FSH-regulated gene expression in mouse granulosa cells as revealed by ChIP- and RNA- Sequencing

Follicle stimulating hormone (FSH) regulates ovarian follicular development through a specific gene expression program. We analyzed FSH-regulated transcriptome and histone modification in granulosa cells during follicular development. We used super-stimulated immature mice and collected granulosa cells prior to and 48h after stimulation with equine chorionic gonadotropin (eCG). We profiled the transcriptome using RNA-sequencing (N=3/time-point) and genome wide trimethylation of lysine 4 of histone H3 (H3K4me3; an active transcription marker) using chromatin immunoprecipitation and sequencing (ChIP-Seq; N=2/time-point). Across the mouse genome 14,583 genes had an associated H3K4me3 peak and 63-66% of these peaks were observed within < 1kb promoter region. There were 72 genes with differential H3K4me3 modification at 48h-eCG (absolute log fold change >1; FDR<0.05) relative to 0h-eCG. Transcriptome data analysis showed 1463 differentially expressed genes at 48h-eCG (absolute log fold change > 1; FDR<0.05). Among the 20 genes with differential expression and altered H3K4me3 modification, Lhcgr had higher H3K4me3 abundance and expression, while Nrip2 had lower H3K4me3 abundance and expression. Using ChIP-qPCR, we showed that FSH regulated expression of Lhcgr, Cyp19a1, Nppc and Nrip2 through regulation of H3K4me3 at their respective promoters. Transcript isoform analysis using Kallisto-Sleuth tool revealed 875 differentially expressed transcripts at 48h-eCG (b>1; FDR<0.05). Pathway analysis of RNA-seq data demonstrated that TGFB signaling and steroidogenic pathways were regulated at 48h-eCG. Thus, FSH regulates gene expression in granulosa cells through multiple mechanisms namely altered H3K4me3 modification and inducing specific transcripts. These data form the basis for further studies investigating how these specific mechanisms regulate granulosa cell functions. Two biological replicates were analyzed for each of 2 conditions (0h and 48h eCG) using H3K4me3 antibody in ChIP-Seq. For RNA-Seq, three biological replicates were analyzed for each of 2 conditions (0h and 48h eCG)