Functional consistency and molecular evolution: a comprehensive characterization of primary endometrial cancer-associated fibroblasts during in vitro expansion.

BACKGROUND

Cancer-associated fibroblasts (CAFs) are critical components of the tumor microenvironment in endometrial cancer (EC). Primary isolation and culture of CAFs serve as the standard method for investigating their biological roles. However, it remains unclear whether prolonged in vitro culture alters the intrinsic properties of CAFs, potentially affecting experimental reproducibility. This study aims to evaluate the phenotypic and transcriptomic evolution of primary EC-derived CAFs across different passages.

METHODS

CAFs were isolated from human EC tissues using enzymatic digestion and differential adherence. Cells at passages 1 (P1), 5 (P5), and 10 (P10) were compared. Biological phenotypes, including proliferation, migration, apoptosis, senescence, and cytokine secretion, were assessed. The effects of CAF-conditioned medium (CM) on the proliferation and invasion of EC cell lines (ISHIKAWA and HEC-1 A) were evaluated. Furthermore, RNA sequencing (RNA-seq) was performed to uncover molecular alterations, followed by functional enrichment and subtype analyses. Additional validation experiments, including ROS detection, NF-κB activation analysis, H₂O₂-induced oxidative stress modeling, and NAC or TPCA-1 intervention, were conducted to investigate the mechanism underlying passage-associated inflammatory drift.

RESULTS

CAFs maintained a consistent spindle-shaped morphology and expression of specific markers (Vimentin/FAP) from P1 to P10. Functional assays demonstrated no significant differences in proliferation, migration, or senescence rates across passages. Crucially, CAFs at all passages exhibited comparable efficacy in promoting the proliferation, migration, and invasion of EC cells. However, transcriptomic profiling revealed profound reprogramming, with 1,300 core genes differentially expressed in P5/P10 compared to P1. While pathways related to extracellular matrix remodeling (myCAF signature) remained stable, establishing the molecular basis for sustained invasion support, passaged CAFs progressively exhibited reduced antigen-presenting signatures and acquired a pro-inflammatory iCAF-like phenotype. Mechanistically, late-passage CAFs showed increased ROS accumulation, enhanced P65 phosphorylation and nuclear translocation, and upregulated inflammatory mediators, including IL6, CXCL1, CXCL2, and CCL2. H₂O₂ treatment partially mimicked this inflammatory phenotype in P1 CAFs, whereas NAC or TPCA-1 attenuated inflammatory activation in P10 CAFs.

CONCLUSIONS

Primary endometrial CAFs retain core tumor-supporting phenotypes and myofibroblastic characteristics up to passage 10, validating their utility as reliable in vitro models for invasion and metastasis studies. However, high-passage CAFs exhibit a culture-associated inflammatory transcriptional drift driven, at least in part, by ROS/NF-κB activation, together with reduced immunogenic features. These findings provide essential guidelines for the standardization of CAF-based research models and caution against using high-passage CAFs for immune-interaction or inflammation-related studies.