Current advances and challenges in the development of Acinetobacter vaccines

Abstract

Acinetobacter baumannii is a major cause of healthcare-associated infections worldwide with high morbidity and mortality. The clinical treatment of A. baumannii infections has become increasingly difficult because of the rapid emerging of multidrug and extremely drug resistant strains. Thus, there is an urgent need for the development of novel intervention strategies to combat this multidrug-resistant pathogen. Vaccine is one of the most effective medical measures for infection control and is likely to overcome the development of multidrug resistance by A. baumannii. Here we discussed the recent advances and potential challenges in development of A. baumannii vaccines with a focus on the 3 most important steps in the preclinical vaccine development: antigen selection, immune correlates of protection, and animal models for efficacy evaluation.

Keywords:

• Acinetobacter
• Vaccine
• antigen selection
• correlate of protection
• animal model

Acinetobacter baumannii is a ubiquitous gram-negative, extracellular bacterium and a major cause of healthcare-associated (hospital- and community-acquired) infections (HAIs) worldwide.^1,2 A. baumannii is also commonly involved in war wound or natural disaster-related infections.³ The infection causes a variety of diseases, ranging from pneumonia, meningitis, to serious blood or soft tissue infections, and is often associated with high morbidity and mortality.^4,5 The clinical treatment of A. baumannii infections has become increasingly challenge and difficult because of the rapid emerging of multidrug and extremely drug resistant strains.⁶ In addition, A. baumannii often forms biofilms that are resistant to host defense and antimicrobial treatment. Thus, there is an urgent need for the development of novel intervention strategies to combat and control multidrug-resistant A. baumannii infection.^3,5

Vaccine is one of the most effective medical measures for infection control, and is likely to overcome the development of drug resistance by the pathogen. Currently, there is a paucity of effective antibiotics available or at the late stage of development for the treatment of multidrug-resistant A. baumannii infections. Thus, development of safe and effective A. baumannii vaccines for active and passive immunization would have a major impact on the burden of A. baumannii infections. Initially spearheaded by Pachon and McConnell,^7-9 several laboratories around the world have begun to develop A. baumannii vaccines for targeted populations and a number of experimental vaccine studies have been reported.^10-19 The needs and economic benefit for an Acinetobacter vaccine, the scientific progress and potential challenges in clinical trials in the development of A. baumannii vaccines have been recently reviewed in several excellent papers.^2,20-22 However, despite those efforts and advances, the progress in the development of A. baumannii vaccines has been generally behind the pace of vaccine development for other nosocomial pathogens (such as Clostridium difficile, Staphylococcus aureus etc.)^23,24 and to date no Acinetobacter vaccine has entered Phase I clinical trial, suggesting the relative challenges in the development of safe and efficacious A. baumannii vaccines. Here we discussed the recent advances and potential challenges in development of A. baumannii vaccines with a focus on the 3 most important steps in the preclinical vaccine development: antigen selection, correlates of protection, and animal models for efficacy evaluation.

Choice of vaccine antigens

The initial studies on the A. baumannii vaccine used inactivated whole cells or cell components (OMVs and OMCs) as the immunogen.^7-9 The results from those studies have convincingly demonstrated the feasibility for the development of A. baumannii vaccines. Immunization of animals (mostly mice) with those vaccines induces antigen-specific antibody responses and protects against the subsequent challenge with varying clinical and ATCC typing strains of A. baumannii with significantly reduced tissue and blood bacterial burdens and its associated inflammatory responses.^{7-9,11-13,1-17,19,25} Although those vaccines are highly effective, their clinical applications are limited by the potential regulatory and safety concerns. Therefore, identification of antigens that best stimulate the protective immune response offers significant advantages in the vaccine safety and manufacturing process.

Conjugation of O-antigens or capsular polysaccharides (CPS) to a protein carrier have been successfully used as effective licensed vaccines against several important bacterial pathogens such as Streptococcus pneumoniae, Haemophilus influenza type b, and Neisseria meningitides.^26,27 It is therefore not entirely surprised that several initial studies on A. baumannii vaccine development have focused on glycoconjugate vaccine approach.^10-12,15 Those studies demonstrated the effectiveness of passive protection against systemic and respiratory A. baumannii challenges with specific antisera, polyclonal and monoclonal antibodies^11,15 and provided the proof of concept for development of glycoconjugate A. baumannii vaccines.

