Comparison of Whole Genome Sequencing and Repetitive Element PCR for Multidrug-Resistant Pseudomonas aeruginosa Strain Typing

Abstract

Hospital-acquired infections pose significant costly global challenges to patient care. Rapid and sensitive methods to identify potential outbreaks are integral to infection control measures. Whole-genome sequencing (WGS)-based bacterial strain typing provides higher discriminatory power over standard nucleotide banding pattern-based methods such as repetitive sequence-based PCR (rep-PCR). However, integration of WGS into clinical epidemiology is limited by the lack of consensus in methodology and data analysis/interpretation. In this study, WGS was performed on genomic DNA extracted from 22 multidrug-resistant Pseudomonas aeruginosa (MDR-PA) isolates using next-generation sequencing. Resulting high-quality reads were analyzed for phylogenetic relatedness using a whole-genome multilocus sequence typing (wgMLST)-based software program and single-nucleotide variant phylogenomics (SNVPhyl). WGS-based results were compared with conventional MLST and archived rep-PCR results. Rep-PCR identified three independent clonal clusters of MDR-PA. Only one clonal cluster identified by rep-PCR, an endemic strain within the pediatric cystic fibrosis population at Texas Children's Hospital, was concordantly identified using wgMLST and SNVPhyl. Results were highly consistent between the three sequence-based analyses (conventional MLST, wgMLST, and SNVPhyl), and these results remained consistent with the addition of 74 MDR-PA genomes. These WGS-based methods provided greater resolution for strain discrimination than rep-PCR or standard MLST classification, and the ease of use of wgMLST software renders it clinically viable for analysis, interpretation, and reporting of WGS-based strain typing.