EFFECT OF ORAL UREA SUPPLEMENTATION ON THE ENDOMETRIAL TRANSCRIPTOME OF MARES

At ovulation detection (D0), oral treatment with urea was initiated and continued until D7. Mares received a treatment or control diet (n= 11 mares/group) in a crossover design. The treated group received urea (0.4 g/kg body weight) mixed with sweet feed and molasses, the control group received sweet feed and molasses alone. Blood samples were collected daily, one hour after feeding, for BUN determination. Uterine and vaginal pH were evaluated with an epoxy pH probe. Endometrial biopsies were taken transcervically one hour after the last feeding on D7. RNA sequencing of the endometrium of a subset of mares (n=6/group) was conducted. A uterine biopsy was collected from the base of the uterine horn seven days after the start of the treatment (n= 6 per treatment). Total cellular RNA was extracted from endometrial samples using a RNeasy Mini Kit (Qiagen, Gaithersburg, MD). A total of 1 μg of RNA was treated with DNase I (Ambion Inc., Austin, TX) for 30 minutes at 37℃ to remove genomic DNA according to manufacturer’s instructions. Paired-end reads with 150 nucleotides in length were produced. A TruSeq Stranded mRNA library prep kit (Illumina, San Diego, CA) was used to prepare libraries for mRNA sequencing. Libraries were loaded onto an Agilent DNA 1000 chip and validated on an Agilent 2100 Bioanalyzer (Agilent). Quantitation was performed with the Illumina Library Quantification Kit, ABI Prism qPCR Mix from Kapa Biosystems. Libraries were run on a NextSeq 500 v2 (Illumina) 300cycles High Output kit in a 2x150 base pairs with paired-end reads. The Fastq files were evaluated for read quality using FastQC 0.11.4. Subsequently, Trim Galore 0.4.1 was used for adapter and read quality trimming (Phred score threshold of 30). Reads were mapped to the Equus caballus reference genome (EquCab 3.0) using the software STAR 2.5.3a, then annotated with the equine reference annotation from NCBI using Cufflinks 2.2.1. Fragments per kilobase per million (FPKM) were used to determine the expression level of genes. Lastly, we used Cuffdiff 2.2.1 to calculate differentially expressed genes (DEG) between samples from the control and urea groups. Significance level was set at FDR-adjusted p-value of the test statistic < 0.1 using a Benjamini-Hochberg correction.