Multilocus sequence analysis of Pasteurella multocida demonstrates a type species under development.

The aim of the present study was to use multilocus sequence typing (MLST) of a diverse collection of Pasteurella multocida with regard to animal source, place and date of collection, including all available serovars of Carter, Heddleston, Little & Lyon, Namioka, Cornelius and Roberts, to further investigate the evolution of this species with a focus on two lineages, A (P. multocida subsp. multocida and P. multocida subsp. gallicida) and B (P. multocida subsp. septica), previously reported. Isolates of P. multocida (n = 116) including reference strains of major serotyping systems were investigated by MLST based on partial sequences of the genes adk, est, gdh, mdh, pgi, pmi and zwf, and 67 sequence types (STs) were observed. Phylogenetic analysis of these concatenated sequences confirmed the separation of groups A (41 STs, 71 isolates) and B (22 STs, 38 isolates) out of the 67 STs. All Carter serovars, 12 Heddleston serovars, all three Little-Lyon types, six out of seven Namioka serovars, all five Roberts types and all four Cornelius serovars were allocated to the A group, while group B included the remaining four Heddleston serovars, 6, 7, 8 and 13, in addition to Namioka type 8 : A. The overrepresentation of reference strains of serotyping systems in the A group contrasts with the high number of isolates obtained from diseased birds in the B group, the effect of which should be addressed in future vaccine development. Isolates from birds (25) dominated the B group, which also included four isolates from Felidae, whereas group A included isolates from all types of hosts. The evolutionary implications of the lack of capsular type D, pig and bovine isolates in group B, as well as its association with Aves and Felidae that also applied to the whole Rural Industries Research and Development Corporation (RIRDC) MLST database, need further investigation. The combination of rpoB and 16S rRNA gene sequence comparison as well as the developed PCR test assigned isolates to lineage A, represented by the type strain of P. multocida subsp. Multocida, or lineage B represented by the type strain of P. multocida subsp. septica. It was not possible to circumscribe either the A or B lineages with a set of conserved phenotypic characters, calling into question the validity of subspecies within P. multocida. Phylogenetic analysis carried out on individual MLST genes showed deviations as to single or multiple genes for 17 % of group A and 43 % of group B, indicating that lineage A probably developed from lineage B, and that major changes are ongoing. From a genotypical point of view, we conclude that P. multocida subsp. gallicida represents an artificial unit.