Strategies in the development of vaccines to prevent infections with group A streptococcus

Abstract

There has long been interest and demand for the development of a vaccine to prevent infections caused by the Gram-positive organism group A streptococcus. Despite numerous efforts utilizing advanced approaches such as genomics, proteomics and bio-informatics, there is currently no vaccine. Here we review various strategies employed to achieve this goal. We also discuss the approach that we have pursued, a non-host reactive, conformationally constrained minimal B cell epitope from within the C-repeat region of M-protein, and the potential limitations in moving forward.

Keywords: :

• group A streptococcus
• J8-DT
• vaccination
• rheumatic fever
• rheumatic heart disease

Introduction

The history of the development of vaccines to prevent group A streptococcus (GAS) infections is one with many promises and many disappointments. It is one where, nevertheless, proponents have passionately progressed the needs for such a vaccine and where many novel approaches and strategies have been tested. Today, however, there is no vaccine.

Industry has flirted with the idea of a vaccine as they have weighed the plusses, from a commercial perspective, of a vaccine to prevent streptococcal tonsillitis, with the risks that any vaccine may actually cause disease. Rheumatic fever (RF) and rheumatic heart disease (RHD) are autoimmune conditions that can follow infections with GAS (see below) and immunizing children with streptococcal antigens and adjuvants may thus be a bridge too far for industry. Still, the need exists. A conservative estimate of the deaths due to GAS infections and the autoimmune complications that can follow, principally RHD, is 500 000 per year.¹ However, in many countries, RF is poorly diagnosed or ignored as a possible cause of illness. In many African countries, for example, a child presenting with fever and lethargy is considered to have malaria and is usually treated as such since blood parasitemia is highly prevalent and often not considered informative to the diagnosis. RF can thus be missed and the consequences can be dire. Thus, the true incidence of GAS pathology may be much higher.

GAS causes death principally by deep tissue sepsis and rheumatic heart disease. However, it is estimated that there are over 600 million cases of non-life threatening pharyngitis annually and over 100 million cases of pyoderma annually. If untreated, pharyngitis and possibly also pyoderma,² can progress to RF and RHD-conditions where GAS cannot be found in the affected tissues and where pathology occurs as a result of mis-directed immune responses that originated from the preceding GAS infection but which now are recognizing human tissue, such as heart valves. It is thus critical that any vaccine approach minimizes the possibility of inducing auto-reactive B and T cell responses. This is an obvious impediment to vaccine research, particularly since in an early study using a crude antigen prepared from GAS organisms, three “at-risk” individuals developed RHD.³ However, this is in contrast to other studies where GAS was administered with no reported serious adverse events.^4,5

In 1979, the FDA regulated the exclusion of GAS from vaccine products after reviewing and considering the findings of the independent advisory Panel on “Review of Bacterial Vaccines and Bacterial Antigens.”⁶ This exclusion was lifted in 2006.⁷

A further significant impediment to vaccine development is the fact that there is significant antigenic strain variation with GAS. Immunity to any particular strain is thought to be due to the development of antibodies to the N-terminal segment of the surface M protein, a virulence factor for the organism.⁸ These strain-specific antibodies can neutralize the pathogenic effects of the M protein, but they can only act against the specific strain that induced them. Unfortunately, there are over 180 different emm types, based on nucleotide sequences of the M protein-encoding gene.^9,10 This enormous sequence diversity has led to two separate strategies—the current leading strategy of combining multiple common serotypic determinants into recombinant polypeptides^11-13 and other strategies of using either conserved regions of the M protein or conserved non M proteins or fragments thereof as vaccine candidates.^14-17

While the M protein, being a major virulence factor, is an obvious target for vaccine consideration, non-M protein vaccine candidates have also been identified (such as virulence factors including proteases^18,19 or adhesins²⁰) or by ‘reverse vaccinology’ approaches of using genomics, proteomics and immunomics to identify conserved, surface expressed molecules recognized by serum from infected individuals.^21-23