However, glycoconjugate approach is likely to be complicated by the anticipated variability of O-antigen and CPS displayed by different A. baumannii strains and isolates. Although the early studies by Traub et al. identified nearly 40 A. baumannii serovars,^28,29 their prevalence among the clinical isolates is largely unknown. A recent study showed that a monoclonal antibody against A. baumannii K1 CPS recognizes 13% of the 100 A. baumannii strains tested,¹⁵ confirming the relative diversity of the surface carbohydrate antigens of this pathogen. Thus, further epidemiological studies would be needed for determining whether there are highly prevalent or specific disease associated serotypes, which would make the best vaccine candidate. In this regard, poly-N-acetyl-β-(1–6)-glucosamine (PNAG) is a surface exopolysaccharide produced by many different bacterial pathogens including A. baumannii.¹⁰ By conjugating a synthetic that mimics PNAG to tetanus toxoid (TT), Bentancor et al. showed that antiserum to 9Glc-NH(2)-TT was highly opsonic against multiple unrelated clinical isolates of A. baumannii that synthesize various levels of surface PNAG.¹¹ More importantly, treatment of mice with those antisera significantly reduced lung and blood bacterial burdens as compared to the placebo-treated mice,¹¹ suggesting that PNAG may be a candidate antigen for inducing protective antibodies with a broad serotype coverage. The other strategies to provide a broad coverage of the different serotypes include the use of multivalent glycoconjugate vaccinations against highly prevalent A. baumannii serotypes, as in the case of S. pneumoniae^10,15,27 or the use of protein antigens.

The genome of A. baumannii encodes more than 1500 proteins,³⁰ any of which alone or in combination can theoretically serve as targets for immune recognition. The feasibility to develop protein antigen-based A. baumannii vaccines was further supported by a recent study, which showed that immunization of mice with an A. baumannii lpxD mutant strain that completely lacks LPS gene induced a comparable protective immunity to that induced by the parental strain with intact LPS.²⁵ To date, a number of proteins have been identified and used in experimental A. baumannii vaccines for active or passive immunization (Table 1). Among those, OmpA is regarded as the most promising one because OmpA is an important virulence factor in the pathogenesis of A. baumannii infection and is highly immunogenic in mice and humans.^13,14,31 At the amino acid level, OmpA is highly conserved (>80%) among different clinical isolates collected from different infectious sites, but shares minimal homology to the human proteome.¹³ OmpA is also highly immunostimulatory in vitro and in vivo by enhancing surface expression of CD80 and CD86 and major histocompatibility complex class I and II molecules of dendritic cells (DCs), inducing phenotypic maturation of murine splenic DCs, and promoting Th1 immune responses.³² Indeed, immunization with recombinant OmpA formulated in alum adjuvant induces partial protection against lethal systemic A. baumannii challenges in diabetic mice ¹³ although others found that OmpA induced little protection in a mouse model of A. baumannii pneumonia.¹⁸ The reasons for those discrepancies are currently unknown but the differences in the OmpA refolding, immunization strategies, route of challenge, and animal models are likely to contribute to the observed differences. Therefore, it is unlikely that OmpA alone will provide sufficient protection against A. baumannii, and identification of additional novel protective antigens remains an important task for A. baumannii vaccine development.

Table 1. Antigens used in experimental A. baumannii vaccines for active or passive immunization

Inactivated whole cells^8,16,25

Outer membrane complexes (OMCs)^7

Outer membrane vesicles (OMVs)^9,19

Outer membrane protein A (OmpA)^13

Membrane transporter Ata^12

Biofilm-associated protein Bap^17

K1 capsular polysaccharide^15

Membrane associated polysaccharide poly-N-acetyl-β -(1–6)-glucosamine (PNAG)^11

Phospholipase D^34

Iron regulated outer membrane proteins^33

Upon recently, the majority of the A. baumannii antigens identified (Table 1) were derived from traditional pathogenesis studies.^{11-13,15,17,33} Therefore, it is encouraging to note that A. baumannii researchers have started to apply reverse vaccinology and immunoproteomics approaches to facilitate the identification of novel antigen candidates.^31,34,35 By combining in silico prediction and comparative genome analysis with in vitro proteomic approaches, Moriel et al. identified 42 surface-exposed and secreted antigens from A. baumannii that could be used as potential vaccine targets.³⁵ Similarly, Fajardo Bonin and colleagues used human sera from A. baumannii-infected patients and health individuals to screen against A. baumannii outer membrane proteins and identified 6 immunoreactive proteins: OmpA, Omp34kDa, OprC, OprB-like, OXA-23, and ferric siderophore receptor protein.³¹ Since those proteins are highly expressed on the bacterial surface and are implicated in the virulence, they are interesting candidates to be included in future A. baumannii vaccine development. As in the development of vaccines for other pathogens, most of the antigens identified by those high throughput platforms have not been evaluated in vivo for their immunogenicity and protection potential.^31,34,35 In a paper published in a recent issue of this journal,¹⁸ Chiang et al. combined the advantages of reverse vaccinology, bioinformatics, and immunoproteomics and identified 13 novel genes from 2752 homologous core genes of A. baumannii as potential vaccine candidate antigens. More importantly, the authors selected and tested 3 of those antigens and found that those antigens were highly immunogenic and provided partial protection (60%) in a mouse model of A. baumannii pneumonia.¹⁸

