Integrating Computational And Experimental Approaches To Build Predictive Models Of The Smg1 Kinase Network In Glioblastoma Stem Cells

Glioblastoma multiforme (GBM) is the most common and most aggressive tumor of the central nervous system, with a mean survival of only 12-14 months post diagnosis and treatment. A major reason for the poor prognosis of GBM is attributed to a small population of refractory GBM stem cells (GSC) that have been shown to initiate and sustain tumor growth. Current evidence suggests that GSC preferentially reside in hypoxic niches in the tumor that contribute to maintenance of its stem cell properties such as self-renewal and resistance to therapy. In order to identify targets for achieving effective tumor remission, we have performed an RNAi kinome screen to identify specific kinases that enable the survival of GSCs in hypoxic microenvironment. Only 40% of the hits were common between the two cell lines screened, emphasizing heterogeneity of tumor response that necessitates personalized targeted therapy for GBM. The SMG1 kinase was identified in this screen as a gene that is preferentially inhibitory to one of our GSC lines under hypoxia (1% oxygen). SMG1 is a member of the phosphoinositide-like family of kinases that includes ATM, ATR and mTOR. SMG1 is a potential target for cancer therapy to counteract tumor cell growth and resistance to DNA damaging chemotherapeutic agents due to its role in nonsense mediated decay (NMD) and DNA damage response (DDR) pathways. Our results are therapeutically significant since we found additive growth inhibitory effect of SMG1 knockdown in GSC in the presence of DNA damaging drug Temozolomide (TMZ), the current standard of care for GBM. The Upf1 is an essential NMD factor that is phosphorylated by SMG1 and is also required for DNA damage induced cell cycle S phase progression. Interestingly, our results show that the growth inhibitory effect of SMG1 knockdown correlated with the induction of Upf1 in GS7-2 cells specifically under hypoxia. We assessed the role of downstream mediators in NMD pathway (ATF-4, MAP4K12, SMG7 and p-EIF2alpha) and DDR pathway (p53, p-p53 (Ser 15), p21, ATM, ATR, Chk1, p-Chk1, Chk2, p-Chk2 and MRN complex) to identify critical components of the pathways that contribute to the differential growth inhibition effect. We have used the Cell Collective modeling software platform (BMC Systems Biology, 6, 96) to model the SMG1 signaling network by integrating our data with results in the literature in order to simulate cellular growth and survival of GSC under hypoxia and TMZ induced DNA damage. Our model predicts that the preferential growth inhibitory role of SMG1 inhibition in GS7-2 cells under hypoxia specifically is due to the role of NMD pathway in regulating autophagy and this is currently being tested experimentally in multiple cell lines. The long-term goal of this study will be to develop therapeutic predictions for targeting SMG1 or other kinases from the screen for a given tumor and thereby develop personalized therapies for treating GBM. Citation Format: Sejuti Sengupta, Steven K. Young, Surbhi Goel-Bhattacharya, Brent H. Cochran. Integrating computational and experimental approaches to build predictive models of the SMG1 kinase network in glioblastoma stem cells. [abstract]. In: Proceedings of the AACR Special Conference on Translation of the Cancer Genome; Feb 7-9, 2015; San Francisco, CA. Philadelphia (PA): AACR; Cancer Res 2015;75(22 Suppl 1):Abstract nr A2-65.