Differing molecular response of young and advanced maternal age human oocytes to IVM.

study questionWhat effect does maternal age have on the human oocyte's molecular response to in vitro oocyte maturation? SUMMARY ANSWER: Although polyadenylated transcript abundance is similar between young and advanced maternal age (AMA) germinal vesicle (GV) oocytes, metaphase II (MII) oocytes exhibit a divergent transcriptome resulting from a differential response to in vitro oocyte maturation. WHAT IS KNOWN ALREADY: Microarray studies considering maternal age or maturation stage have shown that either of these factors will affect oocyte polyadenylated transcript abundance in human oocytes. However, studies considering both human oocyte age and multiple stages simultaneously are limited to a single study that examined transcript levels for two genes by qPCR. Thus, polyadenylated RNA sequencing (RNA-Seq) could provide novel insight into age-associated aberrations in gene expression in GV and MII oocytes. STUDY DESIGN, SIZE, DURATION: The effect of maternal age (longitudinal analysis) on polyadenylated transcript abundance at different stages was analyzed by examining single GV and single in vitro matured MII oocytes derived from five young (YNG; < 30 years; average age 26.8; range 20-29) and five advanced maternal age (AMA; ≥40 years; average age 41.6 years; range 40-43 years) patients. Thus, a total of 10 YNG (5 GV and 5 MII) and 10 AMA (5 GV and 5 MII) oocytes were individually processed for RNA-Seq analysis. PARTICIPANTS/MATERIALS, SETTINGS,

methodsPatients undergoing infertility treatment at the Colorado Center for Reproductive Medicine (Lone Tree, CO, USA) underwent ovarian stimulation with FSH and received hCG for final follicular maturation prior to ultrasound guided oocyte retrieval. Unused GV oocytes obtained at retrieval were donated for transcriptome analysis. Single oocytes were stored (at -80°C in PicoPure RNA Extraction Buffer; Thermo Fisher Scientific, USA) immediately upon verification of immaturity or after undergoing in vitro oocyte maturation (24 h incubation), representing GV and MII samples, respectively. After isolating RNA and generating single oocyte RNA-Seq libraries (SMARTer Ultra Low Input RNA HV kit; Clontech, USA), Illumina sequencing (100 bp paired-end reads on HiSeq 2500) and bioinformatics analysis (CLC Genomics Workbench, DESeq2, weighted gene correlation network analysis (WGCNA), Ingenuity Pathway Analysis) were performed. MAIN RESULTS AND THE ROLE OF CHANCE: A total of 12 770 genes were determined to be expressed in human oocytes (reads per kilobase per million mapped reads (RPKM) > 0.4 in at least three of five replicates for a minimum of one sample type). Differential gene expression analysis between YNG and AMA oocytes (within stage) identified 1 and 255 genes that significantly differed (adjusted P < 0.1 and log2 fold change >1) in polyadenylated transcript abundance for GV and MII oocytes, respectively. These genes included CDK1, NLRP5 and PRDX1, which have been reported to affect oocyte developmental potential. Despite the similarity in transcript abundance between GV oocytes irrespective of age, divergent expression patterns emerged during oocyte maturation. These age-specific differentially expressed genes were enriched (FDR < 0.05) for functions and pathways associated with mitochondria, cell cycle and cytoskeleton. Gene modules generated by WGCNA (based on gene expression) and patient traits related to oocyte quality (e.g. age and blastocyst development) were correlated (P < 0.05) and enriched (FDR < 0.05) for functions and pathways associated with oocyte maturation. LARGE SCALE DATA: Raw data from this study can be accessed through GSE95477. LIMITATIONS, REASONS FOR CAUTION: The human oocytes used in the current study were obtained from patients with varying causes of infertility (e.g. decreased oocyte quality and oocyte quality-independent factors), possibly affecting oocyte gene expression. Oocytes in this study were retrieved at the GV stage following hCG administration and the MII oocytes were derived by IVM of patient oocytes. Although the approach has the benefit of identifying intrinsic differences between samples, it may not be completely representative of in vivo matured oocytes. WIDER IMPLICATIONS OF THE

findingsTranscriptome profiles of YNG and AMA oocytes, particularly at the MII stage, suggest that aberrant transcript abundance may contribute to the age-associated decline in fertility. STUDY FUNDING/COMPETING INTEREST(S): J.M.R. was supported by an Austin Eugene Lyons Fellowship awarded by the University of California, Davis. The Eunice Kennedy Shriver National Institute of Child Health and Human Development of the National Institutes of Health (awarded to P.J.R.; R01HD070044) and the Fertility Laboratories of Colorado partly supported the research presented in this manuscript.