A major goal of cancer research is the identification of tumor-specific vulnerabilities that can be exploited for the development of therapies that are selectively toxic to the tumor. We show here that the transcriptional coactivators peroxisome proliferatoractivated receptor gamma coactivator 1␤ (PGC1␤) and estrogen-related receptor ␣ (ERR␣) are aberrantly expressed in human colon cell lines and tumors. With kinase suppressor of Ras 1 (KSR1) depletion as a reference standard, we used functional signature ontology (FUSION) analysis to identify the ␥1 subunit of AMP-activated protein kinase (AMPK) as an essential contributor to PGC1␤ expression and colon tumor cell survival. Subsequent analysis revealed that a subunit composition of AMPK (␣2␤2␥1) is preferred for colorectal cancer cell survival, at least in part, by stabilizing the tumor-specific expression of PGC1␤. In contrast, PGC1␤ and ERR␣ are not detectable in nontransformed human colon epithelial cells, and depletion of the AMPK␥1 subunit has no effect on their viability. These data indicate that Ras oncogenesis relies on the aberrant activation of a PGC1␤-dependent transcriptional pathway via a specific AMPK isoform.A third of all human cancers, including a substantial percentage of colorectal, lung, and pancreatic cancers, are driven by activating mutations in Ras genes. Activating K-Ras mutations are present in 35 to 40% of colon tumors and are thought to be both drivers of tumorigenesis and determinants of therapeutic regimens (1). Therapeutic disruption of Ras function has been clinically ineffective to date, but investigation of Ras pleiotropy continues to yield a diversity of downstream effectors with obligate roles in the maintenance and adaptation of Ras-driven tumors to changing environments. The Raf-MEK-extracellular signal-regulated kinase (ERK) signaling pathway is essential for the oncogenic properties of mutated K-Ras (2). However, numerous potent and specific MEK inhibitors have been developed yet have failed to demonstrate single-agent efficacy in cancer treatment (3). As a molecular scaffold of the Raf-MEK-ERK kinase cascade (4, 5), kinase suppressor of Ras 1 (KSR1) is necessary and sufficient for Ras V12 -induced tumorigenesis (4), mouse embryo fibroblast (MEF) transformation (5, 6), and pancreatic cancer growth (7) but dispensable for normal development (4). KSR1 is overexpressed in endometrial carcinoma and is required for both proliferation and anchorage-independent growth of endometrial cancer cells (8). Except for minor defects in hair follicles, KSR1 knockout mice are fertile and develop normally (4).This observation predicts that small molecules targeting KSR1 and functionally related effectors should preferentially target Rasdriven tumors while leaving normal tissue largely unaffected. More generally, this observation demonstrates that tumor cells, while under selective pressure to adapt to inhospitable environments and proliferate without constraint, will adopt strategies that, while advantageous to that singular purpose, create...

No abstract

Multiple studies have revealed that Ras-driven tumors acquire unique vulnerabilities by adapting cellular mechanisms that promote uncontrolled proliferation and suppress apoptosis. Targeting these vulnerabilities provide opportunities to develop novel, efficacious cancer therapeutics that lack the harmful side effects accompanying current therapies. RNA interference (RNAi) of the molecular scaffold Kinase Suppressor of Ras 1 (KSR1), which modulates ERK activation downstream of oncogenic Ras, selectively kills malignant, Ras-driven cancer cells, but does not kill immortalized, non-transformed human colon epithelial cells (HCECs). With the exception of a minor hair follicle defect, KSR1-/- mice are fertile and phenotypically normal, suggesting that KSR1 is not required for normal cell survival and that Ras-driven and KSR1-dependent pathways may yield valuable new targets for therapeutic development. To identify targets, like KSR1, that are required for cancer cell survival but not normal cell survival, we used a gene expression-based signature screening approach termed Functional Signature Ontology (FUSION, Potts et al. Sci. Signaling 2013) to screen 15,172 genes in the K-RasD13-bearing human colorectal cancer cell line HCT116. We quantified the functional similarity between KSR1 and each individual gene screened using Euclidean Distance and Pearson Correlation similarity metrics. Additional metrics were added to identify the best targets for biological validation including off-target analysis (siRNA seed sequences), cell viability evaluation, expression analysis, and enrichment analysis. Initial biological validation is completed by assessing cell viability following transient depletion of a screen hit in anchorage-independent and normal culture conditions in HCECs and HCT116s. Due to the similarity between the gene expression signatures, Timeless Circadian Clock (TIMELESS) was identified as being KSR1-like and a potential target. We found that transient TIMELESS depletion decreases cell viability in HCT116 cells under anchorage-independent conditions (47% decrease, p < 0.0001, N = 4). In normal culture conditions, TIMELESS depletion decreases cell viability in HCT116 cells, but not HCECs (HCEC 14% decrease, p > 0.05; HCT116 49% decrease, p < 0.0001, N = 6). TIMELESS is upregulated at the RNA level in colon tumors compared to normal colon tissue (∼2.2 fold, p < 0.0001) (TCGA) and is upregulated at the protein level in three human colon cancer cell lines (HCT116, SW480, SW620) bearing activated Ras compared to HCECs (4-7 fold). Our data indicate that the FUSION screen provides a platform for identifying novel therapeutic targets and demonstrates the potential to identify oncogene-specific vulnerabilities in an unbiased manner. TIMELESS overexpression represents a vulnerability in Ras-driven tumors that will reveal novel and selective targets found in Ras-driven cancers that can be used in the development of selective therapeutics. Citation Format: Beth K. Clymer, Kurt W. Fisher, David L. Kelly, Michael A. White, Robert E. Lewis. TIMELESS is a KSR1-like effector of Ras-driven colon tumorigenesis. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 1252.

