Genomic epidemiology of Mycobacterium bovis infection in sympatric badger and cattle populations in Northern Ireland

Abstract

Bovine tuberculosis (bTB) is a costly, epidemiologically complex, multi-host, endemic disease. Lack of understanding of transmission dynamics may undermine eradication efforts. Pathogen whole-genome sequencing improves epidemiological inferences, providing a means to determine the relative importance of inter- and intra-species host transmission for disease persistence. We sequenced an exceptional data set of 619 Mycobacterium bovis isolates from badgers and cattle in a 100 km2 bTB ‘hotspot’ in Northern Ireland. Historical molecular subtyping data permitted the targeting of an endemic pathogen lineage, whose long-term persistence provided a unique opportunity to study disease transmission dynamics in unparalleled detail. Additionally, to assess whether badger population genetic structure was associated with the spatial distribution of pathogen genetic diversity, we microsatellite genotyped hair samples from 769 badgers trapped in this area. Birth death models and TransPhylo analyses indicated that cattle were likely driving the local epidemic, with transmission from cattle to badgers being more common than badger to cattle. Furthermore, the presence of significant badger population genetic structure in the landscape was not associated with the spatial distribution of M. bovis genetic diversity, suggesting that badger-to-badger transmission is not playing a major role in transmission dynamics. Our data were consistent with badgers playing a smaller role in transmission of M. bovis infection in this study site, compared to cattle. We hypothesize, however, that this minor role may still be important for persistence. Comparison to other areas suggests that M. bovis transmission dynamics are likely to be context dependent, with the role of wildlife being difficult to generalize.

Whole-Genome sequencing in routine Mycobacterium bovis epidemiology – scoping the potential

“…Clusters 4.1 and 4.2, while being separate after the application of the cut-off, were still closelyrelated; indeed the splitting of these clusters was most likely caused by the 2021 ovine sample in cluster 4.2 having a longer branch length that other isolates. These two closely-clusters (alongside clusters 4.3 and 4.4) are consistent with sympatric cattle and badgers sharing closely-related M. bovis by virtue of bidirectional transmission [15,17,20]. The ovine isolates, which all come from the same flock, are divided across clusters 4.1 and 4.2 and are closely-related to earlier sampled bovine isolates from a neighbouring herd.…”

“…In this preliminary study using a relatively small number of M. bovis genomes, we report on the application of WGS to a small number of widely dispersed Northern Irish isolates from a common lineage of M. bovis, defined by a single spoligotype and MLVA genotype/ spoligotype combination. We have previously observed that, owing to the clonality of M. bovis, these common MLVA/spoligotype variants (genotypes) represent ancestral nodes deep in the pathogen's phylogeny [20,27], defining monophyletic lineages that can be further characterised by WGS. Here, we contrast the utility of MLVA/spoligotyping methods and WGS for disease tracing within one such common lineage.…”

“…This is a general feature of bacterial pathogens when compared to faster evolving viruses, which can leave a more substantial trail of mutations to facilitate finer scale disease tracing [21]. The polytomies that result from such genetically homogeneous, short temporal window outbreak data can affect the robustness of the Bayesian phylogenetic trees used to infer transmission dynamics [20,22] and it is therefore accepted that in such settings genomic data from MTBC organisms can have relatively limited utility [23,24]. However, it is evident that when phylodynamic methods are successfully applied to M. bovis, this is best enabled by using multiple samples from the same locale, collected from multiple hosts over a wide temporal window of 10-30 years [12][13][14][15][16][17][18][19][20].…”

“…Recently, much effort has been dedicated to the use of WGS to inform on transmission dynamics within the M. bovis epi-system in a variety of international locations [12][13][14][15][16][17][18][19][20]. The specific goal in these studies has been to try to unpick the Gordian knot of the relative importance of intra-and inter-species transmission in complex, sympatric, multi-host livestock and wildlife populations.…”

“…The polytomies that result from such genetically homogeneous, short temporal window outbreak data can affect the robustness of the Bayesian phylogenetic trees used to infer transmission dynamics [20,22] and it is therefore accepted that in such settings genomic data from MTBC organisms can have relatively limited utility [23,24]. However, it is evident that when phylodynamic methods are successfully applied to M. bovis, this is best enabled by using multiple samples from the same locale, collected from multiple hosts over a wide temporal window of 10-30 years [12][13][14][15][16][17][18][19][20]. This facilitates detection of robust molecular clock behaviour, which is essential for inferring transmission dynamics [25,26].…”