Worldwide Population Structure, Long-Term Demography, and Local Adaptation of Helicobacter pylori

Abstract

Helicobacter pylori is an important human pathogen associated with serious gastric diseases. Owing to its medical importance and close relationship with its human host, understanding genomic patterns of global and local adaptation in H. pylori may be of particular significance for both clinical and evolutionary studies. Here we present the first such whole genome analysis of 60 globally distributed strains, from which we inferred worldwide population structure and demographic history and shed light on interesting global and local events of positive selection, with particular emphasis on the evolution of San-associated lineages. Our results indicate a more ancient origin for the association of humans and H. pylori than previously thought. We identify several important perspectives for future clinical research on candidate selected regions that include both previously characterized genes (e.g., transcription elongation factor NusA and tumor necrosis factor alpha-inducing protein Tipα) and hitherto unknown functional genes.

Keywords: adaptation; human pathogens; neutral evolution.

Figure 1

(A) Plots of individual assignments to clusters according to BAPS and DAPC analysis, using K = 3 (A1) clusters: black, Africa2; blue, AfricaEu; red, EuroAsia and K = 4 (A2) clusters: black, Africa2; blue, Africa1; red, EuroAsia; and orange, AsiaAmerica. (B) Scatterplots of the discriminant space (components 1 and 2), using K = 4 (B1), K = 5 (B2). (C) World map with squares representing individuals colored according to cluster assignments with yellow squares indicating American subcluster (as for K = 5 in DAPC analysis; see Table S1).

Figure 2

Schematic representation of the most likely genealogy inferred for H. pylori worldwide sample. Demographic parameters estimated via coalescent simulations are summarized. T parameters correspond to time of population splits (1 to 4, most ancient to most recent). N_a and N_c parameters indicate effective ancestral and current population sizes, with 0 being the Africa2 population and 5 the America population (most ancient to most recent). R parameters refer to population growth.

Figure 3

Results of the SweeD and OmegaPlus analyses. A comparative representation for a “synthetic” strain of the worldwide sample and one synthetic strain of each population is drawn using a fictitious topology. Selection values are reported on the graph above each synthetic strain, on the y-axis, and genomic position on the x-axis. Omega values are represented with red lines, while alpha values are reported in blue. Since alpha values reach much higher levels than omega values, to make the figure easy to read, we reported both omega and alpha values within a scale from zero to 50, and we indicated alpha values >50 in darker blue. Genes falling into the functional categories explained in the discussion are color coded as reported in the legend, while remaining are in black. This figure was generated using R bioconductor package genoPlotR (Guy et al. 2010).