An update on the MLST scheme for Pasteurella multocida

Pasteurella multocida is a cause of economically important diseases in almost all domestic livestock species, as well as wildlife. While a range of typing methods have traditionally been used, the development of a Multi-Locus Sequence Typing (MLST) scheme in 2010 represented the first standardised, sequence based, Web supported typing scheme. The initial scheme (termed the RIRDC MLST scheme) was based on 63 avian isolates from diseased Australian poultry and three international reference strains, which formed 29 Sequence Types (STs). The MLST database (http://pubmlst.org/pmultocida_rirdc/) now contains data from over 560 isolates that form 220 STs. The use of the scheme in published studies to date has demonstrated some key points: A) the highly clonal nature of haemorrhagic septicaemia (HS) isolates; B) bovine respiratory isolates are typically very distinct from HS isolates; C) evidence of host/niche association (i.e. some STs are associated with specific hosts); and D) the distinct genotype of P. multocida isolates of capsule type B from calf pleuritis and peritonitis cases in New Zealand. The continued use of this MLST scheme by research groups around the world will add to our understanding of the population structure and host associations of this major veterinary pathogen.

Pasteurella multocida is an iconic bacterium that has links back to the beginnings of veterinary microbiology with Louis Pasteur and his pioneering work on a vaccine based on an attenuated strain¹. The bacterium is a Gram negative coccobacillus that is associated with a range of diseases in wild and domestic animals and is also part of the normal oropharyngeal flora of these animals². There are many published typing schemes, both traditional methods such as serotyping and phenotyping, as well as molecular based methods³. Until recently, the generally agreed “gold standard” typing method of Multi-locus Sequence Typing (MLST)⁴ had not been applied to P. multocida. A recent Australian study, completed in 2010⁵, developed the first Web-supported MLST for P. multocida. This article looks at that initial publication and reviews developments since that initial publication.

The development of an MLST scheme for P. multocida was made easier by the existence of an earlier multi-locus enzyme electrophoresis (MLEE) scheme for P. multocida⁶. MLEE schemes were a phenotypic precursor of MLST schemes and were based on differences in amino acid sequences in enzymes detected by starch gel electrophoresis⁶. While cumbersome, the P. multocida MLEE scheme generated novel information – the first evidence that the avian P. multocida population consisted of two quite different clusters⁶ that did not match the earlier proposal of three subspecies – subspecies multocida, gallicida and septica – based on DNA/DNA hybridization data⁷. As the earlier MLEE scheme and the RIRDC MLST scheme were based on the same house-keeping enzymes, it is not surprising that the initial RIRDC MLST scheme (based on 63 Australian avian isolates) recognised the same two major clusters within avian P. multocida⁶. Furthermore, retrospective application of the MLST to a well characterised set of P. multocida isolates from eight fowl cholera outbreaks showed that the RIRDC MLST scheme provided epidemiologically relevant typing results that matched those provided by a range of other genotypic methods⁶.

As a result of both formally published studies and addition of MLST results from unpublished studies, the current RIRDC P. multocida MLST scheme contains 560 P. multocida isolates in 220 STs. The isolates have come from 41 different hosts (including both domestic livestock and wildlife). The largest number of isolates are from cattle (around 25% of isolates) with two other domestic livestock species well represented (pigs at 17% and chickens at 16%). Isolates from wild animals include those from chimpanzees (four isolates), tigers (two isolates) and elephants (two isolates).

The largest single published study using the RIRDC MLST database has resulted from the work of the Moredun Research Institute, which examined 201 P. multocida isolates, 128 of which were bovine respiratory isolates from the UK, France and the USA⁸. The Moredun study found that the bovine respiratory disease isolates were clonal, with 105 isolates belonging to a single clonal complex (CC13)⁸. These respiratory isolates were quite distinct from haemorrhagic septicaemia (HS) isolates⁸. The clonal nature of HS isolates is confirmed within the complete database, with 46 of 49 HS isolates allocated to a single ST – ST122. The other key finding from the Moredun study was the finding that 58 of the 62 STs detected in the study were associated with only one host type⁸.

Perhaps the most unusual use of the RIRDC MLST scheme to date has been the examination of four P. multocida isolates from wild chimpanzees in Côte D’Ivoire that were associated with two severe respiratory disease outbreaks in the same national park in 2004 and 2009⁹. The MLST results showed that two genotypes were involved with the 2004 outbreak, with one of these genotypes re-appearing in the 2009 outbreak, findings also supported by other typing methods.

In another example of the power of MLST, a New Zealand study examined very unusual outbreaks of calf peritonitis and pleuritis associated with P. multocida of capsular type B¹⁰. Until this report, the accepted paradigm was that capsule type B was always associated with HS outbreaks. The use of the RIRDC MLST scheme showed that these atypical serovar B types did not belong to the clone associated with HS (ST122), but rather to a unique ST not previously reported¹⁰.

There is no doubt that our understanding of the population structure and the global epidemiology of P. multocida will increase as more use is made of an array of new and emerging technologies. The RIRDC MLST scheme will be a key plank that helps provide an overarching view of the population structure and possible host associations, particularly as more data are added.