Genomic epidemiology of carbapenem- and/or colistin-resistant Pseudomonas aeruginosa in the Gauteng region, South Africa

Abstract

Carbapenem- and/or colistin-resistant Pseudomonas aeruginosa has emerged as a significant public health threat worldwide due to intrinsic and acquired resistance to carbapenems, colistin and other antibiotics. As there are no whole genome sequencing (WGS)-based genomic epidemiological studies delineating the evolution and resistome in carbapenem- and/or colistin-resistant P. aeruginosa in South Africa, such studies are needed. The study aimed to investigate the molecular characteristics and genomic epidemiology of carbapenem- and/or colistin-resistant P. aeruginosa isolates in the Gauteng region, South Africa using WGS. Eighty-two (82) carbapenem- and/or colistin-resistant P. aeruginosa isolates were obtained from private and public diagnostic laboratories in Pretoria, South Africa and were subjected to short read WGS. Antimicrobial susceptibility testing was routinely performed using the VITEK® 2 system. Raw reads were trimmed and were de novo assembled using SPAdes. The draft genomes were annotated using Prokka and were screened for antibiotic resistance genes (ARGs), plasmid replicons and virulence factor genes (VFGs) using ABRicate. The in silico multilocus sequence typing (MLST) was performed using the mlst software and the pubMLST database. Pangenome analysis was performed using Roary and Scoary. Phylogenetic analysis was performed using Snippy, Parsnp, Gubbins, SNP-sites, RAxML, snp-dists and Microreact. Bayesian Analysis of Population Structure (BAPS) clusters were identified using fastbaps. Genetic environments of carbapenemase genes were analysed using a set of bioinformatics software. Nine isolates were “possible pandrug-resistant” (9/82, 11%), 57 were extensively drug-resistant (XDR) (57/82, 70%), six were multidrug-resistant (MDR) (6/82, 7%) and ten were non-MDR (10/82, 12%). The WGS analysis revealed 20 different sequence types (STs) including high-risk clones like ST111, ST233, ST235, ST357, ST654 and ST773. The P. aeruginosa isolates harboured a wide range of intrinsic ARGs and acquired class A, B and D carbapenemase genes (blaAIM-1, blaIMP-5, blaKPC-19, blaNDM-1, blaVIM-2, blaVIM-4, blaVIM-16 and several blaOXA genes), of which blaVIM-2 was most common. The VFGs for alginate production, quorum sensing, iron acquisition, cell appendages, lipopolysaccharide and protein secretion systems were found. Pangenome analysis revealed that high-risk clones are associated with genes encoding conjugal transfer and phage proteins, multidrug efflux systems and virulence. Phylogenetic analysis revealed that the blaNDM-1-positive ST773 isolates and global ST773 strains form three clades and eight subclades; the blaNDM-1-positive carbapenem-resistant P. aeruginosa (CRPA) isolates and eight blaNDM-1-positive ST773 strains from Hungary, India, Nigeria, South Korea and USA grouped in the same clade. The rest of the CRPA isolates were grouped into three BAPS clusters, of which cluster 1 was spread across three provinces (Gauteng, Mpumalanga and North-West) of South Africa while cluster 2 was confined to the Gauteng region only. The rarely found blaAIM-1, blaKPC-19 and blaNDM-1 genes were found in putative integrative and conjugative/mobilizable elements (ICE/IME). In conclusion, this study presents the first genomic epidemiological study of carbapenem- and/or colistin-resistant P. aeruginosa isolates in the northeastern provinces of South Africa, reporting the genetic environment of blaNDM-1, blaAIM-1 and blaKPC-19. High prevalence rates of XDR phenotypes and dominance of high-risk clones across South African provinces highlights the urgent need for continuous WGS-based surveillance systems and genomic epidemiological studies in South Africa.