PubMed PMID: 2188735. sub-G1 accumulation, in a distinct subset of cell lines. Furthermore, TAK1 inhibition causes G2/M cell cycle blockade and polyploidy in many of the cell lines. MEK plus TAK1 inhibition causes reduced G2/M/polyploid cell figures and additive cytotoxic effects in KRAS/TAK1-dependent cell lines as well as a subset of mutations with serrated carcinoma histology in the N subtype. Both N and WN subtypes bear molecular hallmarks of MEK and TAK1 dependency seen in cell lines. Therefore, N and WN subtype signatures could be utilized to identify tumors that are most sensitive to anti-MEK/TAK1 therapeutics. INTRODUCTION Colon cancers are molecularly and histologically heterogeneous with multiple oncogenic driver mutations promoting tumorigenesis via deregulated MAP kinase, Wnt, BMP and NFB signaling pathway activation. and mutations occur frequently and drive MEK-ERK mitogenic pathway activation. mutations cooperate with inactivating and mutations to hyperactivate deregulated canonical Wnt and TGF-/BMP receptor signaling, respectively, causing accelerated and aggressive tumorigenesis (1-4). KRAS, Wnt, and TGF-/BMP pathways are subject to considerable crosstalk through complex, context dependent mechanisms leading to molecular and histological intra- and Rabbit Polyclonal to BHLHB3 intertumor heterogeneity. This complexity is usually illustrated by global gene expression profiling and molecular subtype classifications (5-8). mutant tumors do not classify Altiratinib (DCC2701) into a unique subtype and display highly diverse molecular signatures. Recently, molecular diversity has been documented in mutant lung cancers, where co-occurring mutations in and generate unique molecular subtypes with selective pharmacological vulnerabilities (9, 10). Identifying subtype-selective vulnerabilities in RAS/RAF pathway-dependent cancers may yield more efficacious therapeutics. Using a transcriptional signature associated with KRAS dependency in colon cancer cell lines, we recognized the TGF- activated kinase (TAK1) as a critical cell survival mediator in KRAS dependent cells (11). We blocked TAK1 kinase activity with an anti-inflammatory agent, 5Z-7-oxozeaenol (5Z-7-oxo), which induces apoptosis in KRAS-dependent cells. In this study, we decided that 5Z-7-oxo has off-target MEK kinase inhibitory activity. This prompted our desire for evaluating the cytotoxic effects of combining MEK and TAK1 inhibition with single agents. mutant colon cancer cell lines exhibit a spectrum of MEK dependencies whereas mutant cell lines are significantly more MEK dependent (12). Furthermore, MEK inhibitor sensitivities can be correlated with unique transcriptional signatures (13). We hypothesized that combining MEK and TAK1 inhibitors would induce additive cytotoxic effects in a KRAS dependent subtype Altiratinib (DCC2701) of colon cancer cell lines. Indeed, previous studies have described effective combination methods with MEK kinase inhibitors to treat KRAS-driven cancers (14-16). TAK1 mediates innate immunity and proinflammatory signaling via regulation of NFB and AP-1 (Jun/Fos) dependent transcriptional programs (17). Autocrine or paracrine proinflammatory signaling drives KRAS-dependent tumor cell survival (18-23). However, the underlying mechanisms and implications of KRAS dependent proinflammatory signaling for treatment of RAS/RAF pathway dependent tumors has yet to be fully determined. In this study, we analyzed MEK/TAK1 dependencies in a comprehensive panel of colon cancer cell lines that display varying molecular and phenotypic characteristics. The overarching goal was to identify definitive molecular correlates of MEK/TAK1 co-dependencies. Given the role of TAK1 in proinflammatory signaling, we investigated the role of the Altiratinib (DCC2701) KRAS-TAK1 axis in regulating inflammatory cytokine expression levels and subsequent effects on MEK/TAK1 dependencies. Finally, we decided whether molecular hallmarks of MEK/TAK1 dependencies correlate with molecular subtype classifications of main tumors from colon cancer patients. MATERIALS AND METHODS Oligonucleotide microarray analyses Robust Multiarray Averaged (RMA) normalized main tumor data from colon cancer patients were utilized for gene expression analyses of the canonical Wnt/NFB signatures and are available through the NCBI GEO database (Affymetrix Human Exon Array – “type”:”entrez-geo”,”attrs”:”text”:”GSE39582″,”term_id”:”39582″GSE39582) (7). All genome-scale datasets were processed and analyzed using R and Bioconductor software packages. A set of genes whose expression correlated significantly with the canonical Wnt target gene was first recognized. Within this list was the most correlated gene with dataset (7). The circulation chart depicts derivation of a canonical Wnt signature using Pearson correlation coefficients to identify genes correlated with expression. The heat map represents gene expression in 3 major subtypes revealed by hierarchical clustering of the 184 mutations, the original six subtype classification by Marisa et al., Wnt/NFB subtypes (W=Wnt-high; N=NFB-high; WN=Wnt+NFB-high) and mismatch repair (MMR) status (d=deficient; p=proficient). (B) Kaplan-Meier curves showing relapse-free survival of patients with tumors classified into N/W/WN subtypes. (C) Boxplots depicting expression of selected canonical Wnt targets (and and and mutations in.