Vaccines Against Mycobacterium tuberculosis: An Overview from Preclinical Animal Studies to the Clinic

More than a decade ago the World Health Organization (WHO) declared tuberculosis (TB) a global emergency and called on the biomedical community to strengthen its efforts to combat this scourge. The WHO predicts that by 2020 almost one billion people will be infected, with 35 million dying from the disease if research for new approaches to the management of this disease is unsuccessful (1). Designing a better TB vaccine is a high priority research goal. This chapter will review the various strategies currently being used to prevent and treat TB. In spite of the numerous new vaccine candidates in clinical trials, and several others in the preclinical pipeline, no clear TB vaccine development strategy has emerged.


Introduction
More than a decade ago the World Health Organization (WHO) declared tuberculosis (TB) a global emergency and called on the biomedical community to strengthen its efforts to combat this scourge. The WHO predicts that by 2020 almost one billion people will be infected, with 35 million dying from the disease if research for new approaches to the management of this disease is unsuccessful (1). Designing a better TB vaccine is a high priority research goal. This chapter will review the various strategies currently being used to prevent and treat TB. In spite of the numerous new vaccine candidates in clinical trials, and several others in the preclinical pipeline, no clear TB vaccine development strategy has emerged. Despite TB control programs, Mycobacterium tuberculosis (Mtb), a facultative bacterial pathogen, remains the most common cause of infectious disease-related mortality for more than 80 years since 1928. BCG lacks the genomic 'Region of Difference' (RD1) which encodes the ESX-1 secretion system, including the immunodominant 6-kDa Mtb antigen ESAT-6, included in the Hybrid 1 (ESAT-6/Ag85) vaccine (described in more detail in a later section of this chapter) and in IFN- release assays (IGRA's) used to diagnose Mtb (16,17). The overriding dogma is that BCG protects against primary childhood TB, but its role in consistently protecting against adult pulmonary disease is minimal (18). Indeed, the efficacy of BCG in several field trials has been variable (19). The suggested reasons for the variability observed include differences in the BCG strains -resulting from inconsistent laboratory culture conditions which caused gene deletions or attenuated organisms (20), poor handling of the vaccine, doses and vaccination schedules in the various field trials (21), interference from environmental mycobacteria (22)(23)(24), and poor nutrition or genetic variability in the populations immunized (25,26). Several analyses have identified genetic changes within some BCG substrains such as in the phoP-phoR system that has occurred along the way since BCG Pasteur was first derived.
Except in cases where infants are HIV-seropositive, BCG is considered safe. This has led to development of other vaccines that either enhance the immune responses resulting from BCG immunization, for example by insertion of specific genes present in virulent M. tuberculosis but which have been lost in the avirulent BCG vaccine -the recombinant forms of BCG (rBCG) -or, more broadly, are capable of boosting the effects of BCG. Recent studies have demonstrated that the new rBCG vaccines are more immunogenic, inducing effector and memory T cells, however one potential concern is that many of these rBCGs encode antigens such as Ag85A, CFP-10 etc. that are immunodominant. Recent data suggest that these antigens are highly conserved and are used by the bacteria as a ploy to cause damage in the lungs resulting in escape of the mycobacteria bacilli and increased transmission. It is important to demonstrate whether the new rBCGs can protect against clinical strains. Furthermore, because BCG is designed to be administered only once, none of the rBCG strategies are likely to yield a successful vaccine superior to what we have now.
Vaccine approaches currently in clinical trials also include altered forms of BCG to increase the effectiveness of the treatment. One of the vaccines, rBCG30, is an engineered form of BCG (rBCG) that over expresses Ag85B (97). It has shown much greater efficacy than the parental Tice BCG vaccine, perhaps due to loss of virulence in the current BCG vaccines, and was shown to increase Ag85B-specific T cell proliferation and IFN- responses in humans (97). Another rBCG in human clinical trials is a rBCG that is a urease-deficient mutant that expresses the lysteriolysin O gene from Listeria monocytogenes (98). Using this approach the vaccine increases phagosomal acidification in the absence of the ureC enzyme, while expressing the lysteriolysin protein, Hly, which requires an acidic pH within the phagosome in order to damage/perforate the phagosomal membrane. This process allows the release of antigen into the cytoplasm and induces macrophage apoptosis, leading to enhanced CD8 + T cell presentation through a cross-priming strategy. Other whole virus vaccine approaches have seen some success against TB. One, based on a recombinant modified vaccinia virus Ankara (MVA) vaccine which expresses the Mtb protein Ag85A, is currently in clinical trials (99). However, the complex nature of TB infections may very well require multiple weapons in our armamentarium. These may include not only the use of multiple Mtb antigens but also vaccines based on other adjuvant and delivery platforms.
