Novel multi‐component vaccine approaches for Burkholderia pseudomallei

Summary

Burkholderia pseudomallei is the causative agent of melioidosis. Historically believed to be a relatively rare human disease in tropical countries, a recent study estimated that, worldwide, there are approximately 165 000 human melioidosis cases per year, more than half of whom die. The bacterium is inherently resistant to many antibiotics and treatment of the disease is often protracted and ineffective. There is no licensed vaccine against melioidosis, but a vaccine is predicted to be of value if used in high‐risk populations. There has been progress over the last decade in the pursuit of an effective vaccine against melioidosis. Animal models of disease including mouse and non‐human primates have been developed, and these models show that antibody responses play a key role in protection against melioidosis. Surprisingly, although B. pseudomallei is an intracellular pathogen there is limited evidence that CD8⁺ T cells play a role in protection. It is evident that a multi‐component vaccine, incorporating one or more protective antigens, will probably be essential for protection because of the pathogen's sophisticated virulence mechanisms as well as strain heterogeneity. Multi‐component vaccines in development include glycoconjugates, multivalent subunit preparations, outer membrane vesicles and other nano/microparticle platforms and live‐attenuated or inactivated bacteria. A consistent finding with vaccine candidates tested in mice is the ability to induce sterilizing immunity at low challenge doses and extended time to death at higher challenge doses. Further research to identify ways of eliciting more potent immune responses might provide a path for licensing an effective vaccine.

Introduction

B. pseudomallei is a facultative, intracellular Gram‐negative bacillus and the causative agent of melioidosis. The organism resides in the soil and water, and infection can occur through multiple routes of exposure, including inoculation, ingestion and inhalation. The incubation period for melioidosis typically ranges from 1 to 21 days, with a median of 9 days; however, latent infections can occur, with disease manifesting decades after exposure 1, 2, 3. Case fatality rates can exceed 40% in certain parts of the world. The signs and symptoms of melioidosis often mimic other diseases (e.g. community‐acquired pneumonias or tuberculosis), resulting in frequent misdiagnosis 4. Pneumonia and bacteraemia are the most common clinical presentations, occurring in approximately 50% of cases 1. Other clinical presentations include ulcers or other skin lesions, gastrointestinal ulceration, sepsis or infections and abscesses involving internal organs (e.g. the spleen, prostate, kidney or liver) 2. In addition, people with certain underlying medical conditions, including diabetes mellitus, alcoholism, chronic lung disease, chronic renal disease, liver disease, haematological malignancy, thalassaemia, cancer, long‐term steroid use and other non‐HIV immunosuppressive disorders, are at greater risk for developing disease 5. Diabetic individuals have a 12‐fold increased risk of melioidosis, and more than half of all melioidosis cases have diabetes 6.

The genome of B. pseudomallei is encoded on two chromosomes and exceeds 7·0 maltose‐binding protein (Mbp). B. pseudomallei possesses numerous virulence factors, including surface polysaccharides, such as capsular polysaccharide (CPS) and lipopolysaccharide (LPS), which are involved in inhibition of opsonophagocytosis and resistance to complement‐mediated killing 7, 8. The organism also utilizes specialized secretion systems, in particular the cluster 3 type III secretion system (T3SS‐3) and cluster 1 type VI secretion system (T6SS‐1) to facilitate survival and growth within the host 8. Additional virulence factors include various secreted proteins (e.g. phospholipases), motility proteins and secondary metabolites 9. B. pseudomallei is also intrinsically resistant to many commonly used antibiotics, including aminoglycosides, penicillins, rifamycins and third‐generation cephalosporins 10, 11, 12.

Predicted and evidenced global incidence of disease

Historically, melioidosis was considered to be geographically restricted to countries in southeast Asia and northern Australia, because these regions have reported the highest incidence and prevalence of the disease, where annual incidence is up to 50 cases per 100 000 people 1, 2; however, the incidence of melioidosis increases in diabetic patients and, for example, in northern Australia it is estimated to be 260 cases per 100 000 13. The appearance of sporadic melioidosis cases and further investigation to identify other endemic areas worldwide has now expanded the global map of the disease to cover more countries within tropical areas 14, including the western hemisphere (Africa and America) 15 and the majority of the Asian continent (Fig. 1) 14, 16. A recent study compiled data from documented global human and animal cases and the presence of environmental B. pseudomallei and used a formal modelling framework to estimate the global burden of melioidosis 14. This analysis estimated that, worldwide, there are approximately 165 000 human melioidosis cases per year, 89 000 of whom die from complications related to the disease. Further, this study estimated that B. pseudomallei is present in the environment in as many as 82 countries 14.

