The nasopharyngeal microbiota of beef cattle before and after transport to a feedlot

Holman, Devin B.; Timsit, Edouard; Amat, Samat; Abbott, D. Wade; Buret, Andre G.; Alexander, Trevor W.

Presently, the metagenome of the NP in cattle has not been characterized. Therefore, in the absence of metagenome sequencing, we used PICRUSt to predict the functional potential for the NP microbiota using 16S rRNA gene sequences. The NSTI (nearest sequenced taxon index) value was used to provide a measure of how well the OTUs in the NP samples were accounted for in the reference genome database. The NSTI for the NP samples was 0.07 ± 0.03 (SD), which is lower than the average NSTI for mammalian microbiomes (0.14), thus indicating that in relative terms, the OTUs in the NP samples in the present study are well represented in the reference genome database [26] . Overall, 74.4% of the reads matched a 16S rRNA gene sequence in the Greengenes database. Following normalization by 16S copy number, the OTUs were assigned to KEGG orthologies (KOs). To determine which OTUs might be excluded due to the use of the closed reference approach required for PICRUSt we also used the open reference OTU picking algorithm with the Greengenes 16S rRNA gene database (data not shown). Only four OTUs had an overall relative abundance of greater than 0.1% and these OTUs were classified only at the family level (Chitinophagaceae, Microbacteriaceae, Moraxellaceae, and Weeksellaceae). When the number of KOs were compared by sampling time, there were significantly fewer KOs at day 14 than in day 0 and 2 samples ( Fig. 6a ; P < 0.05). These KOs were subsequently categorized into KEGG level 2 pathways, with membrane transport, carbohydrate metabolism, and amino acid metabolism as the three most relatively abundant pathways (Additional file 8: Fig. S4 ). There were 15 KEGG level 2 pathways that were differentially abundant among the four sampling times (Table 1 ; P < 0.05), the majority of which are involved in metabolism, with a number of them reduced at day 14 compared to day 0. In particular, there were fewer genes predicted to be involved in amino acid metabolism in the NP microbiome at day 14. Meanwhile, at day 14, genes required for translation, as well as replication and repair, were among those predicted to be more relatively abundant in the NP microbiome. Not surprisingly, given their relative abundance at day 14, Mycoplasma OTUs were the greatest contributors to the replication and repair pathway (data not shown). Bray-Curtis distances were used to compare changes in the predicted functional profile of the NP microbiome over time based on KOs (Fig. 6b) . Samples did group together by sampling time (R = 0.16; P < 0.001), although not as strongly as with OTU-based clustering (Fig. 4 ).