Dissecting the molecular evolution of fluoroquinolone-resistant Shigella sonnei

Abstract

Shigella sonnei increasingly dominates the international epidemiological landscape of shigellosis. Treatment options for S. sonnei are dwindling due to resistance to several key antimicrobials, including the fluoroquinolones. Here we analyse nearly 400 S. sonnei whole genome sequences from both endemic and non-endemic regions to delineate the evolutionary history of the recently emergent fluoroquinolone-resistant S. sonnei. We reaffirm that extant resistant organisms belong to a single clonal expansion event. Our results indicate that sequential accumulation of defining mutations (gyrA-S83L, parC-S80I, and gyrA-D87G) led to the emergence of the fluoroquinolone-resistant S. sonnei population around 2007 in South Asia. This clone was then transmitted globally, resulting in establishments in Southeast Asia and Europe. Mutation analysis suggests that the clone became dominant through enhanced adaptation to oxidative stress. Experimental evolution reveals that under fluoroquinolone exposure in vitro, resistant S. sonnei develops further intolerance to the antimicrobial while the susceptible counterpart fails to attain complete resistance.

Fig. 1

The phylogenetic structure of CenAsiaIII Shigella sonnei. The figure displays the maximum likelihood phylogeny of 395 S. sonnei sequences belonging to the CenAsiaIII clade. The tree is rooted on the most closely related sequence from the S. sonnei global sequencing study. The branch colour scheme indicates bootstrap support for the corresponding branch, ranging from low to high (red to black). The shading covered on the phylogeny indicates taxa harbouring quinolone resistance determining region (QRDR) mutations (see key). The orange diamond indicates the internal node leading to the major fluoroquinolone-resistant (FQr) clone. The magenta triangle denotes the position of taxon 2012–02037, which was subjected to long-read sequencing (see Methods). The rings show the associated information for each taxon, from the innermost to the outermost in the following order: (1) susceptibility testing for ciprofloxacin (black: resistance; brown: susceptible); (2) original geographical source; (3) presence of different plasmids; (4) presence of ESBL bla_CTX-M genes; (5) presence of mphA. The horizontal scale bar indicates the number of nucleotide substitutions per site

Fig. 2

Temporal phylogenetic reconstruction of CenAsiaIII Shigella sonnei overlaid with pseudogenes. A maximum clade credibility phylogenetic reconstruction of 144 CenAsiaIII S. sonnei is shown. Asterisks indicate posterior probability support >80% on major internal nodes. The red diamond indicates the internal node leading to the major fluoroquinolone-resistant (FQr) clone. Grey arrows designate branches on which certain quinolone resistance determining region (QRDR) mutations occurred. Lineage-defining pseudogenes and gene loss events are indicated above corresponding branches. Tip colours show the original geographical isolation regions of taxa (see key). The light orange and light yellow boxes highlight the European and Southeast Asian clonal expansions of CenAsiaIII, respectively. The dark dashed line separates the two populations (Pop1 and Pop2) in CenAsiaIII, defined by Bayesian hierarchical clustering

Fig. 3

Phylogeographical inference of CenAsiaIII Shigella sonnei. Stochastic mapping was used to estimate a the number of transitions between different geographical states: Africa, Southeast/East Asia, Europe, and South Asia, with transitions between states that are less than one considered non-significant and not represented in the plot, and b the proportion of time spent in each geographical state. In both panels, boxplots summarise results of 966 successful subsamplings of the phylogeny described in Fig. 1, and stochastic mapping was performed for each with 100 simulations