Abstract
BACKGROUND: Brucellosis is the most common bacterial zoonoses worldwide. Bovine brucellosis caused by Brucella abortus has far reaching animal health and economic impacts at both the local and national levels. Alongside traditional veterinary epidemiology, the use of molecular typing has recently been applied to inform on bacterial population structure and identify epidemiologically-linked cases of infection. Multi-locus variable number tandem repeat VNTR analysis (MLVA) was used to investigate the molecular epidemiology of a well-characterised Brucella abortus epidemic in Northern Ireland involving 387 herds between 1991 and 2012. RESULTS: MLVA identified 98 unique B. abortus genotypes from disclosing isolates in the 387 herds involved in the epidemic. Clustering algorithms revealed the relatedness of many of these genotypes. Combined with epidemiological information on chronology of infection and geographic location, these genotype data helped to identify 7 clonal complexes which underpinned the outbreak over the defined period. Hyper-variability of some VNTR loci both within herds and individual animals led to detection of multiple genotypes associated with single outbreaks. However with dense sampling, these genotypes could still be associated with specific clonal complexes thereby permitting inference of epidemiological links. MLVA- based epidemiological monitoring data were congruent with an independent classical veterinary epidemiology study carried out in the same territory. CONCLUSIONS: MLVA is a useful tool in ongoing disease surveillance of B. abortus outbreaks, especially when combined with accurate epidemiological information on disease tracings, geographical clustering of cases and chronology of infection.