Statistical analyses were performed utilizing a Wilcoxon rank sum test on log10 transformed gIC50 values, comparing median values for wild type vs

Statistical analyses were performed utilizing a Wilcoxon rank sum test on log10 transformed gIC50 values, comparing median values for wild type vs. expression studies exhibited that this BET/MEK combination uniquely sustains down-regulation of genes associated with mitosis, leading to prolonged growth arrest that is not observed with either single agent therapy. These studies spotlight a potential to enhance the clinical benefit of BET and VGR1 MEK inhibitors and provide a strong rationale for clinical evaluation of BET/MEK combination therapies in malignancy. Introduction BET proteins (BRD2, BRD3, BRD4, and BRDT) modulate expression of genes involved in ALW-II-41-27 cell growth and oncogenesis by binding to acetylated chromatin via their bromodomains, which in turn recruit downstream effectors that promote transcription. Selective BET inhibitors, such as I-BET151 and the clinical molecule GSK525762 (I-BET762; I-BET)1,2, abrogate binding of BET proteins to acetylated chromatin thereby inhibiting BET-dependent transcription1,2. BET inhibitors exhibit broad anti-proliferative activity in malignancy models3. Although several mechanisms have been implicated in the efficacy of BET inhibitors including transcriptional suppression of oncogenes3, there is no consensus and it is likely that mechanisms vary, thus making the identification of predictive biomarkers hard. Although BET inhibitors show broad activity in many malignancy types, within each you will find resistant models. Understanding the basis of BET inhibitor sensitivity and resistance is important to inform the clinical development of BET inhibitors as monotherapies and to identify rational combinations. To this end, we analyzed genetic data from a large set of cell lines treated with GSK525762 to identify biomarkers of sensitivity and resistance. From these studies, we recognized KRAS mutations as a significant predictor of resistance to BET inhibition. This led us to hypothesize that combinations with inhibitors of RAS signaling, such as MEK inhibitors, may further improve upon BET inhibitor efficacy. Indeed, we observed broad synergistic effects for BET/MEK combinations across malignancy models, which we attribute to profound and sustained inhibition of MEK/ERK signaling that is specifically observed with the combination ultimately leading to growth arrest and cell death. Results RAS mutations are novel biomarkers of resistance to GSK525762 To identify genetic predictors of sensitivity or resistance to BET inhibitors we first examined the anti-proliferative activity of GSK525762 in ~230 malignancy cell lines. Hematologic malignancy cell lines were highly sensitive (low growth IC50 values and net cell death) to GSK525762, whereas solid tumor models exhibited a wide range in sensitivity (gIC50 13?nM to ?29.3?M; partial, cytostatic, and cytotoxic responses), providing an opportunity to compare the genetic profiles of a large number of sensitive and resistant lines (Fig. 1a, b, Supplemental Table S1). Using publicly available data for 19,312 genes with protein-changing mutations, we performed unbiased analyses of genetic predictors of sensitivity or resistance to GSK525762 based on gIC50 values. These analyses recognized 634 genes with protein-changing mutations that correspond to resistance or sensitivity to GSK525762 (Wilcoxon rank sum test mutations were among the top five ALW-II-41-27 most significant (mutations among the more resistant cell lines (median gIC50 mutant?=?1667?nM vs WT?=?550?nM; Fig. ?Fig.1c).1c). Analysis of mutations at the amino acid level further recognized G12 missense mutations as significantly (mutations and resistance was observed in colorectal malignancy ALW-II-41-27 (CRC) cell lines, where mutations are frequent (mutations in 13/22 lines; 59%) (Fig. ?(Fig.1d),1d), indicating that the association with resistance is not driven by the general responsiveness of individual tumor types. Finally, mutations were significantly depleted (compared to parent cells, suggesting that activated RAS/MEK/ERK signaling may be associated with adaptive resistance to BET inhibition in TNBC14. Consistent with these reports, RTK and immediate early gene up-regulation and increased p-ERK1/2 are observed in NCI-H510 cells after GSK525762 treatment (Fig. ?(Fig.4b,4b, Supplemental Table S8, Supplemental Determine S20). Based on these results, we suggest that NCI-H510 cells become more dependent on MEK/ERK signaling for survival following GSK525762 treatment, therefore explaining the synergistic growth inhibition and cytotoxicity observed for BET/MEK combinations in this model.