Abstract Guide - The Gulf Coast Consortia
Trainee Symposium on Cancer Research in Texas January 30, 2015 Poster Presenters (listed alphabetically) Last Name, First Poster # Training Program Home Institution Poster Title Alexander, Angela 26 MDACC Cyclin E/CDK2 is a Novel Therapeutic Target in Inflammatory Breast Cancer Araujo, Patricia 62 UT San Antonio Directing Musashyi1 Role in Chromatin Remodeling and its Consequences to DNA Repair and Radio-Resistance Ashcraft, Keith 63 UT San Antonio Deletions in the Long Arm of Chromosome 18 in Metastatic Prostate Cancer Atkinson, Erin 27 MDACC NPSD4: A New Player in the DNA Damage Response Baek, Guem 78 UT Southwestern MCT4 Defines a Global Subtype of Pancreatic Cancer with Poor Prognosis and Unique Metabolic Dependencies Bozeman, Ronald No Poster UT Southwestern Biological Effects on Lung Cancer Susceptible K-Ras1a1Mice Irradiated with an Acute Dose of Protons Compared to Protracted Proton Provided as a Solar Event Simulation Bui, Thanh 58 UT Houston Oral Hygiene, Oral Health, and Oral HPV Infection Chang, Katherine 64 UT San Antonio Targeting the CD44/EMT Phenotype for Improving Response to Therapy in Pancreatic Cancer Colosimo, Dominic 81 UT Southwestern Discoipyrroles: A Multifaceted Approach to Understanding Novel Marine Natural Products Dawson, Emily Packard 1 BCM A Critical Role for Cyclin D1 in a Testicular Germ Cell Tumorgenesis De Angelis, Carmine 2 BCM AP-1 as a Key Mediator/Target of Endocrine Resistance in Breast Cancer Deng, Yilun 65 UT San Antonio The Tumor Suppressor TMEM127 is a Lysosomal Protein that Associates with LAMTOR1, a Component of the mTORC1 Activation Complex at Lysosomal Surface Dextrase, Katherine 28 MDACC Characterization of Photoacoustic Thermography for ImageGuidance and Monitoring of Photothermal Ablations Donkor, Moses 48 UT Austin Tumor-Induced Regulation of T Cell Fe Gamma Receptor Expression Doosten, Iman 29 MDACC Dutchak, Paul 79 UT Southwestern Deregulated Cyclin E Expression Mediates Resistance to Aromatase Inhibitors in Postmenopausal Breast Cancer Patients Regulation of Amino Acid Metabolism and Hematopoiesis by mTORC1 Negative Regulator NPRL2 1 Trainee Symposium on Cancer Research in Texas Last Name, First Poster # January 30, 2015 Training Program Home Institution Poster Title Fan, Yu 15 MDACC MuSE: Somatic Evolution Estimation for Mutation Calling in Sequencing Data of Matched Tumor-Normal Samples Ferrati, Silvia 59 UT Houston An Innovation and Flexible Approach for Cancer Cell Normalization with Cell-Derived Biovesicles Gallolu, Sachith No Poster UT Southwestern The Effect of with No Lysine (WNK) Pathway on Autophagy Grimes, Allison 66 UT San Antonio Genetic Markers for Chemotherapy-related Pancreatitis in Childhood ALL Hsu, Ya-Ting 67 UT San Antonio Co-regulation of EpICD/Lef-1 Gene Targets Correlate with Tumor Progression in Advanced Endometrial Cancer Huang, Le 30 MDACC Jiang, Yu (Sherry) 49 UT Austin Impact of Differentiation Status of Kidney Progenitors in Wilms Tumorogenesis Robust Strand Exchange Reactions for Sequence-specific Detection of Nucleic Acid Amplicons and its Applications in Point-of-Care Diagnostics Kaoud, Tamer 50 UT Austin Inhibition of the TRPM7 Kinase Domain Inhibits Breast Cancer Cell Migration and Invasion and Tumor Metastasis Kelly, Aaron 3 BCM Data Integration Across Multiple Proteomic Platforms Leads to a Novel Prognostic Signature for Osteosarcoma Patients Kendall, Genevieve No Poster UT Southwestern Zebrafish Modeling of PAX3-FOX01 Driven Rhabomyosarcoma Kim, Sun 68 UT San Antonio Biophyical Study of TGF-β: Betaglycan Interactions for Determining the Mechanism of Betaglycan’s Potentiation of TGF- β Signaling Kirk, Brian 16 UH Ko, How-Wen 31 MDACC Krishnan, Samaya 69 UT San Antonio Detection of Long Non-coding RNA Structural Motifs Involved in PRC2 Functions Role of GSK3β-Mediated EZH2 Phosphorylation in Breast Cancer Stem Cells in Tumorogenesis Signifcance of Proto-Oncogene PELP1 in Error-Prone Alternative-NHEJ Pathwayq Kurtova, Antonina 4 BCM Blocking Cancer Stem Cells from Recurrent Wound-Induced Repopulation Confers Chemosensitivity Laurant, Jon 51 UT Austin Uncovering Rules Governing Functional Replacement Between Humans and Yeast Lee, Jaehyuk 17 UH Comparison Response of siRNA Treatment by Metabolic Changes in Ovarian Cancer by Nuclear Magnetic Resonance Leung, Marco 32 MDACC SNES: Single-Cell Exome Sequencing 2 Trainee Symposium on Cancer Research in Texas Last Name, First Poster # January 30, 2015 Training Program Home Institution Poster Title Leung, Wai Chung (Justin) 52 UT Austin Histone H2A Variants and DNA Repair Lim, Byung Joon 53 UT Austin Improvement of Electrochemical DNA Sensors (E-Sensors) for Application to Genetic Testing of Hereditary Cancers Li, Xu 18 UH Li, Xueling 19 UH A Standardized Scoring System for Affinity Purifications/Mass Spectrometry Data The Genome-wide Functional Modulatory Network in Epithelial-Mesenchymal Transition Li, Zhenlong 60 UT Houston A Multiscale Computational Platform for the Rational Design of Folate-Receptor-Specific Drug Nanocarriers Lin, Justin 5 BCM Mechanisms of Maf1-Mediated Repression of Oncogenetic Transformation Liu, Jun 70 UT San Antonio An in vitro Inflammation-Induced Acinar-to-Ductal Metaplasia Model for Primary Human Pancreatic Acinar Cells Logan, Monica 6 BCM Role of Dicer Haploinsufficiency in Aggressive Endometrial Cancer in Mice and Women Long, Byron 20 UH An Image Processing Algorithm for Counting Neurons and Measuring Neurite Lengths Lu, Hengyu 7 BCM Lu, Mingyang 21 UH McKenzie, Jodi 33 MDACC High-throughput Engineering and Functional Annotation of Cancer Fusion Genes Modeling EMT Genetic Regulatory Circuits in Metastatic Cancer Topoisomerase 1 Inhibitors Enhance the Efficacy of Immunotherapy in Melanoma Madan, Simran 8 BCM Whole Exome Sequencing of Human Osteosarcomas Mancha, Anna 71 UT San Antonio The Novel Role of 11β-Hydroxysteroid Dehydrogenases in NonMelanoma Skin Cancer Manton, Christa 34 MDACC Caspace-Dependent Effects of the Proteasome Inhibitors Bortezomib and Marizomib in Glioblastoma Manjunath, Hema 82 UT Southwestern Evaluating the Tumor Suppressor Activity and Mechanism of the let-7 miRNA Family in vivo Mgbemena, Victoria No Poster UT Southwestern Identification and Characterization of BRCA1 Modifiers Miller, Jason 83 UT Southwestern Zwitterionic Aminolipids for siRNA Delivery Min, Jaewon No Poster UT Southwestern Human Telomerase Overhangs in Telomerase-negative Cells are Generated by Multi-step Processes 3 Trainee Symposium on Cancer Research in Texas January 30, 2015 Montgomery, William 54 UT Austin Discovery and Development of Potent Anti-cancer Compound via a Diverted Total Synthesis Mukherjee, Seema 35 MDACC Progression of Small Cell Lung Cancer (SCLC) Nagaraja, Archana 36 MDACC Sustained Adrenergic Signaling Activates Pro-inflammatory Prostaglandin Network in Ovarian Carcinoma Napoli, Marco 37 MDACC Pharmacologic Inhibitions of the ANp63/DGCR8 Axis as a Novel Therapeutic Strategy for p53 Deficient and Mutant Tumors Neill, Nicholas 9 BCM Mechanisms Underlying the Tumor Suppressive Role of PTPN12 in Triple-negative Breast Cancer Nichols, Brandt 84 UT Southwestern Tumor-activated Meiotic Regulators Modulate DNA Damage Response Onyeagucha, Benjamin 72 UT San Antonio Targeting MiR-2x6 Improves the Efficacy of Paclitaxel in Human Triple Negative Breast Cancer Padanad, Mahesh 80 UT Southwestern Acyl-CoA Synthetase Long-chain Family Member 3 is Required for Mutant KRAS-driven Lung Cancer Peng, David 38 MDACC MiR-200 Modulation of ECM in Lung Cancer Invasion and Metastases Puig, Stephanie 30 MDACC Anti-cancer Drugs to Treat Cancer Pain Prakash Srivastava, Priyanka 22 UH Preferred Orientation of Oncogenetic K-Ras in Membrane Ren, Junyao No poster UT Southwestern IKKβ is an IRF5 Kinase that Instigates Inflammation Roh, Whijae 40 MDACC Genomic and Immune Profiles of Melanomas Treated with Immune Checkpoint Blockade Sammons, Rachel 55 UT Austin An in vitro System for Discovery of Inhibitors that Target the DRecruitment Site of ERK No Poster MDACC Identification and validation of Novel Drug Targets for the Eradication of KRAS-Independent Pancreatic Stem Cells Sanchez, Nora Sareddy, Gangadhara 73 UT San Antonio Selective Estrogen Receptors β Agonists as Novel Therapeutic Agents to Treat Glioblastoma Satterfield, Laura 10 BCM MIR-130b Targets ArhGAP1 Increasing Cde42 Activity and Metastatic Potential in Ewing Sarcoma Cells Schibler, Andria 41 MDACC COMPASS Regulates the Spindle Assembly Checkpoint through Histone H3 K4 Methylation 4 Trainee Symposium on Cancer Research in Texas Last Name, First Poster # Scott, Aaron January 30, 2015 Training Program Home Institution Poster Title 11 BCM Feasibility of a Symptom Tracking Smartphone Application Sehgal, Vasudha 23 UH Integrating Functional Hypothesis Generation for Identifying Pan Cancer miRNA Biomarkers Sreekumar, Amulya 12 BCM Determining the Role of Wnt and p53 Pathways in Mammary Stem Cell Fate Determination Srivastava, Jaya 56 UT Austin Twist1 Regulates Keratinocyte Stem Cell Proliferation and Migration is Required for Skin Tumor Formation Stanley, Dennis 74 UT San Antonio Dosimetric Effect of Proton Beam Energy on VMAT Treatment Plan Quality Due to Body Habitus for Advanced Prostate Cancer Su, Chun-Hui 42 MDACC Targeting Neutrophil Elastase in Breast Cancer Metastasis Tashakori, Mehmoosh 43 MDACC in Vivo Significance of Mdm4 and p73 Interaction in Development and Tumorigenesis Thakkar, Kaushik 44 MDACC TRIM24 Links Epigenetics and Metabolism in Breast Cancer Progression Thiabaud, Gregory 57 UT Austin Texaphyrin-based Platinum Conjugates: Combining Targeting and Therapy Tripathi, Swarnendu 24 UH Correlation Between Oncogenic Mutations, Signaling Pathways and Efficacy of Platinum Based Drugs Against Colon Cancer Updegraff, Barrett 85 UT Southwestern Sleeping Beauty Mutagenesis for the Functional Identification of Human Melanoma Metastasis Genes Vega, Anthony 86 UT Southwestern Quantitative Analysis of Cortical Actin-Membrane Microcluster Interactions Veland, Nicolas 45 MDACC PRMT6 Overexpression Causes Global Loss of DNA Methylation in Mouse Embryonic Stem Cells Walker, Christopher 46 MDACC Investigation of Detection Method Bias in Measuring Metabolic Fluxes with Hyperpolarized Magnet Resonance Agents Williams, LeTerrica 13 BCM Targeting the Tumor Vasculature with Antigen-specific T Cells Wu, Sherry Yen-Yao 25 MDACC Evoking Potent RNAi Response using Novel2’-OMePhosphorodithioated Modified siRNAs Yang, Dong 47 MDACC Synergistic Inhibition of PARP and Wee1 Kinase in Triplenegative Breast Cancer Yu, Xiaojie 75 UT San Antonio Preclinical Evaluation of MiR-195 as a Therapeutic Agent in NSCLC 5 Trainee Symposium on Cancer Research in Texas Last Name, First Poster # January 30, 2015 Training Program Home Institution Poster Title Yuan, Bin 76 UT San Antonio Mobilizing Erβ Antitumor Activity Through a Phosphotyrosine Switch Zavadil, Jessica 77 UT San Antonio Improving Therapy for Hepatocelluar Carcinoma by Combining a HDAC Inhibitor with an Alkylating Agent Zhang, Ning 87 UT Southwestern Analysis of Telmerase Position Effect Based on 3D-Fish Zhu, Wenyi 14 Baylor College of Medicine Daam2 Promotes Tumor Proliferation and Migration in Glioblastoma 6 Trainee Symposium on Cancer Research in Texas January 30, 2015 Dawson, Emily Packard – BCM Poster #1 A Critical Role for Cyclin D1 in Testicular Germ Cell Tumorigenesis Dawson E*, Lanza D*, Benton SM, and Heaney,J D.† Department of Molecular and Human Genetics, Dan L. Duncan Cancer Center, Center for Reproductive Medicine, Baylor College of Medicine *These authors contributed equally. Corresponding Author† Jason D. Heaney, PhD, Department of Molecular and Human Genetics, Baylor College of Medicine, One Baylor Plaza, MS BCM 225, Houston TX, 77030, 713-798-1778, Email: [email protected] Testicular germ cell tumors (TGCTs) are the most frequent solid tumor diagnosed in boys and young men. TGCTs arise from fetal germ cells whose pluripotent capacities facilitate tumor differentiation into a variety of somatic cell types. Current treatment regimens for TGCTs cause long-term side effects including hearing loss, cardiovascular disease, cognitive impairment, and infertility. Thus, improvements in early diagnosis and treatment options remain important and the social, emotional, and medical costs remain high. In the 129 inbred mouse model of human TGCTs, tumors initiate between embryonic days (E)13.5 to 15.5, which coincides with the mitotic:meiotic switch: germ cells of both sexes lose pluripotent capacity, female germ cells (oogonia) commit to meiosis, and male germ cells (gonocytes) enter mitotic arrest. We previously demonstrated that gonocyte proliferation, retention of pluripotency, and aberrant expression of genes associated with pre-meiotic oogonia and adult spermatogonia, including cyclin D1 (Ccnd1), at E15.5 were directly related with increased tumor risk. Based on the known oncogenic potential of Ccnd1 overexpression, we hypothesized that aberrant expression of Ccnd1 significantly contributes to TGCT initiation by inducing a breakdown in the mitotic:meiotic switch, which causes gonocytes to continue proliferating and retain pluripotency. Using a Ccnd1 knockout on the high TGCT risk 129-Chr19MOLF/Ei (M19) background, we found that Ccnd1 deficiency significantly reduced TGCT incidence by 62% (p<0.001) compared to wild-type and heterozygous animals. Ccnd1 deficiency reduces the number of proliferating (KI67-positive) and pluripotent (NANOG-positive) gonocytes. Ccnd1 expression in post-migratory embryonic germ cells is normally restricted to oogonia and is transiently expressed from E12.5 to E15.5, prior to their entry into meiosis. Expression of Ccnd1 from E12.5 to E15.5 in TGCT-susceptible gonocytes suggests that a signal normally restricted to the developing ovary is aberrantly active in the TGCT-susceptible testis. Retinoic acid (RA) signaling normally precedes Ccnd1 expression in embryonic oogonia, while RA is normally catabolized by CYP26B1 in the embryonic testis. Curiously, we find an altered expression pattern of embryonic germ cell sex-specific genes (e.g. Ccnd1, Nanos2, and Nodal) in TGCT-susceptible gonocytes. Therefore, we tested whether gonocyte expression of Ccnd1 is induced by an ectopic RA signal. Culturing embryonic testes from TGCT-resistant or TGCTsusceptible mice in medium containing RA induced CCND1 expression in gonocytes. We hypothesize that TGCT-susceptible gonocytes respond to both an abnormal RA signal and normal gonocyte developmental signals, which disrupt gonocyte entry into mitotic arrest and lead to TGCT initiation. These experiments reveal the treatment potential of inhibiting CCND-kinase activity for patients with TGCTs, a strategy which will target cancer cells without significant adverse effects to normal tissue. Baylor College of Medicine Comprehensive Cancer Training Program RP140102 -PI: Dr. Jeffrey Rosen, Baylor Research Advocates for Student Scientists (BRASS) 7 Trainee Symposium on Cancer Research in Texas January 30, 2015 De Angelis, Carmine – BCM Poster #2 AP-1 as a Key Mediator/Target of Endocrine Resistance in Breast Cancer De Angelis C1, Nardone A1, Giuliano M1, Fu X1, Trivedi M1,2, Osborne CK1 and Schiff R1,3 1 Lester & Sue Smith Breast Center, Dan L. Duncan Cancer Center, and Department of Medicine, Baylor College of Medicine 2 Department of Clinical Sciences and Administration, University of Houston College of Pharmacy 3 Department of Molecular and Cellular Biology, Baylor College of Medicine Corresponding author: De Angelis C, Lester & Sue Smith Breast Center, Dan L. Duncan Cancer Center, and Department of Medicine, Baylor College of Medicine, One Baylor Plaza, BCM 600, Houston, TX 77030, USA. Email: [email protected] Resistance to endocrine therapy represents a major impediment to successful treatment of patients with ER+ breast cancer (BC). Multiple adaptive and compensatory pathways in response to endocrine therapies modulate the estrogen receptor (ER) expression and/or its transcription activity leading to endocrine resistance. Targeting a single ‘escape’ pathway is not sufficient to overcome endocrine resistance in most patients and results in only modest improvements in patient outcomes. Identifying and inhibiting a central regulator on which these resistance pathways converge may represent a more effective strategy to overcome endocrine resistance. Preclinical and clinical data from our laboratory and others have indirectly suggested the activator protein-1 (AP-1) transcription factor as the potential master regulator that links multiple resistance pathways. The goal of this study is to validate and understand the mechanistic role of the AP-1 pathway in endocrine resistance and to investigate whether this pathway can provide unique biomarkers as well as novel therapeutic targets for improving treatment outcomes in BC. Parental (P) and endocrine-resistant derivatives (tamoxifen-resistant, TamR and estrogen deprivationresistant, EDR) of a panel of well-characterized and genetically diverse ER+/HER2- cell lines (MCF7L, MCF7 RN, ZR75-1, ZR75-B, BT483, CAMA-1, T47D) were used. The pINDUCER lentiviral system was used to inhibit AP-1 with an inducible dominant-negative (DN) c-Jun that lacks the transcriptional activation domain of the protein. Protein levels of the AP-1 pathway components were determined by immunoblotting. The AP-1 transcriptional activity was measured by a luciferase reporter gene assay. In vitro growth was assessed using methylene blue. Clonogenic assays were scored using the Gelcount instrument (Oxford Optronix Ltd). We found that MCF7-TamR cells had higher levels of c-Jun and c-Fos, the two main subunits of the AP-1 heterodimer, compared to the P cells. Consistently, the upregulated AP-1 in TamR cells was transcriptionally active, turning on higher AP-1 reporter luciferase expression compared to P cells. Importantly, c-Jun or c-Fos expression knockdown led to remarkable cell growth inhibition especially in the TamR cells. AP-1 blockade by -DN-c-Jun induction in three different clones of our acquired TamR derivative of MCF7 cell line resulted in a marked inhibition of cell growth and of colony formation. Increased AP-1 expression and activity was also observed in additional endocrine resistant models, including T47D-TamR and ZR75-1-EDR. Experiments testing the effects of AP-1 blockade on these preclinical models are in progress. Our preliminary results further support the potential role of AP-1 as a mediator of endocrine resistance and provides a strong rationale for my ongoing research, which will validate the role of AP-1 in endocrine resistance across genetically diverse preclinical models (Aim 1), determine the mechanistic aspect of AP1-dependent gene transcription in endocrine resistance (Aim 2), and provide new therapeutic strategies to improve outcome of endocrine therapies for breast cancer patients (Aim 3). This study was supported by CPRIT training grant (RP140102) – Baylor College of Medicine, a Breast Cancer Research Foundation grant, and Susan G. Komen for the Cure Promise Grant PG12221410. 8 Trainee Symposium on Cancer Research in Texas January 30, 2015 Kelly, Aaron – BCM Poster #3 Data Integration Across Multiple Proteomic Platforms Leads to a Novel Prognostic Signature for Osteosarcoma Patients Kelly AJ1,2,3; Flores RJ2,3; Nakka M2,3; Li Y2,3 and Man TK1,2,3 1 SCBMB Program, 2Dept. of Pediatrics, BCM, 3Texas Children’s Cancer and Hematology Centers Corresponding Author: Dr. Tsz-Kwong Man, Associate Professor, Department of Pediatrics, BCM Located at Feigin Center, 10th floor, 1102 Bates Street, Houston, TX Osteosarcoma (OS) is the most common malignant bone tumor in children and adolescents. Although only 20% of patients have detectable metastasis at diagnosis, many patients later relapse contributing to a dismal outcome. Identification of prognostic biomarkers besides metastasis detected through conventional imaging techniques will facilitate risk-stratification at diagnosis to determine the most appropriate treatment option to improve patients’ outcomes. Circulating biomarkers present an attractive opportunity for the identification of biomarkers since blood contains both tumor and host factors and only minimally invasive procedures are needed. Since most of the current studies focus on using a single platform for biomarker identification, the main goal of this study is to test if integrating biomarker candidates from multiple proteomic platforms can develop more robust prognostic models. Based on our preliminary results, biomarker candidates identified from three different proteomic platforms were investigated in this study: ELISA for specific serum proteins; Luminex beads assay for cytokines and chemokines; and high-density protein microarray (ProtoArray) for autoantibody detection with 20 candidates chosen for additional validation with a Luminex assay. The prognostic significance of the candidates was examined in a multi-institutional OS cohort that consists of 250 serum samples obtained from the Children’s Oncology Group (COG). Variables were selected based on their Log-rank p-values with respect to overall survival, and 5-fold cross validation (5-CV) was used to evaluate the prognostic significance in a multivariate Cox Proportional Hazard model after variable combination. By using the three different platforms individually, we found that patients with high SAA and low PF4 levels (ELISA, p=0.014, HR=1.68), a high serum level of CXCL10 (Luminex, p=0.0086, HR=1.22), or a high p27 autoantibody level (Luminex, p=0.050, HR=1.27) had significantly poorer prognosis. We developed a novel prognostic index by combining all of these proteins into a model and showed that the combined model had much better prognostic significance than the individual proteins (5-CV: p=7.1e-4, HR=1.39) and was further significant as a binary predictor, where the decision point was chosen by Martingale residuals (Kaplan-Meier, p=9.0e-4, HR=2.17). The 5-year survival rate of the “low-risk” group is 76% (95% CI, 67% to 86%) and the “high-risk” group is 50% (95% CI, 43% to 60%). The index remained to be significant after controlling for the known prognostic factor, i.e. initial metastasis, suggesting that it is an independent prognostic factor (index: p=0.0019, HR=2.11; Met-dx: p=1.3e-4, HR=2.22). Additionally, the survival of the high risk patients determined by the index who also had initial metastasis with 5-year overall survival of 35% (95% CI, 24% to 51%) was much lower than the rest of the patients (p=6.18e-8, HR= 3.03). Future directions include testing different multivariate selection procedures using cross-validation. With promising preliminary results, we will focus on machine learning approaches such as random forest, stepwise BIC, superPC, and LASSO regression. In addition to survival, we will evaluate the correlations with other clinical variables, such as response to chemotherapy, which could impact future treatment regimens. Altogether, our results indicate that development of prognostic models based on multiple proteomic platforms could be used to risk-stratify osteosarcoma patients, so more appropriate treatment options could be used in the future to improve their outcomes. This project was supported by CPRIT Cancer Biology Fellowship Program (RP140102). The study was supported by NIH/NICHD--5R01HD074553, CPRIT--RP101335-P02, and Baylor's Antibody-based Proteomics Shared Resource -- NCI Cancer Center Support Grant (P30CA125123). 9 Trainee Symposium on Cancer Research in Texas January 30, 2015 Kurtova, Antonina – BCM Poster #4 Blocking Cancer Stem Cells from Recurrent Wound-Induced Repopulation Confers Chemosensitivity Kurtova AV, Xiao J, Mo Q, Pazhanisamy S, Krasnow R, Lerner SP, Chen F, Roh TT, Lay E, Ho PL and Chan KS Corresponding author: Chan KS, Department of Molecular and Cellular Biology, Department of Urology, Dan Duncan Cancer Center, Baylor College of Medicine, 1 Baylor Plaza, Houston, TX, 77030, Email: [email protected]. Acquired chemoresistance remains a major clinical issue in the management of advanced solid cancers. While cytotoxic chemotherapy is effective in debulking tumor masses initially; in some patients tumors become progressively unresponsive after multiple treatment cycles. The causes of chemoresistance are complex; here we approach this problem from a new angle by studying the repopulation of residual surviving cancer cells between chemotherapy cycles and its contribution to progressive development of chemoresistance. Currently the identity of repopulating cancer cells following chemotherapy is unknown, and the underlying molecular mechanisms that initiate tumor repopulation remain poorly understood. In the present study we use bladder cancer as a model and report that quiescent cancer stem cells (CSCs) are unexpectedly recruited to proliferate and repopulate residual tumors in response to chemotherapy-induced damage. This phenomenon is similar to how normal resident tissue stem cells mobilize to wound sites for tissue repair. We further investigate whether blockade of this wound-induced CSC repopulation can provide an innovative approach to abrogate chemoresistance. Our previous findings demonstrated that cytokeratin 14 (CK14) marks the most primitive bladder cancer cells and abundance of CK14+ cancer cells in patients correlates with poor survival. Here, we followed the standard clinical chemotherapy regimen with multiple cycles and gap periods for recovery of normal tissues. While one cycle of gemcitabine and cisplatin effectively reduced tumor growth in vivo, a generalized expansion of CK14+ CSCs occurred in residual tumors during these gap periods between chemotherapy cycles. Further analysis revealed the induction of a “wound-response” gene signature in residual tumors and active recruitment of quiescent CSCs into proliferation in response to chemotherapyinduced damage. We demonstrated that prostaglandin E2 (PGE2) released by neighboring dying cancer cells could induce CSC expansion in a paracrine manner. This undesirable CSC expansion could be abrogated by a PGE2 neutralizing antibody and Celecoxib, an FDA approved COX2 inhibitor that blocks PGE2 signaling. In vivo administration of Celecoxib blocked the induction of “wound-response” gene signature and significantly attenuated progressive development of chemoresistance in xenograft tumors, including primary xenografts derived from a patient who failed chemotherapy. Importantly, analysis of matching pre- and post-chemotherapy cancer tissues in human bladder carcinomas (n=20) that are resistant to chemotherapy validated the enrichment of a “wound-response gene signature”, including PTGS2 (that encodes COX2). This confirms the generalization of PGE2/COX2 signaling in the development of clinical chemoresistance. These results revealed a new mechanism by which CSCs contribute to chemoresistance via repopulating residual tumors between chemotherapy cycles. Repopulation was initiated by dying cells that induced a “wound response”, and a proliferative response of CSCs to repair damages induced by chemotherapy. Therapeutic intervention with Celecoxib effectively blocked this process and improved chemotherapeutic response in bladder tumors, supporting further validation in other solid cancers. This project was supported by CPRIT pre-doctoral fellowship RP101499 (A.V.K.), CPRIT training grant RP140102 for SMART program (T.T.R.), AUA Research Scholar Award (P.L.H.), NCI grants CA129640, CA175397, V Scholar Award, Dan L Duncan Career Award, and Bladder Cancer Partnership (K.S.C.) 10 Trainee Symposium on Cancer Research in Texas January 30, 2015 Lin, Justin – BCM Poster #5 Mechanism of Maf1-Mediated Repression of Oncogenic Transformation Lin J1, Rohira A1 and Johnson D1 1 Department of Molecular and Cellular Biology, Baylor College of Medicine Maf1 is a repressor of RNA transcription which is conserved in eukaryotes from yeast to humans. It has been shown to regulate cellular biosynthesis genes that promote oncogenic transformation by directly repressing RNA polymerase (Pol) II and III transcription, and indirectly represses RNA Pol I transcription through TBP repression. While the mechanism of Maf1 repression of RNA Pol III (to which it binds directly) has been studied in detail, the means by which Maf1 directly represses Pol II transcription, and the mechanisms by which this repression is regulated are as yet unclear. However, we have identified a key residue, S123, which appears to be essential to the regulation of both Maf1 stability and repressive ability. We intend to use genomic and proteomic approaches to further investigate the role of this residue in regulating RNA Pol II transcription. Mass spectrometry analysis of Maf1-interacting proteins is being used to determine key components of RNA Pol II transcription machinery that are targeted by Maf1. Preliminary findings have identified peptides from the CDK8-containing Mediator complex submodule, a key component of RNA Pol II transcription initiation complexes. In parallel, we will also perform ChIPseq analysis, to obtain a better understanding of Maf1 localization across the genome. This data may then be compared to the genome-wide localization of RNA Pol II and other regulators of mRNA transcription, as well as to epigenetic modifications. To do so, we are using CRISPR to introduce an epitope tag to the endogenous locus of Maf1 for this analysis. Together, these data will enable us to piece together a better understanding of how Maf1 specifically represses RNA Pol II genes that regulate oncogenesis. This work is supported by CPRIT grant RP140102 and NIH grant 7R01CA074138-16/17. 