Abstarct: Background: Cancer-associated fibroblasts (CAFs) are a major stromal component of high-grade serous ovarian carcinoma (HGSOC), yet their functional programs and spatial organization remain incompletely defined and inconsistently reported across studies. Methods: We analyzed single-cell RNA-seq data from 7 treatment-naïve HGSOC tumors and integrated these findings with spatial transcriptomics from 4 HGSOC specimens. CAFs were re-clustered into functional programs, trajectories were inferred, and cell–cell communication was modeled. CAF states were mapped onto tissue sections using deconvolution and single-cell spatial mapping. Primary HGSOC CAFs and TGF-β–activated MRC5 fibroblasts (MRC5-CAFs) were used for collagen gel contraction and Western blot validation with perturbation of FGF7/FGFR2 signaling. Results: We identified three CAF programs: matrix-associated CAFs (mCAFs), metabolic-associated CAFs (meCAFs), and proliferative CAFs (pCAFs). Trajectory analyses suggested a continuum from pCAF to mCAF to meCAF. Communication analyses highlighted an FGF7–FGFR2 axis enriched in mCAFs. Spatial analyses indicated CAF-enriched regions were spatially segregated from immune-enriched regions. In vitro, recombinant FGF7 increased AKT phosphorylation and enhanced CAF contractile activation, whereas FGF7 neutralization or FGFR2 inhibition reduced AKT phosphorylation and contractility. Conclusion: Integrative single-cell, spatial, and functional analyses implicate an FGF7–FGFR2–PI3K/AKT axis in maintaining the contractile activation of CAFs in ovarian cancer, providing a rationale for targeting CAF signaling and ECM remodeling. Keywords: ovarian cancer, cancer-associated fibroblasts, FGF7, extracellular matrix, single-cell transcriptomics, spatial transcriptomics
Feng et al. (Wed,) studied this question.