A post-exposure vaccine, to be used in healthy individuals infected with Mtb or those recently exposed to MDR-TB, could also reduce the probability of going on to develop TB disease. It could work by limiting bacteria that cause TB or MDR-TB, that are residing in a dormant state, by preventing reactivation and/or by reducing the chance of reinfection by exogenous Mtb. Finally, a therapeutic vaccine could function alone, or alongside antibiotic regimens, for individuals with active TB disease and could potentially shorten the treatment period.

Immune responses required for development of a successful TB vaccine…
Advances in our knowledge of resistance to Mtb have emerged since the pioneering work of Mackaness (1960's, 1070's) who demonstrated a dependence on cellular immunity against mycobacterial infection (100,101). Another key advancement to the development of vaccines against Mtb was made by Orme and Collins (1980's), who were the first to show that transfer of immunity against Mtb could be achieved with antigen-specific CD4 and CD8 T cells, and that metabolically active mycobacteria secreted key immunologically relevant antigens (102)(103)(104)(105)(106). A major new idea in the mid-1980's, that has shaped the development of vaccines against many different pathogens, was that of Mosmann with the discovery that there were two types of helper CD4 T cells: Thelper 1 and Thelper 2 cells, that secrete either IFN or IL-4 respectively (among other cytokines) (107). More recently, Sallusto et al. have defined memory T cell subsets which can be functionally separated based on their surface receptors, which further advance testing the capability of vaccine induction of long-lived immune responses (108,109). Although our understanding of an effective immune response against Mtb is far from complete, some fundamentals have been identified, resulting in a number of TB vaccines that are now being tested in humans. Several of these advances in our knowledge of the host's resistance to Mtb are discussed in the remainder of this chapter.
Mycobacteria bacilli usually enter the host through aerosol droplets of 1-3 M inhaled to the lung alveoli. Some bacilli remain in the lungs and evade adaptive immunity to persist in the lungs, often for the lifetime of the host, and some are transported to draining lymph nodes where dendritic cells (DC) prime T lymphocytes. Mtb undergoes an initial period of uninhibited growth within non-activated host macrophages (110). Cell mediated immunity (CMI) characterized by the expansion of antigen-specific T-lymphocytes that attract monocytes/macrophages to inhibit bacillary growth through the production of cytokines, plays a key role in the control of TB. Persistence of Mtb inside of mononuclear phagocytes and DCs during all stages of infection can occur via many mechanisms including downregulating major histocompatibility complex (MHC) class II expression or presentation www.intechopen.com

Vaccines Against Mycobacterium tuberculosis:
An Overview from Preclinical Animal Studies to the Clinic 373 (111), neutralizing the phagosomal pH, interference with autophagy, and by inducing the production of immunosuppressive cytokines such as interleukin (IL)-10 and tumor growth factor beta (TGF-) (112)(113)(114)(115). Mtb can also inhibit apoptosis through prostaglandin production (116) and can invade the cytosolic compartment (117). Recent data also showed that of the large number of CD4+ effector T cells recruited to the lungs of infected mice, few are stimulated to produce IFN- (118).
The hallmark of CMI to Mtb infection is the formation of solid granulomas from aggregates of mononuclear phagocytes and polymorphonuclear granulocytes in the lung with a center of infected macrophages surrounded by a marginal zone of lymphocytes (119,120). The protective role of granulomas is confinement of bacilli in a space that is lacking in vascularity and alveolar air, preventing both replication and dissemination to other sites. CD4 + T cells traffic to the lung within 7-14 days following infection and produce IFN- (142,143). Depletion of CD4 + T cells prior to Mtb infection leads to increased bacterial burden and shortened survival (138) and depletion of this subset in latently infected animals leads to rapid reactivation (144). In sublethally-irradiated mice, passive transfer of CD4 + T cells mediates reduced susceptibility to Mtb infection (145). In contrast, CD4-and MHC Class IIdeficient mice are extremely susceptible to Mtb. Finally, clinical conditions that impair CD4 + T cell immunity, such as HIV infection, dramatically increase the likelihood of developing active TB.
