Three epitope-mapping procedures were used to identify B-cell epitopes on Betanodaviruses: neutralisation escape mutant sequence analysis, phage display, and pepscan. Betanodaviruses have emerged as major pathogens of marine fish. These viruses are the aetiological agents of a disease referred to as viral nervous necrosis (VNN), which affects many species of fish that are economically valuable to the aquaculture industry. The identification of betanodavirus B-cell epitopes will facilitate the rational development of vaccines to counter VNN. A panel of mouse monoclonal antibodies (MAbs) was produced using hybridoma methodology for use in each of the epitope mapping procedures. These antibodies were characterised in Western blotting, ELISA, and virus neutralisation tests. Rabbit polyclonal sera, and serum samples from nodavirus-infected fish were also used for pepscan analyses. Attempts to produce betanodavirus neutralisation escape mutants, using plaque assay or limiting dilution based methods, were not successful. Two phage libraries expressing random peptides of seven (Ph.D.7™) or twelve (Ph.D.12™) amino acids in length as fusions to the coat protein were used to identify the ligands recognised by MAbs directed against betanodavirus. Neither of these phage libraries yielded conclusive results. Phage clones containing tandem inserts were obtained after MAb selection from library Ph.D.7™. Extensive screening and nucleotide sequence analysis of MAb-selected clones from library Ph.D.12™) failed to yield a consensus sequence. Pepscan analyses were performed using the recently developed suspension array technology (SAT). This was used to map the recognition sites of MAbs and serum samples onto a panel of overlapping synthetic peptides (12mers) that mimicked the betanodavirus coat protein. The results of pepscan analyses required careful interpretation due to the binding of antibodies and serum samples to multiple peptides. However, three regions of the nodavirus coat protein were identified as containing B-cell epitopes: amino acids 1-50, 141-162, and 181-212. These results are discussed in relation to previous studies of immune responses to betanodaviruses, and to the future development of betanodavirus vaccines and diagnostic reagents.

Investigations detailed in this thesis constitute a part of the continuing programme of research undertaken in our laboratory on the riboflavin carrier protein (RCP) with partic­ular reference to identification and synthesis of neutralizing antigenic determinants, design of relevant epitope mimetics with improved immunogenic characteristics and relationship between their secondary structures and immunological properties. The riboflavin carrier protein is elaborated as a reproductive stratagem to ensure ade­quate vitamin deposition in the developing oocyte in the chickens. The protein is scrupu­lously conserved through evolution in terms of physico chemical and immunological char­acteristics from fish through birds to mammals, including primates. In rodents and sub­human primates immunization with the heteroantigen viz., chicken egg white RCP leads to functional neutralization of the endogenous maternal protein resulting in curtailment of early pregnancy. Thus, the crucial role of RCP in maintenance of pregnancy is established and the protein identified as a potential candidate vaccine for immimocontraception. Fur­ther studies with the reduced and carboxymethylated (RCM) RCP as the immunogen re­veal that antibodies induced by RCM-RCP are equally effective in bioneutralization of the endogenous protein. So it can be surmised that the native folded structure of RCP is not obligatory for eliciting bioneutralizing antibodies. In an attempt to identify functionally relevant regions of the protein, a panel of monoclonal antibodies (MAbs) have been raised and characterized. One of the MAbs viz., 6J32Ci2 could bring about early fetal resorp-tion when injected to mice with confirmed pregnancies. These results prompted a detail molecular immunological approach to understand underlying mechanisms. The principal aims of the present investigations include: (1) identification of neutralizing epitopes; (2) synthesis of peptidyl sequences incorporating these determinants; (3) an understanding of the structure, antigenic and immunogenic characteristics of these peptides; (4) correlation of conformational and antigenic characteristics; (5) rational design and synthesis of peptide analogs with greater propensity to assume predicted secondary structures; (6) analysis of conformation dependency of peptide antigens and the importance of such conformation in generating an optimal B-cell response; (7) the efficacy of the antibodies elicited by these Peptide antigens in neutralizing endogenous protein with the ultimate aim of designing synthetic vaccines. Chapter 1 of this thesis deals with a general introduction summarizing the current status of knowledge regarding the chemistry and biology of RCP as well as synthetic pep­tides as potential immunogens. Chapter 2 outlines details of the experimental procedures adopted. Chapter 3 describes the results of investigations on the C-terminal fragment (residues 200-219) of cRCR The main consideration in selecting this sequence for the design of a potential peptide-based vaccine relied on the epitopic specificity of the neu­tralizing MAb 6S2C12. Epitope mapping using the Pepscan method revealed that the monoclonal antibody recognizes a core sequence corresponding to residues 203-210 of the cRCP. A 21-residue synthetic peptide (C-21) comprising this epitope was synthesized and antibodies elicited to the peptide conjugated to two different carriers, namely diphtheria toxoid and purified protein derivative (PPD) for T-cell help. In both active and passive immunoneutralization experiments, the peptide specific neutralizing antibodies interfered with the biological function of the protein and hence either protected from pregnancy or caused early fetal resorption in rodents as well as in sub-human primates. The conforma-tional properties of the peptide in aqueous buffers were analyzed from circular dichroism which revealed the absence of any ordered structure in the native C-21 peptide. Theoreti­cal predictions of secondary structure suggested a propensity for an t*-helical structure for this fragment in the native protein. Therefore, influence of the helix-promoting solvent, vizM 2,2,2,trifluroethanol (TFE) on the C-21 peptide was investigated. Addition of TFE resulted in spectral changes with negative bands at 208 and 222 nm and a positive band at 190 rim which are typical of an a-helix. To gain more information on the conformational