Evolution of the Bovine TLR Gene Family and Member Associations with Mycobacterium avium Subspecies paratuberculosis Infection

Fisher, Colleen A.; Bhattarai, Eric K.; Osterstock, Jason B.; Dowd, Scot E.; Seabury, Paul M.; Vikram, Meenu; Whitlock, Robert H.; Schukken, Ynte H.; Schnabel, Robert D.; Taylor, Jeremy F.; Womack, James E.; Seabury, Christopher M.

Median joining haplotype networks (Figures 2,3 ,4, Figure S1 ; Table S4 ) constructed for all 10 genes revealed that: 1) The specialized B. t. taurus beef and dairy breeds cannot be genetically discriminated despite an average polymorphism density (266 SNPs + 4 indels; see Table 2 ) of one variable marker per 158 bp; 2) Haplotype sharing occurs among B. t. taurus and B. t. indicus breeds at all 10 loci; 3) Shared haplotypes were often the highest frequency haplotype(s) within a network; 4) Despite haplotype sharing between the subspecific lineages, the 250 Kyr divergence [36] between B. t. taurus and B. t. indicus was evident in most, but not all, haplotype networks (i.e., TLR1-7, TLR10). With very few exceptions (i.e., TLR3 Network 1, TLR4, TLR10), the high frequency network nodes demonstrating subspecific haplotype sharing often included at least two indicine sires. Using summary data derived from the median joining networks (Table S4) , we estimated the relationship between the total number of discrete TLR haplotypes predicted (TLR1-10) in seven major U.S. taurine beef breeds [37] (Angus, Charolais, Gelbvieh, Hereford, Limousin, Red Angus, Simmental), and four U.S. taurine dairy breeds (Braunvieh, Brown Swiss, Holstein, Shorthorn), and found a significant correlation (r = 0.71, P#0.0224). This correlation was driven by the large number of haplotypes predicted to be shared among the beef and dairy breeds. For the investigated beef breeds, we predicted 84 discrete haplotypes across all 10 TLR loci, and at least 60 (71.4%) were predicted to be shared with the four dairy breeds. However, we also detected disparities between the numbers of haplotypes predicted for TLR4 and TLR5, with the dairy breeds possessing 3.8X and 2.3X more discrete haplotypes for these loci, respectively, than did our beef cattle. Exclusion of these two outlying loci resulted in a nearly perfect correlation (r = 0.98, P,0.0001) between the numbers of discrete haplotypes predicted in beef and dairy breeds across the remaining TLR loci. Interestingly, the single haplotype possessing the TLR5 putative nonsense mutation was almost exclusively predicted in Holstein cattle ( Figure S1 , TLR5 Node Q; n = 53 Holstein, n = 1 Braford).