Tumor cells induce the cancer associated fibroblast phenotype via caveolin-1 degradation: Implications for breast cancer and DCIS therapy with autophagy inhibitors

Loss of stromal caveolin 1 (Cav-1) is a novel biomarker for cancer-associated fibroblasts that predicts poor clinical outcome in breast cancer and DCIS patients. We hypothesized that epithelial cancer cells may have the ability to drive Cav-1 downregulation in adjacent normal fibroblasts, thereby promoting the cancer associated fibroblast phenotype. To test this hypothesis directly, here we developed a novel co-culture model employing (i) human breast cancer cells (MCF7), and (ii) immortalized fibroblasts (hTERT-BJ1), which are grown under defined experimental conditions. Importantly, we show that co-culture of immortalized human fibroblasts with MCF7 breast cancer cells leads to Cav-1 downregulation in fibroblasts. These results were also validated using primary cultures of normal human mammary fibroblasts co-cultured with MCF7 cells. In this system, we show that Cav-1 downregulation is mediated by autophagic/lysosomal degradation, as pre-treatment with lysosome-specific inhibitors rescues Cav-1 expression. Functionally, we demonstrate that fibroblasts co-cultured with MCF7 breast cancer cells acquire a cancer associated fibroblast phenotype, characterized by Cav-1 downregulation, increased expression of myofibroblast markers and extracellular matrix proteins, and constitutive activation of TGFβ/Smad2 signaling. siRNA-mediated Cav-1 downregulation mimics several key changes that occur in co-cultured fibroblasts, clearly indicating that a loss of Cav-1 is a critical initiating factor, driving stromal fibroblast activation during tumorigenesis. As such, this co-culture system can now be used as an experimental model for generating "synthetic" cancer associated fibroblasts (CAFs). More specifically, these "synthetic" CAFs could be used for drug screening to identify novel therapeutics that selectively target the Cav-1-negative tumor micro-environment. Our findings also suggest that chloroquine, or other autophagy/lysosome inhibitors, may be useful as anti-cancer agents, to therapeutically restore the expression of stromal Cav-1 in cancer associated fibroblasts. We discuss this possibility, in light of the launch of a new clinical trial that uses chloroquine to treat DCIS patients: PINC (Preventing Invasive Breast Neoplasia with Cholorquine) [See http://clinicaltrials.gov/show/NCT01023477].