Malaria remains an important cause of morbidity and mortality among children in sub-Saharan Africa despite recent reductions in malaria incidence in some parts of Africa. Current control tools face threats such as the emergence of drug resistant parasites and insecticide resistant mosquitoes. A malaria vaccine is needed to complement and/or replace existing tools in order to achieve better malaria control and eventually eliminate the disease. RTS,S/AS01E is the most clinically advanced pre-erythrocytic malaria vaccine candidate and is currently being tested in a phase III trial. The short-term efficacy of RTS,S/AS01E is known but the duration of protection is unknown. Furthermore, although RTS,S is protective, it is unclear which immunological assays predict efficacy: hence there are no known correlates of vaccine-induced protection against clinical malaria. In a randomized controlled trial, I assessed the efficacy of RTS,S/AS01E in children (5-17 months old) residing in Kilifi, Kenya, over 4 years of follow-up and determined the correlates of protection against clinical malaria. In order to examine the effect of variations in malaria exposure on vaccine efficacy, I developed an individual marker of malaria exposure calculated as distance-weighted prevalence of malaria infection within 1 km radius of every child. Over 4 years of follow-up, RTS,S/AS01E had an efficacy of 29.9% (95%CI: 10.3% to 45.3%, p=0.005) and 16.8% (95%CI: -8.6% to 36.3% p=0.18) against first and all malaria episodes, respectively (by intention to treat analysis). Vaccine efficacy waned over time and with increasing malaria exposure. RTS,S/AS01E efficacy was 43.6% (95% CI, 15.5 to 62.3) in the first year but was -0.4% (95% CI, -32.1 to 45.3) in the fourth year. Vaccine efficacy was 45.1% (95%CI 11.3% to 66.0%) among children with lower than average malaria exposure index, but 15.9% (95%CI -11.0 to 36.4%) among children with higher than average malaria exposure index. Despite waning in efficacy, RTS,S/AS01E averted 65 cases of malaria per 100 vaccinated children, with more cases averted among the children in the higher malaria-exposure cohort (78 cases per 100 vaccinated children) than those the low exposure cohort (62 cases per 100 vaccinated children). RTS,S/AS01E induced high titres of anti-CS protein antibodies and CD4+ T cell but not CD8+ T cell responses. Anti-CS antibody titres and the frequency of TNF-α producing CD4+ T cell responses were independently associated with protection from clinical malaria, and the combination of both anti-CS titers and TNF-α producing CD4+ T cell response satisfied the Prentice criteria for surrogate markers of protection. There was no association between avidity of RTS,S-induced anti-CS protein antibodies and protection from clinical malaria. Conclusions: RTS,S/AS01E efficacy against all episodes is 16.8% over the 4 years of follow-up. The vaccine efficacy wanes over time and with increasing malaria exposure. RTS,S/AS01E-induced TNF-α producing CD4 T cell and anti-CS protein antibody responses were independently associated with protection from clinical malaria. Anti-CS avidity did not predict protection from clinical malaria. Long-term follow-ups of malaria vaccine trials are essential in the evaluation of the longevity of vaccine efficacy.

