Effects of ovarian stimulation on endometrial gene expression profile.

Controlled ovarian stimulation (COS) used in assisted reproduction techniques (ART) produces lower implantation rates per embryo transferred as compared to natural and ovum donation cycles, suggesting a suboptimal endometrial development. Endometrial alterations have been observed by histological and biochemical techniques during decades. The recent developments in functional genomics have provided objective tools to analyze the endometrium in natural cycles and evaluate the impact of COS protocols in endometrial development. COS cycles that use GnRH agonists and antagonists have been analyzed in detail during the window of implantation to establish the differences with respecting to the natural cycle [1-5]. Even more, it has been demonstrated that endometria from natural cycles follow different genomic patterns compared to COS cycles in the transition from the pre-receptive (days LH/hCG+1 until LH/hCG+5) to the receptive phase (day LH+7/hCG+7). Specifically, it has been demonstrated the existence of a two-day delay in the activation/repression of two clusters composed by 218 and 133 genes on day hCG+7 versus LH+7 [6]. To further increase our understanding of the pathways governing endometrial receptivity, we have expanded these studies by comparing the gene expression profile of the human endometrium troughout the early-mid secretory phase in natural cycles vs controlled ovarian stimulated (COS) cycles. We have analyzed endometrial samples collected from healthy fertile cycling ovum donors (aged 23-39), that underwent either a natural cycle (n = 25) or stimulated cycles (n = 25) at days LH+1, LH+3, LH+5, LH +7 and LH+9 (n = 5 per time point), no progesterone (P) supplementation was administered. RNA was extracted and labeled cDNA were hybridized onto the GeneChip HG_U133A (Affymetrix) for the comparisons. Gene expression data were analysed using different methods such as clustering or selection of differentially expressed genes, as implemented in the GEPAS (http://www.gepas.org). We analyzed their corresponding temporal functional profiles. For that end, we used the first day as reference and we compared each subsequent day to this reference time by a gene set enrichment analysis, as implemented in the FatiScan tool of Babelomics. This method allows us to trace the functional blocks (GO, KEGG pathways, etc.) that were significantly up- and down-regulated on each day of the WOI. Our results demonstrate that the receptivity transition from pre-receptive (LH+1, LH+3 and LH+5) to receptive (LH+7, LH+9) endometria displayed a differential profile of biological processes over and under expressed, pathways statistically over and under represented and differential gene clustering. Furthermore, these results indicate that gene expression profiling of the endometrium at the time of implantation is different between natural cycles specifically at LH+7 and COH cycles at hCG+7 and similarities are recovered thereafter. Based on this basic data, influences of endometrial function on IVF outcomes will be reviewed as well state of the art of human endometrial receptivity markers. [ABSTRACT FROM AUTHOR]