characteristics of this peptide, it was considered worthwhile to stabilize the native peptide in an a-helical conformation based on simple rational design principles. Towards this end, four analogs of the parent peptide were synthesized and helix stabilization was sought to be achieved by introducing either salt bridges or back-bone conformational constraints such as by incorporating a-amino isobutyric acid at appropriate positions. In all the analogs, the core sequence, recognised by the neutralizing MAb 6B2C12 was maintained intact to ensure induction of antibodies capable of recognizing the native protein. CD spectral analysis of the analog peptides indicated that all the engineered peptides had varying degrees of enhanced helicities as compared to the parent peptide. The immunogenicity of each analog was studied by to the relevant peptide-diphtheria toxoid conjugates and analyzing their reactivities with the native protein by direct and competitive ELISA. The results revealed that these engi­neered conformational analogs axe highly immunogenic eliciting high titers of anti-protein antibodies. The relative affinities of these antibodies to bind cRCP were investigated. The antibodies to peptide analogs had higher affinities for the native protein and a positive correlation was found between the helical content of the peptide antigen in question and the relative affinity of corresponding antibody. The antibodies directed to all the peptide analogs could block the function of RCP resulting in early embryonic resorption when ad­ministered to pregnant mice. An interesting pattern of immunological cross-reactivity has been observed with the native and designed peptides. Antibodies raised to constrained helical analogs could bind the C-21 peptide which is structurally flexible. In contrast, the antibodies raised to the flexible native peptide antigen were inefficient in recognizing the structured peptides. The ability of all the peptide antibody to bind the native protein has been interpreted in terms of a conformationally flexible C-terminus region in cRCP. Chapter 4 details investigations on a 21-residue peptide (N- 21) from the N-terminiis (4-24) of the protein. Selection of this peptidyl sequence relied on theoretical prediction of potential sequential determinants on RCP other than at C-terminus as well as on the outcome of immunoneutralisation experiments using antibodies to egg yolk RCP which lacks the relevant C-terminal determinants. The structure of this peptide in solution was analyzed by two dimensional NMR and CD. NMR experiments revealed the presence of two structured regions in the peptide. Diagnostic nuclear Overhauser effects characteristics of reverse turns or short frayed helical segments over residues 3-9 and 18-21 of the peptide were obtained. CD spectra showed the presence of a strong, negative band at 204 nm over a wide range of solvent conditions, a feature which has been interpreted in terms of a "polyproline Il-like" segment encompassing residues 11-16 which corresponds to an interesting (X-Pro)^ repeat in the N-21 sequence. Specific antibodies were generated to this peptide as a conjugate with diphtheria tox­oid. Administration of the antipeptide antibodies could neutralize the protein in vivo as demonstrated by early embryonic loss in pregnant mice. In limited experiments the anti­peptide antibodies showed propensity to protect bonnet monkeys from pregnancy over a few consecutive ovulatory cycles when titres are maintained elevated by periodic boosting. To address the relationship between peptide structures and antigenicity, epitope mapping of this antipeptide antibodies as well- as the polyclonal antibodies to native RCP was undertaken using the Pepscan method. The results reveal that antigenic regions correspond well to conformationally well-defined elements of structure with the polyproline II-like seg­ment being a common antigenic determinant on both the peptide and the native protein. These observations are suggestive of the involvement of both the N and C-terminal regions of RCP in terms of its binding to putative plasma membrane receptors.

Foot-and-mouth disease (FMD) is a highly contagious and economically important disease that affects even-toed hoofed mammals. The FMD virus (FMDV) is the causative agent of FMD, of which there are seven clinically indistinguishable serotypes. Three serotypes, namely, South African Territories (SAT)1, SAT2 and SAT3 are endemic to southern Africa and are the most antigenically diverse among the FMDV serotypes. A negative consequence of this antigenic variation is that infection or vaccination with one virus may not provide immune protection from other strains or it may only confer partial protection. The identification of B-cell epitopes is therefore key to rationally designing cross-reactive vaccines that recognize the immunologically distinct serotypes present within the population. Computational epitope prediction methods that exploit the inherent physicochemical properties of epitopes in their algorithms have been proposed as a cost and time-effective alternative to the classical experimental methods. The aim of this project is to employ in silico epitope prediction programmes to predict B-cell epitopes on the capsids of the SAT serotypes. Sequence data for 18 immunologically distinct SAT1, SAT2 and SAT3 strains from across southern Africa were collated. Since, only one SAT1 virus has had its structure elucidated by X-ray crystallography (PDB ID: 2WZR), homology models of the 18 virus capsids were built computationally using Modeller v9.12. They were then subjected to energy minimizations using the AMBER force field. The quality of the models was evaluated and validated stereochemically and energetically using the PROMOTIF and ANOLEA servers respectively. The homology models were subsequently used as input to two different epitope prediction servers, namely Discotope1.0 and Ellipro. Only those epitopes predicted by both programmes were defined as epitopes. Both previously characterised and novel epitopes were predicted on the SAT strains. Some of the novel epitopes are located on the same loops as experimentally derived epitopes, while others are located on a putative novel antigenic site, which is located close to the five-fold axis of symmetry. A consensus set of 11 epitopes that are common on at least 15 out of 18 SAT strains was collated. In future work, the epitopes predicted in this study will be experimentally validated using mutagenesis studies. Those found to be true epitopes may be used in the rational design of broadly reactive SAT vaccines / Life and Consumer Sciences / M. Sc. (Life Sciences)