Population genetics of Streptococcus pneumoniae: the response to antibiotic and vaccine pressure

Streptococcus pneumoniae (pneumococcus) is a pathogen and a commensal of the upper respiratory tract, which is a leading cause of child mortality. The development of a successful vaccine against pneumococcal carriage and disease and the ongoing challenge of antimicrobial resistant strains mean that there is an imperative to understand how the pneumococcal population responds to vaccine and antibiotic pressure. Strains of pneumococci with reduced susceptibility to penicillin and other antibiotics have emerged and are a cause of concern, though their clinical impact is unclear. I perform a meta-analysis to examine the impact of antibiotic non-susceptibility on the risk of mortality and show that a meningitis infection with penicillin non-susceptible pneumococci increases the risk of mortality two fold. I then examine why some serotypes of pneumococcus are more likely to carry resistance than others. Using a mathematical model I generate the hypothesis that serotypes with a long duration of carriage are more likely to have a high prevalence of antibiotic resistance than those with a short duration of carriage. Using maximum likelihood estimation, I show that in children less than two years of age, penicillin resistance only occurs in those serotypes whose duration of carriage is nineteen days or more. Having considered the circumstances under which antibiotic resistance carries a selective advantage in a pneumococcal serotype, I then consider the effect that the spread of an advantageous gene has on the genetic diversity in a generalised bacterial population. Most simple models predict that when a novel allele arises in a bacterial population that is fitter than other alleles at that locus, it and its descendents will increase in frequency and sweep to fixation in the population. In the absence of recombination, all genetic diversity at all loci other than those within the sweeping genotype is lost. I consider whether asymmetric recombination can prevent the loss of diversity over the whole genome. I show that asymmetric recombination, when occurring at rates estimated to date from bacterial populations, is not frequent enough to prevent the extinction of alleles from the wild-type population. Finally, I analyse sequence data sampled from carried pneumococci to examine the impact of vaccination on genetic diversity in a pneumococcal population, an example of a selective event in a pneumococcal population