Discovery of a Cryptic Vaccine Epitope

Our approach has been to identify a concealed epitope on the M-protein that is poorly recognized in the human host but which can nevertheless lead to the induction of antibodies that can recognize and destroy GAS. We believe that an advantage of our approach is that hidden or ‘cryptic’ epitopes are highly conserved, as they are not under immune pressure. The surprising factor is, though, that although cryptic epitopes may not be recognized as a result of infection, antibodies induced against cryptic epitopes in isolation may recognize the organism. The cryptic epitope that has most interested us is contained within a 28-amino acid peptide referred to as “J8” as described later. The central 12 amino acids of J8 are GAS-derived (from the “C3” repeat region of the M protein—a region almost buried in the capsule of the organism). The flanking sequences are designed to enable the 12 amino acid critical GAS sequence to maintain its α-helical folding, which is required for its antigenicity. The strategy to identify this cryptic epitope may be applicable to vaccine development for other organisms.^24,25

Our specific approach was to focus on the carboxyl-terminal region of the M protein, a major virulence factor for the organism. There was evidence that this relatively large but highly conserved carboxyl-terminal segment of the protein was capable of inducing protection in a mucosal challenge model.²⁶ It was curious, however, that antibodies to the region were shown by others to be non-opsonic,^27,28 in contrast to antibodies to the serotypic N-terminal segment. Our initial goal was to ask whether antibodies to much smaller peptide fragments within this region might be opsonic. We thus produced a panel of nine overlapping 20-mer synthetic peptides spanning the C3 repeat region and asked whether antibodies produced in mice to these individual peptides would kill GAS in vitro in the presence of human neutrophils and complement. This was a standard “Lancefield” indirect opsonophagocytosis assay²⁹ with one important exception, that being that we used ‘stationary’ phase organisms with a pre-incubation step and not ‘log’ phase organisms.³⁰ One peptide in particular, referred to as P145, was capable of inducing antibodies that could opsonize stationary phase organisms. When log phase organisms were used we observed significantly less opsonic activity possibly because of enhanced expression of hyaluronic acid (HA) capsules in log phase organism.³¹ Moreover, we observed that antibodies to P145 could opsonize multiple serotypic strains of GAS in stationary phase,^17,32 in keeping with the conserved nature of this region. This peptide then formed the basis of our on-going approach to vaccine development.

We observed that the peptide was cryptic. Thus only 30% of Aboriginal children living in a highly streptococcal-endemic area of northern Australia had P145-specific antibodies, although this rose to 90% by adulthood.³³ It thus seemed that many years of streptococcal exposure were required for the development of this immune response. However, once developed, these human antibodies were capable of killing the organisms similarly to the murine antibodies induced by peptide vaccination. More recently we have shown in the mouse model that sequential exposure to multiple GAS strains does not induce these antibodies (M Pandey et al., unpublished). Given the very poor immunogenicity of P145 in its native state, it was not surprising to us that P145 is highly conserved. At the time of our initial studies the data suggested that it was completely conserved within different strains although recent examination of hundreds of different strains has shown that there are minor variants of P145,³⁴ which are cross-reactive.³⁵

Vaccine Safety

As mentioned, GAS infection, if untreated, can lead to the serious autoimmune disease, rheumatic fever (RF), in a small subset of the population (about 3%).³⁶ Evidences suggest that streptococcal specific antibodies and T cells may contribute to this pathology.^37-40 In the case of RHD it has been suggested that antibodies to the streptococcal sugars damage the endothelial surface of the heart valves enabling the ingress of streptococcal-specific immune cells, including CD4 and CD8 T cells^41-43 which are activated by cross-reactive cardiac antigens leading to inflammation and scarring.^40,44 T cells taken from the excised heart valves of acute RF patients have been analyzed for specificity and some have been shown to respond to epitopes from the M protein,^45,46 raising potential concerns that vaccines containing those epitopes may induce pathology. However, there is no reason to believe that any streptococcal antigen could not be involved in pathogenesis of RHD assuming it was cross-reactive with heart epitopes. Thus, safety concerns span all streptococcal vaccine candidates. However, while heart cross-reactive T cell clones have been generated from the peripheral blood of RHD patients, clones have also been generated from the blood of individuals with no evidence of heart pathology.⁴⁷ The functionality of the T cells is likely to be as important as their specificity and so far only limited analyses have been performed on T cells isolated from RHD valve specimens.⁴⁸

