ECM stiffness primes Cancer-associated Fibroblasts to drive breast cancer aggressiveness

Introduction: : The tumor microenvironment (TME) of breast cancer is a dynamic entity that critically influences initiation, progression, and therapy resistance [1]. Cancer-associated fibroblasts (CAFs) are one of the most abundant stromal cell types in the TME and have emerged as central regulators of tumor growth and immune modulation [2]. The heterogeneity of breast CAFs presents a significant challenge to improving therapeutic outcomes [2–4]. Undefined CAF subtypes can drive tumor growth along distinct axes, while others exhibit tumor-suppressive properties [5]. One of the least understood aspects of CAF biology is how the mechanical environment (e.g., ECM stiffness) influences their interactions with cancer cells. To address this gap, we developed a tunable 3D ECM biomimetic model that recapitulates the stiffness range of healthy and malignant breast tissue and investigated how stiffness directs CAF phenotypic heterogeneity and, in turn, modulates breast cancer to identify mechano-regulatory pathways for therapeutic targeting.

Materials and

Methods: : Materials and Methods
Matrix stiffness: GelMA (gelatin methacryloyl) hydrogels are used and photocrosslinked with a Ru/SPS photo initiator under varied light exposure to obtain discrete stiffness (soft: 1 min – 4kPa; stiff: 9 min-35kPa), with stiffness measured using a rheometer.
Cell culture: Primary human breast CAFs (1×10^5 cells/gel) were encapsulated in soft (4 kPa) or stiff (35 kPa) 3D GelMA hydrogels and then cultured for four days. On day 4, constructs were transferred into co-culture inserts along with MCF-7 monolayers at the bottom for a three-day indirect co-culture.

Results, Conclusions, and Discussions:: 1. Morphometric analysis of CAFs under 3D stiffness (Fig. 2)
Breast CAFs were mechanically primed in 3D hydrogels spanning soft and stiff ranges for 4 days of incubation. In soft matrices, CAFs became more spread and flattened, exhibiting significantly higher volume, flatness, and surface area. In contrast, in stiff gels, CAFs adopted a slender, elongated morphology with decreased volume and surface area but increased aspect ratio, as shown by F-actin (green) and DAPI (blue) staining and morphometric analyses. Collectively, these observations indicate that ECM stiffness orchestrates CAF shape plasticity, underscoring the mechanotransduction role in CAF phenotypic diversity within the TME.
2. In vitro evaluation of Breast cancer progression (Fig. 3)
After priming, CAFs were placed in co-culture inserts with MCF-7 monolayers at the bottom and co-cultured for three days for gene-expression analyses. MCF7Stiff CAFs exhibited a pronounced upregulation of DNA-damage response genes (ATM, BRCA2), cell-cycle regulators (CDK2, CCNE1), and EMT/invasion markers (SNAI2, PLAU), alongside inflammatory mediators (PTGS2, IL6), compared to MCF7Soft CAFs. In accordance, MCF7Soft CAFs modestly elevated developmental and angiogenic genes (VEGFA, SLIT2) and increased cell cycle arrest gene (CDKN1A), reflecting their reduced proliferation, and a less aggressive breast cancer phenotype. Stiff-primed CAF reprograms MCF7 toward a highly proliferative, DNA-repair–proficient, invasive/EMT, and therapy-resistant phenotype mirroring the hallmarks of aggressive breast cancer.
Conclusions
In this study, we propose a mechanistic insight into breast cancer progression along the stromal-dynamic microenvironmental axis using a 3D GelMA-based ECM mimicking model with tunable, physiologically relevant stiffness. Our findings establish that ECM stiffness primes CAFs to adopt distinct morphologies and phenotypic functions, thereby directing their subtypes. Accordingly, these mechanically primed CAFs potently modulate breast cancer phenotypes. In soft matrices, CAFs adopt a spindle-shaped morphology that drives moderate MCF-7 proliferation and differentiation. In contrast, CAFs in stiff matrices become rounded and induce an aggressive, EMT- and DNA repair–enhanced, therapy-resistant phenotype in MCF-7 cells. Together, these results highlight mechano-regulatory CAF–cancer axes as promising therapeutic targets. Regulating CAF phenotypic diversity along the mechanosensing axis may offer a more effective in vitro readout of MCF7 therapeutic resistance and invasiveness.

Acknowledgements and/or References (Optional):: Funding and financial support - National Research Foundation of Korea (NRF – 2021R1A2C3008408).
GelMA synthesis and processing - Mechanobiology Institute, National University of Singapore (NUS), Singapore
Hydrogel Rheological characterization - Fluid and Interface Laboratory, Mechanical Engineering, KAIST
GelMA freeze drying process - Neutron scattering laboratory, Nuclear and Quantum Engineering, KAIST.