Figure 1.

Countries where melioidosis is reported to have been acquired (coloured in blue), evidenced through publication of case reports.

A subsequent study evaluating the dissemination and evolution of B. pseudomallei has identified distinct evolutionary bacterial lineages after sequencing and comparing more than 250 B. pseudomallei strains isolated between 1935 and 2013 from humans with melioidosis or environmental samples, and representing countries across Australia, Asia, Africa and Central and South America 17. This genomic global analysis confirmed that the Australian continent represents an early reservoir for B. pseudomallei, with the bacteria being first transmitted to southeast Asia with subsequent dissemination to South and East Asia 17. Further, the study also confirmed that B. pseudomallei isolates from African origin were transmitted to Central and South America between 1650 and 1850, which correspond to the period representing the slave trade. Although this and other studies have provided a better understanding of the prevalence of melioidosis cases worldwide, and in many cases have demonstrated endemicity in countries where the pathogen was only sporadically detected, the full extent of the disease is still not known. Recent efforts to educate health professionals in countries where melioidosis has been reported 18 is expected to result in a steady increase in the awareness and better diagnosis and identification of the pathogen. However, major roadblocks exist that prevent the effective determination of the burden of melioidosis, and those include diversity in the clinical manifestations, lack of basic microbiological facilities in several endemic areas, the lack of conventional bacterial identification methods and awareness and knowledge among health‐care professionals and poor disease surveillance systems 2, 16. Overall, the evidence indicates that introduction of a vaccine in endemic populations could result in significant reduction in the number of human melioidosis cases.

Animal models of disease for testing vaccines

Several animal models have been used or developed to study different aspects of the pathogenesis of B. pseudomallei and recapitulate a variety of clinical aspects of melioidosis disease. Animal models range from several strains of mice, rats and hamsters to invertebrate models and to non‐human primates 19. However, most of the virulence studies as well as the evaluation of candidate vaccines have been restricted to murine models of disease, using a wide variety of routes of infection/vaccination and different ranges of challenge doses. Interestingly, efforts to develop and test an effective melioidosis vaccine candidate in the past decade (see Table 1) have mainly used two distinct but well‐characterized mouse strains (BALB/c and C57BL/6) 20, 21.

Table 1.

Multi‐component vaccine platforms recently developed for Burkholderia pseudomallei. The vaccine platforms described here are not prioritized. They are listed in the order that they occur in the text. Control groups used in each study are not detailed in the table, but should be taken into account when interpreting survival outcomes

