Background: Multilocus sequence typing (MLST) is well-established for Pasteurella multocida but remains undeveloped for Pasteurella canis. We established MLST for P. canis using isolates from humans and companion animals in Japan and Korea to gain insights into its population biology. Methods: We analyzed 39 and 22 isolates from companion animals and humans, respectively. We selected seven housekeeping genes—adk, aroA, deoD, gdhA, g6pd, mdh, and pgi—used in P. multocida MLST. Primer pairs for PCR amplification and sequencing were designed based on conserved sites in 10 whole-genome sequences. We determined frag-ment sequences, variable sites, allelic profiles, and sequence types (STs) of each isolate. A phylogenetic tree of concatenated sequences was constructed using the goeBURST algo-rithm to identify STs and clonal complexes (CCs). ompAM, encoding outer membrane protein A, was genotyped for molecular characterization. Results: The sequenced fragment lengths and allele numbers of the seven genes wereHallym University College of Medicine, 424, 451, 483, 439, 429, 419, and 440 bp and 16, 13, 15, 18, 22, 19, and 18, respec-tively. ST1–ST47, including CC2, CC10, CC18, CC31, and CC33, were diversely distributed among the isolates from different hosts/countries. In the seven-gene phylogenetic tree, apart from P. multocida, all isolates clustered together. goeBURST diagrams revealed di-verse ST distributions among different hosts (animal/human) and countries (Japan/Ko-rea/others). We found clusters 1–4 in ompA genotyping, indicating that MLST discrimination is higher than ompA typing discrimination. Conclusions We established MLST for P. canis isolates from humans and companion ani-This is an Open Access article distributed under mals in Japan and Korea, thereby providing a robust tool for population biology studies.

Background: Comparative analysis of virulence factors (VFs) between Pasteurella canis and Pasteurella multocida are lacking, although both cause zoonotic infections. We determined the virulence-associated genome sequence characteristics of P. canis and assessed the toxin gene prevalence unique to P. canis among clinical isolates of P. canis and P. multocida. Methods: We selected 10 P. canis and 16 P. multocida whole-genome sequences (WGSs) from the National Center for Biotechnology database. The VFanalyzer tool was used to estimate P. canis-characteristic VFs. Amino acid sequences of VFs were compared with multiple-aligned sequences. The genome structure containing P. canis-characteristic and adjacent loci was compared to the corresponding P. multocida genome structure. After designing primer sequences and assessing their accuracy, we examined the gene prevalence of the P. canis-characteristic VFs using PCR among clinical isolates of P. multocida and P. canis. Results: Using VFanalyzer, we found virulence-associated cytolethal distending toxin (cdt)A–cdtB–cdtC loci common to all P. canis WGSs that were not found in P. multocida WGSs. Similarities in the multiple alignments of CdtA–CdtB–CdtC amino acid sequences were found among the 10 P. canis WGSs. Shared or similar loci around cdtA–cdtB–cdtC were identified between the P. canis and P. multocida genome structures. The PCR-based cdtA–cdtB–cdtC prevalence differed for P. canis and P. multocida clinical isolates. Conclusions P. canis-specific cdtA–cdtB–cdtC prevalence was identified among clinical isolates. These three loci may be unique toxin genes and promising targets for the rapid identification of P. canis in clinical settings.