A potential new species of porcine Actinobacillus defined by multi-locus sequence analysis

In recent years the submission of Actinobacillus-like strains sourced from the respiratory tract of pigs to our laboratory has increased. Our usual diagnostic method for identification of non-routine organisms is 16S rDNA sequencing. However, this method does not have the discriminatory power to separate very closely related species in the Actinobacillus complex. This has prompted a project to improve the identification of Actinobacillus-like strains by using multi-locus sequencing analysis (MLSA). The recN, rpoA and thdF housekeeping genes were chosen for this analysis due to their proven ability to predict whole-genome DNA-DNA similarity.

For a total of 36 field isolates, identified by the 16S rDNA sequencing as unusual species – A. porcitonsillarum, A. minor, A. porcitonsillarum/minor complex, A. porcinus, A. indolicus and A. rossi – three genes, recN, rpoA and thdF, were sequenced and aligned to publicly available data of type strains within the genus Actinobacillus to make multiple alignments of DNA sequences constructed by Clustal in Geneious version 8.0.5. The genome similarity index was calculated according to Kuhnert and Korczak (2006). Two species specific PCR for Haemophilus parasuis were used to exclude this species from the study.

None of the strains were positive in the species specific H. parasuis PCR. A potential new species, consisting of 17 isolates, was identified with a genome similarity index of 0.56 to the closest related type strain – A. indolicus (similarity index > 0.4 and > 0.9 indicating the same genus and species, respectively; Table 1). The type strains of H. parasuis and A. indolicus formed a group with nine isolates. A further seven isolates did not fit into a group due to lack of congruence of the thdF gene phylogeny with recN and rpoA and their identity remains uncertain. The analysis also pointed to the inadequacy of the 16S rDNA identification as none of the strains identified as A. porcitonsillarum/minor were confirmed as such. It also highlighted that no single gene by itself can be used for identification.

Table 1.  Sequence of isolates and type strains were aligned with CLUSTAL. Values of greater than 96% belonged to the same species, values between 84 to 96% were uncertain and values below 84% had no similarity

Further work is in progress to look at the pathology associated with the 17 strains identified as a potential new species, to determine the significance of these organisms for the pork industry.