Immune correlates of protection against A. baumannii infections

Although substantial efforts and progresses have been made in last 5 y on the development of A. baumannii vaccines, further improvement in the vaccine-induced protective efficacies in terms of surviving rates and tissue and blood bacterial burdens may be needed before a vaccine candidate can enter clinical trial. This, in turn, will require a better understanding of the mechanisms of protective immunity and the potential correlates of protection.

As a Gram-negative, extracellular bacterium, it is generally recognized that neutrophils, macrophages, complements, and specific antibodies are needed for effective control and elimination of A. baumannii. Indeed, several studies have convincingly demonstrated the critical role of neutrophils, to a lesser extent, macrophages, in the host innate immunity against both systemic and respiratory infections with varying A. baumannii strains in mice.^36-43 The critical role of neutrophils in vaccine-induced protection against intranasal A. baumannii infection has also been demonstrated in a recent study.¹⁶ On the other hand, the role of complements in host innate and acquired immunity against A. baumannii appears varied between different experimental models and different vaccine preparations.^11-13,15

The much anticipated role for specific antibodies in the vaccine-induced protection against A. baumannii infection is well supported experimentally: (1) passive transfer of whole sera from immunized or convalescence animals, polyclonal or monoclonal antibodies against the whole A. baumannii cells or its cell components recapitulated protection against A. baumannii challenges^11-13; (2) the levels of anti-OmpA antibodies are correlated with survival in mice¹³; (3) opsonization of K1-positive A. baumannii strains, but not K1-negative strains, with a specific monoclonal antibody significantly increased neutrophil-mediated bactericidal activity in vitro ¹⁵; and (4) immunization of B cell-deficient mice failed to induce a protective immunity against subsequent A. baumannii challenges.¹⁶ On the other hand, it is less certain how the antibody-mediated protection against A. baumannii works. Some studies demonstrated that immune sera enhance opsonophagocytic, but not complement-mediated, killing of A. baumannii ¹³ while others found the reverse is the case.^11,12 It was also somewhat surprised to note that the vaccine-induced protection was not compromised in FcRγ−/− mice.¹⁶

The contribution of other arms of immune systems (such as T cells) in the vaccine-induced protection against A. baumannii infection is largely unknown. Increasing evidence from the immunological research and vaccine development suggests the importance of CD4 T cells in vaccine-induced protection against other extracellular bacterial infections such as S. pneumoniae and Bordetella pertussis.^44,45 In addition, it has recently been suggested that A. baumannii may invade and survive intracellularly in the respiratory epithelial cells during its infectious process, and intracellular antibacterial innate immune receptors, Nod1 and Nod2, and their adaptor Rip2 play critical roles in the sensing and clearance of A. baumannii by human airway epithelial cells through the production of ß-defensin 2⁴⁶. Although those observations have important implications in the immunopathogenesis of respiratory A. baumannii infection, their contribution in host defense against A. baumannii infection remains to be confirmed in vivo.

Much of the current A. baumannii vaccine research has focused on the systemic routes of immunization and challenge.^7-9,13 Since A. baumannii infection causes a wide array of clinical diseases from soft tissue infection, pneumonia, and bacteremia to sepsis and Acinetobacter pneumonia is a leading cause of the morbidity and mortality of the infection,⁴⁷ the potential significance of mucosal immunity in protection against A. baumannii-associated pneumonia requires further studies and attention. In this regard, KuoLee et al. have recently shown that intranasal immunization of mice with an inactivated whole cell vaccine protects against respiratory infection by a hypervirulent clinical isolate of A. baumannii with significant reduction in the lung bacterial burdens and extrapulmonary dissemination.¹⁶ In addition, it appears that systemic active or passive immunization is effective in protection against both mucosal (pneumonia) and systemic (bacteremia and sepsis) A. baumannii infections,^11,19 raising the possibility that a strong and appropriate type of systemic protective immunity may be sufficient for the control of bacterial replication in the lung.