Immunization					Challenge
Vaccine platform	Antigens	Adjuvant(s)	Route	Mouse strain	Strain	Route	Dose	Survival	Reference
Glycoconjugate	CPS‐Hcp1	Alhydrogel + CpG	s.c.	C57BL/6	K96243	aerosol	1·6 × 103	100% at day 35	46
	CPS‐TssM							80% at day 35
	CPS‐CRM197							67% at day 35
	OPS‐BSA	Alhydrogel + CpG	s.c.	BALB/c	K96243	i.p.	4 × 104	0% at day 14	44
	CPS‐BSA						8 × 104	50% at day 35
	CPS + BSA‐LolC						8 × 104	70% at day 35
	LolC						8 × 104	0% at day 35
	Synthetic CPS + TetHc	MPL/Sigma adj. system	i.p.	BALB/c	K96243	i.p.	9 × 104‐1 × 105	67% at day 35	47
	OPS + Campylobacter jejuni AcrA	+/– Alum	s.c.	BALB/c	K96243	i.n.	2 × 103	40% at day 12	48
								0% at day 21
	LPS‐TetHc	–	i.p.	BALB/c	K96243	i.p.	4 × 104	81% at day 29	49
	LPS only							62% at day 29
	LPS and TetHc							75% at day 29
	AuNP‐Hcp1‐LPS	Alhydrogel + poly(I:C)	s.c.	C57BL/6	K96243	i.n.	1 × 105	10% at day 35	50
	AuNP‐HA‐LPS							20% at day 35
	AuNP‐FlgL‐LPS							90% at day 35
	AuNP‐Comb‐LPS							100% at day 35
Subunit	Chronic antigens	Sigma adj. system	i.p.	BALB/c	K96243	i.p.	7 × 104	30% at day 50	57
	Chronic + LolC							40% at day 50
	Chronic + CPS							50% at day 50
	CPS							10% at day 50
Nano/microparticle	OMVs	–	i.n.	BALB/c	1026b	aerosol	1 × 103	20% at day 14	58
			s.c.					60% at day 14
	OMVs	–	s.c.	BALB/c	K96243	i.p.	2 × 104	100% at day 21	59
	Lysate/MP + Resiq/MP	–	s.c.	BALB/c	1026b	i.p.	1 × 106	12% at day 26	60
	Lysate + Resiq/MP
	Lysate + Resiq/MP	Alum						8% at day 26
		–						16% at day 26
Whole bacteria	Inactivated	CLDC	i.n.	BALB/c	1026b	i.n.	7·5 × 103	100% at day 40	61
	B. pseudomallei
	Live, attenuated (ΔaroC)	–	i.p.	C57Bl/6	A2	i.p.	6 × 103	80% at day 150	62
	Live, attenuated (ΔpurM)	–	s.c.	C57Bl/6	1026b	i.n.	1 × 104	100% at day 60	64
	Live, attenuated (ΔrelA–spoT)	–	i.n.	C57Bl/6	576	i.n.	1 × 103	100% at day 30	65
	Live, attenuated (Δasd)	–	i.n.	BALB/c	1026b	i.n.	4 × 103	100% at day 80	66
	Live, attenuated	–	i.n.	C57Bl/6	K96243	aerosol	1·7 × 103	100% at day 27	67 ref
	(ΔtonB, hcp1)
	B. thailandensis E555	–	i.p.	BALB/c	K96243	i.p.	6 × 106	100% at day 35	68

CPS = capsular polysaccharide; CLDC = cationic liposomes complexed with non‐coding plasmid DNA; CpG = cytosine–phosphate–guanine; BSA = bovine serum albumin; LPS = lipopolysaccharide; OMV = outer membrane vesicles; OPS = O‐polysaccharide; TetHc = Hc fragment of tetanus toxin; i.n. = intranasal; i.p. = intraperitoneal; s.c. = subcutaneous.

BALB/c mice are a popular strain in melioidosis research because it represents a susceptible host to infection and recapitulates an acute form of the disease 19, 22. This mouse model has been extensively used to study mechanisms of virulence as well as performing preliminary screening of candidate vaccines or therapeutics, but the outcome of the disease depends on the different routes of infection/vaccination used [e.g. intraperitoneal (i.p.), intranasal (i.n.) or aerosol] and the B. pseudomallei strains used for the challenges 22, 23. Further, BALB/c are considered the prototypical T helper type 2 (Th2)‐biased mouse strain because they mount a rapid and robust Th‐2 like response, while their adaptive Th1 response is not as efficient or long‐lasting 24. Therefore, it has been proposed that the BALB/c mice could be more appropriate to perform virulence studies and characterization of bacterial strains and virulence factors, because infection in these animals results in an acute disseminated infection that mimics some of the features of human melioidosis 23. Regardless of those caveats, melioidosis vaccine development in the years 2010–16 heavily utilized the BALB/c mouse model (Table 1) and the outcomes were promising, with different vaccines conferring a wide variety of protection, ranging from 40 to 100%.