11 Trainee Symposium on Cancer Research in Texas January 30, 2015 Logan, Monica – BCM Poster #6 Role of Dicer Haploinsufficiency in Aggressive Endometrial Cancer in Mice and Women Logan M1, Wang X1, Khatri S1, Xu J2, Liu D2, Broaddus R3, Coarfa C4 and Hawkins SM1 1 Department of Obstetrics and Gynecology, Baylor College of Medicine 2 Department of Biochemistry and Molecular Biology, Baylor College of Medicine 3 Department of Pathology, The University of Texas MD Anderson Cancer Center 4 Department of Molecular and Cellular Biology, Baylor College of Medicine Corresponding author: Hawkins SM, Department of Obstetrics and Gynecology, Baylor College of Medicine, One Baylor Plaza, BCM-611, Houston, TX 77030, Email: [email protected] Endometrial cancer, with 52,000 new cases diagnosed annually in the U.S., is the second most common cancer in women behind breast cancer. Activation of the PI3K pathway through loss or mutation of the tumor suppressor PTEN accounts for >75% of these tumors. Fortunately, most women will have excellent prognosis with a 5-year survival >80%. Unfortunately, late-stage and recurrent disease carries a 5-year survival of 16%. Decreased DICER expression is associated with both poor prognosis and recurrent endometrial cancers. DICER is the RNAse responsible for processing the precursor miRNA hairpin to generate 2 single-stranded mature miRNA forms, the complementary miRNA-5p and miRNA-3p forms. Based on our interests in DICER processing and miRNA function, our laboratory mined TCGA (The Cancer Genome Atlas) endometrial cancer datasets to discover that 16% of PTEN mutant (PTENmut) endometrial cancers also contained heterozygous mutations in DICER (DICER+/-). Translationally important, gene signatures associated with PTENmutDICER+/- tumors showed significantly worse survival compared to PTENmut only (P=0.02). To investigate the molecular mechanism of DICER haploinsufficiency in endometrial cancer, we created both in vivo (mouse) and in vitro (human endometrial cancer cell line) model systems of DICER haploinsufficiency. For our mouse model system, we generated Pten deleted Dicer haploinsufficient conditional knockout (cKO) mice through expression of Cre recombinase under the control of the progesterone receptor (Pgr). PgrCre/+ deletes floxed alleles where Pgr is expressed (i.e. uterus). Our Dicerf/f (f=flox) mice have Lox P sites flanking exon 23, which contains the second RNAse III domain of Dicer while our Pten f/f mice have Lox P sites flanking the exon 5-containing phosphatase domain. For our in vitro model, we used CRISPR/Cas9 technology to delete one allele of the second RNAse III domain of DICER in the PTEN-null human endometrial cancer cell line, Ishikawa. On a Pten deleted background, Kaplan-Meier survival analysis showed protection from loss of 2 alleles of Dicer and more aggressive disease from loss of 1 allele of Dicer (P=0.005) compared to Pten cKO. Median survival for Dicer haploinsufficient mice was 150 days, 240 days for Pten cKO, and 315 days for double cKO. By 12 weeks, invasion through the myometrium occurred in 85% of cases for Dicer haploinsufficient tumors compared to 30% for Pten cKO. In vitro, heterozygous deletion of DICER resulted in a 65% increase in cellular proliferation. In summary, our findings suggest that DICER is a haploinsufficient tumor suppressor in endometrial cancer. Our model systems will allow us to understand why loss of 1 allele of DICER leads to biologically aggressive endometrial tumors and discover novel therapeutic targets for aggressive endometrial cancers or other cancers that are DICER+/-. Finally, our model systems will serve as preclinical models for testing novel therapeutic molecules or discovery of noninvasive diagnostic markers for endometrial cancer. This project was supported, in part, by CPRIT Training Grant # RP140102. 12 Trainee Symposium on Cancer Research in Texas January 30, 2015 Lu, Hengyu – BCM Poster #7 High-throughput Engineering and Functional Annotation of Cancer Fusion Genes Lu H1,2, Hadjipanayis A3, Pantazi A3, Dogruluk T2, Neill NJ2, Kucherlapati R3, and Scott KL1,2 1 Integrative Molecular and Biomedical Sciences Program, Baylor College of Medicine 2 Department of Molecular and Human Genetics, Baylor College of Medicine 3 Division of Genetics, Department of Medicine, Brigham and Women's Hospital Corresponding author: Scott KL, Dept. of Molecular and Human Genetics, Baylor College of Medicine One Baylor Plaza, Houston, TX 77030, Email: [email protected] Next generation sequencing (NGS) technologies are rapidly being incorporated into the clinic to facilitate decisions on cancer patient care. Recognizing this, large-scale efforts by The Cancer Genome Atlas (TCGA) and others are generating a compendium of genomic aberrations found across major cancer types with the goal of identifying new therapeutic targets. The challenge now is to find ways to identify functional “driver” aberrations, as targeting driver events or their activated pathways offers the greatest hope of improving patient outcomes. Oncogenic transcript fusions resulting from chromosomal rearrangements represent an important class of such events. The successful targeting of fusion oncoproteins such as BCR-ABL and EML4-ALK with imatinib and crizotinib, respectively, provide strong rationale for comprehensive testing of oncogenic fusion genes. Unfortunately, the functional interrogation of fusion genes is complicated by the large quantity identified, inability to accurately predict those with driver activity, and significant technical roadblocks preventing fusion gene construction for biological assays. To circumvent these bottlenecks, we developed novel technologies permitting (1) highthroughput fusion gene construction using a novel multi-fragment DNA recombineering strategy with our platform of >35,000 human open reading frame gene clones, and (2) rapid lentiviral delivery of fusion genes to generalized and context-specific cell models to identify those with in vitro and in vivo driver activity and responsiveness to available therapeutics. As proof-of-concept, we used this approach to engineer known fusion oncogenes (BCR-ABL, EML4-ALK, and ETV6-NTRK3) and validated their transforming ability using our various driver screening systems, demonstrating our ability to rapidly deliver fusion genes with functional activity. In a pilot screen of fusion genes identified in pan-negative melanoma (i.e., devoid of signature mutations in BRAF and NRAS) by TCGA, we identified multiple uncharacterized BRAF and RAF1 fusion events among others that strongly activate MAPK signaling and exhibit potent transforming activity. Moreover, cells carrying the BRAF and RAF1 gene fusions revealed a marked sensitivity to RAF and MEK inhibitors, a finding that may indicate use of these agents for patients whose tumors harbor these events. We are now scaling these efforts for the comprehensive analysis of uncharacterized gene fusions, ultimately allowing functionalization of thousands of fusion events across diverse cancer types. These systems will reveal the highest priority fusion gene targets to enroll in deep mechanistic biology studies, drug discovery and development programs ultimately leading to personalized treatment strategies. This project is supported by NIH/NCI 1U01CA16839401 awarded to Scott KL, part of NCI Cancer Target Discovery and Development (CTD2) network, and Baylor College of Medicine Comprehensive Cancer Training Program (RP140102) funded by Cancer Prevention and Research Institute of Texas (CPRIT). 13 Trainee Symposium on Cancer Research in Texas January 30, 2015 Madan, Simran – BCM Poster #8 Whole Exome Sequencing of Human Osteosarcomas Madan S1,2, Bae Y 1, Chen S1, Campeau P1, Lu JT1,4, Gibbs R1,3, Liang H 5and Lee B1 1 Department of Molecular and Human Genetics, Baylor College of Medicine 2 Interdepartmental Program in Translational Biology and Molecular Medicine, Baylor College of Medicine 3 Human Genome Sequencing Center, Baylor College of Medicine 4 Department of Structural and Computational Biology and Molecular Biophysics, Baylor College of Medicine 5 The University of Texas MD Anderson Cancer Center Corresponding Author: Dr. Brendan Lee, Department of Molecular and Human Genetics, 1 Baylor Plaza, R815, Houston, TX 77030. Email: [email protected]. Osteosarcoma (OS) is the most common primary bone malignancy. OS incidence has a bimodal distribution with patients diagnosed as children or adolescents or as adults over 70 years of age. Despite intervention by surgery followed by combination chemotherapy, the 5-year survival rate is 68% while the overall survival rate is a mere 10-15% due to recurrence and metastasis. Developing targeted therapy to improve survival would require better understanding of the genetic drivers of the disease. So far, it is known that germ-line mutations in tumor suppressor genes such as RB1, p53 and RECQL4, increase the risk of osteosarcoma development. However, beyond these three pathways not much is known about other genetic drivers of tumorigenesis. In this study, we performed whole exome sequencing on paired tumor and blood samples from 6 osteosarcoma patients to identify novel somatic mutations. Mutations identified from the 6 paired samples were compared to mutations identified from our previous whole exome sequencing of different cohort of 10 tumor samples. After filtering out germ-line mutations, known dbSNPs, and variants found in normal population through the 1000 genomes project, we found 68 novel variants (SNPs and INDELs). Of the 68 variants, we were able to successfully validate 14 by Sanger sequencing including mutations in ATXN1, AKAP6, ATRX, ANKRD11, HECW1, OPCML, ESR1, KCNH8, KCNN3, CACNA1F, MED13, UNC13C and ALG13. Functional studies of these candidates are now underway to determine their roles in tumor initiation and progression. This study highlights the power of whole exome sequencing for identifying novel, somatic mutations that could be primary drivers of disease. This project is supported by a CPRIT pre-doctoral fellowship to Madan, S. 14 Trainee Symposium on Cancer Research in Texas January 30, 2015 Neill, Nicholas – BCM Poster #9 Mechanisms Underlying the Tumor Suppressive Role of PTPN12 in Triple-Negative Breast Cancer Neill NJ1, Sun T2, Westbrook TF1,2,3,4 1 Department of Molecular & Human Genetics, Baylor College of Medicine 2 Verna & Marrs McLean Department of Biochemistry & Molecular Biology 3 Department of Pediatrics, Baylor College of Medicine 4 Dan L. Duncan Cancer Center, Baylor College of Medicine Corresponding author: Neill NJ, Department of Molecular & Human Genetics, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, USA, Email: [email protected] Triple-negative breast cancer (TNBC) is a common and aggressive subtype of breast cancer that is newly diagnosed in approximately 220,000 women annually. TNBC is refractory to current targeted therapies and despite recent efforts to characterize TNBC genomes and epigenomes, a major barrier to developing TNBC therapies is the paucity in our understanding of the molecular drivers of TNBC. Identifying the signaling networks whose dysregulation drives TNBC would have enormous impact on our understanding of the disease and how we treat afflicted patients. Using a forward genetic screen, we recently identified the tyrosine phosphatase PTPN12 as a tumor supressor in TNBC (Sun, Cell 2012). Our preliminary data indicate PTPN12 is compromised in many epithelial cancers, including more than 70% of TNBCs, making it the second most frequently inactivated tumor suppressor in TNBC (behind p53). Loss of endogenous PTPN12 leads to transformation of human mammary epithelial cells (HMECs) and mammary carcinogenesis in mouse models. Moreover, restoring PTPN12 function dramatically impairs tumor progression and metastasis in TNBCs. These and other studies suggest PTPN12 functions as a suppressor of human TNBC. Our objectives are to elucidate the mechanisms by which PTPN12 suppresses human TNBC and discover new vulnerabilities of PTPN12-deficient TNBCs that can be used as therapeutic entrypoints. To define such vulnerabilities, we used an unbiased genetic approach to identify genes that are essential for the growth of PTPN12-deficient cells. By combining this genetic information with quantitative proteomics data, we have identified several cellular processes that are dysregulated in PTPN12-deficient cancers and selectively required to support the growth of these tumors. We are currently exploring these pathways as novel therapeutic entrypoints that can be exploited to treat patients with TNBC. This project was supported by CPRIT training grant RP140102. 15 Trainee Symposium on Cancer Research in Texas January 30, 2015 Satterfield, Laura – BCM Poster #10 MIR-130b Targets ArhGAP1 Increasing Cdc42 Activity and Metastatic Potential in Ewing Sarcoma Cells Satterfield LS1,2,3, Kurenbekova L1, Shuck R1, Donehower L2,3 and Yustein J1,2,3 1 Faris D. Virani Ewing Sarcoma Center 2 Interdepartmental Cell and Molecular Biology 3 Dan L. Duncan Cancer Center, Baylor College of Medicine Corresponding Author: Satterfield, LS, Department of Pediatrics, Baylor College of Medicine, 1102 Bates Avenue, Room 1070.08, Houston, TX 77030, Email: [email protected] Osteosarcoma (OS) and Ewing’s sarcoma (EWS) are the most common bone tumors in the pediatric population. While significant progress has been made with towards increasing patient outcomes, our success in eradicating these two malignances is significantly less than most other pediatric malignancies. One reason for continued treatment failure is due to high risk disease states such as the presence of metastatic lesions. Metastasis is of the utmost clinical relevance as it accounts for more than 90% of solid tumor deaths. Improvements in treatment regimens have been stagnant over the past 20-30 years most likely due to our lack of understanding their molecular pathogenesis. Overall survival rates are approximately 65-70% for localized disease and less than 20% for aggressive, disseminated states. Thus, the clinical need to understand high risk disease, such as metastasis is significant in order to develop novel therapeutics. miRNAs are highly conserved short 18-22 single stranded non-coding RNAs that have the ability to repress target translation and affect key cell signaling pathways often deregulated in cancer. Recently, miRNAs have been implicated in numerous cancers; however the functional role of miRNAs in sarcomas is limited. In order to identify novel alterations in miRNAs, we have utilized our analysis of microRNA expression from localized and metastatic tumors derived from novel genetically engineered mouse models of osteosarcoma. Among the miRNAs most significantly dysregulated, miR-130b was shown to be significantly upregulated in the metastatic lesions. Furthermore, it has been shown that miR130b is overexpressed in OS and EWS patient samples, and this event is significantly associated with poor patient outcome in the latter sarcoma. Furthermore, we have verified miR-130b overexpression in both mouse and human sarcoma cell lines and in metastatic EWS tumor samples obtained from Texas Children’s Hospital. Functional analysis shows overexpression of miR-130b increases migration and invasion in vitro. In order to identify novel targets of miR-130b contributing to metastasis, we performed gene expression analysis. We identified Cdc42GAP as a novel target of miR-130b. Cdc42GAP is a negative regulator of Cdc42 which hydrolyzes the active GTPase. ELISA analysis of Cdc42 activity reveals that Cdc42 activity is increased with overexpression of miR-130b. Further microarray and pathway analysis suggest a mechanism by which activation of Cdc42 leads to positive regulation of the MAPK cascade. Subsequent western blot analysis reveals that overexpression of miR-130b leads to the activation of SAPK/JNK and subsequent phosphorylation of c-JUN suggesting activation of AP-1 transcription factor. Taken together, these findings suggest that overexpression of miR-130b promotes sarcoma cell migration and invasion through targeting of Cdc42GAP and modulation of the Cdc42 and SAPK pathways. Funding: CPRIT Training Grant 16 Trainee Symposium on Cancer Research in Texas January 30, 2015 Scott, Aaron – BCM Poster #11 Feasibility of a Symptom Tracking Smartphone Application Scott AR1,3, Naik AD2,3, Suliburk JW1,3 and Berger DH1,3 1 Michael E. DeBakey Department of Surgery, Baylor College of Medicine 2 Department of Medicine, Baylor College of Medicine 3 Center for Innovations in Quality, Effectiveness and Safety, Michael E. DeBakey VA Medical Center Corresponding Author: Scott AR, Michael E. DeBakey Dept. of Surgery, Baylor College of Medicine One Baylor Plaza MS390, Houston, TX 77030; Email: [email protected] Limited communication and care coordination following discharge may contribute to complications for surgical oncology patients. Smartphone applications (“apps”) offer a new mechanism for communicating with patients and directing their care. It is unclear, however, whether or not patients are willing and able to use apps as part of their surgical care. To better understand patient factors which could prevent app use in a surgical setting, we performed a feasibility study on an app designed to facilitate self-care following colorectal surgery. This was a prospective mixed-methods feasibility study performed at an urban public safety net hospital. Following colorectal surgery, patients were approached for enrollment and offered a smartphone app which uses previously validated content to provide recommendations based on symptoms. Patients were asked to use the app daily for 14 days after discharge. Demographics and usability data were collected at enrollment. The System Usability Scale (SUS) was used to measure usability. The SUS was repeated at follow up and then we performed a structured interview covering domains such as ease of use, willingness to use, and utility of use. Chart and app log review identified phone calls and ER visits related to surgery. We screened 126 patients, enrolled 20 (16%), completed follow up interviews with 15 (12%), could not contact 4 (3%) and have pending follow up with 1 (1%). Reasons for non-enrollment included: lack of a suitable device (26 patients, 21%), willingness to participate (27, 21%), language barriers (13, 10%), inclusion criteria (21,17%), and a variety of other reasons (19, 15%). The 30-day unplanned ER visit rate was 42% (8/19), with an 11% (2/19) readmission rate. The app addressed 63% (5/8) of the presenting complaints in the ER, but no patients reported those complaints in the app. The app was used once or not at all by 7/19 patients (37%); the remaining 12 (63%) used it a median of 7 times (IQR 6-22). Median SUS scores were >90th percentile at first use (raw score 95, IQR 82-99) and follow up (95, 78-98). Seven of the 15 patients who were interviewed (47%) reported daily app use. Feeling ill/tired was the most common reason for less frequent use (4/8, 50%). Fourteen out of 15 patients interviewed reported being able to fill out the app themselves, 12 denied difficult or confusing questions, 13 felt the app fit into their daily routine, and none of the patients felt there were too many questions. Nine trusted the app’s recommendations, five weren’t sure, and one did not trust the app. Smartphone based interventions have the potential to improve care coordination and patient perceptions of communication in the management of colorectal cancer. Patient barriers to app use include device availability, interest, compliance when feeling ill, and incomplete or inaccurate symptom reporting. Use patterns in this study fell short of goals outlined at enrollment, suggesting the need for highly engaging apps. Further study is needed to find ways to overcome these barriers as well as methods of integrating apps into surgical oncology care pathways. This project was supported by CPRIT Training Grant RP140102. 17 Trainee Symposium on Cancer Research in Texas January 30, 2015 Sreekumar, Amulya – BCM Poster #12 Determining The Role Of Wnt and p53 Pathways in Mammary Stem Cell Fate Determination Sreekumar A and Rosen J1 1 Department of Molecular and Cellular Biology, Baylor College of Medicine The mammary gland is unique in that the most of its development occurs postnatally. Proliferative bursts in the mammary gland occur during puberty and pregnancy and are stem cell driven, allowing stem cell studies to be conducted in adults in vivo. We use mammary stem cells (MaSCs) as a model to study the balance between stemness and differentiation. In this context, we probe asymmetric cell division (ACD), the process by which a stem cell divides to produce a stem cell and a differentiated cell. ACD is often disrupted in tumours. Strikingly, the basal-like subset of breast cancers shows an enriched stem-like signature suggesting a role for ACD. These tumours show frequent Wnt hyperactivity and p53 loss-offunction mutations, two pathways that have been invoked to regulate ACD in other systems. We are, therefore, interrogating the role of Wnt and p53 pathways in maintaining ACD in normal and tumourigenic mammary gland contexts. We hypothesize that p53 and Wnt are critical regulators of ACD in MaSCs, perturbation of which skews towards symmetric cell divisions (SCD) in mammary tumours resulting in an expanded cancer stem cell (CSC) pool. To address this, we propose to study MaSCs using the following models: • • Rapidly growing pubertal terminal end buds (TEBs) containing cap stem cells Basal-like breast cancer (BLBC) model with an enriched stem cell signature Contrary to current dogma, we find that cap cells in the body layer are largely apoptotic, potentially ruling them out as candidates for bipotent stem cells. Additionally, based on known markers of ACD such as NuMA and LGN, in situ cap cells show occasional ACD. These cap cells constitute a subset of Wnt signalling positive populations as identified by lentiviral transduction of a Wnt reporter into mammary epithelial cells. Finally, global loss of p53 in this Wnt active stem cell population results in mislocalization of ACD markers such as LGN and NuMA, suggesting a role for p53 in perturbation of ACD, which could constitute an early event in tumourigenesis. In the future, we propose to expand the perturbation studies in the Wnt and p53 pathways to study the functional role of ACD in the mammary gland. We will follow this up with studies in MMTV-Wnt1 early hyperplasias as well as established BLBC models to study the contribution of ACD to tumourigenesis. Acknowledgement: This work is supported by CPRIT grant number 140102 and NCI-CA16303. 18 Trainee Symposium on Cancer Research in Texas January 30, 2015 Williams, LeTerrica – BCM Poster #13 Targeting the Tumor Vasculature with Antigen-Specific T Cells Williams L1, Kakarla S1, Phung T2, Rowley D3, Kalra M1 and Gottschalk S1 1 Center for Cell and Gene Therapy, Baylor College of Medicine 2 Department of Pathology, Texas Children's Hospital 3 Department of Molecular & Cell Biology, Baylor College of Medicine Corresponding author: LaTerrica Williams, Center for Cell and Gene Therapy, Texas Children's Hospital, Houston Methodist Hospital, Baylor College of Medicine, 1102 Bates Street, Suite 1770, Houston, Texas, 77030, Email: [email protected] T-cell immunotherapy with genetically modified T cells expressing chimeric antigen receptors (CARs) has shown promise in preclinical models as well as early clinical studies. However, patients with solid tumors often due not respond as well as patients with hematological malignancies. This lack of efficacy for solid tumors is most likely due to several factors including a) emergence of immune escape mutants, and b) inability of tumor-specific T cells to recognize and destroy the vascular bed of solid tumors, which is critical for their malignant growth. The aim of this project is now to generate CARs specific for tumor endothelial markers (TEM) 8, and evaluate their antitumor activity in preclinical animal models. To achieve this, we generated retroviral vectors encoding TEM8-specific CARs. RD114-pseudotyped retroviral vectors were used to genetically modified T cells with the respective retroviral vector. In coculture assays, T cells expressing TEM8-specific CARs recognized target cells in an antigen-dependent fashion as judged by the secretion of pro-inflammatory cytokines (IFN-γ and IL-2), and by their ability to kill antigen-positive target cells. Importantly, TEM8-specific CARs recognized human as well as murine TEM8, which will enable the in vivo testing of TEM8-specific CAR T cells in murine xenograft as well as immune competent animal models In conclusion, we have developed TEM8-specific CAR T cells, and in the future are planning to assess their anti-vasculature and anti-tumor activity in preclinical animal models by themselves or in combination with T cells that target tumor cells. If successful, this approach may improve current T-cell immunotherapies for solid tumors. This project was supported by grants CPRIT Baylor College of Medicine Comprehensive Cancer Training Program (RP140102), NCI-P01 (2 P01 CA094237), and Cookies for Kid’s Cancer 19 Trainee Symposium on Cancer Research in Texas January 30, 2015 Zhu, Wenyi – BCM Poster #14 Daam2 Promotes Tumor Proliferation and Migration in Glioblastoma Zhu W 1,2, Lee HK1, Loturco J4, Mohila C5 and Deneen B3 1 Center for Cell and Gene Therapy, Baylor College of Medicine 2 Integrative Molecular and Biomedical Sciences Graduate Program, Baylor College of Medicine 3 Department of Neuroscience, Baylor College of Medicine 4 University of Connecticut 5 Texas Children's Hospital Corresponding author: Deneen, Benjamin, Department of Neuroscience, Baylor College of Medicine, Houston, TX 77030 Glioblastoma multiforme (GBM) is one of the most common and aggressively malignant high-grade gliomas in humans. GBM tumor as are comprised of astrocytes and oligodendrocytes. Genes involved in the development of these glial cells have been liked to GBM tumorigenesis. Considering tumorigenesis is currently viewed as a convergence of genetic mutation and developmental context, we want to understand how the mechanisms which govern glial cell fate decisions contribute to GBM formation. The Wnt signaling pathway has well defined functions in progenitors self-renewal during development. This function is dysregulated in several cancers, yet surprisingly poorly defined in GBM. Given our finding that Daam2 plays a key role in Wnt signaling in the developing and CNS, we hypothesize that Daam2 is a key component of the Wnt signaling pathway which is dis-regulated in GBM. Indeed, we found that Daam2 is highly expressed in GBM. Subsequent analysis of Daam2 function in GBM revealed that it promotes tumor proliferation and migration. The way we approached this is to take advantage of In Utero Electroporation to generate GBM in vivo and then manipulate Daam2 expression. By using this way we was able to generate GBM on WT mice within three weeks. In the meantime, we also perform In Utero Electroporation on Daam2 knockout mice to analyze its loss of function phenomenon. Currently, I am working to precisely define its molecular mechanism. We surprisingly find that Daam2 is not only involved in Wnt signaling pathway in GBM, but also interactions with PI3K. These studies represent the first characterization of Daam2/Wnt signaling/ PI3K function in GBM. This project is supported by CPRIT training grant and Sontag foundation. 