BackgroundMultilocus 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.MethodsWe 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 fragment sequences, variable sites, allelic profiles, and sequence types (STs) of each isolate. A phylogenetic tree of concatenated sequences was constructed using the goeBURST algorithm to identify STs and clonal complexes (CCs). ompA, encoding outer membrane protein A, was genotyped for molecular characterization.ResultsThe sequenced fragment lengths and allele numbers of the seven genes were 424, 451, 483, 439, 429, 419, and 440 bp and 16, 13, 15, 18, 22, 19, and 18, respectively. 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 diverse ST distributions among different hosts (animal/human) and countries (Japan/Korea/ others). We found clusters 1-4 in ompA genotyping, indicating that MLST discrimination is higher than ompA typing discrimination.ConclusionsWe established MLST for P. canis isolates from humans and companion animals in Japan and Korea, thereby providing a robust tool for population biology studies.