In recent years (2016–18), as the preliminary screening of melioidosis vaccines has advanced, testing now has been shifted to the use of C57BL/6. This mouse strain is considerably more resistant to acute infection and it has been recommended as a more suitable model for chronic infection 19, 25. C57BL/6 are considered to be the prototypical Th1‐type mouse strain, which is important for a long‐lasting immune protective response against disease 23, 24. It is important to consider that the innate immune response of macrophages is different between BALB/c and C57BL/6 mice, which could have an impact on the development of Th1 and Th2 adaptive immunity 24. Upon infection, it is reported that C57BL/6 mice reduced the number of detectable bacteria in the target organs and may remain asymptomatic for months before spontaneous disease occurs 26. This long‐term latent infection is considered to mimic chronic human illness. Therefore, inability of the host to completely resolve the infection may lead to a fatal outcome in both acute and chronic B. pseudomallei infections. Consequently, C57BL/6 mice are considered a useful model for vaccine testing, because these mice can attain full protection (Table 1) and sterilizing immunity – two parameters that are important to advance the vaccine candidate to larger animal models and human clinical trials.

In that sense, as the development of vaccines continues evolving, other animal models are needed to test efficacy and safety of the viable candidates. The Steering Group on Melioidosis Vaccine Development (SGMVD) was created to advise the scientific community about the needs and requirements to develop an effective vaccine that can target populations that naturally acquire melioidosis as well as for biodefence purposes 25. The SGMVD has recommended that if the vaccine will be used against naturally acquired melioidosis, the addition of a diabetic mouse model to the testing vaccination scheme will be recommended because populations suffering from diabetes, particularly type 2, are at increased risk of acquiring melioidosis 2, 25. In addition, efficacy studies in a non‐human primate (NHP) model may be required prior to advance the vaccine to human clinical trials. There is a report of the testing of a melioidosis vaccine candidate in rhesus macaques 27 and glanders vaccine candidates have also been tested in this species 28, 29. Other NHP models for meliodosis, for example involving marmosets, have been reported 30, 31 but not yet used to evaluate vaccine candidates.

Cost‐effectiveness of a vaccine

The costs of developing a vaccine and implementing a vaccination programme are significant. For vaccines which are to be used in high‐income countries, and assuming that there is a clear benefit associated with a vaccination programme, these costs can usually be borne by the health‐care systems. However, melioidosis is a disease that occurs predominantly in low‐income or developing countries, meaning that careful consideration needs to be given to the financial viability of a vaccination programme. It is possible that melioidosis vaccines developed for biodefence purposes might be repurposed, and this could significantly reduce the cost of developing a public health vaccine. The SGMVD report published in 2015 25 addressed this question, and concluded that there are likely to be some differences in the requirements for melioidosis vaccines for biodefence and to prevent naturally acquired infection. A biodefence vaccine will need to protect healthy people from inhalational exposure. In contrast, a vaccine to protect against natural infection will need to protect both heathy and immunocompromised (diabetic) hosts, and may need to protect against infection caused by skin inoculation an inhalation. It is possible that the dose received after exposure to a deliberate release may be higher than that received from natural exposure. It may also be appropriate to test the combined efficacy of vaccines and anti‐microbial drugs for post‐exposure therapy.

Notwithstanding the possibility that a biodefence vaccine can be repurposed for public health use, the value of a vaccination programme, assessed as quality‐adjusted life years, can be calculated. This study has been carried out based on the incidence of melioidosis, mortality rate and demographic features of at‐risk populations in Thailand 32. The investigators concluded that vaccination would be cost‐effective if used in high‐risk populations such as diabetics (Fig. 2). The vaccine was cost‐effective even when only partial protection (>50%) was assumed. Vaccines could therefore play a role in reducing the global incidence of melioidosis, but further work is required to establish whether this cost–benefit analysis would apply to regions of the world other than Thailand.

Figure 2.

Cost‐effectiveness of a melioidosis vaccination in Thailand assessed on the basis of cost ($0–25 per course), protective efficacy (PE; 10–100%), annual incidence (25 or 150 cases per 100 000 people) and duration of PE (3 or 10 years). In this modelling the vaccine was predicted to be used only in individuals with major risk factors (diabetes mellitus, chronic lung disease or chronic kidney failure) predisposing them to disease. Areas in blue indicate situations when vaccination would be cost‐effective. Figure redrawn and modified from 32.