20 Trainee Symposium on Cancer Research in Texas January 30, 2015 Fan, Yu – MDACC (Program Based at UH) Poster #15 MuSE: Somatic Evolution Estimation for Mutation Calling in Sequencing Data of Matched TumorNormal Samples Fan Y1, Xi L2, Wheeler DA2 and Wang WY1 1 Department of Bioinformatics and Computational Biology, The University of Texas MD Anderson Cancer Center 2 Human Genome Sequencing Center, Baylor College of Medicine Corresponding author: Wang WY, Department of Bioinformatics and Computational Biology, The University of Texas MD Anderson Cancer Center, 1400 Pressler St., Floor 4, FCT4.6000, Houston, TX 77030, Email: [email protected] Cancer is a disease process involving somatic mutational events. The use of next-generation sequencing on matched tumor-normal sample pairs is critical for discovery of somatic variation. However, accurate detection of somatic mutations remains a challenge. Here, we present a Bayesian phylogenetic method, MuSE, for describing the evolution from the reference allele to the tumor and the normal allelic composition at a single nucleotide position. Our proposed method incorporates the probability of sequencing errors and computes the unknown allele frequencies, multiple alternative alleles and the rates of nucleotide transition/transversion. By calculating the evolutionary distances and computing the dynamic cutoffs for d, we classify variants into: somatic, germ-line and reversal to the homozygous reference. We include filters that consider the sequence context surrounding the point mutations to further reduce the false positive rate. We validated the performance of MuSE using both a virtual-tumor benchmarking approach, and applied it to analyzing 66 pairs of chromophobe renal cell carcinoma (KICH), 91 pairs of adrenocortical carcinoma (ACC) and 197 pairs of liver hepatocellular carcinoma (LIHC) exome sequencing data that are part of the Cancer Genome Atlas (TCGA) project. In particular, we thank Dr. John N. Weinstein from MDACC and Dr. Richard A. Gibbs from the BCM HGSC. This research was funded by a training fellowship from the Keck Center of the Gulf Coast Consortia, on the Computational Cancer Biology Training Program, Cancer Prevention and Research Institute of Texas (CPRIT) RP140113, PI - Rathindra Bose. And it was also supported by the NCI through grant U24 CA143883 02S2 (to J.N.W.) and the Integrative Pipeline for Analysis & Translational Application of TCGA Data (GDAC), NIH/NCI grant 5U24CA143883-04 (to J.N.W.) Kirk, Brian – BCM (Program Based at UH) Poster #16 21 Trainee Symposium on Cancer Research in Texas January 30, 2015 Detection of Long Non-coding RNA Structural Motifs Involved in PRC2 Functions Kirk B1 , Menlove K 2, Jianpeng Ma 1,2,3* and Wang Q1* 1 Verna and Marrs McLean Department of Biochemistry and Molecular Biology, Baylor College of Medicine 2 Graduate Program of Structural and Computational Biology and Molecular Biophysics, Baylor College of Medicine 3 Department of Bioengineering, Rice University *Corresponding authors: Qinghua Wang, Rm 366A, One Baylor Plaza, Houston, TX 77030, Email: [email protected] Long noncoding RNAs (lncRNAs) function in many cellular processes such as the recruitment of chromatin-modifying machinery. Experimental characterization of structural RNA elements is costly and time consuming. Computational prediction based on detecting conserved structural motifs may provide precious new insight into the structure and function of lncRNA molecules. Here we have developed a protocol to predict structural motifs by generating ensembles of predicted RNA secondary structures with pseudoknots, converting these structures into networks or dual-graphs, and locating statistically overrepresented sub-graphs in these networks. Using a cross-linking immunoprecipitation dataset of Polycomb Repressive Complex 2 (PRC2) binding lincRNA sequences from Khalil, Guttman et al. (2009), we have identified motifs that are statistically overrepresented in that dataset versus a set of randomly selected lincRNA sequences. Nine of the top-forty motifs presented are found in one or more human lncRNAs (HOTAIR, RepA of Xist, and H19) known to bind PRC2 components. KM and BDK were partially supported by the NLM Training Program in Biomedical Informatics through the Keck Center of the Gulf Coast Consortia (NLM Grant No: T15LM007093). BDK is currently supported by a postdoctoral fellowship from the Keck Center Computational Cancer Biology Training Program of the Gulf Coast Consortia (CPRIT Grant No. RP101489). 22 Trainee Symposium on Cancer Research in Texas January 30, 2015 Lee, Jaehyuk – MDACC (Program Based at UH) Poster #17 Comparison Response of siRNA Treatment by Metabolic Changes in Ovarian Cancer by Nuclear Magnetic Resonance Lee J1, Millward NZ1, Rupaimoole R2, Sood AK2 and Bhattacharya PK1 1 Department of Cancer Systems Imaging, The University of Texas MD Anderson Cancer Center2Department of Gynecologic Oncology and Reproductive Medicine, The University of Texas MD Anderson Cancer Center Corresponding author: Pratip Bhattacharya, Department of Cancer Systems Imaging, The University of Texas MD Anderson Cancer Center - Unit 1907, 1881 East Road, Houston, TX 77054, U.S.A, E-mail: [email protected] Altered energetics distinguishes transformed tumor tissues from normal tissues. More than two thirds of all diagnosed cases of ovarian cancer occur after the disease has progressed to stage III of IV. Due to this problem, the survival rates of women diagnosed with some type of ovarian carcinoma—as over 90% of all ovarian cancers are—has remained virtually unchanged for the last 30 years since the introduction of platinum-based treatments. This research is focused on accessing responses of new treatment using siRNAs for ovarian cancer based on metabolic changes using high resolution nuclear magnetic resonance (NMR) which is a powerful metabolomics tool to determine the metabolic profiles of cancerous cells and tumor tissues. In this research, proton NMR spectroscopy revealed the metabolic profile of the cancer cells, media and tumor tissue in one experiment and be further confirmed using 2D NMR spectroscopy techniques and LCMS-MS. All organic metabolites (for example lactate, acetate, glutamate) with chemically unique hydrogens will have particular resonance(s) in 1D proton NMR spectroscopy. Using these spectral peaks, the extracted metabolites from the biological samples were determined. In addition, employing external standards the concentration of the identified metabolites were determined. In all samples discussed below, 1D 1H proton spectroscopy was performed with water suppression on a 500 MHz Bruker Avance III HD NMR equipped with a Prodigy BBO cyroprobe. The cyroprobe increases the sensitivity of the measurement two to three fold. We compared expression of measured metabolites from ovarian cancer cells and tissues after two different siRNA treatments and compared that from untreated tissues as well. Significant differences were found in expression of several metabolites such as lactate, alanine, acetate, glutamate and creatine due to the two different treatments which may point to the metabolic pathway these drugs target and their efficacy. NMR based metabolomics provides an excellent biomolecular method to measure not only the efficacy of therapy but also comparing between treatments to measure the drug response in ovarian cancer. This method can help to investigate the effective metabolic pathway and target metabolites for better understanding and future drug discovery. This research was funded by a training fellowship from the Keck Center of the Gulf Coast Consortia, on the Computational Cancer Biology Training Program, Cancer Prevention and Research Institute of Texas (CPRIT) RP140113, PI-Pratip Bhattacharya. 23 Trainee Symposium on Cancer Research in Texas January 30, 2015 Li, Xu – MDACC (Program Based at UH) Poster #18 A Standardized Scoring System for Affinity Purification/Mass Spectrometry Data Xu Li1, Benjamin White2, Wenqi Wang1, Rudy Guerra2* and Junjie Chen1* 1 Department of Experimental Radiation Oncology, The University of Texas MD Anderson Cancer Center 2 Department of Statistics, Rice University Affinity purification coupled with mass-spectrometry (AP/MS) has revolutionized the field of proteomics. With the advances in MS technology and bioinformatics, we are gradually overcoming the obstacles of generating raw spectra and mapping them to correct proteins. We are now increasingly facing the complication of our own success: how to make sense of a list of hundreds, if not thousands, of putative associated proteins? How to utilize them effectively in directing in-depth functional studies? Current data filtration methods still produce a considerable amount of false-positive and false-negative identifications. Moreover, they are inadequate in assigning meaningful confidence scores to all the interactions, especially these relatively weak but very specific interactions which could be functional important. Here we present a newly developed Minkowski distance-based probabilistic scoring system, which has greatly improved the efficiency of label-free AP/MS data analysis. In this system, standardized probability scores are assigned to interactions in a data-driven manner. In addition, two models have been proposed to integrate different types and scales of datasets: (1) a knowledge-driven model using available AP/MS datasets as controls, and (2) a knowledge-free model using customized negative controls. Compared to other algorithms, this new data analysis system is more accurate in picking up true positives and eliminating false positives in AP/MS data sets. In addition, using this system we have also achieved high inter-laboratory data reproducibility. This new algorithm is able to improve the efficiency for analyzing AP/MS datasets, which will help us greatly in directing in-depth functional studies and to establish a unified human protein interactome in the near future. This research was funded by a training fellowship from the Keck Center of the Gulf Coast Consortia, on the Computational Cancer Biology Training Program, Cancer Prevention and Research Institute of Texas (CPRIT) RP140113. 24 Trainee Symposium on Cancer Research in Texas January 30, 2015 Li, Xueling – MDACC (Program Based at UH) Poster #19 The Genome-wide Functional Modulatory Network in Epithelial-Mesenchymal Transition Li X1,3, Tian B2,3,4, Kudlicki AS1,3, 4* and Brasier AR2,3,4 1 Department of Biochemistry and Molecular Biology 2 Department of Internal Medicine 3 Institute for Translational Sciences 4 Sealy Center for Molecular Medicine, University of Texas Medical Branch Corresponding authors: Kudlicki AS, Dept. of Biochemistry and Molecular Biology, UTMB, Email: [email protected], Brasier AR, Department of Internal Medicine, UTMB, Email: [email protected] Epithelial-mesenchymal transition (EMT) is an epigenetic and transcriptional reprogramming event implicated in cancer progression. Identifying the involved transcription factors (TFs) and predicting how the TF activities are modulated will be instrumental to reveal the molecular mechanism of EMT. TFs may be activated or repressed by modulator (M) proteins that bind directly to the TFs, proteins involved in the upstream pathways or non-coding RNAs through different molecular mechanisms, resulting in target gene (TG)-specific expression changes in six distinct modes (Babur et al, 2010; Li et al, 2014): enhancement, attenuation or inversion of the inhibition or activation activity of transcription factors. This modulatory network can be inferred from gene expression data representing diverse cellular states (Babur et al, 2010; Li et al, 2014). By considering the triplets (M, TF, TG) rather than the pairs (TF, TG), we expand the regulatory network into a modulatory network that carries more biologically relevant information. Here we predicted the modulatory network to include all enriched transcription factors in epithelialmesenchymal transition (EMT). Differentially expressed genes (DEGs) from immortalized cells EMT of small airway epithelial were identified by RNA-Seq and were used as the candidates of modulators and target genes. Enriched human transcription factors were identified based on hypergeometric test by comparing the DNA-binding and DEGs and used as TF candidates. Involved genes # TFs # Modulators # Target Genes # up DEGs in EMT # down DEGs in EMT Number 50 457 2087 833 1254 We predicted 52992 significant M-TF-TG interactions involving 50 transcription factors (See the table for gene composites in the network). We characterized the modulatory network of EMT by clustering the modulators and by comparing their distribution on the TGs and TFs based on the parameters of the main effect of the modulators. By enrichment analysis, we also characterized the prevalent action modes (enhancement, attenuation, and inversion) of the specific groups of modulators, such as transcription factors, chromatin modifiers, non-coding RNA, RNA-binding proteins, proteins involved in alternative splicing, kinases, and cytoskeleton associated proteins. The functional enrichment of different action modes was further characterized. The enriched action modes are consistent with their known molecular functions and modulation level. Our results reveal the modulatory network involved in EMT and will provide a valuable repertoire of hypotheses for further studies of the specific processes and pathways involved in EMT. Babur O, Demir E, Gonen M, Sander C, Dogrusoz U (2010) Discovering modulators of gene expression. Nucleic Acids Res 38: 5648-5656. Li X, Zhao Y, Tian B, Jamaluddin M, Mitra A, Yang J, Rowicka M, Brasier AR, Kudlicki A (2014) Modulation of gene expression regulated by the transcription factor NF-kappaB/RelA. J Biol Chem 289: 11927-11944. This research was funded by a training fellowship from the Keck Center of the Gulf Coast Consortia, on the Computational Cancer Biology Training Program, Cancer Prevention and Research Institute of Texas (CPRIT) RP140113, PI - Rathindra Bose. 25 Trainee Symposium on Cancer Research in Texas January 30, 2015 Long, Bryon – RU (Program Based at UH) Poster #20 An Image Processing Algorithm for Counting Neurons and Measuring Neurite Lengths Long BL1 and Qutub AA1 1 Department of Bioengineering, Rice University Corresponding author: Long B, Department of Bioengineering, Rice University, 6500 Main Street, Suite 135, Houston, Texas 77030, Email: [email protected] The use of neural stem cells (NSCs) is a promising avenue for repairing damage to the brain and nervous system. Recent research has revealed the role of a variety of promising methods to control NSC differentiation – an important part of developing neural stem cell therapies. In order to assist in the evaluation of NSC assays, we have developed image processing software that counts the number of neurons that have neurites with lengths 3 times the diameter of their neuron bodies (as parameterized). Our algorithm takes a pair of DAPI and Tuj grayscale images and starts by performing two thresholding operations in order to identify neuron bodies and neurites. Neurite masks are then skeletonized into single pixel chains that allow neurite lengths to be calculated. A consequence of the double thresholding is the incorporation of neuron body perimeters into the neurite masks; this is corrected by detecting closed curves that contain neuron bodies. The resulting neurite skeleton is then treated as a graph where: 1) vertices are the end points, branch points, and points of intersection of the skeleton and 2) edges are the skeletal sections connecting vertices. From this graph, individual neurites are identified and associated with neuron body clusters. The algorithm was implemented in Matlab using the Image Processing Toolbox. It successfully identifies and counts neurons with long neurites as parameterized. Our work allows NSC assays to be evaluated in a uniform and reproducible manner and facilitates data collection from high-throughput experiments. This research was funded by a training fellowship from the Keck Center of the Gulf Coast Consortia, on the Computational Cancer Biology Training Program, Cancer Prevention and Research Institute of Texas (CPRIT) RP140113, PI - Rathindra Bose. 26 Trainee Symposium on Cancer Research in Texas January 30, 2015 Lu, Mingyang – RU (Program Based at UH) Poster #21 Modeling EMT Genetic Regulatory Circuits in Metastatic Cancer Lu M1*, Jolly MK1*, Huang B1, Hanash SM2, Levine H1, Onuchic J1 and Ben-Jacob E1 3 1 Center for Theoretical Biological Physics, Rice University 2 Red and Charline McCombs Institute for the Early Detection and Treatment of Cancer, University of Texas MD Anderson Cancer Center 3 School of Physics and Astronomy and The Sagol School of Neuroscience, Tel-Aviv University Cancer metastasis is responsible for more than 90% of cancer deaths. Yet, understanding epithelial mesenchymal transition (EMT) during cancer metastasis remains a major challenge in cancer biology. It is now established that cells use genetic regulatory circuits to make functional decisions of whether to undergo EMT or not. In this study, we constructed a theoretical model of the circuitry involved in the EMT. The core regulatory unit for the decision consists of two highly interconnected chimeric modules the miR-34/SNAIL and the miR-200/ZEB mutual-inhibition feedback circuits. We developed a theoretical framework for modeling microRNA-based circuit and applied it to study the chimeric modules. We showed that the miR-34/SNAIL module functions as a noise-buffering signal integrator, and the miR200/ZEB module functions as a three-way switch, allowing not only for the epithelial and mesenchymal phenotypes, but also for a hybrid phenotype with mixed epithelial and mesenchymal characteristics. We further studied EMT in a multi-cell environment by coupling the EMT circuit to cancer-related signaling pathways and by including cell-cell communications. Our model explains recent data on the observation of unusual intermediates with specialized cell behavior, such as collective migration, and phonotypical heterogeneity of the EMT observed in various lung cancer cell lines. 27 Trainee Symposium on Cancer Research in Texas January 30, 2015 Prakash Srivastava, Priyanka – UTHSC (Program Based at UH) Poster #22 Preferred Orientation of Oncogenic K-Ras in Membrane Prakash P and Gorfe AA Integrative Biology and Pharmacology, University of Texas Health Science Center at Houston Corresponding author: Alemayehu A Gorfe, Integrative Biology and Pharmacology, University of Texas Health Science Center at Houston, 6431 Fannin St., MSB 4.108, Houston, TX 77030, Email: alemayehu.g.abebe.uth.tmc.edu Ras is a lipid-modified GTPase that acts as a molecular switch by cycling between active and inactive conformational states and is involved in a plethora of cell signaling pathways. Somatic mutations in Ras are associated with a variety of cancers and are found in ~15% of human tumors such as pancreatic, colorectal, lung cancer to name a few. Of the three major human Ras isoforms H-, N- and K-Ras, cancers associated with mutant K-Ras are the most lethal. Ras attaches itself to the inner leaflet of the plasma membrane in order to perform its biological function. Characterization and targeting of hot-spot residues involved in the interaction of Ras with the plasma membrane could be of therapeutic relevance and may yield isoform-specific drugs. To this end, we performed microsecond-level atomistic molecular dynamics (MD) simulations of full-length, oncogenic K-Ras bound to a heterogeneous membrane. MD-derived populations from these simulations reveal K-Ras residues interacting directly with the membrane, predominantly in two different modes. Probe-based MD simulations on these two predominant orientations of membrane-bound K-Ras are underway to further identify putative ligand-binding sites. We will discuss these results in terms of their potential usefulness for anti-cancer drug design. This research was funded by a training fellowship from the Keck Center of the Gulf Coast Consortia on the Computational Cancer Biology Training Program, Cancer Prevention and Research Institute of Texas (CPRIT) RP140113, PI – Rathindra Bose. PP is grateful to the Texas Advanced Computing Center (TACC) for their computational resources. 28 Trainee Symposium on Cancer Research in Texas January 30, 2015 Sehgal, Vasudha – MDACC (Program Based at UH) Poster #23 Integrating Functional Hypothesis Generation for Identifying Pan Cancer miRNA Biomarkers Sehgal V1, Seviour EG1, Moss TJ1, Azencott R2, Mills GB1, and Ram PT1 1 Department of Systems Biology, UT MD Anderson Cancer Center 2 Department of Mathematics, University of Houston Corresponding author: Ram PT, Department of Systems Biology, UT MD Anderson Cancer Center 7435 Fannin Street, Houston, TX 77054, Email: [email protected] The regulation of gene expression is vital for the maintenance of homeostasis within cells, and the disruption of these regulatory mechanisms frequently results in diseases such as cancer. MiRNAs play a crucial role in the maintenance of homeostasis via regulating the expression of their target genes. As such, the dysregulation of miRNA expression has been frequently linked to cancer. However, the mechanisms by which the majority of miRNA affect cancer progression, and thereby patient survival, are currently unclear. A wide array of bioinformatical tools have been used to generate hypotheses for experimental verification. Here, we propose a workflow that robustly integrates many existing bioinformatical techniques with functional networks analysis tools in order to generate hypotheses which have relevance for many different cancer types. These ideas can be applied to multiple types of omics data, including miRNA, mRNA and protein expression data. The workflow identifies markers which are associated with improved or worsened survival in cancer patients. Robustness of the survival analysis is ensured by a strengthened p-value computed through intensive random resampling. In addition to the discovery of potential miRNA prognostic markers, we characterize them by chromosome location and genomic stability, and construct the associated miRNAmRNA functional networks. Among the functional pathways involved in cancer progression, we can thus identify the pathways associated with our selected. Our approach demonstrates how biomarkers identified by survival analysis techniques can be integrated within a functional network analysis framework to identify good biomarkers candidates in the development of therapeutic strategies for diverse cancer types. This research was funded by a training fellowship from the Keck Center of the Gulf Coast Consortia, on the Computational Cancer Biology Training Program, Cancer Prevention and Research Institute of Texas (CPRIT) RP140113, PI - Rathindra Bose. 29 Trainee Symposium on Cancer Research in Texas January 30, 2015 Tripathi, Swarnendu – UH Poster #24 Correlation Between Oncogenic Mutations, Signaling Pathways and Efficacy of Platinum Based Drugs Against Colon Cancers Tripathi S1, Belkacemi L2, Cheung MS1 and Bose RN2 1 Department of Physics, University of Houston 2 Department of Biology and Biochemistry, University of Houston Corresponding author: Bose RN, Dept. of Biology and Biochemistry, University of Houston, 369 Science & Research Building 2, Houston, TX 77204-5001, Email: [email protected] The purpose of this study is to correlate genetic mutations, amino-acid variants, signaling pathways with platiunum based drug activity to shed light on personalized treatment of cancer. Platinum anticancer drugs such as cisplatin, carboplatin, and oxalipaltin are widely used to treat a variety of cancers. Although DNA is the molecular target for these platinum therapies, their efficacies, toxicities, and resistance mechanisms vary widely among various categories and sub-categories of cancers. To comprehend this great variability, an integrated analysis was performed to determine the impact of somatic mutations on protein functions, signaling pathways, and drug activity (sensitivity or resistance) among the US National Cancer Institute (NCI) 60 human tumor cell lines based on Z-scores to predict a priori treatment outcome using CellMiner (http://discover.nci.nih.gov/cellminer). Specifically, somatic mutations of significantly mutated genes from the cancer genome atlas (TCGA) were analyzed along with the driver genes (oncogenes and tumor suppressor genes) from catalogue of somatic mutations in cancer (COSMIC) for total 188 genes that belong to more than 20 different signaling pathways. The functional impact of individual amino-acid variant for each gene and its correlation with the activity of carboplatin, cisplatin and oxaliplatin for each cancer cell line were explored. Particular attention was given to colon cancer for which nearly 40% of tumors are known to have mutated KRAS (Kirsten rat sarcoma viral oncogene homolog) gene. Resulting analysis revealed that colon cancer cell lines with KRAS mutations for codon 12 (G12Vmutant from SW620 cell line) and codon 13 (G13D-mutant from HCT116 and HCT15 cell lines) correlated with the sensitivity to oxaliplatin. Conversely, oxaliplatin resistant HCC2998 colon cancer cell line did not show any correlation with mutated KRAS for codon 146 (A146T-mutant). Notably, all the colon cancer cell lines were resistant to both carboplatin and cisplatin with no correlation to the KRAS mutants. Based on our integrated analysis we further predicted gene networks related to oxaliplatin activity for colon cancer. The network includes the epidermal growth factor (EGFR) signaling pathway that involves PIK3CA, PIK3CG and MTOR from the phosphoinositide 3-kinase (PI3K)/mammalian target of rapamycin (mTOR) (PIK3/mTOR) pathway, and JAK3 from Janus kinase-signal transducer and activator of transcription (JAK-STAT) pathway in addition to KRAS from Mitogen-activated protein kinases (MAPK) pathway. We conclude that instead of considering all mutations of a given gene in the same way to assess their clinical significance, it may be beneficial to categorize them into different classes based on their functional impact and efficacies towards the anti-cancer drugs for personalized treatment. Similar analytical approach is being extended to non-small cell lung and ovarian cancers where platinum therapies are widely used. This research was funded by a training fellowship from the Keck Center of the Gulf Coast Consortia, on the Computational Cancer Biology Training Program, Cancer Prevention and Research Institute of Texas (CPRIT) RP140113, PI - Rathindra Bose. 30 Trainee Symposium on Cancer Research in Texas January 30, 2015 Wu, Sherry Yen-Yao – MDACC (Program Based at UH) Poster #25 Evoking Potent RNAi Response Using Novel 2’-OMe-Phosphorodithioated Modified siRNAs Wu SY1,*, Yang X2,*, Gharpure K1, Hatakeyama H1, Egli M3, McGuire M1, Rupaimoole R1, Miyake T1, Taylor M1, Pradeep S1, Nagaraja AS1, Sierant M4, Rodriguez-Aguayo C5,6, Lopez-Berestein G5,6, Ram P7, Nawrot B4, and Sood AK1,5,8 1 Department of Gynecologic Oncology, The University of Texas MD Anderson Cancer Center 2 AM Biotechnologies LLC 3 Department of Biochemistry, Vanderbilt University, School of Medicine 4 Department of Bioorganic Chemistry, Centre of Molecular and Macromolecular Studies, Polish Academy of Sciences, Sienkiewicza 5 Center for RNA Interference and Non-Coding RNA, MDACC 6 Department of Experimental Therapeutics, MDACC 7 Department of Systems Biology, MDACC 8 Department of Cancer Biology, MDACC *These authors contributed equally to this work. Corresponding author: Sood AK, Department of Gynecologic Oncology, The University of Texas M.D. Anderson Cancer Center Unit 1362, PO Box 301439, Houston, TX, 77230, Email: [email protected] Objectives Improving small interfering RNA (siRNA) efficacy in target cell populations remains a critical challenge to bringing siRNA therapy into the clinic. There is currently an unmet need to develop a reliable strategy to globally enhance siRNA stability and potency. Methods Here, we report a novel chemical modification, consisting of phosphorodithioate (PS2) and 2’-O-Methyl (2’-OMe) MePS2 on a single nucleotide, that significantly enhances potency and resistance to nuclease degradation for a variety of siRNA sequences. Results We show a 3.5-fold improvement in gene silencing in tumors following systemic delivery of MePS2-modified siRNAs using DOPC nanoliposomes compared to unmodified counterparts. We found that this enhanced potency stems from an unforeseen increase in loading of siRNAs to the RNA-induced silencing complex (RISC), likely due to the unique interaction between 2’OMe and PS2 moieties. We subsequently demonstrate the therapeutic utility of MePS2 siRNAs in orthotopic mouse models of chemoresistant ovarian cancer, focusing on targeting GRAM domain containing 1B (GRAMD1B), a protein whose role in taxane resistance is first reported here. Efficient silencing of GRAMD1B (>80%) was achieved in tumors following systemic delivery of these MePS2modified siRNAs and a synergistic anti-tumor effect was observed when combined with paclitaxel treatment. Conclusions Given that limited success has been achieved thus far toward broadly enhancing siRNA potency with chemically modified siRNAs, our finding represents an important step forward in bringing siRNA therapeutics into the clinic. This research was funded by a training fellowship from the Keck Center of the Gulf Coast Consortia, on the Computational Cancer Biology Training Program, Cancer Prevention and Research Institute of Texas (CPRIT) RP101489, TRIUMPH (Translational Research in Multi-Disciplinary Program) Postdoctoral Fellowship (CPRIT, RP101502), Foundation for Women’s Cancer, and Ovarian Cancer Research Fund, Inc. 31 Trainee Symposium on Cancer Research in Texas January 30, 2015 Alexander, Angela – MDACC Poster #26 Cyclin E/CDK2 is a Novel Therapeutic Target in Inflammatory Breast Cancer Alexander A1*, Karakas C1, Chen X1, Carey JPW1, Gong Y2, Alvarez R3, Ueno NT3 and Keyomarsi K1 1 Department of Experimental Radiation Oncology, University of Texas MD Anderson Cancer Center 2 Department of Pathology, University of Texas MD Anderson Cancer Center 3 Department of Breast Medical Oncology, University of Texas MD Anderson Cancer Corresponding author: University of Texas MD Anderson Cancer Center, Department of Experimental Radiation Oncology, 1515 Holcombe Boulevard, Unit 066, Houston, TX 77030 Inflammatory breast cancer (IBC) is a rare, highly aggressive form of breast cancer, accounting for 2-5% of all breast cancers in the US and 10% of breast cancer deaths. Unfortunately, until recently this disease has not been studied extensively at the molecular/cellular level despite the compelling unmet need to develop more effective therapies. In spite of receiving aggressive multidisciplinary therapy (neoadjuvant chemotherapy, modified-radical-mastectomy and radiation), only about 40% of IBC patients survive 5 years. The goal of this project is to identify new targets that can be exploited pharmacologically either in combination with current therapies or other novel agents. Previous work has identified tumor-specific low molecular weight forms of cyclin E, termed LMW-E that are preferentially found in triple-negative-breast-cancers (TNBC) and serve as a poor prognostic biomarker. As a result of cleavage, LMW-E preferentially accumulates in the cytoplasm where it exerts its pleiotropic oncogenic functions including deregulating the cell cycle. To establish the clinical relevance of targeting the cyclin E/CDK2 axis in IBC, a pilot study was performed to determine the incidence of LMW-E. We performed IHC on 11 post-chemotherapy surgical specimens from MDACC. All tumors expressed cyclin E, 8 of which had predominantly cytoplasmic staining, indicative of LMW-E expression. These results have now been validated in 2 TMAs containing over 100 additional IBC cases and 49 core biopsies, representing chemo-naïve tissue. Results revealed that 80% of samples had cytoplasmic cyclin E staining and the remainder had nuclear staining. Therefore we conclude that cyclin E is a valid target in chemoresistant IBC, which led us to perform in vitro experiments to determine sensitivity to CDK2 inhibitors and design a combination treatment strategy. Using a panel of breast cancer cell lines including IBC cells, we examined the sensitivity to Dinaciclib using both short-term and long-term assays. The long-term high-throughput survival assay (HTSA) we developed, was used to examine drug combinations for potential synergies using isobologram analysis. We have demonstrated that treatment with Dinaciclib followed by anthracyclines (ie epirubicin) results in synergistic cytotoxicity, whereas the reverse sequence does not. Interestingly, this synergy is not observed when utilizing the microtubule-targeting drug paclitaxel in either sequence. A structurally distinct CDK inhibitor, Meriolin 5 also induced susceptibility to epirubicin. Using FACS analysis, we demonstrated that similar to non-IBC cell lines, CDK inhibitors induce a G2/M arrest, which is further increased by sequential treatment with epirubicin. When analyzing TNBC cells, basal subtypes had a pronounced cell death response (sub-G1) that continued even after drug removal, whereas mesenchymal and AR+ TNBC cell lines re-entered the cell cycle. Furthermore, combination sequential treatment leads to cleavage of caspase 3 and annexin V positivity. Since Dinaciclib induced sensitization to DNA damaging chemotherapies, we next asked whether Dinaciclib could potentiate radiation. We adapted the HTSA to evaluate this sequence, and our results showed significant radiosensitization of both IBC cell lines. Taken together these studies provide solid proof of concept that targeting CDK2 is a promising strategy in treating IBC tumors that express cyclin E in various settings. Beyond linking a novel kinase inhibitor to a novel and valid target in IBC, the HTSA employed has demonstrated that clear sequence-specificity exists for certain pairs of drugs, implying that future clinical studies should be designed taking into account the mechanism of action of drugs that target the cell cycle. Funding sources: CPRIT-TRIUMPH fellowship (to AA), Department of Defense Breast Cancer Research Program Postdoctoral Fellowship (to AA), NIH RO1 grants (to KK), State of Texas “Rare and Aggressive Breast Cancer Research approrpriation to the Morgan Welch Inflammatory Breast Cancer Clinic. 