Tracking the establishment of local endemic populations of an emergent enteric pathogen

Abstract

Shigella sonnei is a human-adapted pathogen that is emerging globally as the dominant agent of bacterial dysentery. To investigate local establishment, we sequenced the genomes of 263 Vietnamese S. sonnei isolated over 15 y. Our data show that S. sonnei was introduced into Vietnam in the 1980s and has undergone localized clonal expansion, punctuated by genomic fixation events through periodic selective sweeps. We uncover geographical spread, spatially restricted frontier populations, and convergent evolution through local gene pool sampling. This work provides a unique, high-resolution insight into the microevolution of a pioneering human pathogen during its establishment in a new host population.

Keywords: drug resistance; enteric disease; genomics; phylogeography.

Fig. 1.

Tracking S. sonnei after entering and establishing successive founder populations across Vietnam. North-orientated map of Vietnam shows the location of the study sites in Ho Chi Minh City (HCMC), Khanh Hoa (KH), and Hue; proposed pattern of geographical spread is indicated with arrows, and National Highway 1 is shown in white. Maximum-likelihood (ML) phylogeny for the Vietnamese S. sonnei is shown to the right, with divergence dates for major clones labeled (black internal nodes). Colors indicate location of isolation: red, HCMC; blue, Hue; and purple, KH. Where distinct locally evolving populations were detected, the local phylogenies are represented next to the city; blue and purple arrows indicate where these localized Hue and KH phylogenies fit within the larger tree.

Fig. 2.

The dynamics of S. sonnei microevolution in HCMC. (A) Bars indicate the frequency of S. sonnei isolated in HCMC as a proportion of total culture-confirmed Shigellosis cases; line plots indicate the proportion of S. sonnei resistant to nalidixic acid and ceftriaxone. (B) Phylogenetic tree for HCMC S. sonnei isolates, showing estimated dates for major sweeps, acquisition of pDPT1 colicin plasmid, and fixation of key nalidixic resistance mutations (gray groups) and pKHSB1 plasmid conferring resistance to third-generation cephalosporins (yellow group). (C) North-orientated map of central HCMC, with highlighted waterways, showing the GPS locations of 53 S. sonnei patient residences identified between 2009 and 2010 (gray, sweep 3/gyrA-87Y; yellow, sweep 4/gyrA-87Y/CTX-M-15; white, presweep 3). Heat map indicates the spatial distribution of 1,419 diarrheal disease episodes in HCMC; colors of the heat map indicate the probability of all-cause diarrhea per 0.02 km², ranging from 2 × 10⁻⁶ (dark blue) up to 1.67 × 10⁻⁴ (red). (Scale bar: 5 km.)

Fig. 3.

Colicin plasmids in the VN S. sonnei clone. (A) Schematic representation of the 6,826-bp E5 colicin-encoding plasmid pDPT1 orientated from the hypothesized origin of replication. ORFs are strand orientated and colored according to function as follows: green, plasmid replication and mobility; red, colicin activity; brown, colicin immunity; pink, colicin lysis. The gray region highlights the area of the plasmid used to create plasmid pDPT3 containing an inserted kanamycin-resistance gene (blue). (B) Schematic representation of the 5,262-bp Js colicin encoding plasmid pDPT2 orientated from the hypothesized origin of replication. ORFs are strand-orientated and colored as above. The gray region highlights the area of the plasmid used to create plasmid pDPT4 and containing an inserted chloramphenicol-resistance gene (blue). (C) Example of colicin extracts growth inhibition experiments. (Left) Lawn of S. sonnei MS128 (E5−/Js−) challenged with colicin extracts from 15 different E5+ isolates (squares), Js+ isolates (triangles), and PBS (circle). (Right) Lawn of S. sonnei MS83 (E5−/Js−) challenged with colicin extracts from 15 different E5+/Js+ isolates and PBS (circle). (D) Tabulated summary of >600 colicin extract growth inhibition experiments with 27 different bacterial isolates.