Contagious bovine pleuropneumonia: a rationale for the development of a mucosal sub-unit vaccine.

Contagious bovine pleuropneumonia (CBPP) remains a major cattle disease in Africa with serious socio-economic consequences. Its eradication requires the development of improved vaccines. Knowledge on this disease and its causing agent, Mycoplasma mycoides subsp. mycoides biotype Small Colony (MmmSC), has been progressing significantly in the last years, opening new areas for vaccine design. Advances were achieved in the understanding of the protective immune responses to MmmSC infection and immunopathological mechanisms allowing the pathogen to escape the host immune response. Based on sequencing and genomic studies, some virulence factors and metabolic pathways were unraveled leading to the identification of potential MmmSC vaccine candidates. Based on these findings, this review presents a scientific strategy to design multi-component sub-unit vaccines for mucosal delivery as the most promising approach for efficient long-term protective vaccines to prevent CBPP.

Introduction and rationale

Contagious bovine pleuropneumonia (CBPP) remains one of the major cattle diseases in Africa. It is caused by Mycoplasma mycoides subsp. mycoides biotype Small Colony (MmmSC). CBPP is responsible for heavy economic losses due to mortality, loss of weight, reduced working ability or fertility. A recent study evaluated the total economic cost of CBPP in Africa (direct and indirect production losses and disease control costs) at 44.8 million euros [1]. Accurate data, requiring efficient epidemiological surveillance networks, are however lacking [2]. CBPP is included in the Office International des Epizooties (OIE) list of pathologies requiring official declaration (http://www.oie.int). Infected countries are thus excluded from international trade.

CBPP was eradicated during the twentieth century from most of the world, except limited areas in Europe and Asia, based on a program combining slaughtering, control of cattle movement and vaccination. However, in Africa, the sanitary measures used on other continents to successfully eradicate CBPP are impracticable since cattle raising rely on nomadism and transhumance. The disease was reported in 17 countries in 2001 and in 20 countries in 2004 (http://www.oie.int/hs2/). In 1994, in Bostwana, stamping out the whole cattle population in the affected zones was the ultimate choice of the veterinary services, because vaccination with the available T1 vaccine failed to stop the spread of the disease. This strategy proved very efficient and enabled Bostwana to regain its freedom status, but it remains too expensive to be applied in other African countries [2].

To achieve eradication of the disease from Africa, the development of improved vaccines remains the only realistic approach. The T1-based vaccines currently in use are of low efficacy and require annual, costly vaccination campaigns [3]. These vaccines, made from freeze-dried broth cultures of live attenuated MmmSC strains (T1/44 and T1SR), present severe limitations such as the short duration of immune protection, although longer for the T1/44 strain compared to the T1/SR, and persistent side effects, mainly for the T1/44 strain [3]. Several studies have suggested a possible improvement of their efficacy through the combined use of antibiotics, changes in vaccine formulation, or the use of a prime-boost strategy [2], [4]. However, this might be too expensive or insufficient to overcome the intrinsic limitations of the vaccine. Based on immunological knowledge, another approach could rely on new routes of vaccine delivery. Indeed, it is well known for mucosal diseases, such as CBPP, that mucosal delivery would significantly improve the efficacy of the host immune response [5]. So far, however, this approach has not received the attention it deserves. There is, therefore, a need for improved vaccines that could overcome the T1 vaccine limitations and trigger long-term protection and which, coupled to effective surveillance systems and control of animal movements, would represent the most relevant tool to achieve eradication of CBPP from Africa.

The proof of feasibility of triggering a long-term protective immune response against CBPP by vaccination stems from the observation that cattle recovering from an MmmSC infection are then protected against new MmmSC infections [3], [6], [7]. Recovering cattle, therefore, not only succeed in controlling the MmmSC infection by mounting an appropriate immune response, but also retain in their lymph nodes an MmmSC-specific immune memory able to trigger an anamnestic protective response [8]. This objective is also supported by the experiments made by Willems in 1852 showing that subcutaneous inoculation of CBPP-infective lymph protected cattle against an MmmSC challenge [9]. These findings demonstrate that a long-term protective immune response against MmmSC can be naturally developed and most probably also vaccine induced in cattle.