The above observations raise concerns regarding the safety of any streptococcal vaccine. The approach that we undertook to minimize the risk of a vaccine based on P145 inducing pathology was to define a minimal B cell epitope (capable of inducing protective antibodies) while not stimulating T cells. We defined a 12-mer epitope within P145, referred to as J8i, that was a B cell epitope but did not stimulate T cells in the different mouse strains that we examined.¹⁷ Conversely, other minimal epitopes within P145, principally J2i, stimulated T cells but were not recognized by monoclonal antibodies generated against P145. J8i thus became the focus of vaccine design. However, unfortunately J8i was too small to maintain its helical structure which was required for its antigenicity and as such a technology was developed to fold J8i as a helix⁴⁹ resulting in J8, a 28-mer synthetic peptide of which only the central 12 amino acids (J8i) are derived from GAS sequences, with the flanking sequences derived from a non-streptococcal peptide. The sequence and structure of J8 is illustrated in Figure 1.

Figure 1. Overlapping peptides synthesized from the parent peptide P145. Sequences derived from the M protein of group A streptococcus are underlined. Flanking sequences (not underlined) are derived from the yeast DNA binding protein GCN4 (A). A graphic representation of vaccine candidate J8-DT with J8 peptide conjugated to diphtheria toxoid (DT) molecule. A ratio of DT: J8 (1:8–14 molecules) (B).

To render J8 immunogenic in an outbred population, it was then conjugated to diphtheria toxoid, giving the vaccine the name, “J8-DT”. This vaccine generates specific memory B cells (MBC) and long-lasting Ab responses⁵⁰ which prevent streptococcal sepsis in mice.⁵¹ T cell help is required for activation of MBC but can be provided by naive T cells responding directly to GAS at the time of infection (Fig. 2). Thus individuals whose T cells do not recognize the short synthetic peptide in the vaccine will be able to generate a protective and rapid memory Ab response at the time of infection. The vaccine can also prevent GAS pyoderma in an animal model that we developed to closely mimic human GAS pyoderma (M Pandey, unpublished). J8, when embedded in protein vesicles (without DT)⁵² or presented as lipopeptides⁵³ can also induce salivary IgA and protection from intranasal GAS challenge. Using the rat valvulitis model referred to above, J8 has been shown not to induce pathology (M Batzloff et al., unpublished).

Figure 2. Adoptive transfer and assessment of protective efficacy of J8-DT memory and naïve B and/or T cells: To investigate synergistic effect of memory and naïve B and T cells in protection against GAS, SCID mice (n = 5 per group) were transfused with combinations of B and T cells from J8-DT-immunized or naïve BALB/c mice. Cohorts of mice were transfused with J8-DT B cells/naive T cells, naïve B/J8-DT T cells, J8-DT B and T cells or naive B and T cells, J8-DT or naïve splenocytes or J8-DT B/T cells. Immunized mice received the vaccine 14–16 weeks before their cells were transferred. Twenty-four h post-adoptive transfer, the mice received an IV infection with 50 000 CFU of GAS M1 strain. The bacterial bio-burden in the blood was monitored and bacteremia on day 12 post-infection is shown. Results are shown as means ± standard errors of the means for at least 5 mice in each group. Significance determined by two-way analysis of variance (ANOVA) where **p < 0.01, ***p < 0.001. (Fig. adapted with permission from Pandey et al.; 2013⁵⁰)

Way Forward and Summary

As mentioned, there has not been consistent industry support for the development of GAS vaccines. Thus development will depend on funding from Government and NGOs. It is thus pleasing to note that the New Zealand and Australian Governments have pledged significant funding to complete pre-clinical evaluation of three leading candidates (M protein based and non-M protein based), with the goal of taking the most promising candidates through to clinical trials. A major motivating force for these Governments is the very high incidence of streptococcal-related pathology within the indigenous populations of these two countries.

In summary, attempts to develop a vaccine to prevent GAS infections are not new, but as more is learnt about strain diversity and as novel concepts are developed to identify new antigens and present them to the immune system in a safe manner it is likely that an efficacious vaccine (or vaccines) will be developed in the near future.

Acknowledgments

This work was supported by the National Heart Foundation of Australia, National Health and Medical Research Council (Australia), the Prince Charles Hospital foundation, the Cooperative Research Centre for vaccine Technology, the National Institute of Health (USA) and the Cooperative Research Centre for Aboriginal Health.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.