Multi‐component vaccines

There has been tremendous progress over the last decade in the pursuit of an effective vaccine against melioidosis. The identification of immunoreactive B and T cell antigens from human melioidosis patients as well as animal studies has greatly facilitated the selection of target antigens and rational design of candidate vaccines 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43. Certain bacterial antigens, such as capsule polysaccharide (CPS), FliC, BopE, AhpC and many outer membrane proteins (Omp), are routinely recognized as highly reactive with patient immune sera or peripheral blood mononuclear cells (PBMCs), and many of these antigens confer partial protection when used as individual subunit vaccines 7. However, it is becoming increasingly evident that a multi‐component vaccine, incorporating one or more protective antigens, will probably be essential for complete protection against B. pseudomallei due to its sophisticated pathogenicity as well as strain heterogeneity. Fortunately, a number of multivalent platforms have emerged that are highly effective in experimental animal models of melioidosis. These include glycoconjugates, multivalent subunit preparations, outer membrane vesicles (OMV) and other nano/microparticle platforms and live‐attenuated or inactivated bacteria (Table 1, Fig. 3). Some of the multi‐component vaccines for melioidosis published during the past 10 years (2009 to present) are discussed next.

Figure 3.

Candidates which have been evaluated as meliodosis vaccines.

The CPS has emerged as a leading vaccine antigen 44. The polysaccharide is a homopolymer of unbranched 1 → 3 linked 2‐O‐acetyl‐6‐deoxy‐β‐D‐manno‐heptopyranose and is a major virulence determinant in B. pseudomallei. Importantly, B. pseudomallei appears to express only a single capsule serotype, suggesting that a CPS‐based vaccine could confer protection against all B. pseudomallei strains. While CPS purified from B. pseudomallei induces protective immunity against melioidosis in mice, the immune response and protection conferred by CPS, a T‐independent antigen, is improved by conjugation to a carrier protein that promotes T cell help 45. In a recent study, B. pseudomallei CPS was purified and covalently linked to recombinant CRM197, a non‐toxic mutant of diphtheria toxin. Immunization of C57BL/6 mice with CPS‐CRM197 produced high‐titre immunoglobulin (IgG) and opsonizing antibody responses against the CPS component. When mice were vaccinated with a combination of CPS‐CRM197 and recombinant Hcp1, 100% of the mice survived a lethal inhalational challenge with B. pseudomallei 46. In another study, intraperitoneal immunization with chemically synthesized CPS conjugated to tetanus toxoid promoted IgM and IgG serum antibodies. While the CPS‐specific antibody titres were considerably less than that induced by native CPS, BALB/c mice were still significantly protected from B. pseudomallei challenge, with 66% of mice surviving until the study end‐point at day 35 47.

In addition to CPS‐based conjugates, immunization with the B. pseudomallei O‐polysaccharide (OPS II) conjugated to a carrier protein AcrA from Campylobacter jejuni provided similar levels of protection to immunization with whole‐cell killed bacteria against a highly lethal challenge 48. In another study, LPS conjugated to the Hc fragment of tetanus toxoid provided better protection (81% survival) than LPS alone (62% survival) 49. LPS glycoconjugates have also been successfully incorporated onto the surface of gold nanoparticles (AuNP), a strategy that may enhance uptake and processing of the conjugate by professional antigen‐presenting cells. Mice immunized with AuNP–FlgL–LPS alone or with a protein combination (FlgL, Hcp1 and haemagglutinin) demonstrated up to 100% survival and reduced lung colonization following a lethal challenge with B. pseudomallei 50.

Like the polysaccharide antigens, a number of B. pseudomallei proteins have been shown to provide various levels of protection against experimental challenge and may be useful for incorporation into multi‐component vaccine formulations. These include outer membrane proteins and proteins associated with the T3SS‐3 and T6SS‐1 51, 52, 53, 54, 55, 56. Due to the complex intracellular lifestyle of B. pseudomallei, inclusion of antigens expressed during the chronic stage of infection may be important for obtaining complete protection. Three proteins, BPSL1897, BPSL2287 and BPSL3369, which were up‐regulated in the lungs and spleens of chronically infected mice, were pooled with the surface OmpA (BPSL2765) protein to create a chronic antigen formulation. Mice immunized with these antigens mixed with LolC or CPS demonstrated better survival compared to mice immunized with the individual subunit antigens, LolC or CPS, alone 57.