Rational development of optimized vaccines requires a good understanding of the immunological bases of protection against infection, as well as an understanding of the immunopathological mechanisms underlying the disease and the identification of the components of the pathogen involved in both these aspects. Recent studies have led to improved knowledge of CBPP, which leads to defining more relevant vaccine strategies against the disease. This review gives a rapid overview of most recent results and discusses rational strategies for the development of optimized sub-unit vaccines. Because the safety of vaccines is paramount and because live vectored vaccines, although powerful, raise some concerns regarding safety, stability, potential reversion to virulence or effectiveness in the presence of pre-existing immunity, the trend in vaccinology now focuses on the design of safer and better-defined sub-unit vaccines.

Rational strategy to develop sub-unit vaccines to prevent CBPP

A vaccinal strategy is determined by (1) the protective immune parameters to trigger; (2) the immunopathological mechanisms to control; and (3) the antigens from the pathogen to include in the vaccine.

Protective immune parameters

The objectives of a vaccine are to elicit, or at least prime the relevant immune responses allowing the vaccinated host to quickly mount an efficacious protective response when in contact with a new infection. Identifying the MmmSC-specific immune protective mechanisms is thus a major prerequisite to the development of efficient vaccines. To this aim, recent studies were implemented to characterize the immune response elicited in cattle after natural MmmSC infection. Comparative analyses of the MmmSC-specific humoral and cellular immune responses were performed between animals recovering or dying from the MmmSC infection. Cattle recovering from an MmmSC infection are known to develop long-term immunity [6], [7]. Characterization of the MmmSC-specific immune mechanism generated, therefore, during the primary immune response and persisting in the regional lymph nodes of these recovered animals will thus unravel the basis of immune protection against CBPP and the MmmSC-specific memory response, responsible for the protective anamnestic response. Understanding the mechanisms operative in the development of these memory T-cells is a central goal in the development of effective vaccination strategies.

The results of these studies revealed that cattle recovering from an MmmSC infection, compared to animals with acute infection, were characterized by (1) a stronger and persisting local IgA response [10] and (2) a higher MmmSC-specific CD4+ T-cell response with IFNγ production detected in blood and persisting in the respiratory lymph nodes several months after infection [8], [11]. Data on the MmmSC-specific local cell-mediated immune response, although very valuable to unravel the complete protective mechanisms against CBPP, are not yet available due to the difficulty to assess this aspect in cattle. Nevertheless, a key finding is the persistence of the MmmSC-specific IFNγ-secreting CD4+ T-cells in the draining lymph nodes of recovered cattle several months after infection. Indeed, studies of several viral or bacterial diseases indicate that while a pool of effector T-cells is expanded during active infection, only memory cells remains detectable after infection resolution [12]. Both arms of the immune system are, therefore, playing a role in protection against CBPP, the local humoral response relying on secretory IgAs and the cellular response based on CD4+ T-cells of the Th1-like type. Consequently, both represent the main target to be elicited or primed by a vaccine with the induction and expansion of a subset of MmmSC-specific CD4+ memory T-cells as the required goal to develop long-term protection against CBPP.

CBPP being a respiratory disease whose pathogen generally remains restricted to the lungs, a key role for the local humoral response was predictable. Immunoglobulin A is, indeed, the main element of the humoral response to provide protection against microbial antigens entering mucosal surfaces and represents an important first line of defense against invasion of deeper tissues [5]. The IgA response might play a direct role by inhibiting MmmSC growth and colonization and/or blocking some yet unknown virulence factors. A CD4+ T-cell response was also expected since MmmSC is an extracellular pathogen [13]. Extracellular pathogens are generally taken up by antigen presenting cells (APC) and processed in association with the class II molecules of the major histocompatibility complex (MHC) for presentation to the CD4+ T-lymphocytes. CD4+ T-cells should contribute to protection against CBPP mainly through their cytokine network including IFNγ that will enhance phagocytosis and cell killing ability and promote the expression of MHC class II molecules, thus increasing the host cell capacity for antigen processing and epitope presentation to T-lymphocytes [14]. A protective role for the IFNγ-induced IgG2 antibody, only able to promote killing by neutrophils in cattle, is also expected since neutrophil recruitment has been observed in CBPP lung lesions [15], [16], [17]. Besides, CD4+ T-cells play also a key role in facilitating B-cell proliferation and class-switching to give rise to antigen-specific IgAs in mucosal effector sites [5]. Whether Th1 or Th2 cells are beneficial for optimal IgA production is not known as either Th1 or Th2 cells or their combination can support antigen-specific IgA responses [5].

Based on these findings, CBPP vaccines should be able to trigger both a local antibody response and a persistent CD4+, Th1-like T-cell-mediated response to selected MmmSC components.