Mice deficient in IFN-, an effector cytokine which defines Th1-type CD4 + T cells, are highly susceptible to Mtb infection (127,146). These mice fail to produce nitric oxide (NO) synthase (127) and develop a disseminated form of disease, characterized by irregular granulomas and necrotic areas. Patients in whom the gene for the IFN- receptor is mutated are prone to infection with atypical mycobacteria (147). Strong Th1-type, antigen-specific IFN--secreting T cells are found in peripheral blood mononuclear cells (PBMC) from healthy individuals with latent TB infections (LTBI), but are diminished in individuals with pulmonary TB (148,149). Recent results also indicate that CD4+ effector T cells are activated at suboptimal frequencies in tuberculosis, and that increasing effector T cell activation in the lungs by providing one or more epitope peptides may be a successful strategy for TB therapy (150).
The protective role of TNF in the immune response to Mtb was demonstrated in mice with defects in genes for TNF (151,152). Its critical role for humans was also revealed by the occurrence of reactivation TB in rheumatoid arthritis patients who received long-term therapy with anti-TNF antibodies (153). Recently, both IL-23 and IL-17 were shown to be essential in the establishment of protective pulmonary CD4+ T cell responses, along with the concurrent expression of the chemokines CXCL9, CXCL10 and CXCL11 (154,155).
Studies in mice and humans support an important role of CD8 + T cells in TB immunity, particularly during LTBI. Adoptive transfer or in vivo depletion of CD8 + cells demonstrated that CD8+ cells could confer protection against subsequent Mtb challenge, although the effects were less pronounced than those seen with CD4 + T cells (156)(157)(158). Mtb can egress into the cytosolic compartment of infected DCs resulting in direct loading of MHC class I (117). Cross-priming, which involves apoptosis of macrophages infected with Mtb, uptake of vesicles carrying Mtb antigens by nearby DC, and antigen presentation of the vesicular antigens by MHC I to CD8 is an additional mechanism by which CD8+ T cells are stimulated (159). Mice deficient in class I processing and presentation, including deficiencies in 2 microglobulin (160,161), TAP1 (162), CD8 , or Class Ia (K b-/-/D b-/-) (163), are all more susceptible to Mtb infection than wild-type animals. In humans, Mtb-specific CD8 + T cells have been identified in Mtb-infected individuals and include CD8 + T cells that are classically (164)(165)(166)(167)(168)(169), non-classically (170,171), and CD1 restricted (172,173).

Designing a sub-unit vaccine from start to finish…
This section highlights the development of a new subunit vaccine, ID93/GLA-SE, and briefly discusses the other human TB vaccine candidates in the pipeline (see Table I).
Preclinical studies with a new TB subunit vaccine, ID93/GLA-SE, have been conducted and this vaccine is ready for testing in Phase I human clinical studies. This vaccine now joins 14 others, which are currently being tested in humans (Table I). The selection of the proteins for ID93 involved the generation of an Mtb protein library based on H37Rv proteins that were within the known immunogenic EsX and PE/PPE classes, between 6 and 70 kDa and with low homology with the human genome (less than 30%) (174). A comprehensive analysis was then performed on over 100 potential candidate antigens selected based on genome mining and expression as recombinant proteins. These candidate antigens were then down-selected based on IFN- production from human PBMCs in patients that were PPD(+) and which were non-responsive in PPD(-) patient samples. In combination with the TLR9 agonist, CpG ODN 1826, the vaccine candidates were then tested for efficacy in the C57BL/6 mouse aerosol model of Mtb infection. The ID93 fusion protein consists of four selected Mtb proteins: Rv3619, Rv1813, Rv3620, and Rv2608 (the cumulative molecular weights of each individual protein define the "93" in ID93). Three of the proteins are associated with Mtb virulence (Rv2608, Rv3619, and Rv3620) and one with latency (Rv1813). Rv2608 is a member of the PE/PPE family, Rv3619 and 3620 are in the EsX family of proteins and Rv1813 is expressed under hypoxic conditions (174). Similar to other fusion proteins, including Mtb72f, Ag85B-ESAT6, Ag85B-TB10 and H56, the fusion of more than one Mtb antigen leads to increased vaccine efficacy. Another similarity of these subunit vaccines is the need for an adjuvant to elicit maximum efficacy.
