Design of a Microphysiologic Ovarian Follicle Culture System Using Multiomics Technology

Abstract

Life-saving anti-cancer treatments cause irreversible damage to the ovarian follicular reserve, resulting in infertility for many patients. Development of a translational follicle culture system is stunted by a lack of knowledge on the complex paracrine signaling driving early follicle development. Ovarian follicles are the functional units of the ovary, each containing the oocyte surrounded by specialized hormone-producing cells. During the reproductive lifespan, quiescent primordial follicles are activated in cohorts to join the growing pool. Early-stage follicles are found in the outer 1mm of the ovary (the cortex), while growing follicles expand toward the inner part of the ovary (the medulla) each taking on a multilayered architecture containing the oocyte surrounded by cumulus granulosa, and outer layers of mural granulosa and theca cells separated from the cumulus-oocyte complex by a fluid-filled antrum. The spatial organization of the ovary robustly correlates with its reproductive and endocrine functions but the cellular diversity of the ovary and its complex structures have been difficult to study, largely due to the scarcity of tissue from healthy donors. Here we performed spatial transcriptomics (ST) and single-cell RNA sequencing (scRNAseq) of ovaries from five reproductive-age women. Using ST, we discovered transcriptional heterogeneity of genes involved in extracellular matrix remodeling (VIM, COL1/8/12/14/16/18, TIMP1/2) and hormone signaling (NR4A1, CEBPD, STAR) in the ovarian cortex. We created a panel of 76 oocyte-specific genes using antibody-guided RNA isolation and developed 96- and 46-gene panels for granulosa and theca cells from antral follicles, providing the first spatially-driven confirmation of these markers. We also confirmed the gradient of cumulus cell phenotype, where cells near the oocyte showed high expression of protein kinase A signaling genes (AKAP12, MIF, RDX) while those near the antrum showed high expression of Wnt signaling genes (DDIT3, DKK1, LEF1). scRNAseq revealed four major cell types: stromal, immune, and endothelial cells and pericytes, and four immune cell subtypes: macrophages, and mast, T, and natural killer cells, expanding upon previous studies reporting the cellular makeup of the organ as a whole. While scRNAseq experiments failed to identify rare cell types such as granulosa and theca, we developed an unbiased single cell atlas of the ovary and our ST data provided an unprecedented profile of the rare cell types not identified in scRNAseq. Utilizing our group follicle culture system we performed time course analysis of murine folliculogenesis in vitro and uncovered dynamic expression of genes related to meiosis and steroidogenesis that reflect the carefully orchestrated progression of events observed in vivo. We also identified genes related to angiogenesis (Vegfa, Angpt1/2, Ang) and Wnt signaling (Rspo1, Sfrp1/2) as upregulated in conditions leading to mature oocytes, suggesting these factors are crucial for maturation. Using ligand-receptor analysis of follicular somatic cells and oocytes, we discovered dynamic relationships that support previous studies on bidirectional crosstalk within the follicle. Shotgun proteomics validated our transcriptional findings, leading to a panel of pro-angiogenic factors and Wnt agonists that may serve as targets for exogenous supplementation in a translational follicle culture system. In conclusion, these studies have uncovered crucial insights into the regulatory mechanisms governing follicle development. Our research identified key pathways, cellular constituents, and novel markers that significantly advance our understanding of ovarian biology. These findings hold promise for improving follicle culture systems for assisted reproduction and ultimately enhancing the prospects of fertility preservation for a broad range of patients.