The central problem for researchers of HIV-1 evolution is explaining the apparent design of the virus for causing pandemic infection in humans: understanding how HIV-1 spreads is key to halting the pandemic. Current knowledge of how HIV-1 spreads from host to host is based upon experimental observation and indirect inferences informed by theory. The hypothesis of this thesis is that diversity of HIV-1 around the time of transmission is important for viral adaptation to a new human host, rather than intrinsic superiority of particular strains found in infectious fluids from human donor hosts, and that studying recombination is important for understanding this behaviour. To demonstrate the apparent randomness of transmission, I test the null-hypothesis that hard selection accounts for between-host viral divergence in a rare case study of contemporaneous infection. I explain how the experimental data that I have generated and the analyses I have carried out address certain basic assumptions and predictions about HIV-1 transmission and may inform current strategies for vaccine design. Specifically, my approach contributes to the current literature on HIV-1, by investigating an alternative hypothesis to the single virion theory of sexual transmission and by characterizing the role of recombination in a pseudodiploid virus following multiple-infection.

Typhoid disease, caused by infection with S. Typhi, is a significant cause of mortality and morbidity in resource–poor countries. Efforts have been made to generate a new generation of vaccines that are efficacious and can be given to infants, but have been hindered by a poor understanding of the protective immune response to S. Typhi infection, and in particular by the absence of a correlate of protection. Controlled human infection studies (‘challenge studies’) provide a model for investigating infectious diseases and appraising novel vaccines, including in typhoid disease. This DPhil described the development of a human challenge model of typhoid fever using <en>S. Typhi Quailes strain administered to healthy adults in a sodium bicarbonate buffer. The careful characterisation and manufactured of the strain is described. Following ingestion of 10³ CFU of S. Typhi 55% of participants developed typhoid disease, whilst ingestion of 10⁴ CFU gave a higher attack rate of 65%. At this attack rate vaccine efficacy against human challenge should be demonstrable with a modest sample size. Validity of the model in the appraisal of vaccines was demonstrated using Ty21a, a live, oral, attenuated vaccine. Protective efficacy of Ty21a compared to placebo against challenge was 35%, comparable to that observed in some endemic settings, and the estimated protection in the first year after vaccination in Cochrane meta-analysis. Clinical, microbiological and humoral immune responses were investigated in participants challenged during model development. Typhoid disease was associated with a high fever in most, but not all participants, and a range of symptoms. Severity of disease was variable, and included asymptomatic bacteraemia, as well as fever and symptoms in participants in whom bacteraemia could not be demonstrated. Typhoid disease was associated with a strong humoral immune response to the flagellin and lipopolysaccharide antigens of S. Typhi but not the Vi polysaccharide capsule. Humoral immune responses were not demonstrated in participants without typhoid fever. There was a dose-response relationship to the clinical, microbiological and humoral responses with participants challenged with 10⁴ CFU having more marked responses than those challenged with 10³ CFU. Future success of challenge studies relies on the willing participation of healthy adult volunteers. The motivations for participation, and experiences of participants, were appraised by questionnaire. Whilst financial compensation was an important motivator, it was not the sole motivator. Participants were positive about their experiences, and most would participate again.