Fall 2017 Research Projects

 

Studying the Response of Early-stage Non-small-cell Lung Carcinoma to Surgical Treatment: New Predictors, Drug Targets and Biomarkers 

Principal Investigators:

Curtis Harris, Center for Cancer Research, National Cancer Institute
Eytan Ruppin, Department of Computer Science and Center for Bioinformatics & Computational Biology (CBCB), University of Maryland

Ph.D. Student:

Leandro Cruz Hermida, Computer Science, University of Maryland, leandro@leandrohermida.com

Lung cancer is the leading cause of cancer-associated death in men and women in the United States. Currently, the recommended treatment for early-stage lung cancer patients is surgery, which may be followed by chemotherapy in patients with high-risk, margin-negative tumors, as determined by the tumor pathology. Despite therapeutic advances and these current guidelines, up to 30% of early-stage lung cancer patients suffer a recurrence within 5 years of curative surgery. We set here to build novel machine learning predictors for identifying patients with high risk of recurrence following curative surgery, based on a broad range of molecular data measured from each tumor. This will lead to more accurate predictors of patient's response to curative surgery and will also illuminate new therapeutic options (and biomarkers) aimed at reducing the levels of post-surgical recurrence.

2017 Ruppin.png

Omics of cancer cell lines for identifying novel response determinants for precision medicine

Principal Investigators:

Yves Pommier, Developmental Therapeutics Branch, National Cancer Institute
Wojciech Czaja, Department of Mathematics, University of Maryland

Ph.D. Student:

Jeremiah Emidih, Mathematics, University of Maryland, jemidih@math.umd.edu

Cancer is a disease of the genome, and DNA-targeted drugs including cisplatin and topoisomerase inhibitors remain the most widely used anticancer treatments; yet a large number of patients fail to respond to these cytotoxic agents, and it is presently not possible to distinguish responders from non-responders. To precisely use DNA targeted drugs, it is essential to understand the molecular networks that determine treatment responses. The existence of large cancer cell line databases with whole genome and extensive drug response analyses provides novel opportunities to recover elements of these networks, and examine their integrated behavior. Our project is designed to discover and experimentally validate predictive genomic determinants of response to DNA targeted drugs using novel algorithms, with the ultimate aim of rationalizing the precise use of topoisomerase inhibitors and their combination with other drugs for cancer patients.

Flow-chart outline. Our project utilizes integrated data from all major cancer cell line databases (NCI-60, CCLE, CTRP, and GDSC, a total of approximately 1,400 cancer cell lines) with approximately 22,000 anticancer drugs (1) and whole genome analyses (transcriptome, exome sequencing, gene copy number, promoter methylation, microRNA, RNA seq) (2). We will analyze data across available sources with the help of our CellMiner applications (3) to nd novel genomic determinants for DNA-targeted drugs with a focus on topoisomerase inhibitors, DNA repair and chromatin-remodeling pathways (4). Emerging genomic determinants will be prioritized using pathway analyses (5) and tested for validation in isogenic cancer cell lines in the NCI Developmental Therapeutics Branch (6). The validated novel genes will then be further tested in murine models in vivo and included in clinical trials.

Flow-chart outline. Our project utilizes integrated data from all major cancer cell line databases (NCI-60, CCLE, CTRP, and GDSC, a total of approximately 1,400 cancer cell lines) with approximately 22,000 anticancer drugs (1) and whole genome analyses (transcriptome, exome sequencing, gene copy number, promoter methylation, microRNA, RNA seq) (2). We will analyze data across available sources with the help of our CellMiner applications (3) to nd novel genomic determinants for DNA-targeted drugs with a focus on topoisomerase inhibitors, DNA repair and chromatin-remodeling pathways (4). Emerging genomic determinants will be prioritized using pathway analyses (5) and tested for validation in isogenic cancer cell lines in the NCI Developmental Therapeutics Branch (6). The validated novel genes will then be further tested in murine models in vivo and included in clinical trials.


Deciphering key Ras-dependent activation mechanisms by computer simulations and NMR

Principal Investigators:

Ruth Nussinov, Cancer and Inflammation Program, National Cancer Institute
David Fushman, Department of Chemistry and Biochemistry, University of Maryland

Ph.D. Student:

Tsung-Jen Liao, Biophysics, University of Maryland, tjliao@umd.edu

In our (unfunded) application last year, we proposed the highly significant aim of unraveling the crosstalk between Ras/Raf-elicited MAPK and Ras/RASSF5-elicited Hippo signaling by deciphering the enigmatic mechanism of RASSF5 tumor suppressor action; MAPK signaling promotes cell proliferation, whereas Hippo’s abolishes it. During the last year, our large scale simulations at the NCI have helped in crystallizing the hypothesis underlying the first aim [1-3] (Fig. 1) and on-going NMR experiments at UMD (preliminary results below) aim to test it and provide leads to further refine it. Our faster than anticipated progress allows us to also take up the critical aim of Ras activation via SOS-mediated GDP→GTP exchange by resolving the question of how SOS activates Ras (work has initiated). Accomplishing the aims of Ras activation by SOS, and RASSF5 activation (thus Hippo’s signaling) by Ras will provide a broad view on tumor proliferation, helping drug discovery efforts.

Extracellular epidermal growth factor (EGF) binding to the EGF receptor induces association of SHC, GRB2 and SOS, leading to SOS activation. SOS activates Ras by exchanging GDP to GTP. Active Ras can either associate as well as activate Raf(MAPK) or RASSF5(Hippo). Active RASSF5 increases the fluctuations of its SARAH domain which now heterodimerizes with the SARAH domain of the MST kinase leading to kinase domain dimerization. MST kinase auto-transphosphorylation induces Hippo signaling and results in YAP1 phosphorylation. Phosphorylated YAP1 is degraded by the SCF^β-TRCP ubiquitin ligase complex, preventing cell proliferation and suppressing tumor.

Extracellular epidermal growth factor (EGF) binding to the EGF receptor induces association of SHC, GRB2 and SOS, leading to SOS activation. SOS activates Ras by exchanging GDP to GTP. Active Ras can either associate as well as activate Raf(MAPK) or RASSF5(Hippo). Active RASSF5 increases the fluctuations of its SARAH domain which now heterodimerizes with the SARAH domain of the MST kinase leading to kinase domain dimerization. MST kinase auto-transphosphorylation induces Hippo signaling and results in YAP1 phosphorylation. Phosphorylated YAP1 is degraded by the SCF^β-TRCP ubiquitin ligase complex, preventing cell proliferation and suppressing tumor.