32 Trainee Symposium on Cancer Research in Texas January 30, 2015 Atkinson, Erin – MDACC Poster #27 NPSD4: A New Player in the DNA Damage Response Atkinson E and Wang B UT Health GSBS and Department of Genetics, UT MD Anderson Cancer Center Corresponding author: Atkinson E, Dept. of Genetics, UTHealth GSBS, UT MD Anderson Cancer Center, Houston, Texas, Email: [email protected] DNA damage triggers a series of signaling cascades, called the DNA damage response (DDR). The DDR often involves the recruitment of proteins to DNA damage sites for repair of DNA lesions and regulation of cell cycle checkpoints, transcription and other cellular mechanisms necessary for defense against DNA damage. Improper repair of damaged DNA or compromised DDR signaling often leads to genomic instability, which is a hallmark of cancer. During DDR signaling, post-translational modifications play an important role in the recruitment and regulation of repair factors. For example, ionizing radiation (IR) can induce double strand breaks (DSBs), the most cytotoxic of DNA lesions. In response to DSBs, the ATM kinase initially transduces a signaling cascade by phosphorylating many DDR proteins. The ATM pathway is vital for DNA damage repair, with loss of function mutations in ATM resulting in sensitivity to radiation and increased incidence of cancer. Another posttranslational modification, SUMO (Small Ubiquitin-like Modifier) has also been shown to be important in DSB repair. Many repair proteins are SUMOylated and mutation of the SUMO E2 enzyme, UBC9, leads to defective DSB repair. However, there is still much that is unknown about the functional role of SUMOylation in DDR. Through proteomic and mass spectrometry analyses of proteins involved in SUMOylation signaling induced by DNA damage, we have identified a novel protein that is recruited to sites of DNA damage, that we have named New Player in SUMO dependent DNA damage repair 4 (NPSD4). We identified a potential ATM phosphorylation site and two SUMO interacting motifs (SIMs) in NPSD4. We have shown that NPSD4 is recruited to DNA damage sites, specifically DSBs. We have also determined that the SIMs are required for interaction of NPSD4 with SUMO chains. Fluorescence microscopy revealed that, as well as being recruited to DSBs, GFP-NPSD4 forms punctuate foci in the nucleus that colocalizes with the Promyelocytic Leuckemia protein (PML). PML makes nuclear structures termed PML nuclear bodies (PML NBs) where many SUMOylated proteins accumulate, including those involved in DDR, senescence, cell cycle, transcription, carcinogenesis, and virus infection. While the exact role of PML NBs is not known, disruption of PML NB formation is seen in several tumor types. In order to elucidate the mechanistic role of NPSD4, we have again utilized mass spectrometry and proteomic analyses to identify candidate NPSD4 interacting DDR proteins, and have chosen several for further testing. These results support the hypothesis that NPSD4 is involved in maintenance of genomic stability. Further study is needed to characterize the function of NPSD4 in SUMO regulated DNA damage repair. As genomic instability contributes heavily to tumor development, the DDR must be better understood in order to develop treatments that specifically target tumor cells, capitalizing upon the defects they already exhibit in DDR, as well as to understand how tumorigenesis occurs. This project was supported by the American Legion Auxiliary Fellowship and the CPRIT Graduate Scholar Program. 33 Trainee Symposium on Cancer Research in Texas January 30, 2015 Dextrase, Katherine – MDACC Poster #28 Characterization of Photoacoustic Thermography for Image-Guidance and Monitoring of Photothermal Ablations Dextraze KL, MacLellan CJ, Mitcham TM, Melancon MP and Bouchard RR University of Texas at Houston and University of Texas MD Anderson Cancer Center Graduate School of Biomedical Sciences Corresponding author: Dextraze KL, University of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd, Unit 1902, Houston, TX 77030, Email: [email protected] Photoacoustic-ultrasonic (PAUS) imaging utilizes an ultrasound transducer to provide co-registered photoacoustic and pulse-echo ultrasound images. Furthermore, PAUS is capable of measuring temperature non-invasively while simultaneously providing anatomical images. The sub-millimeter resolution and centimeter-order penetration depths achievable with PAUS imaging have the potential to deliver active monitoring of both a targeted tumor microenvironment and nearby healthy tissue during thermal ablation. These characteristics make PAUS imaging a promising new technique for guidance and monitoring during photothermal ablations of solid tumors of the head and neck. In order to assess the potential clinical role of PAUS imaging, the technique was validated using a clinically accepted magnetic resonance thermal imaging approach (MRTI). To facilitate co-registration between modalities, a tissue-mimicking agar phantom was designed which had inclusions of gold nanoshells encapsulating super-paramagnetic iron oxide (SPIO) particles, where gold enhances the PA signal and SPIOs provide negative contrast on MRI. PA images were acquired on a Vevo LAZR (FUJIFILM VisualSonics Inc., Toronto, Ontario) PA-ultrasound small-animal imaging system (21MHz) operating at 710nm. MRTI experiments were performed using a 6-channel flex coil (GE Healthcare, Waukesha, WI) on a 3T MRI scanner (Discovery MR750, GE Healthcare, Waukesha, WI) using a fast multi gradient echo acquisition (16 echoes, 128x128 acquisition matrix, 25.6x25.6cm field of view, 3mm slice thickness, 60ms TR, 20⁰ flip angle, 2.9ms minimum TE and 1.6ms echo spacing). The accuracy and spatio-temporal resolution of PA thermography was cross-validated with both MRTI and a fluoroptic temperature sensor (LumaSense Technologies, Santa Clara, CA) in the custom-designed phantom. A reproducible, temperature-stable dual-modality phantom was created. Matched PA and MRTI thermography data was acquired using the phantom in the temperature range of 36 – 22°C. The PA and MRI temperature estimates agreed well, with similar temperature accuracy of 0.6°C compared to the gold-standard fluoroptic probe. PA thermography images showed improved spatial and temporal resolutions compared to the MRI thermography images. Axial and lateral resolutions of PA images were sub-millimeter with a temporal resolution of 0.2s, which will accommodate precise real-time guidance and monitoring. This study demonstrated that photoacoustic imaging is capable of measuring temperature changes with high spatiotemporal resolution and accuracy, which shows great promise for thermal monitoring of tissue during thermal ablation. The development of photoacoustic thermography as a precise, real-time technique for image-guidance and monitoring of thermal ablations will facilitate adaptive planning that may improve treatment efficacy for patients with unresectable head and neck cancer. This project was supported by CPRIT and the Matthews family. 34 Trainee Symposium on Cancer Research in Texas January 30, 2015 Doostan, Iman – MDACC Poster #29 Deregulated Cyclin E Expression Mediates Resistance to Aromatase Inhibitors in Postmenopausal Breast Cancer Patients Doostan I1, Karakas C2, Moulder S3, Hunt KK4 and Keyomarsi K5 1,5 University of Texas, Graduate School of Biomedical Sciences & 1,2,5 Department of Experimental Radiation Oncology, 3Department of Breast Medical Oncology and 4Department of Breast Surgical Oncology, MD Anderson Cancer Center, Houston, Texas, 77030 Corresponding author: Keyomarsi K, Department of Experimental Radiation Oncology, University of Texas MD Anderson Cancer Center, Houston, Texas, 77030, Email: [email protected] Estrogen is a major stimulus for tumor progression in breast cancer patients and almost 70% of all breast cancer patients are estrogen receptor positive (ER+) requiring anti-hormonal therapy. Aromatase Inhibitors (AIs) have been the mainstay anti-hormonal therapy in breast cancer patients. However, neoadjuvant studies have shown that response rates range between 30% to 70%. In spite of numerous studies aimed at understanding the biology of such resistance, overcoming resistance remains a major challenge in clinic and is the subject of this study. Cell cycle regulators have recently been proposed to be involved in mediating resistance to hormonal therapy. As an example, deregulation of Rb, a key factor in G1 checkpoint of the cell cycle, is linked to Tamoxifen resistance in patients. We have shown previously that deregulated expression of cyclin E, a key regulator in G1 to S transition, is a positive prognostic indicator in breast cancer patients. In vitro studies have shown that full-length cyclin E is cleaved into low molecular weight (LMW-E) forms that are specific to tumor cells, bind to CDK2 more efficiently and are resistant to inhibition by cyclin dependent kinase inhibitors. Considering the fact that anti-estrogens decrease level of cell cycle regulators and the fact that LMW-E deregulates cell cycle checkpoint, we hypothesize that LMW-E expression may render ER positive cells insensitive to growth inhibitory activity of aromatase inhibitors. To address our hypothesis, we engineered aromatase overexpressing ER+ cells to overexpress fulllength and LMW-E isoforms under doxycycline inducible promoter. In order to detect the activity of aromatase enzyme, these cells were cultured in estrogen-deprived media for four days followed by drug treatments. The cells were collected and counted after three days and were subjected to cell cycle analysis. In addition, cell lysates were collected and subjected to western blot analysis and in-vitro kinase assays. To test our hypothesis in vivo, we have also established a xenograft model system by injecting inducible cells into mammary fat pad of ovariectomized nude mice. We will interrogate the response to AIs in mice harboring LMW-E expressing tumors. Moreover, we will examine the correlation between cyclin E status in tumors and response to AI therapy in a cohort of patients that received neo-adjuvant AIs. Our results indicated that AIs inhibited proliferation by arresting the cells at G1 phase of the cell cycle. However, LMW-E induction significantly enhanced proliferation of the cells and promoted cell cycle. Using an MTT based assay we showed that LMW-E expressing cells have higher viability compared to control cells. At the molecular level, AIs decreased CDK2, pCDK2, and Rb levels and attenuated Rb phosphorylation. However, these effects were completely rescued upon LMW-E induction. Moreover, using an in vitro kinase assay we indicated that AIs decreased CDK2 kinase activity while LMWE expression reversed this effect. Knocking down CDK2 or Rb using independent shRNAs disrupted LMW-E mediated resistance further confirming that CDK2-Rb pathway regulates LMW-E response. Taken together, our results suggest that AIs induce cell cycle arrest in ER+ cells. However LMW-E inactivates Rb protein as a tumor suppressor and renders the cells to bypass G1 arraest following AI treatment. This study further indicates that targeting CDK2 can disrupt LMW-E mediated resistance in cell lines and proposes that CDK2 inhibitors combined with AIs could be a potential beneficial therapy for patients who express LMW-E in their tumors. This work was supported by CPRIT/Julia Jones Matthews Cancer Research Scholar training award to Iman Doostan. 35 Trainee Symposium on Cancer Research in Texas January 30, 2015 Huang, Le – MDACC Poster #30 Impact of Differentiation Status Of Kidney Progenitors in Wilms Tumorigenesis Huang L, Mokkapati S, Hu Q and Huff V UT-GSBS, G&D Program & Department of Genetics, UT MD Anderson Cancer Center Wilms tumor (WT) is one of the most common solid tumors in children. It is an embryonic cancer of the kidney and is thought to arise from undifferentiated renal mesenchyme. However, the types of cells in the mesenchyme that can give rise to WTs are unknown. Gene expression analysis of a large panel of WT patients has identified different subsets of WTs that are distinct in their clinical outcomes and gene expression signatures. These subsets have expression of specific genes that correspond to early renal mesenchyme or fully epithelial differentiation markers. These results suggest that WTs arise from transformed cells at different stages of differentiation. Thus, we hypothesized that the differentiation status of the initially mutated cell results in different types of WTs. To test this hypothesis, we generated mouse models of WT by using progenitor cell-type-specific Cre to introduce mutations of Wilms tumor gene 1 (Wt1) or -catenin in a progenitor-specific manner to determine the cellular origin of the tumors. There are two major types of progenitors during mouse kidney development: the nephron progenitor and the stroma progenitor, and respectively they express Six2 (or Cited1) and Foxd1. Therefore, we can use Six2 (or Cited1)-CreERTM or Foxd1-CreERTM to ablate Wt1 or activate β-Catenin. We found that when the mutations were introduced into the nephron progenitors (Six2 or Cited1-CreERT), WTs developed and that different types of mutations in the nephron progenitor resulted in different types of WTs. Tumors with -catenin activation had differentiated epithelial histology, whereas tumors with Wt1 ablation and Igf2 up-regulation had triphasic histology with more blastema. H&E staining of tumor sections demonstrated that the tumors with -catenin activation were composed mainly of epithelial cells forming tubule-like structures surrounded by blastemal and stromal cells. In contrast, tumors with Wt1 ablation and Igf2 up-regulation comprised undifferentiated blastemal, differentiated epithelia, and stroma cells. qPCR and IHC analysis of differentiation lineage markers further confirmed the different differentiation status between these two types of tumors. On the other hand, with the genetic lesions introduced, the stroma progenitors did not give rise to mouse WTs. Moreover, we used a high dose of tamoxifen to ablate Wt1 in a high proportion of progenitor cells to study the effects of Wt1 ablation on kidney development. We found that Wt1 ablation in stroma progenitors (Foxd1+) did not seem to affect kidney development. However, when Wt1 was ablated in nephron progenitors (Six2+), nephrogenesis was blocked. Compared to the control, the mutant kidneys at E19.5 contained fewer mature nephrons and increased number of stroma cells. Furthermore, the nephrongenic zone (NZ), where the progenitors reside, was expanded. And the cells in the expanded NZ expressed the progenitor markers and were proliferative. This suggested that Wt1 ablation in the nephron progenitors prevented the cells from differentiation but maintained the progenitor cell status such as selfrenewal and proliferation in the NZ. In addition, kidneys from 1 month and 2.5 months old mutant mice showed thinner cortex, reduced medulla and multiple fibrotic lesions. These results will help us to understand the cell origin of WT and the mechanisms of WT initiation and progression and provide insights into the function of Wt1 in the nephron and stroma progenitors. This work was supported by a Research Training Award from the Cancer Prevention and Research Institute of Texas (CPRIT RP140106) and by a Julia Jones Matthews Cancer Research Scholar training award to LH. 36 Trainee Symposium on Cancer Research in Texas January 30, 2015 Ko, How-Wen – MDACC Poster #31 Role of GSK3β-Mediated EZH2 Phosphorylation in Breast Cancer Stem Cells and Tumorigenesis Ko HW1,2, Yang CC1,2, Lee HH1,2, LaBaff AM1,2, Wei Y2, Huo L2, Cheng CC2 and Hung MC1,2 1 The University of Texas Graduate School of Biomedical Sciences at Houston 2 Department of Molecular and Cellular Oncology, The University of Texas MD Anderson Cancer Center Corresponding author: Mien-Chie Hung, Ph. D. Department of Molecular and Cellular Oncology, The University of Texas MD Anderson Cancer Center, Box 108, 1515 Holcombe Boulevard, Houston, TX 77030. Email: [email protected]. Background and Objective Tumor recurrence, progression, and metastasis are still major obstacles in anticancer treatment. Cancer stem cell hypothesis and its related tumorigenesis may account for this clinical observation, but the mechanisms involved remain elusive. Enhancer of zeste homologue 2 (EZH2), a subunit of Polycomb repressive complex 2, has been shown to be able to promote cancer progression and breast cancer stem cell expansion. Yet mechanisms for regulating EZH2 expression are still unclear. Glycogen synthase kinase 3β (GSK3 β) is a serine/threonine kinase. It mediates several cellular processes, including tumorigenesis, and has been shown to be involved in regulating CSCs. Interestingly, we observed that EZH2 contains GSK3 β phosphorylation motifs. It implies that EZH2 is a putative GSK3 β substrate. Therefore, we hypothesize that GSK3 β regulates breast cancer stem cells and tumorigenesis through directly mediating EZH2 protein expression. Methods and Results To test our hypothesis, we first examined whether GSK3 β regulates EZH2 protein expression and activity. We found that inhibition of GSK3 β by LiCl upregulates the EZH2 protein level and H3K27 trimethylation. Similarly, knockdown of GSK3 β by siRNA upregulates EZH2 protein expression. On the other hand, in a transient transfection experiment, constitutively active GSK3 β, but not kinase-dead GSK3 β, reduced EZH2 expression. These results suggest that GSK3 β negatively regulates EZH2 and H3K27 trimethylation expression. To determine whether GSK3 β physically interacts with EZH2, we carried out a co-immunoprecipitation experiment. The results demonstrated an association between EZH2 and GSK3 β in HEK 293T cells, indicating that GSK3 β is physically associated with EZH2. GSK3 β is a kinase. To evaluate phosphorylation of EZH2 by GSK3 β, we conducted an IP-EZH2 experiment. The result showed that a phosphorylated EZH2 protein was detected in the presence of GSK3 β kinase, but it diminished in lane without GSK3 β or with GSK3 β plus phosphatase, supporting that GSK3 β may phosphorylate EZH2. Seven potential phosphorylation residues on EZH2 were via mass spectrometry analysis, 4 on N-terminal part and 3 on C-terminal. The in vitro kinase assay validated the phosphorylation of EZH2 by GSK3 β kinase, and it appeared primarily at N-terminal part. EZH2 phosphor-deficient mutants were generated. Kinase assay revealed that the phosphorylation of EZH2 was significantly diminished in two single site mutants, indicating that GSK3 β phosphorylates EZH2 on these two residues. These results suggest that GSK3 β phosphorylates EZH2 on these two sites. To further explore the biological significance of EZH2 phosphorylation by GSK3 β low EZH2 expressing cell line, MCF7, and a noncancerous mammary epithelial cell line, MCF12A, have been stably transfected with EZH2 wild-type and phosphor-deficient mutants. Through Boyden chamber migration assay, we found that EZH2 indeed increased cell migration. The phosphor-deficient mutants caused more cells migrating. This observation was further supported by wound healing assay using time-lapsed microscopy. We also performed 3D culture soft agar assay. Wild-type EZH2 enhanced cell growth in an anchorageindependent manner. Phosphor-deficient mutants led to more and larger colony formation than WT did. These results suggest that GSK3 β nonphosphorylatable EZH2 can promote cell growth and migration. Functional assays related to cancer stem cells will be further pursued. Conclusion GSK3 β may phosphorylate and regulate EZH2 to suppress its oncogenic function in breast cancer. This study is granted by MD Anderson CPRIT Graduate Scholar Training Program. 37 Trainee Symposium on Cancer Research in Texas January 30, 2015 Leung, Marco – MDACC Poster #32 SNES: Single-Cell Exome Sequencing Leung ML1,3, Wang Y1, Waters J1 and Navin NE1,2,3 1 Department of Genetics, University of Texas MD Anderson Cancer Center, Houston, TX 2 Department of Bioinformatics and Computational Biology, University of Texas MD Anderson Cancer Center 3 Graduate Program in Genes and Development, Graduate School of Biomedical Sciences, University of Texas Health Science Center at Houston Corresponding author: Leung ML, Department of Genetics, University of Texas MD Anderson Cancer Center, 6767 Bertner Ave, Unit 1010, Houston, Texas 77030. Email: [email protected] Single-cell sequencing (SCS) methods have the potential to provide great insight into the genomics of rare cells and diverse populations, but are currently challenged by extensive technical errors. These errors include poor physical coverage and high FP and FN error rates, making it difficult to distinguish real biological variants. To address this problem, we have developed an SCS method called SNES that combines flow-sorting of single nuclei, limited multiple-displacement-amplification, low-input library preparation and exome capture to generate high coverage (>90%) data on single mammalian cells. To mitigate error rates, we collected nuclei that are in the G2/M stage of the cell cycle, providing 4 copies of the genome as input material for whole genome amplification (instead of 2 copies). We developed our method using a normal fibroblast cell line (SKN2) in which we can assume that variants in single cells will be highly similar to the population sample. We sequenced the reference population sample and the exomes of 9 single cells in G1/0 stage and 10 single cells in G2/M stage of the cell cycle. Our data suggest that G2/M cells provide several major technical improvements over using G1/0 cells, including decreased allelic dropout rate (21.52%), improved coverage breadth (95.94%). On average, we detect 24 false positive errors per megabases in each single cell genome. We show that these errors occur randomly in each cell, allowing accurate variant detection using two or more single cells. Our data suggest that, regardless of whether G1/0 or G2/M cells are used as input material, major technical improvement are achieved compared to existing single-cell sequencing methods. In summary, we expect that SNES will have broad applications in fields as diverse as cancer research, microbiology, neurobiology, development and in vitro prenatal genetic diagnosis, and will greatly improve our fundamental understanding of human diseases. This project was supported by CPRIT Graduate Scholar Award, Hearst Foundation Award and American Legion Auxiliary Fellowship in Cancer Research. 38 Trainee Symposium on Cancer Research in Texas January 30, 2015 McKenzie, Jodi – MDACC Poster #33 Topoisomerase I Inhibitors Enhance the Efficacy of Immunotherapy in Melanoma McKenzie JA1, Mbofung RM1, Malu S1, Amaria R1 and Hwu P1 1 Department of Melanoma Medical Oncology, University of Texas MD Anderson Cancer Center Corresponding author: McKenzie JA, Department of Melanoma Medical Oncology, UT MD Anderson Cancer Center, 7455 Fannin Street, Unit 904, Houston, TX 77054, Email: [email protected] Melanoma is a highly aggressive form of skin cancer, whose rates of morbidity and mortality are continuously increasing. The development of immunotherapeutic agents like Ipilimumab, have resulted in fundamental advances in the treatment of melanoma. However, long lasting responses are only observed in a small subset of melanoma patients treated with Ipilimumab. This shortfall highlights the need for a better understanding of the molecular mechanisms that govern tumor sensitivity or resistance to immunotherapy. To address this need, paired tumor samples and tumor infiltrating lymphocytes (TILs) from melanoma patients were utilized in an in vitro activated caspase 3-based high-throughput screen, to identify compounds that increase the sensitivity of melanoma cells to T-cell mediated cytotoxicity. The screen consisted of a library of 850 bioactive compounds. One group of compounds that was most able to enhance T-cell killing of melanoma cells was Topoisomerase 1 (Top1) inhibitors including: Topotecan, and Irinotecan. Topoisomerases are a family of DNA enzymes, which are involved in unwinding DNA and relieving torsional strain during replication and transcription. There are several sub-types of topoisomerases, and our interest is focused on sub-type 1 (Top1), which is primarily involved in relieving tension during transcription. Our results show that treatment of melanoma tumor cells with a Top1 inhibitor prior to exposure to autologous TILs, produced a synergistic increase in tumor cell death, as measured by intracellular staining of activated caspase 3, and computed using CalcuSyn. We have also recapitulated this finding in an in vivo model, where a better anti-tumor effect was observed in tumor-bearing mice treated with an antibody against the co-inhibitory molecule Programmed Death Ligand 1 (α-PD-L1) in combination with MM-398 (liposomal formulation of Irinotecan), than in cohorts treated with either antibody or drug alone. These findings suggest synergism between Top1 inhibitors and immune-based therapies in the treatment of melanoma. Further studies looking at both genomic and proteomic changes that occur in response to Top1 inhibition, are now aimed at discovering the molecular mechanisms by which Top1 inhibitors mediate increased sensitivity of melanoma to immunotherapy. Understanding how Top1 inhibitors enhance melanoma killing by immunotherapy will allow for the development of predictive biomarkers, and also augment immune-based therapeutic strategies to ensure durable responses in a larger population of melanoma patients. This work was supported by Research Training Awards from the Cancer Prevention and Research Institute of Texas (CPRIT RP140106) and the UT MD Anderson Cancer Center SPORE in Melanoma grant from the National Cancer Institute (P50 CA093459), Aim at Melanoma and the Miriam and Jim Mulva Research Funds to Jodi A McKenzie. 