The physical properties of viral capsids are major determinants of vaccine efficacy for several picornaviruses which impact on human and animal health. Current picornavirus vaccines are frequently produced from inactivated virus. Inactivation often reduces the stability of the virus capsid, causing a problem for Foot and Mouth Disease Virus (FMDV) where certain serotypes fall apart into pentameric assemblies below pH 6.5 or at temperatures slightly above 37°C, destroying their effectiveness in eliciting a protective immune response. As a result, vaccines require a cold chain for storage and animals need to be frequently immunised. FMDV is a member of the Aphthovirus genus of the Picornaviridae. Globally there are seven FMDV serotypes: O, A, Asia1, C and SAT-1, -2 and -3, contributing to a dynamic pool of antigenic variation. As part of collaboration between the Division of Structural Biology, Oxford University, The Pirbright Institute, Reading University and ARC, Ondespoort, South Africa we sought to rationally engineer thermo-stable FMDV capsids either as infectious copy virus or recombinant empty capsids with improved thermo-stability for improved vaccines. In this project, in silico molecular dynamics (MD) simulations, molecular modelling, free energy calculations, X-ray crystallography, electron microscopy and various biochemical/biophysical techniques were used to design and help characterise the capsids. For the most unstable FMDV serotypes (O and SAT2), panels of stabilising mutants were characterised by MD. Promising candidates were then engineered and shown to confer increased thermo- and pH-stability. Thus, in silico predictions translate into marked stabilisation of both infectious and recombinant empty viral capsids. A novel in situ method was used to determine crystal structures for quality assessment and to verify that no unanticipated structural changes have occurred as a consequence of the modifications made. The structures of the wildtype and two of the stabilised mutants were solved and the antigenic surfaces shown to be unchanged. Animal trials showed stabilised particles can generate a similar or improved neutralising antibody response compared to the traditional vaccines and may therefore lead to a new generation of stable and safe vaccines.

The primary approach to the control of meningococcal disease remains effective vaccination programmes in susceptible populations. Vaccines against serogroups A, C, W and Y offer broad protection against meningococci and both polysaccharide and conjugate quadrivalent vaccines are licensed for use in the UK. Previous studies have assessed the antibody response to meningococcal polysaccharide and conjugate vaccines, but there is limited information on the nature of the B cell response to these antigens. As part of a clinical trial using both polysaccharide (MenACWY-PS) and conjugate (MenACWY-CRM) vaccines in adult volunteers, this DPhil reports the analysis of subsets of antigen specific B-cells produced in response to either vaccine. Prior MenACWY-PS impaired the response to a subsequent dose of MenACWY-CRM. This may be due to MenACWY-PS driving terminal differentiation of antigen specific cells into plasma cells, without replenishment of the memory B cell pool. In addition, despite prior data indicating that it may act as a thymus dependent antigen, the serogroup A polysaccharide component of MenACWY-PS appears to behave in the same way as serogroup C, W & Y polysaccharide components. Antibody molecules recognise and bind to a multitude of conformational epitopes. This variability is enabled by the complexities of immunoglobulin variable domain gene recombination which can generate a vast potential repertoire of unique antibody molecules. However, the diversity of the antibody repertoire is more restricted against specific antigens and within defined B cell subsets. In this DPhil, ‘next generation’ sequencing technologies were used to investigate the diversity of the B cell variable domain before and after vaccination of adult volunteers. Individuals at baseline were found to have distinct antibody repertoires. Vaccination with a Haemophilus influenzae type b (Hib) conjugate vaccine resulted in an oligoclonal antibody response, with enrichment for Hib specific canonical antibody sequences.

The current worldwide epidemic of Mycobacterium tuberculosis infection is a huge global health problem. Widespread BCG vaccination remains a useful tool in combating this epidemic; however, its variable efficacy requires urgent development of novel vaccines against Mycobacterium tuberculosis. Such a candidate vaccine is a serotype 5 adenovirus expressing antigen 85A from M. tuberculosis (Ad85A). In animal models Ad85A confers significant protection when administered intra-nasally. The work in this thesis demonstrates that intra-nasal immunisation with Ad85A results in inhibition of M. tuberculosis growth in the lung early after infection, in contrast to the late inhibition induced by parenterally administered vaccines. Early inhibition correlates with the presence in the lung of a highly activated population of antigen-specific CD8 T cells, maintained for at least 6 months post-immunisation by persistent antigen. For intra-nasal Ad85A to be effective, the vaccine must be delivered into the lower respiratory tract, as immunisation targeting only the nasal-associated lymphoid tissue (NALT) does not result in protection. Following a change of animal facility, the lung immune response to intra-dermal immunisation with Ad85A increased and this route of immunisation now induced protection, though growth of M. tuberculosis was inhibited only late after infection. However, this response and protection can be altered by exposure to environmental mycobacteria. Further experiments showed that simultaneous respiratory and parenteral immunisations (SIM) act additively, where local lung immunity inhibits the growth of M. tuberculosis early after infection and systemic immunity protects later. SIM regimes generate greatly improved protection over either immunisation alone and do not depend on priming and boosting.