RPA identified age 78 years, Charlson score 5, %DLCO < 71%, % FEV1 < 65%, tumor size 3 cm and SUV max 12 as significant cutoffs for adverse PFS outcome. RPA identified %DLCO < 71%, tumor size 3 cm and SUV max 12 as significant cutoffs for adverse TTP. On multivariate analysis, age (pZ0.05), Charlson score (pZ0.03), %DLCO (pZ0.00001), non-adenocarcinoma histology (pZ0.02), and tumor size (pZ0.002) remained predictive of PFS. Correlations with biological markers are currently ongoing. Conclusion: Age 78 years, Charlson score 5, %DLCO < 71%, tumor size 3 cm and non-adenocarcinoma histology were found to be independent predictors of PFS. Patients with %DLCO < 71%, tumor size 3 cm and SUV max 12 had adverse TTP outcomes, suggesting that they may benefit from treatment intensification. A predictive model is being develop and will be presented.

Kinase Suppressor of Ras 1 (KSR1), a molecular scaffold for the Raf/MEK/ERK kinase cascade, is required for transformation and survival of cells bearing oncogenic Ras, but is not required for normal cell survival. KSR1 knockout mice are phenotypically normal, and depletion of KSR1 by RNAi does not kill immortalized, non-transformed human colon epithelial cells (HCECs), suggesting that KSR1 or KSR1-dependent effectors may serve as highly selective therapeutic targets. Our previous work showed that KSR1 regulates expression of the transcription factor peroxisome proliferator-activated receptor gamma coactivator 1 (PGC1β), whose expression is critical to colon cancer cell survival. Using RNAi and pharmacological inhibition, we observed that KSR1 and ERK drive both cap-dependent and cap-independent translation of Myc in human colon tumor cell lines, which increases PGC1β mRNA expression (McCall et al. MCB, 2016). This effect is mediated through the KSR1 and ERK-dependent phosphorylation of 4E-PB1 and phosphorylation-dependent loss of PDCD4 that reverses their inhibitory effect on eIF4E and eIF4A, respectively. Treatment with ERK inhibitor SCH772984 showed that ERK regulates PDCD4 protein expression and 4EBP1 phosphorylation to promote Myc expression in four of the seven colon cancer cell lines. These observations suggest that KSR1-dependent regulation of translation may be a common mechanism used to support tumor cell survival. Using KSR1 as a reference standard, Functional Signature Ontology (FUSION, Potts et al. Sci. Signaling 2013) was used to identify genetic vulnerabilities in human colon tumor cells that are absent in HCECs. Comparison of results from this screen to genome-wide polysome profiling data from cells transformed with Ras and Myc (Truitt et al. Cell 2015) identified 25 candidate mRNAs with altered translational efficiency that are predicted to support the survival of human colon tumor cells. These data suggest that KSR1 and ERK-dependent alteration of the translational landscape is a common strategy used to support colon tumor cell survival. Identification of mRNAs preferentially translated in colon tumor cells may yield novel targets for therapeutic manipulation and may provide unique markers with which to classify colon tumors, predict patient outcome, and select effective treatment. This abstract is also being presented as Poster B36. Citation Format: Eyerusalem M. Lemma, Jamie L. McCall, Beth K. Clymer, David L. Kelly, Michael A. White, Robert E. Lewis. Translational control of human colon tumor cell survival. [abstract]. In: Proceedings of the AACR Special Conference on Translational Control of Cancer: A New Frontier in Cancer Biology and Therapy; 2016 Oct 27-30; San Francisco, CA. Philadelphia (PA): AACR; Cancer Res 2017;77(6 Suppl):Abstract nr PR04.