39 Trainee Symposium on Cancer Research in Texas January 30, 2015 Manton, Christa – MDACC Poster #34 Caspase-Dependent Effects of the Proteasome Inhibitors Bortezomib and Marizomib in Glioblastoma Manton CA1,2, Johnson B1,2, Singh M1, Bouchier-Hayes L3 and Chandra J1,2 1 Department of Pediatrics Research, MD Anderson Cancer Center 2 Graduate School of Biomedical Sciences, University of Texas Health Science Center Houston 3 Department of Pediatrics, Baylor College of Medicine Corresponding author: Joya Chandra, MD Anderson Cancer Center, 1515 Holcombe Blvd, Unit 0853, Houston, TX 77030; Email: [email protected] New therapeutic options for glioblastoma (GBM) treatment are needed, given the current median survival of only 14 months with current treatment. The proteasome inhibitor bortezomib (BTZ) shows efficacy in cancers like myeloma, but its clinical utility in other cancer types is questionable. Newer proteasome inhibitors such as marizomib (MRZ) have unique inhibitory and death inducing properties that have not been well examined in GBM. We assessed the mechanism of death induced by BTZ and MRZ in GBM in order to identify biomarkers of clinical efficacy, anticipate drug resistance, and design rational combination therapy strategies. We evaluated the kinetics of activation of the initiator caspases 2, 8, and 9 and also used chemical and knockdown strategies to inhibit their activity. Caspase 2 was activated early (4 h) following proteasome inhibitor treatment, while caspases 8 and 9 were activated slightly later (8-12 h). Knockdown and chemical inhibition of caspases revealed that both BTZ and MRZ induced death that was dependent on caspase 9, and not the other initiator caspases, in GBM cells. An orthotopic xenograft model of GBM was used to evaluate questions related to the ability of these agents to cross the blood brain barrier and exert functional effects on proteasome targets. Both BTZ and MRZ increased levels of the proteasome substrates p21 and p27 in intracranial tumors, with MRZ exerting slightly stronger effects. We also examined the combination of proteasome inhibition with the histone deacetylase inhibitor vorinostat. Combinations of BTZ or MRZ plus vorinostat potentiated activation of caspase 9 and death in GBM cell lines. Examination of cleavage of lamin A, a caspase substrate, as a marker of apoptosis in brain tumors from mice treated with combinations of BTZ or MRZ and vorinostat showed increased cell death in brain tumors treated with these combinations. Together, this data clarifies the hierarchy of caspase induction by proteasome inhibitors and provides evidence that proteasome inhibitors can reach brain tumors where they exert functional effects and increase death in combination with vorinostat. This data has clinical implications, as it suggests potential biomarkers of drug efficacy that could be examined in clinical trials and reveals information that could help predict drug resistance, such as in brain tumors with molecular characteristics that limit caspase 9driven apoptosis. This project was supported by the NIH/NCI under award numbers: 1. P30CA016672 for the Characterized Cell Line Core and the Research Animal Support Facility; 2. Brain Tumor SPORE Award Number P50 CA127001 (Developmental Project Grant to J.C.). Support was also provided by the Team Connor Cancer Foundation and the Thomas Scott Family Foundation. C.A.M. was supported by a Research Training Award from the Cancer Prevention and Research Institute of Texas (CPRIT RP140106) and a Julia Jones Matthews Cancer Research Scholar training award. 40 Trainee Symposium on Cancer Research in Texas January 30, 2015 Mukherjee, Seema – MDACC Poster #35 Investigating the Role of Poly (Adp-Ribose) Polymerase-1 (Parp-1) in the Development and Progression of Small Cell Lung Cancer (SCLC) Mukherjee S1, Cardnell RJ 1, Diao L2, Heymach JV1, Mills GB3, Wang J2 and Byers LA1 1 Department of Thoracic/Head & Neck Medical Oncology, MD Anderson Cancer Center 2 Department of Bioinformatics & Computational Biology, MD Anderson Cancer Center 3 Department Chair, Department of Systems Biology, Division of Cancer Medicine, The University of Texas MD Anderson Cancer Center Background Small cell lung cancer (SCLC) is a highly lethal malignancy characterized by rapid growth, early metastasis and poor prognosis. We previously reported that poly (ADP-ribose) polymerase 1 (PARP-1), a DNA repair protein, is dramatically overexpressed in SCLC cell lines and patient samples. Our preclinical data demonstrate single agent activity of PARP inhibitors in SCLC cell lines and xenografts. A recently Phase I trial of the single-agent BMN 673 (a PARP inhibitor) showed response rates of 20-30% in recurrent SCLC, a dramatic improvement of existing second line treatments in this setting. Other studies have shown a pivotal role for PARP-1 as a positive cofactor of E2F1-mediated transcription. RB1, however, has been reported to inhibit E2F1 by masking its transactivation domain. Based on these observations, we hypothesize that RB1 loss leads to PARP-1 overexpression via activation of E2F1. In turn, elevated PARP-1, acting as a cofactor of E2F1, can enhance E2F1-mediated transcription of other DNA repair proteins leading to increased cell survival and metastasis. Further, we predict that PARP-1 overexpression itself can also drive these events independent of E2F1 by inducing EMT (Figure 1). Methods PARP-1 knockdown by shRNA in SCLC cells was verified by western blot. The effect of PARP-1 shRNA on cell proliferation and drug sensitivity was evaluated by Cell titer glow assay. Boyden chamber assay was used to study the effect of PARP-1 loss on cell migration. To investigate the effect of PARP-1 knockdown on downstream proteins, reverse phase protein array (RPPA) analysis and western blotting were performed using lysates from PARP-1 shRNA knockdown Figure1. Proposed mechanism of PARP-1 overexpression in SCLC cell lines. SCLC. RB1 loss leads to E2F1 activation and resulting PARP-1 Results shRNA targeting PARP-1 resulted in a overexpression. PARP-1 then activates E2F1 in positive feedback significant decrease in PARP-1 expression in 8 loop, driving the synthesis of additional DNA repair proteins, SCLC cell lines. PARP-1 loss slowed proliferation increased cell survival and metastasis. and migration, and reduced sensitivity to PARP inhibitors (BMN673 and AZD-2281). Proteomic analyses show PARP-1 knockdown resulted in decreased E2F1 expression and other DNA repair protein (e.g. p-ATR, p-Chk2 ATM), in agreement with our hypothesis. Proteomic profiling also showed downregulation of proteins such as vimentin and snail following PARP-1 knockdown. Futher, using RPPA data from our larger 34 SCLC cell line panel, we observed an inverse correlation between RB1 and PARP-1 protein expression. Conclusions and future directions Our results strongly indicate that inhibition of PARP-1 in SCLC by shRNA resulted in suppressed cell proliferation and migration; with decreased expression of EMT related proteins suggesting a role for that PARP-1 play a role in EMT-induced metastasis. PARP-1 attenuation also results in decreased E2F1 expression, in support of our hypothesis. Supported by: CPRIT training grant (SM), LUNGevity Foundation (LAB), and The Sidney Kimmel Foundation for Cancer research (LAB). 41 Trainee Symposium on Cancer Research in Texas January 30, 2015 Nagaraja, Archana – MDACC Poster #36 Sustained Adrenergic Signaling Activates Pro-Inflammatory Prostaglandin Network in Ovarian Carcinoma Nagaraja AN1, Dorniak PL1, Sadaoui NC1, Armaiz-Pena G1, Zand B1, Wu SY1, Allen JK1, Rupaimoole R1, Rodriguez-Aguayo C2, Pradeep S1, Lin T3, Previs RA1, Hansen JM1, Yang P3, Lopez-Berestein G3, Lutgendorf SK5, Cole SW6 and Sood AK1,2,4* 1 Gynecologic Oncology and Reproductive Medicine 2 Center for RNAi and Non-Coding RNA 3 General Oncology 4 Cancer Biology The University of Texas MD Anderson Cancer Center 5 Departments of Psychology, Obstetrics and Gynecology, and Holden Comprehensive Cancer Center, University of Iowa 6 Department of Medicine and Jonsson Comprehensive Cancer Center, University of California, Los Angeles School of Medicine, UCLA Molecular Biology Institute, and Norman Cousins Center *Corresponding author: [email protected] Objectives Catecholamine mediated stress effects are known to induce production of various pro-inflammatory cytokines. However, the mechanisms and functional effects of adrenergic signaling in driving inflammation by cancer cells and cancer-associated fibroblasts (CAFs) are currently unknown. Here, we address the functional and biological consequences of adrenergic-induced COX-2/PGE2 axis activation in cancer cells and effects of adrenergic-induced elevation in recruitment and transformation of CAFs in promoting ovarian cancer. Methods We first analyzed global metabolic changes in tumors isolated from patients with known Center for Epidemiologic Studies Depression Scale (CES-D; depressive) scores and tumoral norepinephrine (NE) levels. Skov3 and HeyA8 cells were used to study gene and protein levels of PTGS2 (cyclooxygenase2), PTGES (prostaglandin E2 synthase) and metabolite PGE2 in vitro and in vivo. To study tumor-specific effects on PTGS2, we used the DOPC delivery system for PTGS2 siRNA. Normal fibroblasts (NoF 151 and NoF 182) and cancer-associated fibroblasts (CAF 148) were treated with conditioned media from cancer cells and norepinephrine respectively. Gene and protein expression of cytokines and changes in migratory potential of fibroblasts were assessed. Results PG levels were significantly increased in patients with high depressive scores (>16). PGE2 was upregulated by 2.38 fold in this group when compared to the low CES-D score group. Exposure to NE resulted in increased PTGS2 and PTGES gene expression and protein levels in Skov3 and HeyA8 cells. Similarly, , PGE2 levels were increased in conditioned medium from Skov3 and HeyA8 cells upon treatment with NE. Treatment with a broad ADRB agonist (isoproterenol) or ADRB2 specific agonist (terbutaline) led to increases in expression of PTGS2 and PTGES as well as PGE2 levels in supernatants. ChIP analysis showed enrichment of Nf-kB binding to the promoter region of PTGS2 and PTGES by 2.4 and 4.0 fold, respectively when Skov3ip1 cells were treated with NE. Importantly, in the Skov3-ip1 restraint stress orthotopic model, silencing PTGS2 abrogated stress mediated effects and decreased tumor burden by 70% compared to control siRNA with restraint stress. Adrenergic signaling was associated with increased levels of α-smooth muscle actin (α SMA), a marker for CAFs. These increases were abrogated when mice were treated with broad beta-blocker propranolol during restraint stress. Treatment of normal fibroblasts with conditioned media from cancer cells treated with NE accelerated the expression of α-SMA indicating a role for NE in driving CAF transformation. This was also associated with increased migration and increased production of pro-inflammatory cytokines by conditioned fibroblasts. Conclusion Increased adrenergic stimulation of cancer cells and CAFs results in a pro-inflammatory microenvironment mediated by prostaglandins and cytokines that drives tumor progression and metastasis in ovarian cancer. This work was supported by a Research Training Award from the Cancer Prevention and Research Institute of Texas (CPRIT RP140106) to AN. This research was funded in part by support from NIH grants (CA109298), the UT MDACC Ovarian Cancer Spore (P50 CA083639), the Betty Ann Asche Murray Distinguished Professorship, and the Blanton-Davis Ovarian Cancer Research Program. 42 Trainee Symposium on Cancer Research in Texas January 30, 2015 Napoli, Marco – MDACC Poster #37 Pharmacologic Inhibition of the Np63/DGCR8 Axis as a Novel Therapeutic Strategy for p53 Deficient and Mutant Tumors Napoli M1, Venkatanarayan A1, Raulji P1, Abbas HA1, Norton W2, Sood AK3, Lopez-Berestein G4, Tsai KY5, Coarfa C6, Gunaratne PH7 and Flores ER1 1 Department of Molecular and Cellular Oncology 2 Department of Veterinary Medicine and Surgery 3 Department of Gynecologic Oncology and Reproductive Medicine 4 Department of Experimental Therapeutics 5 Department of Dermatology, The University of Texas M.D. Anderson Cancer Center 6 Department of Molecular and Cellular Biology, Baylor College of Medicine 7 Department of Biology and Biochemistry, University of Houston Corresponding author: Napoli M, Dept. of Molecular and Cellular Oncology, The UT M.D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, Texas 77030, USA. Email: [email protected] The p63 gene is a member of the p53 family, whose transcription is driven by two alternative promoters, allowing the expression of two distinct isoforms: TAp63 and ΔNp63. In order to determine their roles in tumorigenesis, our group has generated isoform-specific knockout mouse models. TAp63 was found to be a pivotal tumor and metastasis suppressor, whose function is achieved by the regulation of the miRNA biogenesis pathway in a Dicer dependent way. ΔNp63 was shown to be a master regulator of the differentiation of epithelial tissues, such as the epidermis. Np63 is indeed able to directly induce the expression of DGCR8, a co-factor essential in modulating miRNA biogenesis, sustaining in this manner a proper epidermal differentiation. Accordingly, ΔNp63 knockout mice die shortly after birth because of developmental defects. In addition to characterize its role in skin development, our group has recently demonstrated that ΔNp63 may act as an oncogene by counteracting TAp63 tumor suppressive activities in vivo: specific deletion of ΔNp63 in p53-/- thymic lymphomas caused TAp63 reactivation and consequent tumor shrinkage. Based on all these findings, we hypothesized that ΔNp63 might support tumor formation not only by inhibiting TAp63, but also by affecting miRNA maturation in a DGCR8 dependent manner. In line with this hypothesis, we deem that the inhibition of the ΔNp63/DGCR8 axis might be crucial to curb ΔNp63 tumor promoting activities. To this end, we screened a drug library of more than 850 FDA approved compounds looking for inhibitors of the ΔNp63/DGCR8 axis. We found several small molecules that were able to reduce ΔNp63 protein stability and, as a result, to decrease the expression levels of its target gene DGCR8 and that of a specific set of miRNAs, whose biosynthesis is DGCR8-dependent. On the contrary, the levels of DGCR8-independent miRNAs were not affected by the drug treatments. In agreement with our hypothesis, curbing the ΔNp63/DGCR8 axis by either these drugs or inhibitors against the identified miRNAs, reduced cell viability of different cancer cell types of both murine and human origin, without affecting that of normal cells. In vivo analyses of these compounds as well as of these miRNA inhibitors have showed their efficacy in counteracting tumor formation and progression in mice. In summary, we believe that the in vitro and in vivo characterization of the ΔNp63/DGCR8 axis will provide a novel and effective strategy to target tumors relying on ΔNp63 for their expansion, especially the most therapeutically challenging ones devoid of a functional tumor suppressor p53. This work was supported by a Research Training Award from the Cancer Prevention and Research Institute of Texas (CPRIT RP140106) to Marco Napoli. 43 Trainee Symposium on Cancer Research in Texas January 30, 2015 Peng, David – MDACC Poster #38 MiR-200 Modulation of ECM in Lung Cancer Invasion and Metastases Peng DH1 and Gibbons DL1, 2 1 Department of Thoracic/Head and Neck Medical Oncology 2 Department of Molecular and Cellular Biology, The University of Texas MD Anderson Cancer Center Corresponding author: Gibbons DL, 1515 Holcombe Blvd, Unit 432, Houston, TX, 77030 Lung cancer has the highest cancer-related deaths due to distant metastatic disease, correlating with a poor overall 5-year survival rate. Thus, understanding the mechanisms governing metastasis will identify new therapeutic targets and treatments to improve overall patient survival. A mutant KrasG12D;p53R172H murine model of metastatic lung adenocarcinoma was generated to study the metastatic process, which demonstrated that metastatic cancer cells undergo an epithelial-to-mesenchymal transition (EMT) that is regulated by a double-negative feedback loop between the miR-200 family and Zeb1/2. Although the cell intrinsic effects of EMT are important for tumor progression, cancer cell interaction with the extracellular matrix (ECM) also plays a critical role in regulating invasion and metastasis. We have demonstrated that lung cancer cell invasion occurs through interaction with specific ECM components, such as type I collagen, is EMT-driven, and can be attenuated by miR-200 expression. Comparative proteomic profiling data between mesenchymal (low miR-200) and epithelial (high miR200) murine cell lines as well as analysis of TCGA datasets revealed a positive correlation between EMT gene signatures and expression of matrix-related proteins, including collagen, LOX, and LOXL2. The LOX enzymes are involved in crosslinking collagen, leading to ECM stiffening, which has been implicated in driving cancer invasion and metastasis. We hypothesize that miR-200 regulates expression of ECM- associated proteins in cancer cells, thereby altering the cell-matrix interactions that activate signaling pathways to produce invasion and metastases. Masson’s trichrome, immunohistochemistry, and second harmonic imaging of lung tumor tissues from the KrasG12D;p53R172H mice with metastatic disease showed an increase in total collagen deposition, LOX protein expression, and linearized collagen compared to lung tumors from non-metastatic KrasG12D mice. Western blot and qPCR analyses of epithelial and mesenchymal NSCLC cell lines showed increased LOX, LOXL2, and collagen expression in mesenchymal cells. Luciferase reporter assays demonstrated that miR-200 directly binds to LOX 3’-UTR while predictive algorithm analysis of the LOXL2 promoter region reveals several potential Zeb1 and Ets1 transcription factor binding sites, which are direct miR-200 targets. To assess the functional relevance of LOX and LOXL2 in driving lung cancer cell invasion and metastasis, we knocked down LOX and LOXL2 expression with stable short hairpin RNAs and chemically inhibited the catalytic activity of all LOX enzymes using β-aminopropionitrile (BAPN) and DPenicillamine in mesenchymal murine cells. Decreased LOX protein levels and enzymatic activity in mesenchymal murine cell lines significantly reduced cancer cell invasion specifically through a collagen matrix in vitro. In contrast, expression of LOXL2 in epithelial murine lung cancer cells showed a significant increase in collagen invasion. Extending the in vitro functional studies to in vivo subcutaneous transplants in syngeneic mice demonstrated a decrease in metastatic disease when either LOX or LOXL2 were stably knocked down in mesenchymal murine lung cancer cells. Our data show that metastatic lung tumor tissues and cell lines are characterized by increased collagen and LOX proteins, correlating with a decrease in miR-200 levels. Furthermore, we demonstrate that collagen crosslinking by the miR-200-regulated LOX enzymes helps drive cancer cell invasion and metastasis. Results from our studies reveal a mechanism of miR-200 regulation of the ECM during lung cancer progression and provide potential diagnostic and therapeutic targets for metastatic disease. This project was supported by a Research Training Award from the Cancer Prevention and Research Institute of Texas (CPRIT RP140106). 44 Trainee Symposium on Cancer Research in Texas January 30, 2015 Puig, Stephanie – MDACC Poster #39 Anti-Cancer Drugs to Treat Cancer Pain Puig S1 and Gutstein HB1 1 Department of Anesthesiology and PeriOp Medicine, University of Texas MD Anderson Cancer Center Corresponding author: Puig Stephanie., Department of Anesthesiology and PeriOp Medicine, UT MD Anderson Cancer Center, Unit 110, 1515 Holcombe Blvd, Houston, Texas 77030, Email: [email protected] Over 50% of cancer patients and 33% of cancer survivors suffer with severe chronic pain. For centuries, opioids have been the gold standard for treating severe pain. However, with repeated dosing, tolerance to the analgesic effect of opioids develops. The mechanisms underlying opioid tolerance are poorly understood. We recently made the groundbreaking discovery that Imatinib, a chemotherapeutic agent used in the clinic that specifically blocks platelet-derived growth factor receptor (PDGFR) signaling, prevented morphine tolerance from occurring and completely reversed established tolerance. We also found that PDGFR-β signaling is both necessary and sufficient to cause opioid tolerance. The aim of the present study is to evaluate the neural circuitry underlying PDGFR-β modulation of opioid tolerance. Using immunohistofluorescence and confocal microscopic imaging, we have defined the specific types of cells expressing the PDGFR-β, the growth factor PDGF-BB, and the mu opioid receptor in specific cell subtypes of the dorsal root ganglion and dorsal horn of the spinal cord. We are also determining whether the distribution and co-localization of these markers is altered by chronic administration of various opioids. These studies will provide unique insights into the neural circuitry underlying opioid tolerance and should have important implications for the treatment of chronic cancer pain. The authors declare no conflict of interest. This research is funded by a CPRIT grant to H.B. Gutstein and by a TRIUMPH fellowship to S Puig. 45 Trainee Symposium on Cancer Research in Texas January 30, 2015 Roh, Whijae – MDACC Poster #40 Genomic and Immune Profiles of Melanomas Treated with Immune Checkpoint Blockade Roh W1,7, Cooper ZA1,2, Chen P-L3, Zhang J4, Cao Y1, Amaria RN5, Allison JP6, Davies MA5, Hwu P5, Chin L1 and Wargo JA1,2 1 Department of Genomic Medicine, MD Anderson Cancer Center 2 Department of Surgical Oncology, MD Anderson Cancer Center 3 Department of Pathology, MD Anderson Cancer Center, Houston 4 Applied Cancer Science Institute, MD Anderson Cancer Center 5 Department of Melanoma Medical Oncology, MD Anderson Cancer Center 6 Department of Immunology, MD Anderson Cancer Center 7 Program in Cancer Biology, The University of Texas Graduate School of Biomedical Sciences Corresponding author: P. Andrew Futreal, Dept. of Genomic Medicine, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030, USA, Email: [email protected] The diversity, specificity, and memory of immune cells against tumor-associated- antigens are unique features of immunotherapy that can be exploited. Melanoma is particularly responsive to immunotherapy due to the high infiltration of immune cells into tumors and potentially more neoantigens (antigens expressed only in tumor cells) presented due to its high mutation rate. Among several ways to activate T cells for antitumor response, the immune checkpoint blockade (anti-CTLA-4, anti-PD-1), which reverses negative signals of T cell activation, is recognized as breakthrough treatment for advanced-stage melanoma. Though response rates are modest (anti-CTLA-4~15%, anti-PD-1~30%), such responses are often durable. Therefore, there are ongoing efforts to identify biomarkers that predict response to immune checkpoint blockade with the potential to also identify actionable strategies to improve responses. A recent study showed that neoantigen signature was correlated with clinical benefit from anti-CTLA-4. However, the number of neoantigens alone cannot capture the interaction between tumor cells and immune cells well. Our hypothesis is that response or resistance mechanism will be correlated with genomic landscape of tumor cells and immune microenvironment (frequency and proportions of immune cell subsets). The changes in the genomic landscape of tumors and clonality pattern of TILs after therapy will also provide understanding of co-evolution of tumor cells and T cells under immune checkpoint blockade. The primary objective of this proposal is to study genomic and immune profiles and their changes in tumors on immune checkpoint blockade and to correlate these findings with response to therapy. Schematic Diagram of Study Design This project is supported by Cancer Prevention and Research Institute of Texas (CPRIT). 46 Trainee Symposium on Cancer Research in Texas January 30, 2015 Sanchez, Nora – MDACC No Poster Identification and Validation of Novel Drug Targets for the Eradication of KRAS-Independent Pancreatic Stem Cells Sánchez NS, Viale A, Carugo A, Ying H, Kapoor A, Pettazzoni P, Corti D, Rose J, Yao W, Green T, Windham J, Heffernan T and Draetta G The University of Texas MD Anderson Cancer Center at Houston Introduction Pancreatic Ductal Adenocarcinoma (PDAC) is the 4th leading cause of cancer deaths in the U.S. An in-depth understanding of the signaling mechanisms in PDAC is needed to allow the rational development of effective targeted therapies. Using a unique doxycycline (DOX) inducible KrasG12D driven PDAC mouse model (iKras) we have identified a population of tumor cells that survive KRAS extinction in vivo and in vitro and is highly tumorigenic upon re-expression of oncogenic KRAS. These cells behave like bona fide stem cells, are intrinsically more resistant to drugs and environmental stresses, suggesting an important role in drug resistance and tumor relapse after conventional therapies. Using these cells we have developed an in vivo functional genetic screening platform that takes into account the importance of genetic context and microenvironment to systematically investigate the tumorigenic potential of driver genetic-elements of interest (GEOI) in PDAC. Methods We developed an inducible KRAS* mouse model of PDAC that enables pancreas-specific and doxycycline-inducible expression of KrasG12D in a p53 mutant background (p53L/L or L/+)(Ying et al 2012). A subpopulation of cells survives oncogene extinction and contributes to tumor relapse. These cells were extensively characterized including stem cell profile, tumorigenic potential and genome wide transcriptome analysis. ~ 600 genes are differentially regulated in surviving cells. These GEIOs were interrogated in a high-throughput fashion using a pooled virus, gain or loss of function approach in vivo. Results The ability of these cells to form tumors in vivo or spheres in serum free culture is dependent on expression of KRAS. Upon oncogene extinction, the majority of the cells undergo apoptosis. The few surviving cells express CD133, remain quiescent for an extended period of time, and are highly enriched in tumor initiating cells. Through pooled primary screens and individual secondary validations in vivo, over 30 GEIO’s were identified as potential driver GEOI’s in PDAC. Conclusion The unique characteristics of the surviving cells have provided a system we can exploit to identify GEOIs that cooperate with K-RAS in promoting PDAC. Enrichment of genes in these highly tumorigenic cells provides a refined set of relevant GEOIs. Hits resulting from our pooled screening have been validated in mouse. Further validation is planned using human spheroid assays. We anticipate this powerful screening approach will identify several potential druggable targets in PDAC. 47 Trainee Symposium on Cancer Research in Texas January 30, 2015 Schibler, Andria – MDACC Poster #41 COMPASS Regulates the Spindle Assembly Checkpoint through Histone H3 K4 Methylation Schibler A1, Tomida J1, Koutelou E1, Wilson-Pham M2, Wood R1, and Dent S1 1 The Virginia Harris Cockrell Cancer Research Center at The University of Texas MD Anderson Cancer Center Science Park 2 The University of Texas Graduate School of Biomedical Sciences at Houston Corresponding author: Dent S, Dept. of Epigenetics and Molecular Carcinogenesis, University of Texas MD Anderson Cancer Center Science Park, 1808 Park Road 1C, P.O. Box 389, Smithville TX 78957 Email:[email protected] Mitosis is a multi-step process that involves the duplication of cellular machinery and successful separation of these components into two cells. During cellular division, DNA must be partitioned evenly. Saccharomyces cerevisiae divides the duplicated genome by attaching the mitotic spindle to each chromosome in a bipolar fashion. The sister chromatids are then pulled away from each other to opposite sides of the cell before the two nuclei segregate and the cells divide. Misregulation of these events often results in aneuploidy, which is an abnormal number of chromosomes per cell. To prevent aneuploidy, cells activate the Spindle Assembly Checkpoint (SAC) to ensure that the duplicated genome is divided equally. Interestingly, SAC components are commonly mutated in tumors and SAC deactivation has been observed in many types of cancer. Set1 is a lysine methyltransferase enzyme and a component of the COMPASS complex. Currently there are only two known substrates of COMPASS in S. cerevisiae one is histone H3 (a component of the nucleosome). I have found that set1 and COMPASS mutants display a resistance to benomyl. Benomyl functions as a mitotic spindle poison by destabilizing microtubule-kinetochore interactions and activating the SAC. Interestingly, this phenotype is at least partially due to lack of histone H3 lysine 4 methylation. Moreover, the benomyl resistance phenotype displayed by set1 and histone H3 K4 mutants requires components of the SAC pathway indicating a novel cellular function for histone H3 K4 methylation. Lastly, by using recombinant binding assays, we have found SAC components can directly bind histone H3. I have found that COMPASS mediated methylation is required for proper regulation of the SAC. Moreover, COMPASS regulation of the SAC is through histone H3 K4 methylation. Regulation of the SAC pathway through methylation highlights the role Set1 plays in maintaining genomic stability. I hypothesize methylation of histone H3 regulates mitosis through the SAC pathway in a posttranscriptional manner. Deciphering how methylation of histone and non-histone substrates affects the Spindle Assembly Checkpoint will elucidate unknown regulatory mechanisms during mitosis and help identify how mutations to lysine methyltransferases and SAC components contribute to cancer progression. This project was supported by a Research Training Award from the Cancer Prevention and Research Institute of Texas (CPRIT RP140106). 