Characterisation of the genetic diversity of Brucella by multilocus sequencing

Abstract

Background: Brucella species include economically important zoonotic pathogens that can infect a wide range of animals. There are currently six classically recognised species of Brucella although, as yet unnamed, isolates from various marine mammal species have been reported. In order to investigate genetic relationships within the group and identify potential diagnostic markers we have sequenced multiple genetic loci from a large sample of Brucella isolates representing the known diversity of the genus.

Results: Nine discrete genomic loci corresponding to 4,396 bp of sequence were examined from 160 Brucella isolates. By assigning each distinct allele at a locus an arbitrary numerical designation the population was found to represent 27 distinct sequence types (STs). Diversity at each locus ranged from 1.03-2.45% while overall genetic diversity equated to 1.5%. Most loci examined represent housekeeping gene loci and, in all but one case, the ratio of non-synonymous to synonymous change was substantially <1. Analysis of linkage equilibrium between loci indicated a strongly clonal overall population structure. Concatenated sequence data were used to construct an unrooted neighbour-joining tree representing the relationships between STs. This shows that four previously characterized classical Brucella species, B. abortus, B. melitensis, B. ovis and B. neotomae correspond to well-separated clusters. With the exception of biovar 5, B. suis isolates cluster together, although they form a more diverse group than other classical species with a number of distinct STs corresponding to the remaining four biovars. B. canis isolates are located on the same branch very closely related to, but distinguishable from, B. suis biovar 3 and 4 isolates. Marine mammal isolates represent a distinct, though rather weakly supported, cluster within which individual STs display one of three clear host preferences.

Conclusion: The sequence database provides a powerful dataset for addressing ongoing controversies in Brucella taxonomy and a tool for unambiguously placing atypical, phenotypically discordant or newly emerging Brucella isolates. Furthermore, by using the phylogenetic backbone described here, robust and rationally selected markers for use in diagnostic assay development can be identified.

Figure 1

Polymorphic sites detected at the nine loci examined in this study. All polymorphic sites are shown relative to the ST1 sequence. Polymorphic sites are shown, while dots indicate nucleotides identical to ST1. The number of strains possessing each ST is shown at the end of each sequence string while the status of each site as a potential synonymous or non-synonymous change is shown by S or N respectively. The numbers above the gene designations represent the base number in the 4,396 bp concatenated sequence.

Figure 2

Unrooted phylogenetic reconstruction of the relationships between STs. This tree was constructed with the concatenated sequence data of the nine loci (4,396 bp) using the neighbour joining approach. The Jukes-Cantor model, which is based on the assumption that all nucleotide substitutions are equally likely, was used to determine genetic distances The percentage bootstrap confidence levels of internal branches were calculated from 1,000 resamplings of the original data.