Animal models for A. baumannii vaccine development

As in most vaccine research and development, the use of animal models of clinical relevance is critical to the successful vaccine development. To date, the mouse remains the most commonly used laboratory animal species in the evaluation of A. baumannii vaccine efficacy.^7–17,19,25 The mouse models initially developed use either immunosuppressive (such as cyclophosphamide)⁴⁸ or virulence-enhancing agents (such as porcine mucin)⁸ to ensure or enhance the development of a sustained infection. Recently, several groups have successfully developed reproducible A. baumannii infections in immunocompetent and conventional mouse strains.^49,50 In addition, the rat model of A. baumannii pneumonia and wound infection mimics many aspects of human clinical infection.⁵¹ The use of multiple animal models would likely facilitate the development of safe and effective A. baumannii vaccines since immunocompromised animals mimic the situation in human A. baumannii infections, which generally occur in immunocompromised individuals, whereas the use of immunocompetent animals would enable the identification of the key immune components that are important in the host defense against A. baumannii infection.

In addition to the rational selection of appropriate animal models for vaccine efficacy evaluation, it will also be important for the A. baumannii vaccine research community to start standardize the vaccine immunogenicity and efficacy evaluation protocols (such as the intervals between immunization and challenge, and challenge strains, doses, and routes etc.) so that the comparison in the efficacies of the experimental vaccines between different laboratories is possible. In this regard, it was noted that the amount of OmpA used in vaccines has a profound impact on the immunodominant epitope coverage and the immune responses elicited with a low dose inducing a balanced Th1/Th2 response and a high dose promoting a Th2-biased response,¹⁴ and may account for the differences in its protective efficacies observed by different laboratories.^13,18 In addition, a too short interval between last immunization and pathogen challenge may suffer from the potential contribution of innate immunity stimulated by the adjuvant and vaccine components and therefore artificially overestimate the vaccine efficacies.

It is encouraging to note that many A. baumannii vaccine studies have evaluated the protection against not only homologous strain challenges but also challenges by other clinical isolates to estimate the broad protection offered by the vaccine candidate.^11-13,19 However, in most of the studies the serotype and genetic diversities of the challenge strains were not fully characterized. In addition to the well-recognized variations in strain-specific surface carbohydrate antigens as discussed above, the proteomic profiles and biological function of OMVs from different A. baumannii strains are also highly heterogeneous.⁵²

Conclusion

Substantial progresses have been made recently in the development of a safe and effective vaccine to combat infections with multidrug resistant A. baumannii. In this regards, several experimental vaccines using whole cell, polysaccharide or protein antigens have been evaluated for active and passive immunization. They all induced promising antigen-specific antibody responses and conferred varying degrees of protection against systemic or respiratory challenges with homologous and heterologous bacterial strains in experimental animals. However, substantial scientific and technical challenges remain in the development of safe and effective vaccines for A. baumannii. Although active and passive immunization studies have convincingly demonstrated the important role of antibodies in vaccine-induced protection, the mechanisms of such protection and its potential correlates remain unknown. In particular, the role of antibody-mediated opsonisation and bactericidal activities and mucosal IgA responses, all of which are critical in protection against many extracellular bacterial infections, in vaccine-induced host protection against A. baumannii needs further studies. Such knowledge will likely to accelerate the vaccine development. In addition, knowledge on the contribution of other arms of immune systems (such as T cells) in the vaccine-induced protection will be important and may provide clues on the relatively limited protection offered by the current experimental A. baumannii vaccines. Moreover, the course of A. baumannii infection is usually acute and the fate of the infection outcome is suggested to be determined at very early stage of the infection.³⁸ Therefore, any effective A. baumannii vaccine will have to control the bacterial replication and tissue damage at the early onset of the infection. Furthermore, since none of the current vaccine candidates induces a sterile immunity, it is therefore unclear what magnitude of the tissue/blood bacterial burden reduction is needed in animal models to enable a vaccine candidate attractive enough to move into clinical trials.

Compared to other pathogens, relatively smaller numbers of antigen candidates have currently been identified and available for development of A. baumannii vaccines. However, the recent surges in A. baumannii research and our increased understanding in the molecular pathogenesis and virulence factors of the pathogen are likely to expedite the generation of new antigen candidates. Although it was recognized that a vaccine targeting even only a limited number of A. baumannii strains may still be useful,² additional epidemiological data on the serotype prevalence of clinical A. baumannii isolates is important for rational design of glycoconjugate vaccines with broad serotype coverage. Thus, together with the new generation of vaccine development technology (such as reverse vaccinology, immunoproteomics and glycomics etc.), it is high likely that multicomponent A. baumannii vaccines with well-defined composition, high efficacies, broad coverage, and good safety profile will enter the clinical trials in the next 5 – 10 y

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Funding

The Acinetobacter research was supported by the Vaccine Program, National Research Council Canada (NRC) and by a Joint Collaborative Research Project between NRC and Ministry of Science and Technology Taiwan. The view expressed in this article is solely the responsibility of the author and does not necessarily represent the official views of NRC.