48 Trainee Symposium on Cancer Research in Texas January 30, 2015 Su, Chun-Hui – MDACC Poster #42 Targeting Nentrophil Elastase In Breast Cancer Metastasis Su CH1, Liu Y1, Caruso JA2, Karakas C2, Kulmacz RJ3, Keyomarsi K2 and Hunt KK1 1 Department of Surgical Oncology, MD Anderson Center 2 Department of Experimental Radiation Oncology, MD Anderson Center 3 Department of Internal Medicine, University of Texas Health Science Center at Houston Corresponding author: Keyomarsi K, Department of Experimental Radiation Oncology, MD Anderson Center, 1515 Holcombe Blvd, Unit Number: 0066, 77030, Houston, E-mail: [email protected] Hunt KK, Department of Surgical Oncology, MD Anderson Center, Houston TX, 1515 Holcombe Blvd, Unit Number: 1434, 77030, Houston, Email: [email protected] Neutrophil elastase (NE) is a secretory serine protease with its physiologic role in innate host defense. It accumulates and plays a proteolytic role at inflammatory responses. NE activity is regulated by an endogenous protease inhibitor, elafin and α-1 antitrypsin. When NE activity exceeds the endogenous inhibitors, it causes abnormal degradation of extracellular matrix, leading to inflammatory lung disorders. Sivelestat , NE inhibitor, is currently used in Japan for the treatment of acute respiratory distress syndrome due to accumulation of NE. However, the use of sivelestat is limited to respiratory disorders and not currently used for the treatment of neoplasia. In breast cancer, it has been shown that patients with higher levels of NE have poor survival compared to those with lower levels of NE. Our lab found that NE cleaves cyclin E and generates the tumor specific and oncogenic LMW cyclin E, thereby causing cell cycle deregulation. Additionally downregulation of NE in MDA-MB-231, BT549 and MDA-MB157 cells result in the inhibition of cell proliferation, colony formation, motility and invasion. Furthermore, downregulation of NE also results in the inhibition of tumorigenesis in otherwise tumorigenic MDA-MB-231 cells. Therefore, NE inhibitors may have profound anti-tumor activity. We hypothesize that NE is a therapeutic target and inhibition of NE can be used as targeted therapy in treatment of breast cancer. A novel, orally available NE inhibitor, AZD9668, is in the phase II clinical trial for patients with inflammatory lung disease. We have found that treatment of breast cancer cells with AZD9668 delays cell growth, anchorage-independent growth, cell migration, and cell invasion similar to the NE knockdown studies. We have also found that AZD9668 is able to reduce lung metastasis in an in vivo mouse model. Collectively, our results suggest that inhibition NE can result in inhibition of metastasis in breast cancer. This project was supported by the CPRIT training grant. 49 Trainee Symposium on Cancer Research in Texas January 30, 2015 Tashakori, Mehmoosh – MDACC Poster #43 in vivo Significance of Mdm4 and p73 Interaction in Development and Tumorigenesis Tashakori M1, Flores ER2 and Lozano G1 Department of Genetics, UT MD Anderson Cancer Center, Houston Department of Molecular and Cellular Oncology, UT MD Anderson Cancer Center Corresponding author: Lozano G, Department of Genetics, UT MD Anderson Cancer Center, 1515 Holcombe Blvd., Unit 1010, Houston, Texas 77030, Email: [email protected] The tumor suppressor protein p53 is negatively regulated by the Mdm proteins. The significance of Mdm2 and Mdm4 in regulation of p53 was determined from mouse models. The Mdm2-null mice die at E3.5 while Mdm4-deficient mice die at E7.5. Both embryo lethalities are completely rescued by deletion of p53. p73, a member of the p53-family, is a transcription factor with tumor suppressor activity. Isothermal titration calorimetry studies revealed that Mdm4 has higher affinity for p73 than p53. To investigate the physiological importance of the Mdm4 and p73 interaction, we asked whether Mdm4 is a negative regulator of p73 in vivo. To address this question, I studied whether the embryonic lethality of Mdm4-deficient mice is rescued by p73 loss. I intercrossed Mdm4+/∆2 p73+/− mice and performed the rescue study at different developmental time points starting from mid-gestation to embryonic day (E) 9.5. I did not obtain Mdm4∆2/∆2 p73−/− embryos among analyzed embryos, suggesting that deletion of p73 does not rescue Mdm4 deficient embryonic lethality. Additionally, I performed gene-dosage experiment to determine the contribution of both p53 and p73 to the Mdm4-null phenotype. I analyzed 27 pups at postnatal day 7 and I found that Mdm4∆2/∆2 p53+/− p73+/− or Mdm4∆2/∆2 p53+/− p73−/− mice are not viable. At E11.5 and E9.5, I analyzed 27 and 48 embryos, respectively. At each time point, all Mdm4∆2/∆2 p53+/− embryos, regardless of p73 status, were severely runted with an aberrant developmental phenotype while Mdm4+/∆2 p53+/− embryos had normal morphology. These studies showed that loss of one or both alleles of p73 does not rescue the phenotype of Mdm4∆2/∆2 p53+/− embryos. There is a possibility that, in the absence of Mdm4, the strong phenotypic effect of p53 during embryogenesis might mask the effect of p73 deletion. Therefore, we decided to test this hypothesis in an organ-specific manner. Our lab has previously shown that deletion of Mdm4 in the central nervous system (CNS) is embryonic lethal due to porencephaly at E17.5. This phenotype is rescued by deletion of p53. Moreover, mouse studies have shown that p73 is essential for CNS neurogenesis. Therefore, I asked whether the embryonic lethality of Mdm4 deficiency in the CNS is rescued by p73 loss. I crossed Mdm4fx/fx p73+/− mice with Mdm4+/fx Nes-cre p73+/− mice to generate Mdm4fx/fx Nes-cre p73+/− or Mdm4fx/fx Nes-cre p73−/− embryos. Cre recombinase is expressed under control of Nestin promoter at E10.5, and floxed Mdm4 allele (fx) will be deleted in the CNS of the embryo. I analyzed 79 embryos at E14.5, and among them, I obtained Mdm4fx/fx Nescre p73+/− and Mdm4fx/fx Nes-cre p73−/− embryos in Mendelian ratio. I am currently doing histopathological and transcriptional studies on these samples to examine the extent of rescue and possible mechanism. To investigate the Mdm4 and p73 interaction in tumorigenesis, I am monitoring a cohort of mice to compare the onset of tumorigenesis, tumor spectrum, and especially metastatic potential of Mdm4Tg15 p73+/− mice compared to Mdm4Tg15, p73+/− and wild type mice. Thus far, Mdm4Tg15 p73+/− mice have higher ratio of lymphoma compared to Mdm4Tg15 mice (44.4% and 16.6%, respectively). Interestingly, these mice have a massively disseminated phenotype or a second primary tumor compared to Mdm4Tg15 mice with lymphoma. Combined, this study provides a comprehensive in vivo characterization of potential Mdm4 and p73 interactions both in development and tumorigenesis. Since Mdm4 is overexpressed in many tumors and most tumors lack p73 mutations, this study may identify a novel mechanism for inactivation of p73 in human cancer. This project has been supported by CPRIT and Julia Jones Matthews Cancer Research Scholar Award at MD Anderson Cancer Center. 50 Trainee Symposium on Cancer Research in Texas January 30, 2015 Thakkar, Kaushik – MDACC Poster #44 TRIM24 Links Epigenetics and Metabolism in Breast Cancer Progression Thakkar KN1, Jiang S1, Stratton SA1 and Barton MC1 1 Department of Epigenetics and Molecular Carcinogenesis, UT MD Anderson Cancer Center, The University of Texas Graduate School of Biomedical Sciences at Houston, Program in Genes and Development Corresponding author: 1515 Holcombe Blvd, Department of Epigenetics and Molecular Carcinogenesis, UT MD Anderson Cancer Center, Houston, TX The TRIpartite Motif (TRIM) family of proteins has approximately 70 proteins, which are involved in a variety of processes, including regulation of protein stability, transcription, cell proliferation and apoptosis. Our laboratory discovered TRIM24, as an E3ubiquitin ligase of p53, using embryonic stem cells and breast cancer cell lines as model systems. Earlier reports identified TRIM24 as a transcriptional co-regulator of nuclear receptor signaling, directly interacting with retinoic-acid receptor (RAR), thyroid receptor, androgen receptor and estrogen receptor (ER). Since our laboratory’s initial report showing TRIM24 over expression correlates with poor survival of breast cancer patients, several studies reported roles of TRIM24 in multiple cancers such as NSLC, head and neck carcinoma, hepatocellular carcinoma and glioblastoma. We recently reported that TRIM24 expression is deregulated during breast cancer progression and likely early in the process. In this work, I found that the ectopic expression of TRIM24 in immortalized Human mammary epithelial cells (HMECs) greatly increased cellular proliferation (Fig. 1A & 1B) and induced malignant transformation. Subcutaneous injection of TRIM24-HMECs in nude mice displayed significantly higher xenograft volume as compared to their control counterparts (Fig. 1C). Molecular analysis of TRIM24-HMECs revealed a glycolytic and tricarboxylic acid cycle gene signature (Fig. 1D), alongside increased glucose uptake and activated aerobic glycolysis. Using Chromatin Immunoprecipitation (ChIP), I saw TRIM24 binding at the promoters of several glycolytic and TCA cycle genes such as GLUT1, IDH1, IDH2 and c-Myc Consistent with in vitro findings, the Glucose transport pathway was among the top 10 pathways positively correlated with TRIM24 expression in human breast tumors from the TCGA database. Thus, TRIM24 is co-expressed with genes that regulate glucose metabolism in breast tumors, supporting clinical relevance of our findings. Our studies suggest a unique role for TRIM24 in early steps of mammary carcinogenesis that involves reprogramming of glucose metabolism. Mechanistically, we have previously reported that TRIM24 acts as a transcriptional co-regulator by “reading” a specific signature of histone post-translational modifications (H3K4me0-H3K23ac) via a tandem plant homeodomain (PHD) and bromodomain (Bromo) within the C-terminus of TRIM24. Future studies will determine if chromatin association of TRIM24 via the PHD/Bromo-domain is critical for oncogenic transformation linked with metabolic reprograming in cancer. My developed assay systems and results provide the groundwork to test chemical therapeutics that disrupt TRIM24 functions, e.g. Bromodomain inhibitors and TRIM24 inhibitors. The project was supported by grants from Center for Cancer Epigenetics (MD Anderson) and CPRIT Training Grant 51 Trainee Symposium on Cancer Research in Texas January 30, 2015 Veland, Nicolas – MDACC Poster #45 PRMT6 Overexpression Causes Global Loss of DNA Methylation in Mouse Embryonic Stem Cells Veland N1,2, Di Lorenzo A1,2, Rothbart SB3, Strahl BD3, Bedford MT1,2 and Chen T1,2 1 Department of Epigenetics and Molecular Carcinogenesis, The University of Texas MD Anderson Cancer Center 2 Center for Cancer Epigenetics, The University of Texas MD Anderson Cancer Center 3 Department of Biochemistry, University of North Carolina at Chapel Hill Corresponding author: Chen T, Department of Epigenetics and Molecular Carcinogenesis, The Virginia Harris Cockrell Cancer Research Center at Science Park, The University of Texas MD Anderson Cancer Center. Unit 116, 1808 Park Road 1C – P.O. Box 389, Smithville, Texas 78957. Email: [email protected] Methylation of the C5 position of cytosine is a major epigenetic modification, which is mostly restricted to CpG dinucleotides, and it is present in 60-80% of all CpG sites in the mammalian genome. DNA methylation is essential for mammalian development and plays crucial roles in gene regulation, genomic imprinting, and X-chromosome inactivation. Aberrant DNA methylation patterns are observed in cancer, where cells usually exhibit global hypomethylation and regional hypermethylation. The underlying mechanisms for this dysregulation are largely unknown. DNA methylation patterns are faithfully maintained by DNMT1 during each round of DNA replication. Biochemically, DNMT1 preferentially methylates hemi-methylated CpG dinucleotides. During DNA replication, DNMT1 is recruited to replication foci by the multi-domain protein UHRF1 to methylate the newly synthesized DNA. Moreover, UHRF1 binds to the N-terminal tail of histone H3 and the binding is significantly reduced when arginine 2 of H3 (H3R2) is methylated, suggesting that H3R2 methylation may regulate DNA methylation by altering the recruitment of the UHRF1-DNMT1 complex to replication foci. Interestingly, PRMT6, which mediates H3R2 methylation, is frequently overexpressed in cancer cells. Based on these observations, we hypothesize that PRMT6 acts as a negative regulator of DNMT1 and that abnormal H3R2 methylation caused by PRMT6 upregulation contributes to global DNA hypomethylation in cancer. Results in HEK 293 cells showed that PRMT6 overexpression leads to increased H3R2 methylation, as well as disassociation of UHRF1 from chromatin. To determine whether PRMT6 regulates DNA methylation, we established stable PRMT6-expressing clones in mouse embryonic stem (mES) cells, an ideal system for studying DNA methylation, because they do not require DNA methylation to survive and proliferate (unlike somatic cells). Preliminary data clearly indicate that PRMT6 overexpression indeed results in global loss of DNA methylation in mES cells. Together, these preliminary results suggest that DNMT1 recruitment to chromatin is modulated by H3R2 methylation. Upregulation of PRMT6 may contribute to global DNA hypomethylation in cancer. This project was supported by CPRIT Rising Star awarded to Chen T, Student Scholarship from Center of Cancer Epigenetics M.D. Anderson Cancer Center, and MD Anderson CPRIT Graduate Scholar Award and Julia Jones Matthews Cancer Research Scholar Award awarded to Veland N. 52 Trainee Symposium on Cancer Research in Texas January 30, 2015 Walker, Christopher – MDACC Poster #46 Investigation of Detection Method Bias in Measuring Metabolic Fluxes with Hyperpolarized Magnet Resonance Agents Walker CM1 and Bankson JA1 1 Department of Imaging Physics, The University of Texas MD Anderson Cancer Center Corresponding author: James A. Bankson 1881 East Road Houston, TX 77054, Email: [email protected] Magnetic resonance spectroscopy (MRS), a technique that can detect the chemical makeup of tissue, is generally limited by low signal. A new technique known as hyperpolarization is able to massively increase the MRS signal of select compounds. With hyperpolarized compounds it is possible to detect real time metabolic fluxes, non-invasively. A great deal of interest has been focused on hyperpolarized pyruvate because of altered conversion in cancer known as the Warburg effect. The flux of pyruvate to lactate as measured by hyperpolarized MRS is in stage two clinical trials for prostate cancer. Investigators currently have two parameters to control the detection of hyperpolarized agents by MRS; the flip angle, basically how much of the hyperpolarized compound is sampled, and the repetition time or how often the hyperpolarized compound is sampled. It is not known how those two parameters affect the detection of the metabolic flux. In this study we investigate what affects sequence parameters flip angle and repetition time have on the accuracy of the detected flux of hyperpolarized pyruvate. Using mathematical models we are able to generate synthetic MRS studies of hyperpolarized pyruvate flux. The synthetic data resulting from simulation was processed in the same manner as real world hyperpolarized studies and the quantified metabolic flux was compared underlying fluxes that generated the synthetic data. With these methods we were able to accurately characterize any effect of the detection methods on measured flux under a range of physical, biologic and chemical conditions. Two tissue types were simulated, one with a moderate pyruvate to lactate flux, the other with a high flux. In the low conversion rate simulation accurate measurements was achieved along a band of flip angle and repetition times from two seconds and 20 degrees to seven seconds and 80 degrees. Interestingly, the majority of sequence parameters tested yielded inaccurate measurement of the pyruvate flux giving results with an up to 75% error. A similar trend is seen for the higher flux data; however, the band of accuracy is narrower as well as covering higher flip angles and faster repletion times. These results show that even when performing the most basic MRS studies how much and how frequently the hyperpolarized agents are sampled can drastically change the endpoint of the study, i.e. the detected pyruvate flux. This means that as this technology is moves towards clinical practice investigators will need to keep in mind the way data is acquired to ensure little to no biases are introduced by detection methods. Additionally, comparing hyperpolarized studies with different acquisition parameters or basal conversion rates will need to be done with great care. More advance MRS and imaging strategies of hyperpolarized agents currently under investigation will likely have similar dependencies on acquisition parameters. The simulation environment used in this study will prove to be a powerful tool in characterizing how well and under what conditions these advanced techniques accurately detect metabolic fluxes. Figure 1 Heat maps of percent error in detected pyruvate flux across a range of flip angles and repetition times for low (left) and high (right) flux tissue simulations. This work was supported by a Julia Jones Matthews Cancer Research Scholar training award to [CW] 53 Trainee Symposium on Cancer Research in Texas January 30, 2015 Yang, Dong – MDACC Poster #47 Synergistic Inhibition of PARP and Wee1 Kinase in Triple-Negative Breast Cancer Yang D and Keyomarsi K1 1 Department of Experimental Radiation Oncology, The University of Texas MD Anderson Cancer Center Defined by lack of expression of estrogen receptor (ER), progesterone receptor (PR) and HER2, TNBC represents 15-20% of all breast cancer diagnoses and is prevalent in younger women. In addition to surgery, radiation therapy and chemotherapy, some targeted therapies have been proved effective and become part of standard treatment for certain types of breast cancer dependent on particular biological properties, including inhibition of ER and HER2 signaling. However, hormone and targeted therapies are not available for TNBC patients due to the absence of the three prognostic biomarkers. Even when detected at an early stage, the five-year survival of TNBC tends to be lower than other subtypes of breast cancer. The poor clinical outcomes are due in part to the high recurrence upon treatment and the lack of targeted therapies. Hence, there is an urgent need to develop additional treatment options for TNBC patients based on the underlying biology of TNBC. Several PARP inhibitors are currently under evaluation in multiple breast cancer clinical trials. However they are projected to be limited to patients with BRCA1/2 mutations. Only 10-20% of TNBC patients harbor BRCA1/2 mutations, leaving a vast majority of patients ineligible for PARPi therapy. Since Wee1 kinase inhibitor impairs homologous recombination repair (HRR) and abrogates G2 DNA damage checkpoint, we hypothesize that Wee1 kinase inhibitor can sensitize cancer cells to PARP inhibition through blockage of BER and HRR as well as reduced time for cells to repair DNA. Our initial analysis using high-throughput survival assay (HTSA) in TNBC cell lines supported the hypothesis. The combination treatment of PARP and Wee1 inhibitors were synergistic in most TNBC cell lines we tested, including both BRCA1 wild type and mutant cells. The combination treatment increased sub-G1 and polyploid cell population compared to single drug treatment alone in MDA-MB-231 cells. Consistently, the combination treatment largely increased the amount of apoptotic cells and induced the formation of multinucleated giant cells. We will next interrogate the combination treatment in cell line xenografts and patient derived xenograft (PDX) mouse models. Based on our findings thus far, we propose that the molecular mechanism of synergism between these two classes of pathways is that, blockage of BER by PARP inhibitor and HRR by Wee1 inhibitor, plus inhibition of G2 DNA damage checkpoint by Wee1 inhibitor, accumulate unrepaired DNA damage and eventually lead to cell death through apoptosis. Successful conduct of this proposal by demonstrating the potential of the combination therapy may provide TNBC patients with a targeted treatment avoiding the use of cytotoxic and non-specific chemotherapeutic agents. 54 Trainee Symposium on Cancer Research in Texas January 30, 2015 Donkor, Moses – UT Austin Poster #48 Tumor-Induced Regulation of T Cell Fc Gamma Receptor Expression Donkor MK, Charab W, Triplett K and Georgiou G Department of Chemical Engineering, University of Texas at Austin Corresponding author: George Georgiou, Ph.D. Department of Chemical Engineering, University of Texas at Austin, Austin, Texas, Email: [email protected] The failure of immunosurveillance is linked mechanistically with many factors of which tumor-induced tolerance in T cells is recognized as crucial in many tumor types. Because immunoglobulins of the IgG subclass can induce tumor cell apoptosis and further modulate innate immune responses via Fcγ receptors (FcRs) that bind to the Fc portion of IgGs, IgG-based therapy has become an indispensable modality for cancer treatment. There are 2 kinds of FcγRs: activating receptors that turn on tumor cell-killing machinery of immune cells and inhibitory FcR2B (CD32B) that suppresses tumor cell killing. Although IgGdependent orchestration of tumor-reactive T cell responses has been described, a direct role of FcγRs in the regulation of T cell antitumor function is largely unknown. We detected transcripts of Fcgr2b but not transcripts of other Fcgr by RT-PCR evaluation of FACS-purified mouse T cells. In vitro stimulation of T cells showed that Fcgr2b mRNA and protein co-localization with the T cell marker TCR increased with the concentration of anti-CD3 antibody. To determine whether FcRIIB expression is regulated in vivo, we immunized wild type mice with OVA protein intraperitoneally and found that mesenteric lymph node CD4+ and CD8+ T cells expressed high levels of surface FcRIIB compared to T cells found in other tissues. In addition, OVA-reactive CD8 TcraTCRb transgenic OT-I T cells adoptively transferred into mice immunized with OVA revealed that T-cell induction of FcRIIB is tightly linked with cell division. When assessed for the effect of tumor development on the regulation of T-cell FcRIIB expression, we found that T cells in transplantable B16 melanoma and 4T1 mammary tumors expressed high surface levels of FcRIIB. Significantly, CD4+ and CD8+ T cell induction and activation of inhibitory FcRIIB ITIM in spontaneous autochthonous MMTV-Neu mammary adenocarcinoma were associated with T cell effector function in tumors. These findings reveal a previously unrecognized FcRIIB expression mechanism in T cells that is contingent upon T activation and suggest the possibility that antigen-IgG immune complexes regulate T cell responses similar to other immune cell types. We are currently delineating the function of FcRIIB signaling in the regulation of T cell cancer immunosurveillance. This work is supported by a Cancer Prevention Research Institute of Texas (CPRIT) postdoctoral training grant (MKD) and the Clayton Foundation for Research. 55 Trainee Symposium on Cancer Research in Texas January 30, 2015 Jiang, Yu (Sherry) – UT Austin Poster #49 Robust Strand Exchange Reactions for Sequence-specific Detection of Nucleic Acid Amplicons and its Applications in Point-of-Care Diagnostics Jiang YS2, Du Y1, Bhadra S1, Hughes R1, Li B1, Wu YR3, Milligan JN3 and Ellington AD1* 1 Center for Systems and Synthetic Biology, University of Texas at Austin 2 Department of Chemistry, University of Texas at Austin 3 Department of Molecular Biosciences, University of Texas at Austin Corresponding author: Ellington AD, Center for Systems and Synthetic Biology, University of Texas at Austin, 2500 Speedway, Austin, Texas 78712, United States, Email: [email protected] Loop-mediated isothermal amplification (LAMP) of DNA is a powerful isothermal nucleic acid amplification method that can generate upwards of 109 copies from less than 100 copies of template DNA within an hour. It is potentially a powerful tool to be used in point-of-care diagnostics. Unfortunately, while the amplification reactions are extremely powerful, quantitative detection of LAMP products remains analytically challenging. In order to both improve the specificity of LAMP detection and to make readout simpler and more reliable, we have replaced the intercalating dye typically used for monitoring in real-time fluorescence with a toehold-mediated strand exchange reaction termed one-step strand displacement (OSD). Due to the inherent sequence specificity of toehold-mediated strand exchange, the OSD reporter could successfully distinguish side-products from true amplicons arising from templates corresponding to the biomedically relevant M. tuberculosis RNA polymerase (rpoB) and the melanoma-related biomarker BRAF. OSD allowed rpoB to be detected in a complex mixture such as synthetic sputum, and also demonstrated single nucleotide specificity in detecting a mutant BRAF allele (V600E) in the presence of 20-fold more of the wild-type gene. Real-time detection of different genes in multiplex LAMP reactions also proved possible. The development of simple, readily designed, modular equivalents of TaqMan probes for isothermal amplification reactions should greatly enable point-of-care diagnostics testing. We now present nascent results on engineering OSD for converting amplicons to glucose molecules that can be read by commercial glucometers. Initial results show robust, sequence-specific detection of LAMP amplicons that originate from as few as 20 target copies of the MERS and Ebola viruses, two high profile public health threats. This project was supported by grants Bill and Melinda Gates Foundation, Defense Advanced Research Projects Agency, National Institutes of Health in conjunction with the Boston University, the Cancer Prevention and Research Institute of Texas. 56 Trainee Symposium on Cancer Research in Texas January 30, 2015 Kaoud, Tamer – UT Austin Poster #50 Inhibition of the TRPM7 Kinase Domain Inhibits Breast Cancer Cell Migration and Invasion and Tumor Metastasis Kaoud TS1, Xie X, Park J2, Tavares CD1, Mitra S2, Cano M1, Tseng CC1, Radwan MF3,4, Bartholomeusz C2 and Dalby KN1 1 College of Pharmacy, The University of Texas at Austin 2UT MD Anderson Cancer Center 3 Faculty of Pharmacy, Minia University, Minia, Egypt 4 Faculty of Pharmacy, King Abdulaziz University, Jeddah, Saudi Arabia Corresponding author: Dalby KN, College of Pharmacy, The University of Texas at Austin, 107 West Dean Keaton, BME, TX 78712, USA. Email: [email protected] TRPM7 (transient receptor potential melastatin 7) is a non-selective cation channel fused to protein kinase domain at the C-terminal whose activity is linked to the control of actomyosin contractility. TRPM7 mediates adhesion and migration of breast cancer cells and promotes breast tumor metastasis. The lack of cell-permeable inhibitors of the kinase domain represents a barrier to understand the kinase function. Herein, we discovered the first small molecule (KD-1) that targets TRPM7 kinase activity. Mg2+ starvation, which promotes TRPM7 kinase activity, induces phosphorylation of eEF2. Treatment of Mg2+starved HEK293 cells with KD-1 decreased eEF2 phosphorylation, consistent with TRPM7 kinase activity suppression in-cells. KD-1 decreased the binding of Myosin IIB to TRPM7 in HEK293 and MDA-MB-231 cells. And when MDA-MB-231 cells were treated with increasing doses of KD-1, no change in cell viability was seen. Interestingly, KD-1 inhibited MDA-MB-231 cells migration and invasion that is reportedly regulated by TRPM7 kinase activity. Finally, the bioluminescent signals (to assess metastasis) were significantly lower in KD-1-treated mice (25 and 50 mg/kg/day) than in mice treated with vehicle control (P ≤ 0.05, 2-sided t-test.). in conclusion, inhibition of TRPM7 kinase activity may reduce or block breast tumor progression and/or metastasis. Financial support was from grants from the National Institute of General Medical Sciences and the Welch Foundation for KND. TSK acknowledges Cancer Prevention and Research Institute of Texas Postdoctoral Training Award (RP140108). 57 Trainee Symposium on Cancer Research in Texas January 30, 2015 Laurent, Jon – UT Austin Poster #51 Uncovering Rules Governing Functional Replacement Between Humans and Yeast Laurent, JM1, Kachroo, AH1 and Marcotte, EM1 1 Center for Systems and Synthetic Biology, Department of Molecular Biosciences, University of Texas at Austin Corresponding author: Laurent JM, University of Texas at Austin, 2500 Speedway, MBB 3.128, Austin, TX 78712, Email: [email protected] Owing to its ease of handling and rapid growth, the baker’s yeast Saccharomyces cerevisiae is a popular model organism for studying many aspects of eukaryotic biology. Due to the functional conservation still present between many human and yeast proteins, yeast has served an important role in the study of human cancers, especially in regards to control of the eukaryotic cell cycle and elucidation of DNA repair pathways. Inspired by these studies, our lab has been systematically testing which essential yeast proteins are replaceable by their human counterparts, screening for functionality by the human genes’ ability to rescue growth in the absence of the yeast proteins. To date we have tested nearly 500 pairs of 1-to-1 orthologs between the two species, and have observed a ~50% rate of replaceability. We have further investigated what factors determine a proteins ability to replace by assembling a set of quantitative features of the genes or ortholog pairs, including calculated properties of the genes’ sequences (e.g., gene and protein lengths, sequence similarities, codon usage, and predicted protein aggregation potential) and other properties such as protein interactions, mRNA and protein abundances, transcription and translation rates, and mRNA splicing features. We then quantify how well each feature predicts replaceability. Strikingly, sequence similarity only weakly predicts replaceability. Instead, replaceability depends most strongly on gene modules: genes in the same process tended to be similarly replaceable (e.g., the proteasome or sterol biosynthesis) or not (e.g., DNA replication initiation or ribosomal proteins). Our data demonstrate that a substantial portion of conserved yeast and human genes perform much the same roles in both organisms even after >1 billion years of evolution—to an extent that the proteincoding DNA of a human gene can actually substitute for that of the yeast. The strong pathway-specific pattern of individual replacements suggests that simultaneous group-wise replacement of the genes should be feasible, raising the possibility of humanizing entire cellular processes in yeast. Many of the genes that can be replaced have important roles in human disease, including cancer. Such ‘humanized’ strains would simplify drug discovery against human proteins, enable studies of the consequences of human genetic polymorphisms, and empower functional studies of entire human cellular processes in a simplified organism. This project was supported by CPRIT training grant RP140108. 