Global alterations of the epigenetic landscape define endometrial receptivity and provide a hallmark for endometriosis

The endometrium plays an essential role in the female reproductive cycle. It undergoes extensive morphological and transcriptional changes during the menstrual cycle, in response to steroid hormones. For implantation to occur successfully, the endometrium must attain a receptive state which is synchronised with the presence of a competent blastocyst. Failure in the establishment or timing of receptivity is a primary cause of infertility. The endometrium is also the tissue chiefly involved in the aetiology of endometriosis, a condition in which endometrial tissue is found outside the womb where it forms lesions. Endometriosis affects ~10% of women of reproductive age, can cause severe pain and poses an increased risk of infertility. Overall, the molecular regulation required to fine-tune the cyclical changes of the endometrium and that may also play a role in the development of endometriosis remains poorly understood. In this thesis, I set out to investigate how the epigenome may contribute to achieving gene expression changes that underpin the attainment of the window of receptivity as well as the disease state of endometriosis. To this end, in conjunction with histological analyses, I generated a comprehensive set of genome-wide profiles, including transcriptomes by RNA-seq, histone modification profiles by ChIP-seq for H3K4me3, H3K9me3 and H3K27me3, as well as DNA methylation (5mC) and hydroxymethylation (5hmC) profiles by meDIP-seq and hmeDIP-seq, respectively, of endometrial biopsies from control and endometriosis patients. My data reveal considerable changes to the epigenome that occur during the transition from the non-receptive to the receptive state. I identify a cohort of regions marked by H3K4me3 and H3K27me3, known as bivalent domains, that resolve to H3K4me3-only in the receptive state and globally correlate with increased expression of the associated genes. Notably, I show that this subset of bivalent domains that will become activated can be predicted already in the non-receptive state based on marking with 5hmC. These regions are also significantly enriched for Kruppel-like factor 9 (KLF9) and Forkhead box O1A (FOXO1) binding motifs, which are progesterone-regulated transcription factors with essential roles in reproduction. Perhaps the most dramatic epigenetic change is observed in endometrial epithelial cells that acquire high levels of 5hmC during the establishment of the receptive state. This increase of 5hmC correlates with increased gene expression, in particular at genes that are crucial for the attainment of receptivity. These data are also in line with my finding that the majority of gene expression changes observed in the whole tissue can be attributed to epithelial cells. Thus, 5hmC appears to play a key role in the molecular functionality of epithelial cells, which have a high secretory output and are critical for implantation to occur. Finally, I report on alterations to the epigenetic profile of endometrium from patients with endometriosis, a noteworthy discovery. Differential 5mC and 5hmC enrichment is particularly prevalent at genes involved in the Hippo signalling pathway, which regulates core aspects of cell and tissue biology including cell polarity, adhesion, proliferation and apoptosis. These epigenetic changes may potentially predispose endometrial cells of endometriosis patients to exhibit increased levels of adherence and proliferation at ectopic sites, thus contributing to the formation of lesions. In addition, and perhaps more importantly, I identified and validated a set of genomic regions based on their differential DNA (hydroxyl)methylation state that separates patient samples according to their receptivity and disease status. Although these epigenetic probes require independent validation on many more independent patient samples, identification of such a cohort of loci holds great promise for the establishment of an epigenetic biomarker capable of determining receptivity state and as a diagnostic tool for endometriosis. If proven successful, this would be a great advance in the field of reproductive biology and in particular female infertility. My data prompt further investigations into the contribution of the epigenome to the transformation of the endometrium during the normal menstrual cycle and to the aetiology of endometriosis