CORDIS - EU research results

Evolution of Pathogen and Host Diversity

Final Report Summary - DIVERSITY (Evolution of Pathogen and Host Diversity)

I have consolidated a body of theoretical work on the evolution of pathogen population structure, collectively known as Strain Theory, through which I have demonstrated that immune-mediated competition can cause pathogen populations to self-organise into either a stable or sequentially emerging set of discrete antigenic and metabolic types, despite frequent genetic exchange. Through extensive long-term collaborations with experimentalists, I have been able to validate this theory; the model is now available as an R package, MANTIS.

Using Strain Theory, I have shown that the epidemic behaviour of influenza may be driven by immune responses against epitopes of limited variability, thereby challenging the conventional model of antigenic drift. We have now identified one such epitope in the head region of haemagglutinin of subtype H1 and performed vaccine studies in mice (as described the proposal); IP protection has been undertaken in collaboration with Oxford University Innovation. This is a rare example of a theoretical prediction leading to the development of a vaccine.

I have applied Strain Theory to the within-host dynamics of HIV-1 showing (i) that the loss neutralizing antibody induction, due of attrition of the CD4+ T cell population, is more likely to precipitate loss of control of the viraemia rather than any alteration in the quality, induction or effectivity of the CD8+ T cell response and (ii) B-cell epitopes of limited variability may play an important role in duration of infection and could be exploited for developing therapeutic vaccines.

Combining Strain Theory with whole genome analysis, I have shown that targeting particular pneumococcal serotypes through vaccination can cause metabolic and virulence-associated profiles to migrate to non-vaccine serotypes: a phenomenon termed Vaccine-Induced Metabolic Shift, thereby providing a novel explanation for the changes observed in the USA following vaccination. By applying machine learning techniques to pneumococcal whole genomes, we have identifying the GroEL chaperone protein as a primary determinant of lineage structure and consequently a vaccine target which would be less likely to lead to rapid changes in population structure than vaccines targeting the capsular serotype.

Strain Theory shows that immune selection forces pathogen populations to exist as a set of antigenically discrete strains; I have demonstrated that this then drives non-overlapping associations between the HLA loci through which recognition of these antigens is mediated. This result can be exploited to uncover functional relationships between HLA alleles. We have also developed multi-locus models to explain the population structure of killer-cell immunoglobulin-like receptors (KIR). Using multi-locus models, I have also shown that epistatic interactions between two malaria-protective blood disorders – sickle-cell trait and α-thalassemia - are capable of explaining why (i) the sickle cell trait is so uncommon in the Mediterranean despite its long history of malaria selection and (ii) why α-thalassemia has failed to reach fixation in sub-Saharan Africa and maintains contrasting profiles of allele frequencies among neighbouring South Asian tribes.

Finally, we have shown that Strain Theory has broader ecological implications and can be adapted to explain how any set of competing species fills niche space; these finding have important implications for conservation biology.