Immunopathological mechanisms

Understanding the immunopathological aspects of CBPP is also an important step for the rational design of an optimized vaccine. Indeed, stimulating a vaccine-induced protective immune response implies that the host still maintains its ability to mount an efficient anamnestic response when in contact with an MmmSC infection. However, MmmSC appears to possess an efficient mechanism to escape the host immune response, as suggested by its ability to survive in encapsulated lung lesions but also in lymph nodes of infected cattle [18], [19].

To characterize this immunomodulatory property, MmmSC interactions with bovine immune cells were analyzed in vitro. We thus demonstrated that (1) viable MmmSC was able to trigger a time-dependent apoptosis of bovine blood leukocytes whereas heat-killed MmmSC had no such effect [20]; (2) soluble MmmSC-secreted components appeared to play a role in this process [20]; and (3) viable MmmSC strongly depressed the mitogenic activity of ConA on bovine T-cells [21]. This MmmSC-mediated immunosuppressive effect might thus allow the pathogen to escape the host response in vivo by diminishing its ability to mount an immune response to antigenic challenge. Although further studies are necessary to fully unravel this mechanism, these results confirm the importance of taking into account the immunopathological aspects of the disease in the rational development of a vaccine.

Besides this immunosuppressive effect on T-cell responsiveness, MmmSC also appears able to escape the innate and humoral host responses. MmmSC is characterized, indeed, by the presence of a polysaccharide (PS) external pseudo-capsule. Genome sequencing has identified the metabolic pathway leading to the synthesis of the capsular PS [22]. The pseudo-capsule was shown to have a role in virulence and to protect MmmSC from the host antibody response and cell phagocytosis [23], [24].

These data confirm that MmmSC did evolve efficient ways to escape the bovine immune response. These should be controlled to achieve full vaccine efficacy, otherwise the impairment of the host immune system (innate, humoral and cellular) by exposure to MmmSC infection will lead to inefficient vaccine-induced anamnestic responses. To this aim, elucidation of the MmmSC component(s) which mediate these pathological effects will help identifying critical vaccine candidate antigens able to control these mechanisms and thus enhance vaccine efficacy.

Potential MmmSC candidate antigens for a multi-component sub-unit vaccine

Screening of MmmSC components and genetic libraries with sera and bronchoalveolar lavages taken from naturally MmmSC-infected cattle, has already revealed a panel of IgA- and IgG2-specific proteins ([25] Du Plessis, personal communication). Several of these components are membrane lipoproteins (Lpps), which might have a central role in interaction between MmmSC and the host cells and in the inflammatory reaction [25], [26]. Besides the capacity of Lpps to elicit an humoral response, their strong potential to trigger also a cell-mediated response and their major role in the early host response was reported [27], [28]. Some proteins are known to display both B-cell and T-cell epitopes and several vaccines such as those against influenza, malaria or Yersinia pestis are based on this type of antigens [29], [30], [31]. A similar approach to select MmmSC vaccine candidates is currently being undertaken by in vitro assessment of the ability of a panel of antibody-inducing MmmSC antigens to elicit an MmmSC-specific CD4+ Th1-like T-cell response (Dedieu, unpublished data). To this aim, the capacity of four MmmSC Lpps (LppA, LppB, LppC and LppQ) to recall in vitro a cellular immune response from lymphocytes taken from three MmmSC-infected cattle was evaluated. The results showed that LppA was the only MmmSC protein recognized by the MmmSC-primed lymphocytes (Dedieu, submitted). This finding demonstrates that MmmSC-infected cattle, besides their known humoral response against MmmSC LppA, do also develop an LppA-specific T-cell response. The MmmSC LppA, a highly conserved lipoprotein, might be retained as a potential vaccine candidate against CBPP due to its ability to trigger both arms of the host immune response. At this time, however, the protective efficacy of an LppA-triggered immune response has to be further evaluated.

Besides the screening of MmmSC components for their immunogenicity, another approach could target selected MmmSC virulence factors. However, in contrast to other pathogenic bacteria, where virulence relies mainly on adhesins or toxins, no such typical virulence factors have been identified in the genome of the MmmSC reference strain PG1, recently published, nor on the genomes of the 10 other mycoplasmas species that have been completely sequenced [22], [26]. This might be due to their small genome, leading mycoplasmas to drastic savings in genetic resources, reduced to essential functions of life [32]. Mycoplasmas have thus adopted a parasitic mode of life, acquiring macromolecular precursors and high-energy compounds from the host cells.

