, pH 7.4, 137 mM NaCl, 2.7 mM KCl, and 0.05% Tween 20) and the wash buffer was removed after beads were bound to a plate magnet. Serum or BAL was undiluted or diluted 1000-fold in 1x Block BSA solution, and 100 μL diluted serum was added to the beads. Serum and beads were incubated for 2 h with shaking at room temperature. The beads were washed 3 times with wash buffer, and 100 μL of 2 μg/ml of Goat anti-human IgG (Abcam), mouse anti-monkey IgA (Thermo Fisher), or goat anti-monkey IgM (Sigma); goat anti-monkey IgG (Fitzgerald) biotin conjugate secondary antibody diluted in 1 x Block BSA were added to the beads. The secondary antibody and beads were incubated for 1 h with shaking at room temperature and washed 3 times with wash buffer after incubation. Beads were then incubated with 4 μg/ml SAPE (Life Technologies) diluted in 1 x Blocker BSA. The SAPE and beads were incubated for 0.5 hours with shaking at room temperature and washed 3 times with the wash buffer after incubation. Beads were suspended in 1 x Blocker BSA, and the beads were read on a BurkPx system (Luminex), and the median fluorescent intensity (MFI) units per bead region was calculated using xPONENT software (Luminex).

Opsonophagocytosis assay

The assay was performed as previously described with minor modifications52,53. Macaque bone marrow derived macrophages (Cell Biologics) were cultured overnight in a T-25 flask in complete macrophage medium (Cell Biologics), then seeded in a 24-well plate at 2.5 × 105 cells per well. A 16 h culture of B. pseudomallei strain Bp82 was collected by centrifugation at 8000 × g for 10 min, washed twice with phosphate-buffered saline (PBS), and resuspended in PBS to an optical density of 0.75 at 600 nm (equivalent to 1 × 108 colony forming units (cfu)/ml). A 1 ml aliquot of the suspension was incubated at 37 °C for 30 min in the dark with an equal volume of a 2.0 μM solution of 5(-and 6) carboxyfluorescein diacetate succinimidyl ester (CFDA/SE, Molecular Probes), the non-fluorescent diacetate form of CFDA that easily passes through cell membranes to label viable bacteria. Once inside the organisms, the molecule is converted to the fluorescent form known as CFSE [5(-and 6) carboxyfluorescein succinimidyl ester]. After washing two times with PBS to remove excess dye, the bacteria were resuspended in 1 ml culture media (equivalent to 1 × 108 cfu/ml). 75 ul CFSE-labeled bacteria were incubated for 1 hr at 37 C with 25ul of sera collected from OMV vaccinated macaques at baseline (pre-immune) or after boost (post-vax) for paired comparisons. Opsonized bacteria were then incubated with macrophages at MOI 10:1 for 1 hr at 37 C. Flow cytometry was performed using standard methods for cell analysis. Briefly, forward scatter, side scatter (linear) and fluorescence (log) parameters were collected. Instrument thresholding was performed on forward scatter signals. CFSE fluorescence was detected on the LSRFortessa (Becton Dickinson, San Jose, CA) using a 530/30 bandpass filter. From each sample, 30,000 cells were acquired, and CSFE fluorescence intensity of the gated macrophages was measured as a parameter for phagocytosis of labeled bacteria (See Supplementary Fig. 6 for gating strategy). The phagocytosis index (PI) was calculated as the percentage of macrophages that have phagocytized multipled by the mean fluorescence intensity (MFI), as previously described54.

Melioidosis patient population and sample collection

Venous whole blood was collected from 92 culture-confirmed patients with melioidosis from the Darwin Prospective Melioidosis Study41. Patients were subdivided into individuals that succumbed (n = 8) or survived (n = 82) B. pseudomallei infection. Patient samples from individuals that succumbed to infection had limited samples within a narrow time frame after admittance to the hospital (0–134 days). To limit the potential effect of sample collection time, we selected the last sample collected from each patient within this time frame for both clinical groups. Prior to reactivity detection, human serum samples were thawed, mixed, and diluted 1000-fold in 1% BSA in PBS.

Statistical analyses

Statistical analyses were performed using GraphPad Prism version 10.4 (GraphPad Software, San Diego, CA, USA). All statistical tests were two-sided. p-values < 0.05 were considered statistically significant.

Ethics statement

This study was performed in strict accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health (NIH). The protocols were approved by the Tulane University Institutional Animal Care and Use Committee (P0276). The Tulane National Primate Research Center (TNPRC) is fully accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care-International. Rhesus macaques were maintained in ABSL-2 and ABSL-3 housing. Macaques were observed at least three times daily during the acute stage (first week) of infection. Euthanasia of macaques was performed if animals were determined to be under respiratory distress and at the study endpoint using an overdose of pentobarbital under anesthesia. Banked human samples were utilized in this study and obtained from Darwin Prospective Melioidosis Study41 as described below. This human subjects study was approved by the Human Research Ethics Committee of the Northern Territory Department of Health and the Menzies School of Health Research (HREC 02/38 and HREC 2014–2037).

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.