Computational analysis of human enterovirus 71 capsid proteins with the aim of identifying target antigenic sites that induce both humoral and cellular immune responses

Human Enterovirus (HEV-71) is the etiological agent of human hand, foot and mouth disease (HFMD). HEV-71 is typically an asymptomatic virus but recently it has accounted for severe neurological and CNS complications in children, which have resulted in fatalities mostly in the Asian Pacific region. Enterovirus71, a member of the genus Enterovirus of the family Picornaviridae, is characterized by a high degree of genetic variability and polymorphism between the serotypes, which makes it difficult to control with current vaccination strategies. To be effective against HEV-71, the protective immunity needs to be mediated by both antibody and Th (T helper) responses which means that the antigen has to be presented to T cells bound in the groove of an MHC-II molecule and to a lesser extent, the MHC-I molecules. At the start of this project there were no HEV-71 vaccine candidates that could stimulate both T and B cells or induce broad immunity, and none were licensed for use in humans. This project aimed to establish whether short peptide sequences based on HEV-71 capsid proteins would serve as effective vaccine candidates, which would evoke both a protective cellular and humoral immune response. The rationale for this approach was that artificial peptide based epitopes should be highly suitable as vaccine candidates for use in a pediatric population. One focus of this dissertation was to identify and evaluate whether computationally derived HEV-71 epitopes would induce a robust cellular and humoral response in-vivo. First, a screen for sequences with the potential to serve as unique vaccine candidates was performed using an in-silico sequence and structural approach based on peptide location and hydrophobicity. The in-silico analysis identified three potential antibody candidate sequences. B-cell epitopes were selectively identified based on homology modeling of the HEV-71 protein regions were then synthesized incorporating a promiscuous T-cell epitope sequence from the closely related poliovirus. The efficacy of these peptide epitopes was then tested using a variety of in-vitro and in-vivo tests. Mice inoculated with these peptide constructs were capable of generating cross-reacting antibodies against HEV-71, and other picornaviruses. MP-26M mice pups born to mothers vaccinated during pregnancy showed a marginal extension of survival times and a lower incidence of myositis as compared with controls, when fatally challenged with a mouse adapted strain of HEV-71. Since the induction of both humoral and functional cellular immune responses are imperative for viral clearance and the initiation of cognate anti-HEV-71 CD8+ help, the identification of T-cell epitopes was also of paramount importance in designing and optimizing the in-silico search algorithms for vaccine candidates. Promiscuous CD4+ and CD8+ epitopes from the HEV-71 virion (VP1 and VP3) proteins were identified and included the complete set of H2 class I (H2-Kd, H2-Ld and H2-Dd) and class II (I-Ed and I-Ad) molecules. The study assessed the ability of these epitopes to induce broad T cell responses in mice with differing HLA class II molecules, namely; H2d of the BALB/c mouse, the (H2b) of the C57BL/6 mice, and the (H2k) of the CBA mice. As expected, mice varied in their IFN-γ responses in the context of their respective MHC binding affinity. Two out of four CD8+ predicted epitopes produced a significant IFN-γ response whereas three out of four CD4+ predicted epitopes produced significant IFN-γ responses in all mice. Furthermore, mice inoculated with the CD4+ epitopes were able to respond to CD8+ pulsed peptide, indicating that may be a shared structural core motif in MHC peptide recognition. The ability to induce this response in conserved CD4+ and CD8+ T cell epitopes, demonstrates the feasibility of a broad immunological response in a diverse population as applicable to humans. Human immune cytokine responses were also monitored. Although human CD4+ responses to HEV-71 infections have been documented, computational identification and validation of T-cell epitopes has only recently started to be used. Computational methods for predicting and modeling epitopes have improved but there still remains an Achilles heel in CD4+ epitope prediction algorithms due to the degree of polymorphism on the MHC II. For broad population recognition, an epitope must contain conservation against considerable MHC diversity. To test this theory, the work on mice was expanded to include analysis of blood from human participants. The results indicate that our HEV-71 immune-dominant peptides were presented to T-cells and three peptides induced broad immune responses to peptide-stimulated whole blood in thirteen out of twenty volunteers representing multiple human HLA-DR and HLA-DQ types. Two peptides that stimulated the murine (H-2b) and (H-2d) alleles were also able to induce the production of IFN-γ in white blood cells from human participants. Overall, these results make the case that a computational approach is an effective and time saving step in the development of human vaccine candidates and the results obtained with the identified candidates has provided a broader framework for understanding the complexities of MHC epitope interactions and improving peptide based vaccines.