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.

Procedures involving primer design, PCR amplification with gene-specific primers, and standard dye-terminator cycle sequencing (Sanger) of all 10 bovine TLRs have previously been described [23] [24] [25] 60] . For this study, we synthesized gene-specific amplification primers with a unique 10 bp 59 barcode (Roche MIDs) for each of the 10 bovine TLR genes (Table S5) . Thereafter, we standardized all 96 discovery panel DNAs to 50 ng/ml and created three DNA pools, with each pool consisting of 32 elite sire DNAs mixed at equal concentrations. Notably, larger-scale DNA pooling in a human amplicon study supports the accuracy and reliability of this approach when coupled with Roche 454 pyrosequencing [61] . Three bovine DNA pools were used to amplify all TLR targets via barcoded primers (Table S5) , with PCR conditions and thermal parameters as previously described [23] [24] [25] 60] . Targets that were intolerant to the addition of 59 oligonucleotide barcodes for PCR amplification were amplified using standard primers in conjunction with downstream dye-terminator cycle sequencing methods previously described [23] [24] [25] 60] , with one exception: A second set of DNA pools (n = 12) was created, with each pool containing equal concentrations of DNA from eight elite sires derived from the sequencing discovery panel. Importantly, both sets of DNA pools (Sanger and Roche 454) were seeded with one or more reference DNAs that had previously been sequenced and/or SNP genotyped across all 10 bovine TLR genes [23] [24] [25] 60] , which collectively included$12 reference DNAs possessing 216 validated diallelic variants (212 SNPs + 4 indels) [30] . All amplicons were purified using the Qiaquick PCR purification kit (Qiagen, Valencia, CA) as previously described [24, 25] , and the concentrations were estimated by Nanodrop. For preparation of a Roche 454 Titanium fragment library, we standardized all barcoded amplicons to 10 ng/ml and devised a normalization procedure that accounted for differences in amplicon size (Table S1 ). Because the TLR amplicons differed in size, an adjustment was necessary to ensure balanced 454 pyrosequencing results. Specifically, using amplicon size, we computed the mean (bp) and standard deviation (SD; bp) across all PCR targets. Thereafter, any amplicon deviating from the mean by$0.5 SDs in either direction was subject to proportional adjustment within the fragment library (Table S1 ). The direction of adjustment (plus or minus) was determined by the direction of the deviation (i.e., smaller = proportionally less template; larger = proportionally more template; Table S1 ). Because the emulsion PCR process involved in the preparation of Roche 454 Titanium fragment libraries favors smaller fragments, amplicons smaller than the mean by$0.5 SDs must be proportionally reduced in the final library, whereas the opposite is true for larger amplicons. Following normalization, the bovine TLR sequencing library was constructed via random ligation of sequencing adaptors provided with the GS FLX Titanium library kit (Roche Applied Science, Indianapolis, IN). All library preparation, emulsion PCR, quantitation, and sequencing steps followed the manufacturer's protocol (Roche Applied Science).