Transcriptomic signatures in trophectoderm and inner cell mass of human blastocysts classified according to developmental potential, maternal age and morphology.

Selection of high-quality embryos is important to achieve successful pregnancy in assisted reproductive technology (ART). Recently, it has been debated whether RNA-sequencing (RNA-Seq) should be applied to ART to predict embryo quality. However, information on genes that can serve as markers for pregnant expectancy is limited. Furthermore, there is no information on which transcriptome of trophectoderm (TE) or inner cell mass (ICM) is more highly correlated with pregnant expectancy. Here, we performed RNA-Seq analysis of TE and ICM of human blastocysts, the pregnancy expectation of which was retrospectively determined using the clinical outcomes of 1,890 cases of frozen-thawed blastocyst transfer. We identified genes that were correlated with the expected pregnancy rate in ICM and TE, respectively, with a larger number of genes identified in TE than in ICM. Downregulated genes in the TE of blastocysts that were estimated to have lower expectation of pregnancy included tight junction-related genes such as CXADR and ATP1B1, which have been implicated in peri-implantation development. Moreover, we identified dozens of differentially expressed genes by regrouping the blastocysts based on the maternal age and the Gardner score. Additionally, we showed that aneuploidy estimation using RNA-Seq datasets does not correlate with pregnancy expectation. Thus, our study provides an expanded list of candidate genes for the prediction of pregnancy in human blastocyst embryos.