Breast cancer progression cues driven by adherens junction inactivation

Abstract

Invasive lobular breast cancer (ILC) is a subtype of breast cancer accounting for 10-15% of all invasive breast cancer cases. Unfortunately, early detection of ILC is difficult due to its diffuse growth pattern that hinders detection by palpation or mammography. Also, surgical resection is challenging and most ILC tumors are intrinsically resistant to chemotherapy. Follow-up treatment therefore usually consists of endocrine therapy as most ILCs express ER. Resistance to estrogen antagonists however poses an unmet need for alternative treatment. Interestingly, the majority of ILC cases are marked by early inactivation of the core protein of the adherens junction (AJ) protein E-cadherin, an event that is causal to the development of this disease. As a result, ILC represents a relatively homogeneous group of tumors that are likely to respond to targeted intervention based on the effects of AJ inactivation. In this thesis we have studied these breast cancer progression cues that are driven by the inactivation of the adherens junction. As a consequence of E-cadherin loss, the intracellular component of the AJ p120-catenin (p120) is translocated to the cytosol and the nucleus. To study the contribution of p120 to the development of ILC we made use of genetically engineered mouse models to inactivate p120 in an E-cadherin and p53-driven mouse model of human ILC. In these mice, formation of ILC was mostly prevented pointing towards a critical role for p120 in mouse ILC formation. Loss of E-cadherin also results in the acquisition of anoikis resistance. We have now shown that the ability of E-cadherin negative cells to evade anoikis is based on the prevention of transcriptional upregulation of the BH3-only pro-apoptotic protein BMF. BMF is a direct target of FOXO, an established transcription factor that is inhibited by PI3K-AKT activation upon E-cadherin loss. Interestingly, the BH3-only mimetic ABT199 could re-sensitize E-cadherin negative breast cancer cells to anoikis, advocating for the use of these mimetics in ILC treatment. To provide an overview of the oncogenic signaling pathways affected by loss of E-cadherin we performed a reverse phase protein array. This revealed that E-cadherin inactivation is directly linked to increased AKT activation both in the presence and absence of activating mutations in the PI3K pathway. Inhibition of AKT inhibited cell survival and growth, suggesting that E-cadherin negative breast cancer is likely to respond PI3K pathway targeted therapy. Finally, genome-wide analysis of the p120-catenin binding protein and transcriptional repressor Kaiso has identified a CpG-containing consensus sequence (mKBS) in active promotors. Our data points towards a role for Kaiso in highly expressed target genes functioning in key fundamental cancer-related processes including cell cycle regulation and DNA damage response.