RNA-Seq analysis of bovine oocyte transcriptome reveals that differences between heifers and repeat breeders are limited to a few key transcripts

Maternal transcripts are accumulated during oocyte growth and drive early embryonic development; therefore, their characterisation is a relevant factor for predicting fertility. DNA microarrays have been the method of choice for transcriptional profiling, but this method has some limitations when applied to domestic species because it relies upon existing knowledge about genome sequence and offers a limited quantitative evaluation. These limits are overcome by next-generation sequencing technology. The aim of the work was to define a reference standard for bovine fertility determining the list and the level of transcripts stored in fully grown oocytes collected from heifers (H) and to compare this pattern with that of adult repeat breeders (RB). Oocytes were collected by ovum pick-up (OPU) from 5 Italian Dappled Red heifers of 11 to 15 months of age that became pregnant at the following oestrus and from 4 adult cows of the same breed with an age of 4 to 7 years, classified as repeat breeders after they failed to become pregnant for a minimum of 3 consecutive AI. In both groups, oocytes were aspirated from follicles of 4 to 6mm in diameter. Each oocyte was carefully denuded and immediately snap frozen in liquid nitrogen. Oocytes from each animal were pooled together (range 4 to 11) and analysed as a single sample. Total RNA extraction was performed by RNeasy Micro Kit (Qiagen, Valencia, CA, USA). Amplified cDNA, from both mRNA and non-polyadenylated transcripts, was prepared starting from total RNA using the Ovation RNA-Seq System V2 (NuGEN Inc., San Carlos, CA, USA). Purified cDNA was ligated directly into an Illumina sequencing library using TruSeq DNA Sample Prep kit (Illumina Inc., San Diego, CA, USA). Sequencing was performed on Illumina HiSEqn 2000 in the 50-bp long single-read set-up, at a 4-plex of multiplexing level, producing 30 to 40 million reads per sample. Data were annotated using the cDNA ENSEMBL UMD 3.1.67 database. On average, the number of transcripts present in each sample was 15438±766 in H and 15624±768 in RB oocytes. Nineteen thousand one hundred sixty-one transcripts were detected at least in one sample, and 12174 were detected in all samples. The comparison between H and RB showed that 598 transcripts out of 19161 (3.12%) and 437 out of 12174 (3.59%) are expressed at a significantly different level (P<0.05) in the 2 groups. Taking into consideration only the transcripts detected in all the samples, with an expression rate of at least 10-fold different and a P<0.05 we identified 39 genes. Seventeen transcripts were more abundant in RB oocytes, whereas 22 were downregulated. This is the first analysis of the oocyte transcriptome performed with deep sequencing technology. The method enabled us to compile a full list of transcripts that are found in highly competent oocytes. A direct comparison with low-quality oocytes indicated that quantitative differences of transcripts level are limited to a small subpopulation of key transcripts.