Cell‐specific network analysis of human folliculogenesis reveals network rewiring in antral stage oocytes

Abstract

Although previous studies have explored the gene expression profiles of human oocytes and granulosa cells by single‐cell RNA sequencing (scRNA‐seq), the dynamic regulatory network at a single‐cell resolution during folliculogenesis remains largely unknown. We identified 10 functional modules by WGCNA, four of which were significantly correlated with primary/antral oocyte and antral/pre‐ovulatory granulosa cells. Functional enrichment analysis showed that the brown module, which was correlated with antral oocyte, was enriched in oocyte differentiation, and two core subnetworks identified by MCODE were involved in cell cycle (blue subnetwork) and oogenesis (red subnetwork). The cell‐specific network (CSN) analysis demonstrated a distinct gene network structure associated with the antral follicular stage, which was notably different from other developmental stages. To our knowledge, this is the first study to explore gene functions during folliculogenesis at single‐cell network level. We uncovered two potential gene subnetworks, which may play an important role in oocyte function beginning at the antral stage, and further established their rewiring process at intra‐network/whole transcriptome level. The findings provide crucial insights from a novel network perspective to be further explored in functional mechanistic studies.

A single‐cell transcriptomic atlas characterizes cell types and their molecular features in yak ovarian cortex

“…7,10 Afterward, the ovarian follicles go through several developmental processes to reach maturity, including granulosa cell proliferation and differentiation, follicle fluid accumulation in the antrum, and theca cell generation from stromal cells. 11,12 However, the overwhelming majority of the follicles undergo a degenerative process known as atresia and follicle atresia happens at any stages of folliculogenesis and follicle development. 13,14 From two to three million primordial follicles present within a yak (Bos grunniens) ovary at birth, about 210 000 of which remains around puberty, whereas the rest undergoes degeneration and apoptosis before reaching ovulation.…”

“…Upon activation, the primordial follicles migrate toward the ovarian medulla, the inner part of the ovary 7,10 . Afterward, the ovarian follicles go through several developmental processes to reach maturity, including granulosa cell proliferation and differentiation, follicle fluid accumulation in the antrum, and theca cell generation from stromal cells 11,12 . However, the overwhelming majority of the follicles undergo a degenerative process known as atresia and follicle atresia happens at any stages of folliculogenesis and follicle development 13,14 .…”

Regulon active landscape reveals cell development and functional state changes of human primary osteoblasts in vivo

“…Linking genomic regulatory patterns to variations in gene expression at the single-cell level could be robust against drop-out commonly seen in single-cell sequencing data, thereby optimizing the discovery and characterization of cellular states [ 20 ]. Additionally, unlike traditional network construction groups of cells, the CSN method constructs separate networks at the single-cell level, thus the heterogeneity of each individual cell is preserved [ 16 , 19 ]. To this end, our research revealed unique features in the regulon activity landscape and reciprocal gene interactions within each human osteoblast subtype.…”

“…We identified cell heterogeneity among human osteoblasts in vivo, and preliminarily explored transcriptome dynamic changes based on gene expression. Accumulating evidence suggested that many pathophysiological processes are not only controlled by the expression states of one or several molecules but rather by coordinated expression changes of a number of genes [ 16 – 19 ]. Network analysis, accounting for expressional changes of multiple genes simultaneously, provides an opportunity to explore specific biological processes more comprehensively, under conditions resembling their natural biological state.…”

“…Another approach is to construct a cell-specific network (CSN), separate gene regulatory networks for individual cells, which allows for identification of cell type heterogeneity based on multiple genes and their co-expressions [ 16 ]. TF networks and CSN analysis have been efficiently applied to provide potentially impactful insights regarding embryonic development, aging processes, and tumorigenesis [ 16 – 19 ].…”