Abstract 4771: Identification of inhibitors of TWIST1 as a treatment for lung cancer

Abstract

Abstract Although a large fraction of non-small cell lung cancers (NSCLC) are dependent on defined oncogenic driver mutations, little progress has been made in the treatment of patients with the most frequently observed driver mutation, mutant KRAS. Acquired resistance to therapy targeting other driver mutations (EGFR mutant and ALK translocation positive tumors) is all but inevitable. We recently demonstrated that Twist1, a basic helix-loop-helix transcription factor, cooperates with mutant Kras to induce lung adenocarcinoma in mouse models, and that inhibition of Twist1 in murine models and human lung cancer cell lines led to oncogene-induced senescence (OIS) and in some cases, apoptosis. We have also found that TWIST1 is essential for tumor maintenance in human NSCLCs characterized by oncogenic drivers including KRAS mutation, EGFR mutation and c-MET amplification. As TWIST1 is not typically expressed post-natally, therapies directed against TWIST1 may present a more specific and perhaps less toxic therapy. Targeting the TWIST1 pathway represents an exciting and novel therapeutic strategy which may have a significant clinical impact. We employed a combined bioinformatics-chemical approach to identify pharmacologic inhibitors of TWIST1. We used gene expression profiles from several KRAS mutant human NSCLC cell lines following shRNA-mediated TWIST1 knockdown to perform connectivity map (CMAP) analysis. We found that several of our highly ranked compounds had significant growth inhibitory effects in NSCLC cell lines. Furthermore, several of our candidate agents produced dose-dependent inhibition of TWIST1-induced dissemination in a novel 3D organoid dissemination assay. Interestingly, a family of related harmala alkaloids including harmine ranked highly on the CMAP analysis. We found that harmine could not only inhibit growth in several oncogene driver defined NSCLC cell lines through the induction of apoptosis but could also decrease TWIST1 levels through a post-transcriptional mechanism. The growth inhibitory effects of the harmala alkaloids correlated with the ability to degrade TWIST1. Additionally, independent of our CMAP analysis, we identified another compound, salinomycin, which appears to modulate cancer cell growth and resulted in a decrease in TWIST1 levels through a post-transcriptional mechanism as well. Interestingly, salinomycin treatment induced a mesenchymal-epithelial transition (MET) in NSCLC which could be partially blocked in TWIST1 overexpressing NSCLC cell lines. We are currently examining the in vivo efficacy of these agents. In conclusion, we have identified several compounds that inhibit TWIST1 dependent dissemination and in some cases can lead to TWIST1 degradation. Since TWIST1 is essential for not only KRAS mutant NSCLC but more broadly for oncogene driven NSCLC, these studies could lead to the development of a novel class of inhibitors which could have a significant clinical impact. Citation Format: Sarah NH Chatley, Jessica Cades, Neil Neumann, Lucia Mazzacurati, Zachary Yochum, Katriana Nugent, Yoon-Jae Cho, Andrew Ewald, Charles Rudin, Phuoc Tran, Timothy F. Burns. Identification of inhibitors of TWIST1 as a treatment for lung cancer. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 4771. doi:10.1158/1538-7445.AM2014-4771