The adjuvant selected for use with the ID93 vaccine is a synthetic toll-like receptor (TLR4) agonist called glucopyranosyl lipid adjuvant (or GLA). This molecule has been extensively characterized in many biological systems, including mice, guinea pigs, ferrets (unpublished results), hamsters, non-human primates (NHPs) and humans (52,175,176). Early on, the Mtb72F subunit vaccine, in Phase II human clinical trials, included AS02A as its adjuvant. AS02A consists of a biological TLR4 agonist called monophosphoryl lipid A (MPL), derived from Salmonella minnesota mixed with QS21 and an oil-in-water formulation (177).
Other TB vaccine candidates currently in clinical trials include four different categories of vaccines: a) recombinant protein vaccines; b) recombinant live vaccines; c) viral vectored vaccines; and d) whole cell, inactivated or disrupted mycobacterial vaccines (Table 1). The recombinant subunit vaccines will be briefly described below. The M72 (Mtb72F) + AS01 (or AS02A) vaccine was originally developed by Corixa and the Infectious Disease Research Institute (Seattle, WA) and clinical trials are currently being sponsored by GlaxoSmithKline (GSK) and Aeras. This vaccine is a fusion of tandomly linked proteins, Mtb32(C), Mtb39, and Mtb32(N) which showed efficacy in mice, guinea pigs, and NHPs (179)(180)(181) and is currently being evaluated in humans. This vaccine includes an AS01 adjuvant (GSK), which comprises the TLR4 agonist, monophosphoryl lipid A (MPL), QS21 and liposomes. In the first phase I clinical trial, Mtb72F combined with the AS02A adjuvant, which includes MPL, QS21, and an oil-in-water emulsion, the vaccine was locally reactogenic but the adverse events were mostly mild and transient and thus had an acceptable tolerability in humans (177). Immunologically, three doses of the Mtb72F/AS02A vaccine (given at 0, 1 and 2 months) induces both humoral and cellular responses in healthy PPD-negative adults (18-40 years of age); IL-2 and IFN-is elicited in PBMCs by ELISPOT and increased antigen-specific CD4+ T cells expressing CD40L, IL-2, TNF- and IFN-by intracellular cytokine staining (ICS) are also induced.
Another subunit vaccine in development by the same group that developed the Hybrid-1 vaccine is the H56 vaccine which includes a fusion of Hybrid 1 and a latency-associated protein, Rv2660c, which is activated during hypoxic conditions (50). The H56 vaccine, formulated in CAF01, shows a 10-fold reduction in lung bacterial load in the mouse model in a head-to-head comparison with their precursor subunit vaccine, the Hybrid 1 vaccine, containing only Ag85B and ESAT6. In addition, the authors demonstrate that the H56 vaccine is capable of protecting against reactivation when tested after Mtb exposure in a modified Cornell mouse model. HyVac4/AERAS-404 combined with IC31 is also in clinical trials, and includes a fusion of the Mtb antigens Ag85B and TB10.4. Replacement of the ESAT-6 protein with TB10.4 in this vaccine, conserves the use of ESAT-6 for diagnostic purposes (16,190). This vaccine induces polyfunctional CD4 T cells, which express IFN-, TNF- and IL-2, correlating with protective efficacy in the mouse model against Mtb (191) and guinea pig model using a BCG prime/subunit boost strategy (192).

Conclusion
Today, an ambitious portfolio of novel vaccines, drug regimens, and diagnostic tools for TB is being supported by various research funding agencies. Mathematical modeling of TB to evaluate the potential benefits of novel interventions under development and those not yet in the portfolio suggest that: neonatal vaccination with an effective portfolio vaccine would decrease TB incidence by 39% to 52% by 2050, while drug regimens that shorten treatment duration and are efficacious against drug-resistant strains could reduce incidence by 10-27%. Clearly, TB elimination will require one or more effective vaccines. Importantly, new vaccines should have the potential to be effective against clinical strains representing all the major geographical regions.