Escherichia coli O157 Phylogeny: Implications for Virulence and Disease Distribution

ABSTRACT This study brings together information on the phylogenetics and epidemiology of Shiga toxin-producing Escherichia coli (STEC) O157 to examine variation in the virulence and distribution of disease. Previous studies have suggested that genetic characteristics reflected in the phylogeny of STEC O157 strains could explain differences in virulence, specifically incidence of hemolytic uremic syndrome (HUS). However, these studies are based on only a partial phylogenetic tree and employ analysis methods that do not accurately summarize the risk associated with individual STEC O157 strains. In this project, I will test the association between phylogenetic lineage and HUS using an expanded phylogenetic tree and rigorous analytic methods (Aim 1). A potential mechanism will then be tested using mediation analysis to determine the role of Shiga toxin genes (stx) in the association between lineage and HUS (Aim 2). Finally, this study will further deepen understanding of STEC O157 epidemiology by testing whether phylogenetic lineages are geographically segregated and detecting clusters of STEC O157 by lineage (Aim 3). In pursuit of these Aims, I will use a study population of 1,160 culture- confirmed STEC O157 cases reported to the Washington State Department of Health from 2005 to 2014. Phylogenetic lineage and stx subtype for these cases were determined using single nucleotide polymorphism (SNP) typing and Shiga toxin-encoding bacteriophage insertion site (SBI) typing, respectively. The phylogenetic tree for the study population will be drawn, incorporating HUS status to depict how HUS clusters by lineage. Generalized estimating equations will be used to test the association between lineage and HUS while accounting for the correlation among isolates of the same strain. Potential outcomes mediation analysis will be used to test the contribution of different stx subtypes. Clusters of disease due to each lineage will be identified using spatial scan statistics, phylogeography will be employed to explore how STEC O157 in Washington has evolved according to case locations, and spatial segregation will be estimated to determine whether the lineages are randomly mixed throughout the state. The results of this study will expand our understanding of genetic determinants of virulence of STEC O157. It will provide evidence to focus research on the most virulent forms of STEC O157 and test whether stx subtypes are responsible for observed associations. If results are negative, this will suggest that virulence determinants are not correlated with the genetic markers used to establish phylogenetic lineage. This study will also provide knowledge about the geographic distribution of STEC O157 by phylogenetic lineage. Segregation may suggest that locally-maintained bacteria populations are important sources of infection, whereas random mixing of lineages may indicate that dominant sources of infection are diffuse. With positive or negative findings, this study will make an important contribution to the biologic and epidemiologic knowledge of STEC O157.