Abstract A24: Fibronectin-mediated activation of Stat3 facilitates the resistance of triple negative breast cancer to EGFR-targeted therapies

Abstract

Abstract Breast cancer (BC) can be subdivided into at least five genetically distinct subtypes. Those classified as triple-negative BC (TNBC) are the most aggressive and do not respond to existing molecularly targeted therapies due to their lack of estrogen receptor and Her2. In the context of TNBC, expression of EGFR within the primary tumor is strongly linked to decreased patient survival, pointing to EGFR as an effective target in treating TNBC. However, EGFR-targeted therapies do not offer clinical benefit to TNBC patients. The mechanisms that underlie this disconnect between the diagnostic and therapeutic values of EGFR in TNBC are unknown. Signal-transducer and activator of transcription-3 (Stat3) is an established oncogene and the importance of constitutive Stat3 activity has recently been described in the maintenance of a tumor-initiating stem-like phenotype in BC. We establish herein that overexpression of WT-EGFR is capable of transforming normal mammary epithelial cells (MECs) through a Stat3-mediated process. Using this overexpression model, we show that activation of Stat3 can be initiated through EGF-induced signaling and via fibronectin (FN)-mediated transactivation of EGFR. Following the induction of epithelial-mesenchymal transition (EMT) by TGF-β, orthotopic mammary tumors formed by EGFR-transformed MECs progress to a TNBC phenotype and undergo postsurgical recurrence and pulmonary metastasis. This induction of metastasis is associated with a potent inhibition of EGF-induced activation of Stat3 signaling and a corresponding increase is FN-driven Stat3. Using the human TNBC cells lines MDA-MB-231 and MDA-MB-468 we show that unlike EGF-induced activation of Stat3, kinase inhibitors of EGFR (AG1478) or Src (PP2) activity do not block FN-mediated Stat3 activity. Consistent with these data and clinical findings we demonstrate that following the induction of EMT, even cells originally transformed by EGFR fail to respond to erlotinib therapy during the outgrowth of recurrent and metastatic lesions. Concomitant with the diminution of EGFR-mediated Stat3 signaling we used microarray and RT-PCR analyses to show that the IIIc isoform of FGFR1 is potently and stably upregulated following the induction of EMT. Along these lines, we have previously established that a gain in metastatic outgrowth potential is associated with the loss of EGFR expression in the D2-HAN model of metastatic pulmonary outgrowth. In contrast to EGFR, mesenchymal and metastatic D2.A1 cells express robust levels of FGFR1-IIIc. Importantly, FN was still capable of activating Stat3 in these EGFR-null, metastatic TNBC cells. To examine the role of FN transactivation of FGFR:Stat3 signaling in mediating the metastatic outgrowth of TNBC we used an organotypic 3-D culture model that accurately recapitulates the pulmonary microenvironment. Addition of FN to this system potently induces the outgrowth of D2.A1 cells, a process known to require β1 integrin. Genetic depletion of FGFR1 using an shRNA approach or pharmacological inhibition FGFR (PD17304) or Stat3 (Stattic) ablated the ability of FN to induce 3-D outgrowth of D2.A1 cells. Taken together, our studies indicate that EMT empowers TNBC cells with a secondary tumor-initiating phenotype by converting Stat3 activation from a ligand-dependent event to a constitutive event that occurs through FN:β1 integrin-mediated transactivation of FGFR-IIIc within the metastatic microenvironment. Specific targeting of FGFR-IIIc activity and/or its transactivation by FN may hold the key to preventing the outgrowth of metastatic TNBCs that are resistant to EGFR-targeted therapies.