Back

Stromal LOX/FAK/beta catenin pathway locks mammary fibroblasts into a tumor-promoting myCAF state

Gandhi, A.; Beri, D.; Pally, D.; Manjunath, S.; Humcha, S.; Rangarajan, A.; Bhat, R.; Tatu, U.; Kumar, R. V.; Kondaiah, P.

2025-07-17 cancer biology
10.1101/2025.07.11.664481 bioRxiv
Show abstract

BackgroundCancer-associated fibroblasts (CAFs) sustain tumor progression, yet the soluble cues that maintain their myofibroblast (myCAF) state are poorly defined. Transforming growth factor beta (TGF-{beta}) is a canonical CAF activator. This study aims to identify TGF-{beta}-induced secreted mediators that reinforce the myCAF phenotype in breast cancer and map the downstream signaling cascade. Methods and ResultsSecretome profiling of primary patient-derived myCAFs and human mammary fibroblasts (HMF3s) engineered to over-express TGF-{beta}1 revealed 20 extracellular-matrix remodelers shared exclusively by both activated states; lysyl oxidase (LOX) was the top-ranked hit. LOX knockdown abrogated TGF-{beta}-driven -smooth-muscle actin (-SMA) induction, collagen-gel contraction and migration in HMF3s, and reduced constitutive -SMA and {beta}-catenin in myCAFs. Mechanistically, TGF-{beta} upregulated LOX, which activated focal-adhesion kinase (FAK), leading to p38 MAPK- and Akt-mediated Ser9 phosphorylation (inactivation) of GSK3{beta} and consequent {beta}-catenin stabilization. In HCC1806-luciferase orthotopic xenografts, CAFs accelerated tumor growth, whereas LOX-deficient CAFs lost this pro-tumoral effect. ConclusionLOX is a pivotal autocrine effector of TGF-{beta} that locks breast CAFs into a pro-tumoral myCAF state through a LOX/FAK/GSK3{beta}/{beta}-catenin axis. Targeting stromal LOX may disrupt CAF activation and curb breast cancer progression.

Matching journals

The top 11 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.