In recent years the glycan shield of the HIV-1 envelope spike (Env) has emerged as a potential target for microbicide and vaccine design. The densely packed glycans on its surface include an intrinsic population of under-processed oligomannose structures, and a number of lectins and broadly neutralising antibodies (bnAbs) have been isolated which are reactive to these ‘non-self’ glycan structures. The potential value of these agents in therapeutic or vaccine contexts depends upon the prevalence of their glycan targets in nature and their resilience to sequence mutation. Here the prevalence of oligomannose-type glycans on recombinant gp120 was demonstrated across a panel of isolates, revealing subtle cross clade differences. Alanine scanning of all potential N-glycosylation sites (PNGSs) within a model gp120 demonstrated the overall stability of the oligomannose population, but highlighted regions of glycan clusters where individual glycans act to limit the processing of their neighbours. This was formally demonstrated for the N332 ‘site of vulnerability’, where deletion of nearby glycosylation sites led to altered glycan processing at the N332 site. A panel of N332-dependent bnAbs was screened for their ability to tolerate such changes in glycan processing, with differing results. While some displayed promiscuous binding, others were more sensitive to glycan microheterogeneity. Site-specific glycosylation analysis of the PGT135 epitope revealed that an intolerance of certain glycoforms may explain its limited breadth. While a greater understanding of Env glycan microheterogeneity and bnAb promiscuity is required, these findings reveal insights into the architecture of the HIV-1 glycan shield that suggest it is a conserved and robust target for drug and vaccine design.

Tuberculosis (TB), caused by infection with Mycobacterium tuberculosis (M.tb), remains a global health problem and although BCG offers some protection against childhood disseminated disease and other mycobacterial infections, its efficacy against pulmonary TB varies between 0 and 80%. Modified Vaccinia virus Ankara expressing antigen 85A (MVA85A) is a novel TB vaccine designed to boost mycobacterium-specific CD4+ T cell response primed by BCG. MVA85A induces strong interferon (IFN)-γ responses, a cytokine known to be essential for protection following M.tb infection. A strong IFN-γ response is not a correlate of protection and in terms of the adaptive response, interleukin (IL)-17 is emerging as an important cytokine following vaccination as it is thought to help boost IFN-γ production by CD4+ T cells. This thesis shows that MVA85A induces IL-17 in PBMC and whole blood of human BCG – MVA85A vaccinees. This is replicated in mice receiving BCG – MVA85A intranasally. The administration of cholera toxin (CT) with BCG enhances IL-17 and confers improved protection following M.tb challenge, which is partially dependent on IL-17 and on the mucosal route of administration. Since CT is not a suitable adjuvant in humans, an alternative IL-17-inducing pathway was investigated. In human BCG – MVA85A-vaccinated volunteers, blocking the hydrolysing ability of the CD39, an apyrase responsible for hydrolysing pro-inflammatory ATP, enhances IL-17 production. Challenge of BCG – MVA85A-vaccinated CD39-/- mice with M.tb slightly improved the protective capacity of the vaccine, suggesting that a pathway dependent on ATP-driven inflammation may be a target for improving the immunogenicity of a vaccine against M.tb disease. Overall, this thesis has confirmed the important role of IL-17 in vaccine-induced protection against TB disease and identifies a possible target pathway for improvement of a novel vaccine.

