Balancing stromal-induced complexity in 3D ovarian cancer models through heterotypic co-culture with fibroblasts or architected micro-scaffolds

Tumor heterogeneity limits the reproducibility and interpretability of in vitro ovarian cancer models. In this study, we established a comparative framework that intentionally modulates stromal-induced complexity through two complementary approaches: scaffold-free stromal co-culture with patient-derived cancer-associated fibroblasts (CAFs) or fibroblast cell line WS1, and scaffold-based micro-scaffolds printed using two-photon polymerization (2PP) with poly(ethylene glycol) diacrylate (PEGDA). We analyzed four ovarian cancer cell lines (A2780, SKOV3, COV362 and OV7,) stemness/epithelial-mesenchymal transition (EMT), and TGF-β. We then evaluated their 3D spheroid formation and growth dynamics across different seeding densities. In ultra-low-attachment plates, SKOV3, COV362, and OV7 formed round, compact spheroids across different seeding densities, while A2780 produced loose aggregates. Co-culturing with early-passage CAFs or WS1 fibroblast cell line (in 2:1, 1:1, and 1:2 ratios of cancer cells : fibroblast) rescued A2780‘s ability to form spheroids and increased the compactness of OV7 and COV362 in a line- and ratio-dependent manner in co-culture with CAFs. Compared to 2D cultures, 3D spheroids exhibited higher expression of stemness/EMT and angiogenesis related genes. In contrast, the PEGDA scaffolds (with pore sizes of 65, 100, and 130 µm) standardized early attachment and surface coverage for both spheroid-competent line SKOV3, and a line that does not typically form spheroids, A2780. These scaffolds also shifted EMT/stemness gene expression without elevating vascular endothelial growth factor (VEGF) levels compared to scaffold-free 3D cultures. Together, these results provide practical guidance: utilize fibroblast co-culture to study the stromal-induced complexity and cell-to-cell interactions, including the rescue of cell lines that typically do not form spheroids, and employ printed micro-scaffolds to control 3D culture variability.