Erysipelothrix rhusiopathiae can cause erysipelas in animals and erysipeloid in humans. Since its recurrence in 2012, swine erysipelas has caused serious losses within the pig industry in China. The aim of this study was to perform multilocus sequence typing and understand the virulence and antimicrobial susceptibility of E. rhusiopathiae isolates in China. Multilocus sequence typing (MLST) of a total of 120 strains was performed, and as a result, three different sequence types were identified, of which ST48 was the main one. Five isolates of each MLST type were randomly selected to be used to challenge mice. ST48 was associated with a higher virulence. Antimicrobial susceptibility was tested using a microdilution technique and, to analyze the resistance mechanism, six strains were selected for genome sequencing. A comparison of the six genomes indicated the presence of a suspected macrolide resistance gene, namely, Erm(A)-like, in erythromycin-resistant strains, which increased the minimum inhibitory concentration (MIC) of erythromycin against E. coli C600 at least four-fold. In addition, three mutations (gyrA86T-I, gyrA90D-N, and parC81S-I) were observed in the quinolone resistance-determining regions (QRDRs) of gyrA and parC in quinolone-resistant strains. After the gyrA gene with the 86T-I mutation or the parC gene with the 81S-I mutation was transfected into E. coli C600, the MIC of enrofloxacin against this strain increased at least two-fold. Our findings provide a theoretical basis for developing antibacterial drugs and may contribute to the clinical prevention and control of E. rhusiopathiae.

Erysipelothrix rhusiopathiae is a causative agent of erysipelas in animals and erysipeloid in humans. However, current information regarding E. rhusiopathiae pathogenesis remains limited. Previously, we identified two E. rhusiopathiae strains, SE38 and G4T10, which were virulent and avirulent in pigs, respectively. Here, to further study the pathogenic mechanism of E. rhusiopathiae, we sequenced and assembled the genomes of strains SE38 and G4T10, and performed a comparative genomic analysis to identify differences or mutations in virulence-associated genes. Next, we comparatively analyzed 25 E. rhusiopathiae virulence-associated genes in SE38 and G4T10. Compared with that of SE38, the spaA gene of the G4T10 strain lacked 120 bp, encoding repeat units at the C-terminal of SpaA. To examine whether these deletions or splits influence E. rhusiopathiae virulence, these 120 bp were successfully deleted from the spaA gene in strain SE38 by homologous recombination. The mutant strain ΔspaA displayed attenuated virulence in mice and decreased adhesion to porcine iliac artery endothelial cells, which was also observed using the corresponding mutant protein SpaA’. Our results demonstrate that SpaA-mediated adhesion between E. rhusiopathiae and host cells is dependent on its C-terminal repeat units.