Foot-and-mouth disease virus (FMDV) is an economically devastating and highly contagious livestock pathogen. It exists as seven serotypes, comprising numerous antigenically distinct subtypes. The large amount of antigenic heterogeneity has confounded attempts at developing broadly reactive vaccines. In order to overcome this issue the fundamentals of the interactions between the virus and the host humoral immune response must first be understood. Previous work in this area using monoclonal antibody (mAb) escape mutants has identified five antigenic sites for the O serotype and efforts have been made to quantify their relative importance. However, this does not represent a complete picture of serotype O antigenicity. The work conducted in this thesis demonstrates the role of a limited number of dominant substitutions in mediating the antigenic diversity of serotype O Foot-and-Mouth disease virus. Two alternative but complementary methods for identifying epitopes were developed. The first used a mathematical model to analyse newly generated serological and sequence data from 105 viruses, cultured for this purpose (and cross-reacted to 5 reference antisera), in the context of an existing crystallographic structure to identify and quantify the antigenic importance of sites on the surface of the virus. The second approach was purely structural, using existing B cell epitope prediction tools to develop a method for predicting FMDV epitopes using existing crystallographic structures of FMDV. These techniques were validated by the use of reverse genetics, which confirmed the impact on cross reactivity of two predicted novel serotype O antigenic residues, with a further four novel residues identified by looking in depth at the interactions between two genetically close, but antigenically distant viruses. This increased knowledge of the antigenic composition of serotype O FMDV contributes to our understanding of the nature of vaccine efficacy and the breadth of protection, which, in the longer term, will aid in the goal of developing vaccines to better protect livestock from such a highly antigenically variable disease.

For both influenza viruses and hepatitis C viruses, T cell responses to conserved antigens are one strategy for the human host to control the spread of infection. Such T cell responses can be generated with the use of viral vectored vaccines. Initially I show that the viral vectored vaccine MVA-NP+M1 can boost memory T cell responses to influenza A virus in adults aged over 50 years old. However within this group, MVA-NP+M1 had reduced immunogenicity in adults who were aged over 70 years old. The influenza virus-specific T cell responses comprised both CD4 and CD8 T cells, and were capable of secreting multiple Th1 cytokines. I then show that MVA-NP+M1 can be safely co-administered alongside seasonal influenza vaccine. The combination does not interfere with the peak T cell response that normally occurs 1 week following MVA-NP+M1. There was a statistically significant increase in antibodies to the H3N2 strain when the vaccines were co-administered, suggesting that the MVA-NP+M1 can act as an adjuvant. The efficacy of MVA-NP+M1 in humans had been previously evaluated in an influenza virus challenge study. I used a whole blood transcriptome approach to improve the classification of outcomes following influenza virus challenge. For subjects with laboratory-confirmed influenza, individuals with moderate/severe symptoms were found to have a distinct transcriptional signature comprising over 2,000 genes. I used a machine learning algorithm to reduce this variation down to just six genes (CCL2, SEPT4, LAMP3, RTP4, MT1G and OAS3). I validated this finding using expression data from an independently conducted challenge experiment. Data from these six genes was successfully able to predict symptomatic and asymptomatic cases with 89% and 100% accuracy respectively. To induce T cell responses to hepatitis C virus, I used the vaccines ChAd3-NSmut and MVA-NSmut in a prime-boost regimen. While the combination was highly immunogenic in healthy young adults, MVA-NSmut alone was unable to prime immune responses. The magnitude of T cell responses to the vaccine immunogen was correlated with the breadth of the T cell responses to different epitopes. Re-administration of the same two vaccines after a short time interval (8 weeks) did not improve upon previous peaks in T cell response. However with a longer time interval (> 34 weeks), some individuals were able to achieve higher frequencies of virus-specific T cells compared to the first round of vaccines. A whole blood transcriptome approach was used to study gene expression in volunteers vaccinated with ChAd3-NSmut and MVA-NSmut. Vaccination with MVA-NSmut results in a very strong, but relatively short-lived host gene expression signature. In contrast, the transcriptional response seen following ChAd3-NSmut was much less pronounced. A comparison of the functional analysis of gene lists from both vaccines showed that similar pathways were being activated and repressed.

