Bordetella pertussis causes whooping cough in humans. Despite high routine vaccination coverage, a number of countries have experienced increasing whooping cough incidence since the 1990s. While a plethora of candidate explanations for this resurgence remain hotly debated, one of the prevailing views attributes pertussis reemergence to vaccine-driven pathogen adaptation, especially after the switch in vaccines, from whole cell to acellular. Pathogen adaptation is proposed to occur through antigenic divergence between vaccine and circulating strains as well as surfacing of strains with increased pertussis toxin production. To formally test this hypothesis, we obtained whole genome single nucleotide polymorphism (SNP) sequences from the Netherlands, between 1949 and 2010. We inferred phylogenetic relationships using a maximum likelihood framework (RAXML) and mutation rates and ancestral nodes were estimated using Bayesian analysis (BEAST).
Results/Conclusions
We document significant shifts in B. pertussis population throughout the time span of our data indicating pathogen adaptation. There is no evidence for loss of diversity nor any indication of a bottleneck in the population. Our analyses suggest the existence of two distinct clades, with the larger clade arranged along a cline of genetic differentiation. In the Netherlands, the resurgence of pertussis is associated with the expansion of B. pertussis strains with a changed promoter region for pertussis toxin (ptxP3). About 20% of SNP sites within the bacterial genome are polymorphic. Moreover, we detected that changes in genes encoding virulence proteins, in the circulating strains, occurred prior to the switch from whole cell to acellular vaccine. Additionally, our analyses suggested that virulence-associated genes were under selection, as reflected in the topology of the trees. Further, we are developing an integrated approach of phylogenetic analysis with hypothesis testing and quantification of the contribution of a variety of potential predictors. Statistical testing for the candidate predictors is underway. This work provides insight into ways in which pathogens may adapt under several driving forces of selection and suggests ways to improve pertussis control