Bacterial Disease Research Group
Molecular epidemiology and the population and evolutionary biology of bacterial pathogens
The ability to identify isolates of bacterial pathogens is central to clinical microbiology. Until recently the methods that were used to characterise bacterial isolates were unsatisfactory as they indexed genetic variation that was difficult to compare between laboratories. A new approach - multilocus sequence typing (MLST) - has been developed which unambiguously characterises isolates of bacterial pathogens using the nucleotide sequences of internal fragments of seven house-keeping genes. The different sequences at each locus are assigned different allele numbers and each isolate is precisely defined by seven integers, which correspond to the alleles at the seven loci. MLST has been developed for a number of major pathogens in our laboratory and databases containing the allelic profiles of isolates of each species, together with epidemiological data, and interrogation and analysis software, can be accessed via the internet (http://www.mlst.net/).
MLST is now widely used by clinical microbiology laboratories and public health laboratories since it allows them to unambiguously identify and track important strains of major pathogens (e.g. multiply antibiotic-resistant strains of Streptococcus pneumoniae or methicillin-resistant strains of Staphylococcus aureus). However, the availability of the sequences of internal fragments of seven house-keeping genes, from many hundreds of isolates of an increasing number of bacterial pathogens, also provides data that can be used to address the comparative population and evolutionary biology of bacterial species. For example, these data are being used to measure the relative frequency of recombination, compared to point mutation, during the diversification of clones of different bacterial species, and to understand the longer-term impact of recombination on the evolution of bacterial populations.
Many bacterial pathogens are commonly carried asymptomatically and carriage only rarely leads to disease. Careful sampling of populations of such pathogens can be used to understand the nature of virulence and the selective forces that lead to the emergence of virulent strains from among a basically commensal population. A number of studies are underway to understand the nature of virulence in bacterial species, particularly Streptococcus pneumoniae and S. pyogenes.
We are also interested in the effect of vaccines that prevent disease by most, but not all, of the isolates that cause disease. Vaccines of this type (e.g. the multivalent conjugate pneumococcal vaccines) are designed to protect against the serogroups most commonly recovered from disease and assume that those serogroups which less commonly cause disease are less able to do so. The validity of this assumption needs to be examined, as serogroups of S. pneumoniae that are less commonly recovered from invasive disease (or pneumonia or acute otitis media) may be inherently less able to cause disease, or may simply have less opportunity to cause disease as they are infrequently carried. Vaccines with incomplete coverage will exert strong immune selective pressures on bacterial populations, which may significantly alter the serogroups that are carried by vaccinated children, and may increase the extent of disease caused by some non-vaccine serogroups. We are addressing some of these concerns by using well-sampled populations of pneumococci from disease and carriage to measure serogroup- and clone-specific attack rates and the impact of vaccination on carriage.
Figure 1. Emergence of diversification of adaptive lineages from within a recombining bacterial population.
Figure 2. Statistical tests of congruence between loci, using data obtained from multilocus sequence typing, show the lack of any phylogenetic signal in gene trees from some bacterial species.
