Functional evaluation of genetic variants associated with endometriosis near GREB1.

STUDY QUESTION: Do DNA variants in the growth regulation by estrogen in breast cancer 1 (GREB1) region regulate endometrial GREB1 expression and increase the risk of developing endometriosis in women? SUMMARY ANSWER: We identified new single nucleotide polymorphisms (SNPs) with strong association with endometriosis at the GREB1 locus although we did not detect altered GREB1 expression in endometriosis patients with defined genotypes. WHAT IS ALREADY KNOWN: Genome-wide association studies have identified the GREB1 region on chromosome 2p25.1 for increasing endometriosis risk. The differential expression of GREB1 has also been reported by others in association with endometriosis disease phenotype. STUDY DESIGN, SIZE, DURATION: Fine mapping studies comprehensively evaluated SNPs within the GREB1 region in a large-scale data set (>2500 cases and >4000 controls). Publicly available bioinformatics tools were employed to functionally annotate SNPs showing the strongest association signal with endometriosis risk. Endometrial GREB1 mRNA and protein expression was studied with respect to phases of the menstrual cycle (n = 2-45 per cycle stage) and expression quantitative trait loci (eQTL) analysis for significant SNPs were undertaken for GREB1 [mRNA (n = 94) and protein (n = 44) in endometrium]. PARTICIPANTS/MATERIALS, SETTING, METHODS: Participants in this study are females who provided blood and/or endometrial tissue samples in a hospital setting. The key SNPs were genotyped using Sequenom MassARRAY. The functional roles and regulatory annotations for identified SNPs are predicted by various publicly available bioinformatics tools. Endometrial GREB1 expression work employed qRT-PCR, western blotting and immunohistochemistry studies. MAIN RESULTS AND THE ROLE OF CHANCE: Fine mapping results identified a number of SNPs showing stronger association (0.004 < P < 0.032) with endometriosis risk than the original GWAS SNP (rs13394619) (P = 0.034). Some of these SNPs were predicted to have functional roles, for example, interaction with transcription factor motifs. The haplotype (a combination of alleles) formed by the risk alleles from two common SNPs showed significant association (P = 0.026) with endometriosis and epistasis analysis showed no evidence for interaction between the two SNPs, suggesting an additive effect of SNPs on endometriosis risk. In normal human endometrium, GREB1 protein expression was altered depending on the cycle stage (significantly different in late proliferative versus late secretory, P < 0.05) and cell type (glandular epithelium, not stromal cells). However, GREB1 expression in endometriosis cases versus controls and eQTL analyses did not reveal any significant changes. LIMITATIONS, REASONS FOR CAUTION: In silico prediction tools are generally based on cell lines different to our tissue and disease of interest. Functional annotations drawn from these analyses should be considered with this limitation in mind. We identified cell-specific and hormone-specific changes in GREB1 protein expression. The lack of a significant difference observed following our GREB1 expression studies may be the result of moderate power on mixed cell populations in the endometrial tissue samples. WIDER IMPLICATIONS OF THE FINDINGS: This study further implicates the GREB1 region on chromosome 2p25.1 and the GREB1 gene with involvement in endometriosis risk. More detailed functional studies are required to determine the role of the novel GREB1 transcripts in endometriosis pathophysiology. STUDY FUNDING/COMPETING INTERESTS: Funding for this work was provided by NHMRC Project Grants APP1012245, APP1026033, APP1049472 and APP1046880. There are no competing interests.

Non-telomeric epigenetic and genetic changes are associated with the inheritance of shorter telomeres in mice.

Studies using human and mouse cells have revealed some changes to non-telomeric chromatin and gene expression in response to abnormally short telomeres. To investigate this further, we studied the effect of inheriting shorter telomeres on transcription and genetic stability at non-telomeric sites in the mouse. Using multiple generations of Terc knockout mice, we show that inheriting shorter telomeres from one parent increases the likelihood of transcriptional silencing at a non-telomeric green fluorescent protein (GFP) transgene inherited from the other parent. In these cases, silencing must occur at or after zygote formation. In grand-offspring from a G3 Terc (-/-) parent, transgene expression was further reduced and associated with increased DNA methylation and, surprisingly, reduced copy number at the transgene array. In these cases, the transgene had been passed through the germline of a Terc-compromised parent, providing an opportunity for meiotic events. Furthermore, genome-wide microarray analysis of copy number variations revealed greater genetic instability in G3 Terc (-/-) mice than detected in wild-type mice of the same genetic background. Our results have implications for the molecular mechanisms underlying premature-ageing syndromes, such as dyskeratosis congenita. In autosomal-dominant dyskeratosis congenita, progressive telomere shortening is seen as it passes down the generations, and this is associated with anticipation, i.e. the disease becomes more severe earlier. The underlying mechanism is not known, but has been considered to be simply associated with decreases in telomere length. Epigenetic and/or genetic changes at non-telomeric regions could, in theory, be involved.