Signatures of negative frequency dependent selection in colonisation factors and the evolution of a multi-drug resistant lineage of Escherichia coli

Abstract

0Escherichia coli is a major cause of bloodstream and urinary tract infections globally. 2 1 3 3

Limited and strain-specific transcriptional and growth responses to acquisition of a multidrug resistance plasmid in genetically diverseEscherichia colilineages

“…Much of this fitness impact is driven by changes in transcription in the acquiring cell, both the need to transcribe genes on the plasmid, but also global effects on host cell transcription to offset the impact of carrying the plasmid (Buckner et al, 2018;Harrison et al, 2015;Millan et al, 2015). There are very few of these studies examining the impact of acquisition of multi-drug resistance plasmids on cells from genetically diverse strains (Buckner et al, 2018), with most evidence of adaptations that occur as a result of MDR plasmid acquisition stemming from large comparative population genomics studies (Feng et al, 2019;McNally et al, 2016McNally et al, , 2019. Acquisition of the ESBL plasmid varied among strains, although all strains stably maintained the plasmid once they had acquired it, likely due to the presence of two toxin-antitoxin systems on pLL35.…”

“…The MDR clones also carry clone specific alleles of key genes encoding traits involved in human colonisation, such as adhesins and iron acquisition (Feng et al, 2019;McNally et al, 2019;Zong et al, 2018). Comprehensive analysis of ST131 MDR clade C showed it differed from the drug susceptible clade A and B of the lineage in a number of unique alleles of genes involved in colonisation as well as anaerobic metabolism genes (McNally et al, 2019). MDR clones also contain unique intergenic-sequence alleles, which correlate with plasmids carried by strains (McNally et al, 2016) Zong et al, 2018).…”

Designing ecologically optimized pneumococcal vaccines using population genomics

“…The second was from the Maela refugee camp on the Thailand-Myanmar border, comprising 2,336 genomes carried by an unvaccinated population 11,12 . Based on the number of detected serotypes, approximately 3.47x10 9 and 1.05x10 13 13valent PCVs were possible in each, respectively. To enable computationally efficient simulation of these populations' changes in response to arbitrary vaccine designs, the multi-locus NFDS model of S. pneumoniae ecology was reimplemented in a deterministic form using ordinary differential equations 12 (Methods).…”

“…Despite moderate changes in antibiotic use during the epidemiological studies, antibiotic resistance phenotypes were maintained at approximately stable levels in the Massachusetts and Maela carriage populations 10,11,40,41 . Correspondingly, the model assumed that individual AMR phenotypes have stable frequencies in carriage that are maintained by NFDS 12,13,42 , such that elimination of vaccine-type resistant isolates would drive replacement by other non-susceptible genotypes. Given this constraint, we optimised vaccine formulations to minimise multidrug-resistant IPD across infants and adults.…”

“…They hypothesised that multi-locus negative frequency-dependent selection (NFDS) explained these observations, based on functional annotation of the accessory genome and the significantly improved predictive ability of an NFDS model compared with other ecological models 12 . Similar multi-locus NFDS models have also proved informative when applied to trends in Escherichia coli epidemiology 13 , and when reformulated to identify strains likely to invade a vaccine-disrupted population 14 .…”