Comparing PFGE, MLST, and WGS in monitoring the spread of macrolide and rifampin resistant Rhodococcus equi in horse production

Abstract

BackgroundRhodococcus equi (R. equi) infections are endemic in many horse facilities in the United States resulting significant economic loses annually. Currently, there is no commercial vaccine available and the emergence of isolates that are resistant to the current treatment and prophylaxis using antibiotics prompts closer surveillance of this pathogen. ObjectiveThis study compares three different genotyping techniques, Pulsed Field Gel Electrophoresis (PFGE), Multilocus Sequence Typing (MLST) and whole genome SNP-based phylogeny to determine the most accurate method to monitor the spread of macrolide-and-rifampin-resistant R. equi. Methods16 macrolide and rifampin-resistant and 6 susceptible R. equi and their Illumina Miseq whole genome sequences were used in this study. The isolates were sub-typed by PFGE with VspI and a dendrogram based on their similarities generated. Additionally, three phylogenetic trees were constructed using CSI phylogeny on (i) whole genome sequences (WGS), (ii) in silico MLST sequences and (iii) MLST sequences obtained after PCR-amplification and Sanger sequencing. ResultsPFGE identified 18 different genetic profiles and grouped the 22 isolates into 3 clusters independently of their susceptibilities. The phylogenetic trees built from WGS and MLST data showed similar topology, separating the isolates into 2 major clades in accordance with their susceptibility profiles (susceptible and resistant). However, only the trees generated with next generation sequencing data could detect the clonality of the resistant isolates.