Specific CpG Dinucleotides Within Genes Encoding Key Proteins of Steroid Biosynthesis Show Tissue-Specific Methylation Levels That Negatively Correlate with Transcript Abundance in Estrogen-Active Dominant and Pre-Ovulatory Bovine Follicles.

Abstract

The profound morphological reorganization of the somatic cell layers, granulosa and theca, during the transformation from the follicular to luteal stage, is preceded by a well orchestrated modulation of cell type-specific gene expression. In vertebrates, methylation of cytosines in the context of CpG dinucleotides plays an important role in the local organisation of the chromatin and thus the long-term regulation of gene expression via epigenetic mechanisms. The goal of this study was to elucidate possible epigenetic mechanisms behind the LH induced transition from the large dominant estrogen-active follicle to the preovulatory follicle, which is an intermediate stage towards full luteinization. For this, the methylation status and transcript abundance of key genes of steroid biosynthesis, Star encoding steroidogenic acute regulatory protein, Cyp11a1 encoding cytochrome P450 cholesterol side-chain cleavage enzyme, Hsd3b encoding 3-beta-hydroxysteroid dehydrogenase, Cyp17 encoding steroid 17-alpha-hydroxylase, Cyp19 encoding aromatase cytochrome P450 were measured by bisulfite direct sequencing and real-time PCR, respectively, in the granulosa and theca of large dominant and preovulatory bovine follicles. The abundance of several other key transcripts was also quantified: Fshr transcripts, encoding follicle stimulating hormone receptor, Lhr transcripts, encoding LH receptor, and Ghr transcripts, encoding growth hormone receptor. Pcna transcripts encoding the proliferating cell nuclear antigen, were measured as a marker for proliferation activity. The appearance of Ptgs2 transcripts, encoding COX-2 in the granulosa was used as an indicator for the start of prostaglandin biosynthesis prior to ovulation. Gapdh transcripts, encoding glyceraldehyde 3-phosphate dehydrogenase, were measured as a housekeeping control to assess the fundamental metabolic activity of cells. The steroid hormones progesterone and 17-beta-estradiol were monitored to distinguish estrogen-active and estrogen-inactive follicles. Expression analysis revealed that (i) independently of the follicular stage, the gene expression profile was very different in granulosa and theca; (ii) the abundance of several key transcripts was significantly lower in estrogen-inactive, compared with estrogen-active, dominant follicles; (iii) in the granulosa of preovulatory follicles, transcripts encoding steroidogenic enzymes and hormone receptors were largely down regulated, whereas (iv) progesterone and estradiol were found at high concentrations in the follicular fluid. Methylation analysis revealed that (v) neighboring CpGs can be methylated at very different levels; (vi) specific CpGs within the Hsd3b, Cyp11a1, and particularly within the Cyp19 gene show different methylation levels in granulosa and theca that negatively correlate with the respective transcript abundance levels; (vii) methylation levels of CpGs investigated did not significantly change during the transition from the estrogen-active dominant to preovulatory follicular stage. Collectively, our data show that preovulatory follicles have a transient and unique gene expression profile and are clearly different from both the preceding and subsequent (follicular and luteal, respectively) stages. The methylation data show site and tissue-specific, but not differentiation-specific methylation levels and suggest that the methylation of specific CpGs, particularly within Cyp19, precludes transcription. (poster)