The apicomplexan parasite Plasmodium falciparum is the causative agent of the most severe and deadly form of human malaria. The production of an efficacious malaria vaccine is seen as one of the key steps towards the eradication of the disease, however to date only one candidate has progressed to application for licensure. Candidate malaria vaccines target the different stages of the P. falciparum lifecycle through induction of a functional immune response. Vaccines targeting the blood-stage parasite require induction of high titre neutralising antibodies. To achieve this, vaccine regimens have been designed specifically to maximise antibody induction and maintenance in humans. The ultimate test of any candidate vaccine is clinical efficacy and controlled human malaria infection (CHMI) is a powerful tool for measuring this. This model can also be used to study how vaccine induced antigen-specific components of the immune system respond to native antigen exposure in the context of parasitic infection In this Thesis I describe the induction and maintenance of B cell responses, including memory B cells (mBC) and antibody secreting cells (ASC) to the candidate blood-stage malaria antigens MSP1 and AMA1 following vaccination with a variety of regimens and CHMI. These B cell populations along with peripheral blood T follicular helper (Tfh) cells correlate strongly with antibody induction. Within these populations I have identified a number of phenotypically distinct subsets which contribute to a functional response to vaccine and/or parasite antigen. From single cell sorting of ASC at day seven post-boost I have managed to produce the first fully human monoclonal antibodies (hmAbs) specific for AMA1, one of which shows significant growth inhibitory activity (GIA). Despite promise the vaccine candidates MSP1 and AMA1 have been disappointing in terms of human efficacy. In this Thesis I have attempted to provide explanations on a cellular level as to why there is such disparity between pre-clinical and human data and ultimately why these candidates may have failed to provide efficacy. Such work will provide a strong basis for analysing future clinical trials of alternative candidate blood-stage vaccines and allow accurate characterisation of immune correlates and clinical efficacy when it is achieved.

Vaccines have had a profound influence on human health with no other health intervention rivalling their impact on the morbidity and mortality associated with infectious disease. However, the magnitude and persistence of vaccine immunity varies considerably between individuals, a phenomenon that is not well understood. Recent studies have used contemporary technologies to correlate variations in the genome and transcriptome to complex phenotypic traits, and these approaches have started to provide fresh insight into the intrinsic factors determining the generation and persistence of vaccine-induced immunity. This thesis aimed to describe the relationship between genomic and transcriptomic variations, and the immunogenicity of childhood immunisations. Candidate gene and genome-wide genotyping was conducted to evaluate the influence of genetic variants on vaccine-induced immunity following childhood immunisation. Furthermore, contemporary methodologies were used to assess non-coding and coding gene transcript profiles following vaccination, to further dissect the molecular systems involved in vaccine responses. Key findings from this thesis include the description of the first genome-wide association studies into the persistence of immunity to three routine childhood immunisations: capsular group C meningococcal (MenC) conjugate vaccine, Haemophilus influenzae type b (Hib) conjugate vaccine and tetanus toxoid (TT) vaccine. Genome-wide genotyping was completed on over 2000 participants, with an additional 1000 participants genotyped at selected genetic markers. Genome-wide significant associations (p<5×10⁻⁸) were described between single- nucleotide polymorphisms (SNPs) in two genes, CNTN6 and ENKUR, and the persistence of serological immunity to MenC following immunisation of children 6-15 years of age. In addition, genome-wide significant associations were described between SNPs within an intergenic region of chromosome 10 and the persistence of TT-specific IgG concentrations following childhood immunisations. Furthermore, a number of variants in loci with putative involvement in the immune system such as FOXP1, the human leukocyte antigen locus and the lambda light chain immunoglobulin locus, were shown to have suggestive associations (p<1×10⁻⁵) with the persistence of vaccine-induced serological immunity. The fundamental challenge will be to describe functional mechanisms associated with these findings, and to translate these into innovative and pragmatic strategies to develop new and more effective vaccines.