Toxic Fescue Exposure Alters Vaginal Microbial Communities of Crossbred Beef Cows.

Consumption of toxic fescue by beef cattle results in adverse physiological effects such as: reduced reproductive success, vasoconstriction, poor body condition, hyperthermia, decreased prolactin, and reduced hair shedding. The purpose of this study is to characterize bacterial community of the reproductive tract in animals grazing toxic or novel fescue and determine the microbiota’s relationship to host phenotype. One-hundred fall-calving crossbred cows were allocated to graze toxic fescue (toxic: n=50) or novel (non-toxic: n= 50) fescue pastures for five months (March-August). Treatments were blocked by sire breed (Charolais or Hereford) and parity (first, second or third). Animals rotated pastures every other week and pastures were monitored monthly for ergovaline. One novel pasture was contaminated with toxic fescue (ergovaline > 300ppb), and 21 animals were exposed to toxic pasture 2 weeks each month. This group was defined as an alternating treatment (Alt: n=21). Rectal temperatures, hair scores, body condition scores, and body weights were collected monthly. In August, the reproductive tract was swabbed for 16s rRNA-based microbiota analysis. The V4 region was sequenced using the Illumina Miseq platform. Data was processed using mothur v.1.39.5. Selected operational taxonomic units (OTUs), evenness, and diversity indexes were analyzed in Proc Mixed of SAS. Class variables included: treatment, breed, parity, pregnancy, and breed*treatment interaction. Significant difference in vaginal bacterial community structure was observed between toxic and novel treatments based on Bray-Curtis distance (ANOSIM, R = 0.445697, P< 0.001). Toxic communities also differed from those of the Alt treatment (ANOSIM, R= 0.429223, P< .001), however, novel and Alt treatments did not differ (R = -0.0077, P =.536). Charolais exhibited greater (p= 0.0427) Shannon diversity than Herefords, and diversity increased with parity number with no difference (p=0.7312) observed between parities one and two. Smith-Wilson evenness was greater (p=0.0008) for novel animals than toxic animals. Random forest indicated OTU’s from Pseudomonas, Ureaplasma, and Burkhoderia were most predictive of pasture treatment. The suppression of Pseudomonas was the best indicator of toxin exposure. Novel animals hosted 16.5 times more (p=0.0057) Pseudomonas than toxic animals. Ureaplasma was the second most abundant OTU and second most predictive of pasture type. Novel animals hosted 3.8 times more (p=0.0004) than toxic animals. Burkholderia was the most abundant OTU and the third most predictive of treatment. Burkholderia represented 41.74% of the bacterial community in the toxic treatment reproductive tracts and 14.67% in novel animals. Further analysis is underway to determine function and validate discoveries.