Endometrial Gene Expression Predicts Pregnancy Outcome in Brahman Cows

There is a small number of studies describing the uterine biology of Bos indicus cows. We hypothesized that there is a transcriptional signature in endometrial epithelial cells 4 days after estrus (D4) that predicts the ability of the cow to remain pregnant/artificial insemination (AI). Brahman cows were submitted to an estrous synchronization protocol and AI. On D4 cows were submitted to endometrial cytology. Pregnancy/AI was diagnosed on day 30 and endometrial cytology samples were submitted to RNA-seq (n = 32 nonpregnant and n = 32 pregnant). Based on RNA-seq we performed targeted analysis using pathways previously reported in the literature and untargeted analysis using Ingenuity Pathway Analysis. A total of 975 genes were significantly associated (p ≤ 0.1) with pregnancy/AI, 64.9% of them (620/975) showed a negative association. Targeted and untargeted analysis showed a downregulation of Th2 immune response and activation of cholesterol biosynthesis in pregnant cows. Prediction analysis resulted in greater accuracy for the targeted transcripts than the whole transcriptome (0.91 vs. 0.86) but reduced precision (0.64 vs. 0.74). In conclusion, the endometrial receptivity was predominantly marked by an overall reduction in the molecular response. Th2 response and cholesterol biosynthesis are promising pathways to understand uterine biology, and the use of a set of genes rather than a single gene appears to be the future for prediction of pregnancy in bovine.