OMVs are multivalent nanoparticles naturally secreted by Gram‐negative bacteria. B. pseudomallei produces OMVs that contain numerous protective proteins, such as OmpA, and surface polysaccharides, such as LPS and CPS. Immunization with B. pseudomallei‐derived OMVs provided significant protection against both pneumonic (60% survival) and septicaemic (100% survival) murine melioidosis 58, 59. Protection was associated with the induction of both antibody and cellular‐mediated immune responses. The B. pseudomallei‐derived OMV vaccine was also shown to be safe and highly immunogenic in rhesus macaques 27. Synthetic microparticles composed of acetylated dextran have also been used to encapsulate a B. pseudomallei cell lysate. The formulation showed significant but incomplete protection against a lethal challenge in mice 60. However, the microparticle platform could potentially be enhanced with different antigens and/or adjuvant combinations. For example, whole‐cell‐based vaccines typically provide partial protection against B. pseudomallei 20, 21. However, incorporation of a novel adjuvant, cationic liposomes complexed with non‐coding plasmid DNA (CLDC) to a heat‐killed B. pseudomallei vaccine, conferred 100% survival for >40 days 61.

Live‐attenuated bacterial strains have also shown promising results as vaccine candidates. B. pseudomallei strain A2 with a deletion in aroC was attenuated in both BALB/c and C57BL/6 mice, and C57BL/6 mice immunized with the aroC mutant were significantly protected against challenge 62. Similarly, deletion of purM in B. pseudomallei strain 1026b produced a highly attenuated strain (Bp82) that was avirulent in highly susceptible 129/SvEv mice, Syrian hamsters and immune incompetent mice [interferon (IFN)‐γ^–/–, severe combined immunodeficiency (SCID)] 63. Immunization of C57BL/6 mice with Bp82 conferred 100% survival over 60 days against pulmonary melioidosis 64. A B. pseudomallei relA–spoT double mutant displayed a defect in stationary‐phase survival and intracellular replication in murine macrophages and was attenuated in both acute and chronic mouse models of melioidosis. Vaccination of mice with the ΔrelA ΔspoT mutant resulted in full protection against infection with wild‐type B. pseudomallei 65. An asd mutant of B. pseudomallei 1026b was avirulent in BALB/c mice, and animals vaccinated with the mutant were protected against acute inhalation melioidosis 66. In a recent study, immunization of C57Bl/6 mice with a B. pseudomallei strain lacking tonB and hcp1 also provided complete protection against an otherwise lethal aerosol challenge 67. Immunization with closely related B. thailandensis strain E555, which possesses CPS, provided significant protection against B. pseudomallei intraperitoneal (i.p.) challenge 68. However, emerging reports of human infections caused by B. thailandensis should caution the use of unmodified B. thailandensis as a vaccine 69, 70. Rather, attenuated bacterial strains with defined and confirmed genetic mutations that have undergone significant safety testing in immunocompromised animal models should be prioritized over those strains that have not met these criteria.

In summary, there are a number of promising multi‐component vaccine platforms in the pipeline, and a protective multivalent vaccine against melioidosis may be on the horizon. Nearly all the promising vaccine platforms described in this review have precedence for use as human vaccines 71. This is significant, as successful results in relevant animal models could lead to eventual clinical trials in humans. Going forward, it is imperative that promising candidate are rigorously evaluated in NHP models of melioidosis using appropriate immunization routes and with adjuvants that are likely to be accepted as part of a licensed vaccine.

Correlates of protection

There is little published information on the nature of the protective immune response to B. pseudomallei infection in humans, but it is clear that after exposure to the bacterium T cells are primed and serum antibodies are detectable 8.

