Y23 - Linking Form and Function: Morphodynamic and Molecular Profiling of Ovarian Cancer via Bioengineered Cell Culture Platform

Introduction: : Ovarian cancer progression involves dissemination through multicellular spheroids that vary in morphology and invasive behavior. These phenotypic differences suggest underlying molecular heterogeneity, but the associations between physical traits and transcriptional profiles remain poorly defined. Understanding these links is critical for interpreting tumor behavior and informing therapeutic strategies.
We employed a bioengineered silicon-titanium diboride (Si-TiB) micropatterned platform to monitor the dynamics of cellular aggregation in ovarian cancer cells. This platform enables spatially controlled culture, direct imaging access for longitudinal assessment of cell and nuclear morphology, and cytoskeletal organization . Here we present analysis of image-derived phenotypes in conjunction with transcriptomic profiles to better understand the molecular basis of tumor heterogeneity. Notably, the same platform supports patient-derived cancer cells, allowing direct comparison between model cell lines and clinical samples. By integrating morphological assessment with RNA sequencing data, we aim to identify gene expression profiles associated with specific phenotypic states associated to tumor progression. This work bridges bioengineering and molecular oncology to advance mechanistic understanding of ovarian cancer biology.

Materials and

Methods: : Three ovarian cancer cell lines with varying aggressiveness levels —SKOV3 (high), OVCAR3 (moderate), and OVSAHO (low)— were cultured on Si-TiB micropatterned substrates with circular confinement pattern. Cellular, nuclear, and cytoskeletal morphologies were assessed using stereomicroscopy and confocal microscopy.
Quantitative phenotypic metrics including nuclear eccentricity and orientation were extracted using automated image analysis. Confocal immunofluorescence staining included DAPI, F-actin, and vimentin to assess nuclear shape and cytoskeletal alignment. Bulk RNA sequencing was conducted for each cell line, and gene set variation analysis (GSVA) was used to identify molecular pathways associated with observed morphologies, such as epithelial–mesenchymal transition (EMT), cell adhesion and migration. Patient-derived ovarian cancer cells were seeded onto the micropatterned substrates and analyzed to compare early morphologic states with those of established cell lines.

Results, Conclusions, and Discussions:: Longitudinal imaging over days 3, 7, and 9 revealed distinct morphodynamic behaviors across the ovarian cancer cell lines and patient ascites derived cells cultured on the Si-TiB substrate (Figure 1). Distinct phenotypes were observed: OVSAHO show compact monolayers and non-dispersive growth through days 3-9, and OVCAR3 progress from compact monolayers to disorganized, dispersive behavior, with loss of cohesion to the micropattern boundary by day 9, whereas SKOV3 form transition from monolayers to aggregates by day 7, with invasive protrusions observed through day 9, consistent with a collective migration phenotype. Patient-derived cells showed intermediate phenotypes, including both confined and locally migratory behaviors. Confocal imaging at day 7 with Ki-67 revealed proliferation of OVCAR3, SKOV3 and patient-derived samples (Figure 2). A polar histogram of nuclear orientation showed significant differences between cells constrained within circular patterns and those spreading beyond them (p < 0.001), supporting morphological polarization during spatial expansion. F-actin analysis demonstrated a significantly higher number of protrusions in regions outside the patterns compared to patterned areas (p < 0.05), consistent with increased migratory potential. These protrusions, visualized on original F-actin images, highlight cytoskeletal morphology reminiscent of migratory stretching.
Transcriptomic analysis (Figure 3) revealed that OVCAR3 was enriched in EMT- and migration-associated pathways, including PI3K/AKT signaling and integrin-mediated interactions, relative to OVSAHO.
This study demonstrates alignment between morphological phenotypes and molecular subtypes in the ovarian cancer cell lines. The Si-TiB micropatterned culture platform enables mechanistic studies of cancer cell state transitions with potential to support future applications in drug screening, subtype classification, and patient stratification. Future work will integrate transcriptomic data with high-content imaging and statistical modeling to define phenotype-linked transcriptional signatures.

Acknowledgements and/or References (Optional)::