58 Trainee Symposium on Cancer Research in Texas January 30, 2015 Leung, Wai Chung (Justin) - UT Austin Poster #52 Histone H2A Variants And DNA Repair Leung JW and Miller K1 1 Institute of Cellular and Molecular biology, Department of Molecular Biosciences, The University of Texas at Austin Corresponding author: Kyle Miller, Institute of Cellular and Molecular biology, Department of Molecular Biosciences, The University of Texas at Austin, 2506 Speedway, NMS 4.262, Austin, TX and the Dan L Duncan Cancer Center at Baylor College of Medicine. Houston, TX, Email: [email protected] Histone H2A/H2AX ubiquitination is important for cellular functions, including the DNA damage signaling pathway. However, as the most diverse histone family, the functional significance of the noncanonical H2A variants, H2AZ, macroH2A and H2A.Bbd, has been enigmatic. In our current study, we aim to characterize the role of non-canonical H2A variants in genome stability and DNA damage responses. We have identified an acidic patch domain of H2A/variants, which is essential for ubiquitination and DNA damage signaling. We also have mapped the major ubiquitination acceptor residues of each individual H2A variants, which include novel RNF168 ubiquitination sites on the noncanonical H2A variants. Additionally, in order to study the specific role of H2A variants in DNA repair pathway, we have performed an unbiased proteomic screen using mass spectrometry to define individual H2A-variant interactomes. In our preliminary screening, we identified a list of H2AX, H2AZ, macroH2A and H2A.Bbd specific interacting proteins that are recruited to laser-induced sites of DNA damage. Together, our data provides evidence that non-canonical H2A variants participate in a network that is involved in DNA damage signaling and repair pathway in human cells. The Project is supported by Cancer Prevention Research Institute of Texas (CPRIT) 59 Trainee Symposium on Cancer Research in Texas January 30, 2015 Lim, Byung Joon - UT Austin Poster #53 Improvement of Electrochemical DNA Sensors (E-sensor) for Application to Genetic Testing of Hereditary Cancers Lim BJ1, Du Y2, Li B2, Jiang YS2, Sessler JL1 and Ellington AD1,2 1. Department of Chemistry, The University of Texas at Austin, Austin 2. Institute for Cellular and Molecular Biology, Center for Systems and Synthetic Biology, and Department of Chemistry, The University of Texas at Austin Corresponding authors: Sessler JL, Department of Chemistry, The University of Texas at Austin Ellington AD, Institute for Cellular and Molecular Biology, Center for Systems and Synthetic Biology, and Department of Chemistry, The University of Texas at Austin, 105 E. 24th Street – MS A5300, Austin, Texas 78712-1224, Email: [email protected], [email protected] DNA Genetic testing is a medical process to look for genetic changes or mutations on DNA. It enables the diagnosis of the risk of tumors by find inherited mutation on genes from individual patients. Compared with typical fluorescent gene sensing methods, electrochemical DNA sensors (E-sensor) have many advantages to apply to genetic testing such as high sensitivity, low cost, and easiness for miniaturization and multiplexing. In this study, we have tried to improve some short falls of conventional E-sensor systems including low reproducibility. To overcome this, we developed a ratiometric E-sensor using two different electrochemically active probes, methylene blue (MB) and ferrocene (Fc) (scheme 1). The newly designed ratiometric E-sensor shows dramatically improved reproducibility and sensitivity in sensing of human T-lymphotropic virus type I gene. As an application, we also tried single nucleotide polymorphism (SNP) detection with the new E-sensor, and it shows high discrimination between target and mutated genes. Furthermore, multiplexing gene sensing is being developed as another improvement of E-sensor design. Using the introduction of different functional groups on Fc, six kinds of Fc derivatives (called Fc1 to Fc6, respectively) have been chemically synthesized as electrochemical probes and they displayed diverse oxidation potentials as shown below. Currently, introduction of these probes on real oligonucleotide is being performed by chemical synthesis and further application will follow. We are expecting that the multiplexing electrochemical signals enable to detect various tumor-related genes at the same time. Scheme 1. Scheme of ratiometric E-sensor Figure 1. Multiplexing signals from Fc1 to Fc6 This project was supported by the Welch Foundation, the Gates Foundation, the National Security Science and Engineering Faculty Fellowship, the NIH, and the CPRIT. 60 Trainee Symposium on Cancer Research in Texas January 30, 2015 Montgomery, William - UT Austin Poster #54 Discovery and Development of Potent Anti-Cancer Compound Via a Diverted Total Synthesis Granger BA1, Knezevic CE2, Montgomery WL1, Parkinson EI2, Hergenrother PJ2 and Martin SF1 1 Department of Chemistry and Biochemistry, The University of Texas at Austin 2 Roger Adams Laboratory, Department of Chemistry, University of Illinois at Urbana-Champaign Corresponding author: Martin, SF, Department of Chemistry and Biochemistry, The University of Texas at Austin, 105 E. 24th St. Stop A5300 Austin, TX 78712, Email: [email protected] In the course of the total synthesis of (±)-actinophylic acid, the natural product and several advanced intermediates were screened in vitro for ability to induce cell death in a variety of cancer cell lines. Subsequent derivitization of the biologically active intermediates led to the discovery of low µM cytotoxic compounds that do not induce hemolysis at concentrations up to 100 µM. Furthermore, the timing of cell death was determined to be as fast or faster than FDA approved drugs (staurosporine, doxorubicin, paclitaxel, and mitomycin C), with >95% cell death in only 6 h. Preliminary in vivo studies also show that the compounds are fully tolerated in doses up to 40 mg/kg by healthy mice. Investigation into the mechanism of action found that the cell death is a result of apoptosis that is triggered by endoplasmic reticulum stress in the G1 phase of the cell cycle. Subsequent transcript profiling revealed a distinct signature with no high similarity matches, suggesting that these compounds have a unique biological target. In order to facilitate the transition of a compound to clinical trials, work is currently underway to develop a more thorough structure-activity-relationship, find more potent leads, and to determine the biological target via biotin pull-down studies. This research was supported by the National Institute of Health, the Robert A. Welch Foundation, and the Cancer Prevention Research Institute of Texas. 61 Trainee Symposium on Cancer Research in Texas January 30, 2015 Sammons, Rachel - UT Austin Poster #55 An in vitro System for Discovery of Inhibitors that Target the D-Recruitment Site of ERK Sammons RM1, Kaoud TS2, Devkota AK2, Cho EJ3 and Dalby KN2 1 Department of Biomedical Engineering, Cockrell School of Engineering, The University of Texas at Austin 2 Division of Medicinal Chemistry, College of Pharmacy, The University of Texas at Austin 3 Texas Screening Alliance for Cancer Therapeutics, The University of Texas at Austin Corresponding author: Sammons RM, Department of Biomedical Engineering, Cockrell School of Engineering, University of Texas at Austin, 107 W. Dean Keeton St., BME Building, Stop C0850, Austin, Texas 78712, Email: [email protected] Docking sites on mitogen-activated protein kinases (MAPKs) direct protein binding interactions and generate signal specificity. Therapeutically targeting these sites in cancer offers a way to potentially overcome many issues with ATP-competitive MAPK inhibitor design, such as high cellular ATP concentrations and highly conserved active sites among MAPKs. Extracellular signal-regulated kinases 1/2 (ERK1/2) are MAPKs that have a pro-tumorigenic role in numerous cancers and possess two protein docking sites that are distinct from the active site: the D-recruitment site (DRS) and F-recruitment site (FRS). In this study, we have developed a high-throughput fluorescence anisotropy screening to identify small molecule inhibitors that target the DRS of ERK. DRS inhibitors were detected in the high-throughput screening by using fluorescence anisotropy to measure the ability of over 22,000 small molecules to competitively displace a fluorescent peptide from the DRS of inactive ERK2. The top 100 compounds from this screening were validated with a secondary radioactive kinase assay. In this assay, the ability of each compound to prevent the phosphorylation of a DRS-specific peptide by active ERK2 was assessed using [γ-32P]-labeled ATP substrate to measure enzyme activity. The top nine compounds were characterized by in vitro dose-response assays against wild-type and C159S ERK2, a DRS mutant. Inhibition of ERK2 phosphorylation of FRS substrate and JNK2 phosphorylation of c-Jun was evaluated in vitro to determine compound specificity. The potency and specificity of the compounds were also assessed in EGF-stimulated HEK293T cells for their ability to prevent ERK1/2, JNK1/2, and p38α phosphorylation using Western blot. The compounds were tested for irreversible inhibition in vitro through ERK activity tests after exposure to, and removal of the compounds. Of the nine hit compounds found from the screening, four reversibly inhibited the DRS of ERK2 with in vitro IC50 and KI values under 5 µM, and did not show significant inhibition of the FRS or C159S ERK2. Two of these compounds also reduced ERK activation to basal levels in MAPK pathwaystimulated HEK293T cells at concentrations of 5µM. However, in specificity tests these inhibitors demonstrate ability to also target JNK2 and p38α. The fluorescence anisotropy screening developed here can be used to identify potent inhibitors that target the DRS of ERK. These inhibitors can be used to probe ERK signaling events that are mediated by the DRS and can be further optimized to yield highly selective ERK DRS inhibitors with potential therapeutic value. This work was supported by NIH (GM059802, GM106137, and CA167505), the Welch Foundation (F1390 and F-1691), and CPRIT (RP101501, RP 140648, RP140649, and RP140108). 62 Trainee Symposium on Cancer Research in Texas January 30, 2015 Srivastava, Jaya - UT Austin Poster #56 Twist1 Regulates Keratinocyte Stem Cell Proliferation and Migration and is Required for Skin Tumor Formation Srivastava J1, Rho O1 and DiGiovanni J1,2,† 1 Division of Pharmacology and Toxicology, College of Pharmacy, The University of Texas at Austin 2 Department of Nutritional Sciences, College of Natural Sciences, The University of Texas at Austin † Corresponding author: John DiGiovanni. Dell Pediatric research Institute, The University of Texas at Austin, 1400 Barbara Jordan Blvd, Austin, TX 78723, Email: [email protected] Currently, little is known about the role that Twist1 plays in epithelial carcinogenesis. Twist1 is a basic helix-loop-helix transcription factor and is known to play a role in tumor progression through regulation of genes involved in epithelial-mesenchymal transition. Twist1 is a known transcriptional target of signal transducer and activator of transcription-3 (Stat3). In previous studies, Stat3 was shown to play a role in skin tumor progression in the two-stage skin carcinogenesis model, at least in part, by regulating the levels of Twist1. Recent studies using keratinocyte-specific knockout (KO) of Twist1 indicate that Twist1 regulates keratinocyte proliferation during skin tumor promotion. Western blot analysis of lysates from both Twist1 deficient mouse primary keratinocytes and from epidermis isolated from BK5.Cre x Twist1flox/flox mice indicated that Twist1 KO leads to reduced levels of the cell cycle proteins Cyclin E1, E2F1, and Cdk2 and increased expression of p21 following treatment with 12-O-tetradecanoylphorbol13-acetate (TPA). ChIP analysis of epidermal lysates from BK5.Cre x Twist1flox/flox mice revealed that Twist1 bound to the promoter regions of Cyclin E1, E2F1, and c-Myc, indicating direct transcriptional regulation of these cell cycle regulators. Moreover, Twist1 KO in keratinocytes impeded cell cycle progression by reducing the number of cells that advanced to S-phase. Further analyses have demonstrated that deletion of Twist1 either in primary keratinocytes or in vivo leads to an upregulation of p53, which is responsible for the increased expression of p21 observed in these cells. Keratinocyte specific deletion of Twist1 in vivo suppressed epidermal proliferation induced by TPA treatment compared to that observed in wild-type (WT) mice. To further characterize the potential role of Twist1 during epithelial carcinogenesis, we analyzed the impact of Twist1 deletion in vivo on bulge region keratinocyte stem cells (KSCs). Keratinocyte specific deletion of Twist1 in vivo led to a significant reduction in the number of label-retaining cells as well as the number of α6-integrin+/CD34+ cells in the hair follicles of untreated mice. An ongoing two-stage skin carcinogenesis experiment shows that BK5.Cre x Twist1flox/flox mice have significantly reduced tumor development as observed by decreased tumor multiplicity and delayed latency compared WT mice. Furthermore, tumors that developed in BK5.Cre x Twist1flox/flox mice have significantly reduced size compared to tumors on WT mice. Collectively, these findings suggest that Twist1 has a novel role in epithelial carcinogenesis by regulating proliferation and migration of keratinocytes and KSCs. Research supported by CPRIT RP140108, NIH CA76520, and NIH T32 ES007247. 63 Trainee Symposium on Cancer Research in Texas January 30, 2015 Thiabaud, Gregory - UT Austin Poster #57 Texaphyrin-based Platinum Conjugates: Combining Targeting and Therapy Thiabaud G1, Arambula JF1, Siddik ZH2 and Sessler JL1 1 The University of Texas at Austin 2 The University of Texas, MD Anderson Cancer Center One of the major drawbacks associated with current cancer therapies is a lack of targeting selectivity. Could drugs be delivered specifically into cancer tissues in a targeted manner, it could increase treatment efficiency and decrease side effects. Recently, the gadolinium (III) complex of a water solubilized texaphyrin (motexafin gadolinium; MGd) was demonstrated to have several attributes that are attractive in the context of anticancer drug development. These include i) A novel mechanism of action that leads to apoptosis without disrupting DNA, ii) an ability to localize within cancerous lesions (i.e., distributions of greater than 9:1 relative to surrounding tissues as evidenced by MRI enhancements, fluorescence bioimaging and radiolabeling in vivo), and iii) a high relaxivity that permits easy visualization via MRI.1 This combination of features is near-unique to the texaphyrins and makes them, as a class, attractive for use in creating tumor-targeting drugs. We recently adopted a strategy which consists of conjugating a gadolinium texaphyrin to a platinum(II) drug. Two potential drug leads based on this strategy have been developed and the first results in cancer cells have provided support for the validity of the approach.2,3 To generalize the strategy, we have targeted delivery of a platinum(IV) complex instead of a platinum(II) center. The use of platinum (IV) offers the potential of greater stability and reduced lability due to the presence of a higher oxidation state. On the other hand, the platinum (IV) – MGd conjugate can be activated by reduction of the platinum, releasing a active platinum(II) agent.4 Using this approach, it may prove possible to increase the targeting selectivity as a result of the higher stability of the platinum(IV) center before its reductive activation. The synthesis, characterization, and in vitro studies of a first generation platinum (IV)texaphyrin conjugate will be presented. References: 1 R. A., Miller, K. Woodburn, Q. Fan, M. Renschler, J. L. Sessler, J. A. Koutcher, Int. J. Radiat. Oncol. Biol. Phys. 1999, 45, 981–989. 2 J. F. Arambula, J. L. Sessler, Z. H. Siddik, Bioorg. Med. Chem. Lett., 2011, 21, 1701-1705. [1] 3 J. F. Arambula, J. L. Sessler, Z. H. Siddik, Med. Chem. Commun., 2012, 3, 1275-1281. 4 G. Thiabaud, J. F. Arambula, Z. H. Siddik, J. L. Sessler, Chem. Eur. J. 2014, 20, 8942-8947. 64 Trainee Symposium on Cancer Research in Texas January 30, 2015 Bui, Thanh – UT School of Public Health/MDACC (Program Based at UT Houston) Poster #58 Oral Hygiene, Oral Health, and Oral HPV Infection Bui TC,1 Markham CM,2 Vidrine DJ1 and Mullen PD2 1 Department of Behavioral Science, The University of Texas MD Anderson Cancer Center 2 Department of Health Promotion and Behavioral Sciences, School of Public Health, The University of Texas Health Science Center at Houston Correspondence: Thanh Cong Bui, Department of Behavioral Science, The University of Texas MD Anderson Cancer Center, P.O. Box 301439, Houston, Texas 77230-1439, USA; Email: [email protected], [email protected] It is unknown whether oral hygiene may help to reduce oral Human Papillomavirus (HPV) infection, an established cause of oropharyngeal cancers. One major frame of thinking in (HPV) research is that HPV is often considered a sexually transmitted STI; and hence behavioral means of preventing HPV transmission other than safe sex or vaccination remain unexplored or underemphasized. We break through this typical frame of thinking by using two innovation generation tools: broaden our view of HPV as other viruses in general, and analogy. Specifically, analogical to the effect of washing one's hands often in order to prevent flu transmission, we inquire if oral hygiene, particularly after oral sex, may help to reduce oral HPV infection. The goal of our research is to investigate the effect of hygiene practices on preventing HPV transmission and infection. Examining this effect is very important because hygiene practices are modifiable; and hence this will lead to novel, simple, widely useable, and affordable behavioral initiatives to interrupt the oncogenic HPV transmission cycle. Using data from the 2009–2010 National Health and Nutrition Examination Survey, our study was the first to show that poor oral health was an independent risk factor of oral HPV infection, irrespective of smoking and oral sex practices. This article was highlighted as newsworthy for press release by the American Association of Cancer Research, and it has been covered in >30 media hits in the US (e.g., The New York Times, Bloomberg), and worldwide. This result suggested the role of oral hygiene, which is fundamental to oral health, in preventing oral HPV infection. In a subsequent pilot study which aimed to examine the association between oral hygiene practices and oral HPV infection among 126 high-risk women in Vietnam, our results corroborated the association between measures of poor oral health and oral HPV infection. Additionally, we found that higher unadjusted prevalence of oral HPV infection was associated with two measures of oral hygiene, including lower frequencies of tooth brushing per day and gargling without toothbrushing. Although preliminary, these data suggest that oral hygiene may have a role in oral HPV prevention. Thanh Bui was supported by the UTHealth Innovation for Cancer Prevention Research Postdoctoral Training Program, funded by the Cancer Prevention and Research Institute of Texas (CPRIT, grant number: RP101503, PI: Roberta Ness). Thanh Bui is currently supported by The University of Texas MD Anderson Cancer Center, Duncan Family Institute Junior Mentored Faculty Fellowship. The pilot study was supported by the UTHealth Center for International Training on AIDS Research, funded by the National Institutes of Health, Fogarty International Center, AIDS International Training and Research Program (grant D43 TW007669, PI: Palmer Beasley). 65 Trainee Symposium on Cancer Research in Texas January 30, 2015 Ferrati, Silvia – UT Austin (Program Based at UT Houston) Poster #59 An Innovative and Flexible Approach for Cancer Cell Normalization with Cell-Derived Biovesicles Ferrati S1, Gadok AK2, Klein H1, Smyth H1 and Stachowiak J2 1 Department of Biomedical Engineering, The University of Texas at Austin 2 College of Pharmacy, The University of Texas at Austin Corresponding author: Silvia Ferrati PhD, College of Pharmacy, The University of Texas at Austin, 2409 University Ave., Austin, TX, USA 78712-1113; email: [email protected] Gap junction networks, which are found in nearly every tissue of the body, consist of transmembrane channels composed of connexin proteins that connect the cytoplasm of neighboring cells. Overwhelming evidence supports the role of gap junctions as tumor suppressors. Unfortunately, cancerous cells downregulate the expression of these proteins, enabling rapid cell division, increased epithelial to mesenchymal transition, and metastasis. Reintroduction of connexin genes into tumor cells down-regulates expression of multiple oncogenes, thus helping tumor cells regain control of gene expression and reprogram them away from invasion and proliferation. The goal of our project is to develop lipid based nanoparticles for the delivery of connexin to cancer cells. Our hypothesis is that the particles will integrate themselves into the remaining gap junctions among cancer cells, rebuilding a functional junction network. Because traditional incorporation of functional, properly oriented connexin transmembrane proteins in synthetic therapeutic particles is technically challenging, we pursued a novel approach. Biovesicles were directly extracted from the plasma membrane of healthy donor cells that express a high concentration of functional connexin, thus avoiding the need for further functionalization or chemical modification of the particles. The ability of the biovesicles to normalize cancer cells by rebuilding gap junction networks was tested in vitro by measuring the trans-epithelial electric resistance (TEER) of cancer cells monolayers. A549 epithelial lung cells, possessing dysfunctional connexin, were chosen as the cellular model. Our preliminary data demonstrates that the harvested biovesicles were able to reinforce the gap junction network among A549 cells, leading to increased TEER. These encouraging data support the feasibility of producing drug delivery particles from donor cells by blebbing their membrane and employing them as a novel therapeutic vector for cancer cell normalization. This innovative and flexible strategy will potentially allow for the production and customization of novel biovesicles from specifically pre-engineered donor cells, with broad applicability for drug delivery investigations and therapeutics development. The UTHealth Innovation for Cancer Prevention Research Training Program postdoctoral fellowship funded by the Cancer Prevention and Research Institute of Texas (CPRIT, grant number: RP101503, PI: Roberta Ness) for financial support for Dr. Ferrati. 66 Trainee Symposium on Cancer Research in Texas January 30, 2015 Li, Zhenlong –UT Houston Poster #60 A Multiscale Computational Platform for the Rational Design of Folate-Receptor-specific Drug Nanocarriers Li ZL and Gorfe AA1 1 Department of Integrative Biology and Pharmacology, The University of Texas Health Science Center at Houston Nanoscale targeted drug delivery systems, such as ligand-conjugated drug nanocarriers, are promising solutions to reduce the toxic side effects and enhance the therapeutic efficacy of diagnostic/therapeutic anticancer agents. Among various cell-surface targets, human folate receptors (FRs) represent one of the most intensively-explored targeting systems. However, clinical applications of FR-targeted nanocarriers are severely hindered by poor tissue selectivity and tumor permeability. Solutions to both problems are curtailed by lack of structural information about FRs in physiological conditions, hence difficulties in characterizing the specific nanocarrier-cell interaction at the molecular level. In this project we are developing a novel multiscale computational platform that is capable of characterizing nanocarrier-cell interactions and systematically optimizing nanocarrier structural parameters. Specifically, using mesoscopic Dissipative Particle Dynamics (DPD) simulations, we have revealed the general dynamic characteristics of receptor-mediated membrane adhesion of a typical nanocarrier. This enabled us to propose a comprehensive molecular model describing the cellular interactions of FR-targeted nanocarriers by accounting for the structural dynamics of FRs and their membrane environment. Working towards this model, we have unveiled the conformational dynamics of two FR isoforms (FOL1 and FOL2) using principal component and protein structure network analyses of data from atomistic Molecular Dynamics (MD) simulations. Expanding our efforts, we are further employing coarse-grained MD simulations to uncover the spatiotemporal organization of FRs in heterogeneous lipid membranes. Outcomes from both atomistic and coarse-grained MD simulations will be used to investigate the cellular interactions of FR-targeted nanocarriers using DPD simulations. The combined results will establish the molecular basis for optimization of FR-targeted nanocarriers and aid experimental design. The multiscale computational strategy developed in this research will be applicable to other cell targeting systems. The UTHealth Innovation for Cancer Prevention Research Training Program post-doctoral fellowship funded by the Cancer Prevention and Research Institute of Texas (CPRIT, grant number: RP101503, PI: Roberta Ness) funded Dr. Li; the Texas Advanced Computing Center (TACC) supported computational resources. 67 Trainee Symposium on Cancer Research in Texas January 30, 2015 Zhou, Renke – UT School of Public Health/BCM (Program Based at UT Houston) Poster #61 A Innovative Flexible Semiparametric Joint Models for Semi-competing Risks Data with Missing Cause of Informative Terminal Event Zhou R1, Zhu H2, Bondy M1 and Ning J3 1 Duncan Cancer Center, Baylor College of Medicine, Houston; Division of Biostatistics, The University of Texas School of Public Health, Houston 2 Division of Biostatistics, Department of Clinical Sciences, The University of Texas Southwestern Medical Center 3 Department of Biostatistics, The University of Texas MD Anderson Cancer Center Corresponding author: Zhou, R, Duncan Cancer Center, Baylor College of Medicine, Houston, TX 77030, USA, Email: [email protected] Understanding disease process on cancer-related health outcomes has attracted intense clinical, epidemiologic and translational research interest. Despite a high level of research activity on cancer-related outcomes, several critical questions remain unresolved, partly due to lack of appropriate statistical analysis methods addressing data structure and study design. One challenge in analyzing such data is that death dependently censors cancer progression (e.g., recurrence), whereas progression does not censor death. We dealt with the dependent censoring by first selecting a suitable copula model through an exploratory diagnostic approach and then developing an inference procedure to simultaneously estimate the marginal survival function of cancer relapse and an association parameter in the copula model. The additional challenge is missing cause of death and unreliable cause of death information. Therefore, an immediate question is how to analyze the semi-competing risks data in presence of uncertain type of censoring due to missing causes of death. We adopted a novel Expectation–Maximization (EM) algorithm to account for such uncertainty. We showed that the proposed estimators possess consistency and weak convergence, and use simulation studies to evaluate their finite sample performance. The proposed methods were applied to a retrospective cohort study of women diagnosed with American Joint Committee on Cancer pathologic stage I or II breast cancer who were treated at The University of Texas MD Anderson Cancer Center between January 1, 1985 and December 31, 2000. The UTHealth Innovation for Cancer Prevention Research Training Program pre-doctoral fellowship funded by the Cancer Prevention and Research Institute of Texas (CPRIT, grant number: RP101503, PI: Roberta Ness) provided partial support for Dr. Zhou. 68 Trainee Symposium on Cancer Research in Texas January 30, 2015 Araujo, Patricia - UT San Antonio Poster #62 Dissecting Musashi1 Role in Chromatin Remodeling and its Consequences to DNA Repair and Radio-Resistance Araujo PR1, Silva AE1, Tonapi SS1, Uren PJ2, Smith AD2, Bishop AJ1, and Penalva LOF1 1 Department of Cellular and Structural Biology, UT Health Science Center at San Antonio 2 Molecular and Computational Biology Section, Division of Biological Sciences at University of Southern California Corresponding author: Araujo PR, Greehey Children's Cancer Research Institute – UTHSCSA 8403 Floyd Curl Drive, San Antonio, TX 78229, Email: [email protected] Musashi1 (Msi1) is an evolutionarily conserved RBP with well establish roles in neuronal stem cell selfrenewal and differentiation. In the context of tumorigenesis, Msi1 has been described to be highly expressed in gliomas, medulloblastoma, lung, colon and breast cancers and linked to a poor prognosis. Its expression is required for tumor maintenance and possibly initiation. We recently established that a decrease in Msi1 expression makes glioblastoma cells more radio-sensitive: Msi1 KD cells subjected to ionizing radiation showed reduced proliferation as evident from colony formation assays; moreover, an increase in DNA damage was observed via comet assay, γ-H2AX expression levels and 53BP1 ionizing radiation induced foci. Msi1 has been always defined as a translation regulator. However, our recent data shows that Msi1 is often present in the nucleus of tumor samples and cell lines. Moreover, CLIP assays identified a large number of Msi1 interacting sites in intronic regions. We suggest then that Msi1 displays additional regulatory functions in the nucleus that could equally contribute to its role in tumorigenesis. Supporting this idea, among Msi1’s protein partners identified by yeast two-hybrid and mass spectrometry are several chromatin remodeling proteins including Ino80C. The Ino80 complex has been associated with transcription factors involved in cell proliferation, differentiation and embryonic development. Moreover, Ino80 interacts directly with γH2AX during DNA double-strand break repair. We are currently conducting experiments to explore the hypothesis that Msi1 influences radio-resistance and DSB repair via its association with the chromatin remodeling machinery. Funding sources: CPRIT training grant (RP 140105). 