Antibodies appear to be directed against cell surface polysaccharides and against proteins while circulating T cells have been shown to respond to a range of protein antigens 35, 40, 72, 73, 74 There is significant evidence that antibody responses play a key role in protection against melioidosis. The probability of survival of humans who have developed melioidosis has been shown to be correlated with the level of serum antibodies to LPS but not to the level of antibodies towards CPS or flagellin 73. In addition, monoclonal or polyclonal antibodies against LPS confer protection in mouse or rat models of disease 75, 76, 77. Antibody against LPS is reported to be opsonizing, promoting uptake and killing by polymorphonuclear leucocytes 72. It is possible that antibodies to other antigens also provide protection against disease, but the evidence is less compelling. For example, antibodies against CPS can protect mice against experimental disease 75, 77 but in humans antibodies to CPS does not appear to confer protection against fatal disease 73. Similarly, antibodies against flagellin appear to be protective in mice 78 but not in humans 73.

All the available evidence also shows that, in murine models of disease, antibodies provide protection and sterilizing immunity only at low challenge doses of B. pseudomallei 75, 76. This may not be a limitation in situations where the challenge dose encountered is likely to be low, or the vaccines might be used in combination with antibiotics to eliminate bacteria that escape antibody‐mediated clearance.

Because B. pseudomallei is an intracellular pathogen, it might be expected that T cell responses would play a key role in protection against disease. This possibility has been explored by immunizing mice with live attenuated mutants and characterizing the protective immune responses that subsequently develop. Although CD8⁺ T cell responses occur after exposure of mice to the vaccines, they appear to play little or no role in protection against disease 28, 64, 79. However, in humans CD8⁺ T cell responses are associated with increased survival 80. CD4⁺ T cell responses occur in mice immunized with some experimental vaccines 28, 79 and in humans from melioidosis endemic regions 38. However, the finding that HIV infection does not appear to predispose individuals to infection with B. pseudomallei in melioidosis endemic regions suggests that they play little role in protective immunity 81.

Why protective immunity is difficult to evoke

A consistent finding with vaccine candidates tested in mice is the ability to induce protection evidenced as an increased time to death, but the inability to induce sterilizing immunity unless animals receive low challenge doses. Observations that melioidosis is often a chronic or relapsing infection in animals and humans also indicate that exposure to the pathogen does not necessarily induce an immune response capable of clearing the infection. Collectively, these observations suggest that the pathogen is able to avoid the induction of protective immune responses.

Currently, it is not known why it is difficult to induce sterilizing immunity. The ability of the bacterium to grow within host cells and spread directly from infected to adjacent uninfected cells, without exposure to the extracellular milieu, would suggest that cellular responses involving cytotoxic lymphocytes would be critical for clearance. The inability of antibody to provide high‐level protection against B. pseudomallei may also reflect the intracellular lifestyle of the organism. However, as highlighted above, the CD8⁺ responses that are induced after vaccination, even with live attenuated mutants, do not appear to be protective. Manipulation of the major histocompatibility complex (MHC) class I processing pathway is a common mechanism by which viral pathogens avoid the induction of protective CD8⁺ T cell responses 82, 83. Often, this is achieved by blocking parts of the pathway downstream of proteasome processing. However, some viral proteins are known to avoid proteasome processing by mechanisms that are not fully understood 83. Although B. pseudomallei encodes a functional proteasome inhibitor, there is no evidence of manipulation of the MHC‐I pathway occurring in infected cells 84. In contrast, there is evidence from studies in Caenorhabditis elegans that the bacterium might recruit the ubiquitin–proteasome system to degrade a GATA transcription factor required for host immune defence 85. This may be important, because GATA factors play conserved roles in immunity in both insects and mammals.

Conclusions and prospects for an effective and viable vaccine

There is now good evidence that devising a vaccine capable of providing protection against melioidosis is feasible. Although the candidate vaccines which are currently available do not provide sterilizing immunity at anything other than relatively low challenge doses, there is evidence from modelling studies that even low levels of protection would be valuable in reducing the incidence of disease. Additional research is ongoing to improve the efficacy of vaccine candidates by evoking responses to a broader range of antigens. Moreover, it is possible that vaccination would allow therapeutics to be used more effectively because vaccines could extend the window of therapeutic intervention for treatment, and there may be synergistic effects between vaccines and anti‐microbials. However, none of the candidates have yet to be tested in humans, and we believe that there is now a strong argument for selecting the most promising for clinical trials.

Disclosures

None.