Genome sequencing and comparative genomic studies have, nevertheless, led to identify some MmmSC pathogenic mechanisms, such as the glycerol transport system leading to the production of cytotoxic H2O2, as well as metabolic pathways such as the phosphotransferase system for sugar transport or the several ATP-binding cassette (ABC) transporters for transfer of components such as phosphate or oligopeptides [22], [32]. Both represent interesting targets for a vaccine-induced antibody response aiming at controlling MmmSC growth and virulence.

Blocking the glycerol transport membrane proteins GtsABC or the membrane-located glycerophosphate oxidase GlpO by specific antibodies resulted, indeed, in inhibition of glycerol uptake or glycerol metabolism and in significant reduction of H2O2 production and cytolytic activity [32], [33]. Such antibody response might thus control MmmSC growth, as glycerol metabolism is presumably required for triglyceride synthesis. Control of the pathogenicity should also be achieved, since H2O2 is considered as a virulence factor [32], [33], [34]. Similarly, an antibody response blocking membrane components of metabolic pathways such as the ABC transporters will likely limit MmmSC growth and thus host infection. Indeed, due to its drastic parasitic mode, MmmSC is dependent on these ABC transporters to scavenge precursors from the host cells. Therefore, a screening of these potential MmmSC components for their ability to elicit an antibody response, and mainly local IgAs, with a functional role in limiting MmmSC growth and virulence, would lead to promising vaccine candidates.

In parallel, a vaccine-induced MmmSC-specific INFγ-secreting CD4+ T-cell response is also needed. Therefore, MmmSC components able to prime such a response also have to be included. These vaccine candidates should be selected by T-cell screening, as described above, either from the selected antibody-inducing MmmSC antigens or from other purified MmmSC proteins.

The final selection of MmmSC vaccine candidates will include an in vitro safety control evaluating pathological effects such as the inflammatory potential which is a major characteristic of this pathogen.

This strategy should hopefully lead to the selection of appropriate antigens that, included in a safe, multi-components sub-unit vaccine, will be able to trigger an immune response that will reduce bacterial load and the clinical outcome of infection, while also controlling the immunopathological mechanisms involved in disease, thus allowing the establishment of efficient innate and vaccine-induced immune responses able to clear the infection with MmmSC while maintaining an MmmSC-specific immune memory.

Relevant vaccine delivery system and route of delivery

Non-replicating antigens, such as purified proteins, are often of low immunogenicity. The major challenge to an effective vaccine, once the protective antigens are identified, lies, therefore, in their formulation and delivery. First, the choice of a relevant delivery system and the addition of immunostimulating molecules will allow enhancing, modulating and orientating the host immune response. Then, of major importance is the selection of an effective system capable of inducing a significant immune response at a mucosal site. Indeed, in CBPP, MmmSC enters and initiates infection at the mucosal respiratory surface, which constitutes the best target for effective vaccination. As systemic immunization is largely ineffective at providing immunity at mucosal surfaces [35], [36], the mucosal surface should be the choice site for immunization. In fact, mucosal immunization can induce both mucosal and systemic immunity [35].

Mucosal immunization benefits from the existence of a “common mucosal immune system” where immunization at some mucosal inductive sites can activate mucosal B- and T-cells to migrate from these sites and home to mucosal effector sites [5], [35]. Accordingly, although intranasal delivery appears the most relevant, the oral route can also be very promising and may offer an inexpensive (no need for needles, professional health care infrastructure required for injectable preparations, or for a cold chain) and convenient way of vaccination (edible vaccine). The effectiveness of this approach is demonstrated by the panel of vaccines currently licensed for mucosal delivery in domestic animals, e.g. vaccines against BHV, coronavirus or rotavirus in cattle [35]. Although these are live vectored vaccines, a large panel of potent adjuvants (bacterial toxins or cell wall components, CpG oligonucleotides, antimicrobial peptides) and delivery systems (biodegradables microspheres, liposomes, novasomes, VLP, ISCOMs, etc.) are now available for mucosal vaccine delivery [35]. Some of them are able to induce both a mucosal antibody and cellular immune response. Studies in domestic animals, nevertheless, are limited and the potential of mucosal immunization with purified antigens remains unexploited in veterinary medicine [35].

