Reciprocal regulation of hMENA and TGF-β signaling in cancer-associated fibroblasts promotes EMT, immunosuppression, poor prognosis, and ICT resistance in NSCLC

BACKGROUND: Cancer-associated fibroblasts (CAFs) significantly impact cancer progression and CAF subtypes are key determinants of response to immune checkpoint therapy (ICT). The transforming growth factor-β (TGF-β) signaling is a main pathway in protumorigenic activity of CAFs and resistance to ICT. The actin cytoskeleton regulator hMENA plays crucial roles in epithelial-mesenchymal transition (EMT) and regulates pathways critical to antitumor immune response, such as interferon-I and Axl-GAS6. METHODS: We constructed a single-cell atlas of CAFs using integrated public dataset. Experimental data were obtained by biochemical and molecular approaches in CAFs from freshly explanted non-small cell lung cancer (NSCLC) tissues hMENA silenced and in tumor cell lines or peripheral blood mononuclear cells treated with CAF conditioned medium. Multiparametric flow cytometry was used to characterize T cells. A gene signature indicative of ICT response was developed by a machine learning model. RESULTS: Computational analysis indicates that hMENA is primarily overexpressed in a myofibroblast cluster enriched for a TGF-β-activated CAF signature. Experimentally, we showed that TGF-β1 treatment increases hMENA expression and, reciprocally, hMENA/hMENAΔv6 modulate TGF-β1/2 production and secretion and transforming growth factor-β type II receptor expression in CAFs. Functionally, hMENA contributes to TGF-β1-driven CAF phenotype, programmed death-ligand 1 (PD-L1) upregulation, extracellular matrix remodeling and secretion of immunosuppressive cytokines/chemokines. The hMENA-driven TGF-β secretion in CAFs promotes PD-L1 expression and EMT in cancer cells by activating TGF-β signaling. On the tumor cell side, hMENA expression sustains the TGF-β signaling and EMT mediated by hMENA-driven CAFs secretome. This immunosuppressive secretome favors regulatory T cell (Treg) abundance and reduces CD8+ and CD4+ T cell functionality. Finally, based on the hMENA and TGF-β enriched CAF subtype, we developed a 9-gene signature, which in combination with hMENA/hMENAΔv6 correlates with increased Treg abundance and poor prognosis in the Cancer Genome Atlas NSCLC and associates with ICT resistance in Stand Up To Cancer (SU2C) and in phase III clinical trial (OAK) (NCT02008227) datasets. CONCLUSIONS: CAFs associate with TGF-β and regulatory T-cell signatures and correlate with poor patient prognosis and resistance to immune checkpoint therapies, supporting their role as key contributors to an immunosuppressive, ICT-refractory tumor microenvironment.