Some experiments in cattle have shown the potential of the cholera toxin and E. coli heat-labile (LT) enterotoxin as mucosal immunoadjuvants and the efficacy of oral or intranasal immunization with microparticles [35], [37], [38], [39]. ISCOMs also have great potential in the delivery of antigens to the mucosa [35]. However, in the two experiments conducted in cattle with ISCOM vaccines against CBPP, the vaccines were delivered parenterally and failed to protect the animals against an MmmSC challenge [40], [41]. Another possible reason for this failure is that these vaccines were based on either whole inactivated MmmSC or the cytotoxic LppQ protein. In contrast, oral delivery in pigs of another pathogenic mycoplasma, Mycoplasma hyopneumoniae, included in microspheres led to effective protection [42]. Therefore, these new antigen delivery systems might offer promising approaches for mucosal vaccine delivery in large animals.

Advances in the understanding of the protective immune responses to MmmSC infection are providing promising opportunities for the rational development of improved vaccines. Protection was shown to require MmmSC-specific pulmonary IgAs and activation of the local macrophages through stimulation of MmmSC-specific CD4+ IFNγ-secreting T-cells. Both immunological arms could be effectively targeted by a mucosal vaccine. Indeed, induction of protective secretory IgAs was shown to be efficiently achieved after mucosal antigenic stimulation, such as the targeting of the mucosa-associated lymphoid tissues and local lymph nodes [5]. To enhance the priming of MmmSC-specific IFNγ secreting CD4+ T-cells, a Th1-inducing adjuvant will have to be used with a delivery system allowing the vaccine antigens to be presented by the APC MHC class II pathway. This objective could be achieved using microparticles (biodegradable microspheres, chitosan, liposomes, ISCOM, etc.) which can increase host cell phagocytosis, antigen uptake and MHC class II presentation to CD4+ T-cells. Besides, the ability of these particulate systems to also elicit local IgAs has largely been demonstrated [35], [36]. The potency of these delivery vehicles can be further enhanced by incorporating adjuvants or coating with immunostimulating molecules such as lectins or Toll-like receptor ligands, which would reduce the quantity of antigen needed and thus the cost of the vaccine. Various strategies are already under evaluation to enhance particulate vaccine uptake and efficacy [35].

Selection of the most appropriate system for a CBPP vaccine will depend on the nature of the selected MmmSC antigens and on the immunological properties of the available delivery systems. However, in our context of tropical veterinary diseases, where the final objective is vaccine technology transfer to our African partners for local production, other criteria have to be taken into account, such as thermostability, cost effectiveness, manufacturing process and energy requirement, etc. While these are very stringent criteria, one can hope that due to the rapid progress in this field, it will be achievable in the near future. The final choice will then rely on preliminary evaluation by in vitro assays for efficacy testing in the presence of the MmmSC vaccinal candidates. Unfortunately, no small animal models are relevant for CBPP studies and cattle experiments are very costly as they require high security facilities when dealing with tropical diseases. Therefore, although not fully relevant in terms of mucosal immunity, in vitro cellular assays will help determine the capacity of the delivery systems to be efficiently captured by bovine APCs and their efficacy to trigger DC maturation, epitope presentation and MmmSC-specific CD4+ Th1-like T-cell response from autologous T-cells taken from recovering cattle. Cattle experiments will then permit to evaluate the vaccine-induced mucosal immunity and the protective efficacy against CBPP infection.

Conclusion

The objectives of an efficacious CBPP vaccine would be (1) to limit MmmSC colonization, thus reducing host infection and the strong inflammatory process typical of the disease; (2) to control the MmmSC-induced immunosuppression so as to maintain the responsiveness of the host immune system; and (3) to stimulate an MmmSC-specific immune memory response, responsible, in case of a new antigen encounter, of an anamnestic protective response. To this aim, it was shown that the immune parameters that correlate with protection are the mucosal production of MmmSC-specific IgAs and a persistent MmmSC-specific INFγ-secreting CD4+ T-cell response. Accordingly, the MmmSC vaccine candidates will have to be selected on these criteria.

The advances in our understanding of CBPP and the better characterization of the causing agent, MmmSC, together with the improved knowledge in mucosal immunology and vaccine delivery systems, offer very promising approaches for the development of efficient long-term protective mucosal vaccines to prevent CBPP.

Mucosal sub-unit vaccines against MmmSC infection will hopefully limit the shedding of MmmSC, thereby reducing the pathogen load in the environment and consequently reducing the rate of herd infection and transmission of CBPP through the herd. It is hoped, therefore, that the availability of optimized vaccines, coupled with an effective surveillance system and stricter control of animal movements will allow one to control and even eventually to eradicate CBPP from the African continent.