69 Trainee Symposium on Cancer Research in Texas January 30, 2015 Ashcraft, Keith - UT San Antonio Poster #63 Deletions in the Long Arm of Chromosome 18 in Metastatic Prostate Cancer Ashcraft KA, Wilson D, Johnson-Pais TL, and Leach RJ University of Texas Health Science Center San Antonio Corresponding author: Robin J. Leach, Ph.D., Department of Cellular and Structural Biology, University of Texas Health Science Center SA, 7703 Floyd Curl Drive, San Antonio, TX 78229-3900, Email: [email protected] Prostate cancer is the most common form of non-cutaneous malignancy in American men. According to the American Cancer Society, there will be 233,000 new cases in 2014 with 30,000 deaths. Prostate cancer is unlike most cancers in that it grows very slowly. While serum prostate specific antigen (PSA) screenings have been very useful in the detection of prostate cancer, many prostate cancers never progress while others become more aggressive and metastasize. Since cancer is a genetic disease, chromosomal instability is known to contribute to its initiation and progression. Regions of loss on certain chromosomes have been shown to contribute to prostate cancer, and in many cases, a putative tumor or metastasis suppressor gene has been identified within these regions. Previous publications studying allelic imbalance on chromosome 18 found two regions of loss that correlate strongly with prostate cancer metastasis. The genes that are targeted by these chromosomal alterations, however, have remained elusive. Our objective is to confirm the association between chromosome 18 loss and aggressive prostate cancer while also narrowing the critical regions of loss. Using this information, we will identify putative metastasis suppressor genes located within these regions. To confirm the relationship between the loss of chromosomal region 18q with metastatic prostate cancer, we will utilize a unique set of prostatectomy specimens. The specimens are from men with prostate cancer who have been followed for a minimum of five years with an outcome categorized as either No Evidence of Disease (NED) or Metastasis. Using an Agilent custom Comparative Genomic Hybridization (CGH) array, we have tiled DNA oligos across 18q at very high density. By analyzing the copy number variation of chromosome 18 in these patient samples, we can evaluate loss of 18q with outcome of prostate cancer. We will also be able to narrow the critical region containing the important genes driving this relationship. In an effort to find a putative metastasis suppressor gene, we identified the genes located on the previously published region of loss of chromosome 18. One promising candidate gene, Collagen and Calcium Binding EGF domains 1 (CCBE1), is down regulated in cancer samples within both The Cancer Genome Atlas (TGCA) database as well as our own primary patient samples. Furthermore, CCBE1 RNA levels are significantly down regulated in four prostate cancer cell lines when compared to primary prostate epithelial cells (Fig.1). Loss of CCBE1 has also been implicated in both breast and ovarian cancer. We are currently introducing the CCBE1 gene into prostate cancer cell lines using an inducible Tet-On promoter. We will then test CCBE1’s effect on cancer cell migration using the Transwell migration assay. Our preliminary data together with the fact that CCBE1 falls within a region of loss previously reported to correlate with metastasis makes CCBE1 a very promising candidate gene. These studies will help determine the relationship between loss of chromosomal region 18q and prostate cancer outcome. This project was supported by grants: CPRIT (RP140105) and U01CA86402. 70 Trainee Symposium on Cancer Research in Texas January 30, 2015 Chang, Katherine - UT San Antonio Poster #64 Targeting the CD44/EMT Phenotype for Improving Response to Therapy in Pancreatic Cancer Models Chang K1, Chen C1, Zhao S1, Cao L1 and Freeman JW1 1 University of Texas Health Science Center at San Antonio, Department of Medicine, Division of Hematology/Oncology Corresponding author: Freeman JW, University of Texas Health Science Center at San Antonio, Department of Medicine, Division of Hematology/Oncology, 7703 Floyd Curl Dr, Rm 5.214S, San Antonio, TX 78229-3900, Email: [email protected] Heterogeneity of cancer cells within tumors and established cancer cell lines is well known. The impact of this heterogeneity in regards to pathobiology and therapy has not been fully investigated. We found that CD44, a non-kinase transmembrane receptor that binds hyaluronan (HA), shows a varied level of expression in pancreatic ductal adenocarcinoma (PDAC) cell lines. Here we investigated the contribution of CD44 expression on cellular phenotype, tumorigenic properties and response to gemcitabine. In this preliminary study, cells from CFPAC-1 were sorted, by flow cytometry, on the basis of their expression level of CD44 and these cells are referred to as CD44high and CD44low. CD44high cells were characterized for changes associated with epithelial to mesenchymal transition (EMT), including changes in morphology, loss of E-cadherin and increased expression of vimentin. In vitro tumorigenic properties were characterized by invasion assays, sensitivity to gemcitabine, and ability to form tumorspheres. CD44 isoforms were determined by RT-PCR. CD44 expression was knocked down using lentiviral shRNA construct. An orthotopic mouse model was used to determine whether CD44high/EMT cells showed increased tumorigenic properties, including tumor growth and desmoplasia. CD44high cells expressed high levels of predominantly the CD44 standard isoform (CD44s); whereas, CD44low expressed mainly CD44 variant isoform (CD44v). In vitro studies showed that CD44high/EMT cells were more invasive, showed a reduced response to gemcitabine, and formed tumorspheres. Knock down of CD44 expression in CD44high cells reversed the EMT phenotype, decreased invasion and increased response to gemcitabine. In a preliminary study, tumors from mice implanted with CD44high/EMT cells tended to be larger than tumors with more desmoplasia compared to mice implanted with CD44low cells. A follow-up study using larger numbers of mice confirmed that tumors from mice implanted with CD44high/EMT cells grew more rapidly and developed more metastases to the liver when compared to the mice implanted with CD44low cells. CD44high cells expressed predominantly a CD44s isoform, displayed an EMT phenotype, and were more highly invasive in vitro, with increased resistance to gemcitabine. Mice implanted with CD44high cells grew more rapidly in vivo, displayed a greater amount of desmoplasia and showed rapid metastasis to the liver. This study suggests that targeting CD44 will reduce the metastatic phenotype and improved response to therapy in pancreatic cancer. Further studies that combine targeting CD44 with chemotherapy are needed to determine whether this strategy will improve patient survival. The project was supported by CPRIT training grant (RP 140105). 71 Trainee Symposium on Cancer Research in Texas January 30, 2015 Deng, Yilun - UT San Antonio Poster #65 The Tumor Suppressor TMEM127 is a Lysosomal Protein that Associates with LAMTOR1, a Component of the mTORC1 Activation Complex at Lysosomal Surface Deng Y1, Qin Y2, Srikantan S2, Luo A2 and Dahia PLM2 1 Department of Cellular Structural Biology 2 Department of Medicine, University of Texas Health Science Center at San Antonio Corresponding author: Dahia P.L.M, Dept. of Medicine, Cancer Therapy and Research Center, University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, Texas, 78229, USA, E-mail: [email protected] We previously identified TMEM127 as a tumor suppressor gene involved in susceptibility to pheochromocytomas and renal cell carcinomas. However, little is known about the function of this gene. Using cell lines depleted for TMEM127 and mouse embryonic fibroblast (MEFs) derived from a Tmem127 knockout (KO) mouse model, we recently reported that loss of TMEM127 leads to disrupted early-to-late endosomal progression, implicating TMEM127 in endosomal function. In addition, loss of TMEM127 in vivo and in vitro leads to activation of mTOR downstream targets by unknown mechanisms. Here we examined the role of TMEM127 in late endosome/lysosome, which is recognized as an important platform for activation of the amino acid-dependent mTOR complex mTORC1. Using confocal microscopy we found that late endosomal and lysosomal vesicles are larger and more numerous in KO MEFs compared with wild-type (WT) MEFs. Moreover, mRNA and protein levels of several lysosomal genes are upregulated in Tmem127 KO MEFs. Furthermore, TMEM127 extensive co-localization with lysosomal marker LAMP1 is decreased under conditions of cell starvation, suggesting that TMEM127 is dynamically associated with the lysosome. Next, we investigated whether the actions of TMEM127 on the lysosome affect mTOR signaling. It has been established that amino-acid dependent activation of mTOR relies on its coupling to the lysosomal surface. This process involves the assembly of a multi-protein complex, including Rag GTPases, five anchor proteins known as the Ragulator (LAMTOR1-5), and the vacuolar ATPase complex (v-ATPase). We found higher degree of colocalization between mTOR and LAMP2, a lysosomal protein that labels the amino-acid activated mTORC1 complex, in Tmem127 KO MEFs when compared to WT cells. These findings are consistent with an increase in the proportion of lysosomal-bound, active mTOR in the absence of Tmem127. Tmem127 KO MEFs showed higher levels of endogenous LAMTOR1 and vATPase protein than WT MEFs, suggesting that TMEM127 may affect the function and/or stability of these proteins. We further examined the interaction between TMEM127 and the mTOR lysosomal complex by coimmunoprecipitation in 293T cells, and we are able to detect endogenous TMEM127 in the same complex with LAMTOR1/P18, the Ragulator component that functions as a scaffold to this multi-protein machinery. Next, we began to examine the biological relevance of this association. We found that the binding between RagB and RagC in the Rag GTPase complex is not affected by overexpression of TMEM127 under nutrient-full conditions. The effect of TMEM127 expression or absence on the binding/assembly of other components of the mTOR activation machinery under conditions known to modulate mTORC1 signaling is currently being investigated. Taken together, these results suggest that TMEM127 is a lysosomal protein that influences the availability of mTORC1 linked to the lysosome possibly through its association with the Ragulator complex. This work builds on a growing appreciation of the role played by the endolysosome in cancer signaling. This study is supported by CPRIT pre-doctoral training grant (RP140105) and CPRIT individual investigator grant to PLMD (RP110202/140473) 72 Trainee Symposium on Cancer Research in Texas January 30, 2015 Grimes, Allison - UT San Antonio Poster #66 Genetic Markers for Chemotherapy-Related Pancreatitis in Childhood ALL Grimes A1,2, Aguilar MPH1,2, Bansal H2, Chen Y2,3 and Tomlinson G1,2 1 Dept of Pediatrics UT Health Science Center San Antonio 2 Greehey Children’s Cancer Research Institute, San Antonio 3 Dept of Epidemiology & Biostatistics UT Health Science Center San Antonio Corresponding author: Grimes A MD, Dept of Pediatrics UT Health Science Center San Antonio 7703 Floyd Curl Drive, San Antonio, Texas 78229, Email: [email protected] Although ninety percent of children with standard risk acute lymphoblastic leukemia (ALL) remain disease-free five years after treatment, quality of life is greatly affected by treatment-related toxicities. Acute pancreatitis is one of the more severe toxicities, limiting further use of a critical chemotherapeutic agent and delaying subsequent therapy. Host factors, including genetic variability, play a significant role in susceptibility to certain toxicities and are thought to contribute to treatment-related pancreatitis. However, gene studies to evaluate risk for pancreatitis in childhood leukemia have not been performed. The development of pancreatitis during leukemia treatment is associated with L-asparaginase (L-Asp), a key medication in achieving and sustaining remission. Pancreatitis is the most common cause of L-Asp intolerance, noted in 5-18% of children during leukemia therapy. The incidence at our institution falls within this range with 8% developing L-asparaginaseinduced pancreatitis, which can result in prolonged hospitalization, delays in cancer treatment, and numerous complications including secondary diabetes, pancreatic pseudocysts, abscess formation and death. Much speculation exists regarding which underlying genetic factors confer greater risk of developing treatment-related pancreatitis. Children with ALL who develop pancreatitis may harbor mutations in one or more genes, such as those associated with familial, chronic, or idiopathic pancreatitis or those involved in asparagine metabolism which increase vulnerability. Therefore, the overall hypothesis is that underlying gene alterations in some children with ALL lead to acute pancreatitis when exposed to L-Asp therapy. The goals of this project are to determine if mutations in known pancreatitis genes are expressed in leukemic children who develop pancreatitis. In addition, we aim to evaluate pharmacogenomics in pancreatitis risk by isolating mutations in genes involved in asparagine metabolism among children with ALL who develop pancreatitis. Finally, we will explore whole exome sequencing to identify mutations in other genes present in leukemic children with pancreatitis. Blood samples from children with ALL with and without pancreatitis during therapy were evaluated for isolation of total DNA. A custom 18-gene DNA-sequencing panel was designed to include the four genes identified in familial, chronic, recurrent, and idiopathic pancreatitis as well as genes involved in asparagine metabolism. Mutation analysis from pancreatitis-affected children with ALL was compared to leukemic children without pancreatitis. We recognize that alterations in genes which have not been previously investigated may be present in the subset of ALL children who develop pancreatitis, increasing individual susceptibility to this adverse treatment effect. Therefore, using DNA isolated from blood samples from leukemic children, we will use whole exome-sequencing techniques to compare the gene alteration profiles in children who developed pancreatitis versus those who did not. We expect to find mutations in one or several of genes among leukemic children who developed pancreatitis compared to leukemic controls without pancreatitis. Analysis of sequencing data is complete on the first twenty subjects with data on the remaining forty subjects anticipated in early 2015. Thus far, a greater incidence of gene alterations in one of the asparaginase metabolism genes has been demonstrated among case subjects with pancreatitis over treatment-matched controls. These findings, however, are preliminary and results from the remaining subjects are required to determine if this finding persists. Identifying mutations correlated with pancreatitis in childhood ALL will provide an innovative screening method to identify those at greatest risk for pancreatitis prior to treatment. Such knowledge would greatly impact supportive care, early interventions, and alter therapeutic protocols according to toxicity risk. CPRIT Training Grant RP 140105 and Hyundai Hope on Wheels Scholar Grant for support of this project. 73 Trainee Symposium on Cancer Research in Texas January 30, 2015 Hsu, Ya-Ting - UT San Antonio Poster #67 Co-regulation of EpICD/Lef-1 Gene Targets Correlate with Tumor Progression in Advanced Endometrial Cancer Hsu YT1, Wang Y1, Huang YW2, Jadhav RR1, Ruan J3, Jin VX1, Kirma NB 1and Huang TH1 1 Department of Molecular Medicine, Institute of Biotechnology, University of Texas Health Science Center at San Antonio 2 Department of Obstetrics and Gynecology, Medical College of Wisconsin 3 Department of Computer Science, University of Texas at San Antonio Corresponding author: Huang TH, Dept. of Molecular Medicine, Institute of Biotechnology, University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, TX 78229, Email: [email protected] Endometrial cancer is one of the most common malignancies in women. Although the majority of patients are diagnosed in their early stages, 15 to 20 % of patients exhibit high grade and aggressive tumors. This group of patients also has higher chances of developing recurrent tumors after their primary diagnosis, and their survival outcome is less than 30% after 5 years of initial treatment. However the mechanism leads to advanced tumor is still unclear. It has been shown that epidermal growth factor receptor (EGFR) signaling pathway plays major roles in promoting aggressiveness in many tumors. Our in vitro models suggested that EGFR activation may contribute to aggressive malignant growth and epithelial-mesenchymal-transition (EMT). To further explore how EGFR activation leads to tumor aggressiveness, we have focused on EGFR activation of the epithelial cell adhesion molecule EpCAM, which is overexpressed in metastatic tumor cells. Our data suggested that EGFR activation may trigger intra-membrane proteolysis of EpCAM and release its Cterminal fragment EpICD to translocate into the nucleus, leading to transcriptional activation of invasion related genes. Because our previous studies suggested that EpICD forms a transcriptional complex with Lef-1, we hypothesized that WNT/TCF signaling pathway may be involved in EGFR-activated tumor progression. Based on this line of studies, we showed that EpICD together with Lef-1 may play critical roles regulating genes that may function in promoting tumor progression in endometrial cancer. By using chromatin-immunoprecipitation sequencing (ChIP-Seq) we identified genes that are targeted by Lef-1 and EpICD in the presence of activated EGFR. Pathway enrichment analysis revealed that these genes are highly involved in oncogenic pathways (e.g., MAPK/AKT and p53), cell cycle (apoptosis and nucleotide excision repair) and cell mobility (adhesion and tight junctions). We further investigated the correlation of EpICD/Lef-1 binding with gene expression and showed genes with co-binding of EpICD/Lef-1 in their transcription start site (TSS) exhibited up-regulation corresponding to EGFR activation, suggesting EpICD and Lef-1 act as co-regulators to activate the gene expression. With the advantage of CRISPR/Cas 9 gene editing technique we generated EpCAM knockout system to discover the detail mechanism. Our initial data suggested that without the presence of EpCAM, these genes showed reduced level of upregulation and Lef-1 occupancies by EGF stimulation. Also the EpCAM knockout cells exhibited lower invasion ability compared to control cancer cells and these data indicating EpICD/Lef-1 complex is required for regulating gene expression and responsible for promoting invasion in vitro. By using The Cancer Genome Atlas (TCGA), we identified several EpICD/Lef-1 target genes that exhibited differential expression in endometrial cancer patients. Also we discovered in the lesions of mouse endometrium showed higher nuclear expression of Lef-1 and EpICD compared to normal endometrium, supporting our hypothesis that Lef-1 and EpICD play major roles in endometrial cancer and the expression of their target genes are correlated with tumor progression. This project was supported by CPRIT training grant (RP 140105) and NIH (R01 CA172279). 74 Trainee Symposium on Cancer Research in Texas January 30, 2015 Kim, Sun - UT San Antonio Poster #68 Biophysical Study of TGF-β: Betaglycan Interaction for Determining the Mechanism of Betaglycan’s Potentiation of TGF-β Signaling Kim K and Hinck AP1 1 Department of Biochemistry, UT Health Science at San Antonio Corresponding author: Kim K, Dept. of Biochemistry, University of Texas Health Science at San Antonio, Allied Health Building, Room 5.410, Tel: (210) 567-8787, Email:[email protected] Transforming Growth Factor-β (TGF-β) isoforms -1, -2, and -3 are pleiotropic cytokines that regulate normal tissue development and homeostasis, such as cell proliferation, cell differentiation, and expression of matrix proteins. Dysregulation of this important and ubiquitous pathway is implicated in diseases such as cancer and fibrosis, where the diseased cells use TGF-β signaling to their own advantage by altering the cell context. Many cancer cells, for example, dysregulate their cell cycle, which antagonizes TGF-β’s ability to inhibit cell growth. However, TGF-β’s tumor promoting activities remain intact in these cells, which include immune system suppression, Epithelial to Mesenchymal Transition (EMT), and angiogenesis. This, coupled with the overexpression of TGF-β in most cancer cells, drives the growth and invasiveness of many cancers, including those of the breast, brain, and prostate. Our study focuses on understanding the mechanisms of TGF-β receptor assembly and how it is regulated. Betaglycan, also known as the TGF-β type III receptor, is a co-receptor that increases TGF-β2 ligand access to its signaling receptor TβRII, which potentiates TGF-β2 signaling. However, the mechanism by which betaglycan promotes receptor complex and signaling is completely lacking, hindering efforts to develop new therapeutic agents that target TGF-β. In order to understand this mechanism, we are using X-ray crystallography to resolve the atomic structure of TGF-β:betaglycan interaction. Previous Surface Plasmon Resonance (SPR) studies showed that full-length betaglycan binds TGF-β with 1:1 stoichiometry, while allowing one TβRII molecule to be bound to TGF-β. We have isolated the TGF-β3:TβRII:betaglycan orphan domain (BGO) ternary complex (TC), and was able to optimize relatively large well-ordered crystals by trypsinization of the ternary complex. N-terminal sequencing revealed that TGF-β3 was cleaved by trypsin at two specific sites, BGO was cleaved at one site, while TβRII remained intact. We have found that these crystals diffract to 6.5Å. Trypsinization of individual components of the ternary complex showed that TGF-β3 cleavage produces morphologically same crystals as the TGFβ-3 and BGO cleaved crystals, while BGO cleavage does not produce crystals. This suggested that TGF-β3 cleavage is responsible for the improved crystallization. To increase homogeneity of the complex upon cleavage by trypsin, we created TGF-β3 and BGO mutants that will allow typsin to cleave at targeted residues within the TC. Atomic structure will lead us to testing and fully understanding betaglycan’s mechanism in potentiation of TGF-β signaling, which will allow us to pursue novel therapeutic strategies to target TGF-β signaling in disease states. This project is supported by the CPRIT (RP140105) training grant awarded to Kate Kim. In addition, this research is supported by grants from the NIH (GM58670 awarded to A.P.H., CA172886 awarded to A.P.H and L-Z.S., and CA54174 awarded to the U. Texas HSC San Antonio Cancer Therapy and Research Center), the Robert A. Welch Foundation (AQ-1842 awarded to A.P.H.), and the Cancer Prevention and Research Institute in Texas (RP120867 awarded to A.P.H). 75 Trainee Symposium on Cancer Research in Texas January 30, 2015 Krishnan, Samaya - UT San Antonio Poster #69 Significance of Proto-Oncogene PELP1 in Error-Prone Alternative-NHEJ Pathway Krishnan SR, Nair BC, Sareddy GR, Mann M and Vadlamudi RK Corresponding author: Dr. Ratna K Vadlamudi, University of Texas Health Science Center at San Antonio, Department of OB-GYN, 7703 Floyd Curl Drive, San Antonio, Texas, 78229 Estrogen Receptor (ESR1) positive breast cancer is commonly treated with hormonal therapy; however, most patients acquire therapy resistance over a period of time, and this represents a major clinical problem. Recent evidence suggests that the accumulation of chromosomal abnormalities via increased error-prone alternative-non-homologous end joining (A-NHEJ) DNA repair pathway could contribute to breast cancer therapy resistance. However, the mechanism of A-NHEJ regulation by the ESR1 pathway is unknown. Proline, Glutamic acid, and Leucine rich Protein-1 (PELP1), an oncogenic co-regulator of ESR1, is commonly overexpressed in breast cancer, and its deregulation has been implicated in therapy resistance. In this study, we discovered that ESR1 co-regulator PELP1 is phosphorylated at Ser1033 by DNA damage response kinases and phosphorylated PELP1 co-localizes with the -H2AX foci. Using Homologous Recombination (HR) and NHEJ pathway specific reporter assays, we found that PELP1 knockdown decreases NHEJ, with no apparent effect on the HR pathway. Further, using A-NHEJ specific reporter cell line, we found that PELP1 knockdown decreased the frequency of A-NHEJ, while PELP1 overexpression increased the repair frequency. More importantly, metaphase chromosome spreads revealed gross chromosomal abnormalities in PELP1 deregulated cells. Mechanistic studies revealed that PELP1 interacts with Mre11, an exonuclease important for A-NHEJ, and modulates the degree of end resection at the DNA double strand breaks. Mapping studies identified C-terminus of PELP1 as the binding site for Mre11. We rationally screened a random peptide library composed of 10 million peptides and identified a tight-binding peptide to the C-terminus of PELP1 that interferes with PELP1-Mre11 interaction. Using emerging stapled peptide technology, we converted this peptide as a stable cell permeable peptide inhibitor of PELP1 (sPIP3). sPIP3 bound to PELP1 with high affinity, interfered PELP1-Mre11 interaction, reduced A-NHEJ repair frequency and significantly reduced frequency of chromosomal translocations. sPIP3 functioned as a potent cytotoxic agent for breast cancer cells with little activity on normal cells. Further, sPIP3 efficiently inhibited the A-NHEJ pathway and significantly, reduced survival and promoted apoptosis of hormonal therapy resistant breast cancer model cells. Collectively, these findings suggest that ESR1 co-regulator PELP1 plays a critical role in the A-NHEJ pathway and sPIP3 represents a novel drug for the treatment of therapy resistant breast cancer. Funded by CPRIT training grant (RP 140105) 76 Trainee Symposium on Cancer Research in Texas January 30, 2015 Liu, Jun - UT San Antonio Poster #70 An in vitro Inflammation-Induced Acinar-to-Ductal Metaplasia Model for Primary Human Pancreatic Acinar Cells Liu J, Akanuma N and Wang P1 1 Department of Cellular and Structural Biology, The University of Texas Health Science Center at San Antonio Corresponding author: Pei Wang, Department of Cellular and Structural Biology, The University of Texas Health Science Center at San Antonio, San Antonio, TX 78229 Pancreatic ductal carcinoma (PDAC) is one of the most deadly human malignancies, largely due to the lack of methods for early diagnosis and intervention. Development of human PDAC tumorigenesis model would be critical to design new strategies for improving PDAC diagnosis, prevention and treatment. Pancreatitis-induced acinar-to-ductal metaplasia (ADM) has been found to be a key event for KRAS-driven tumorigenesis in PDAC development. Thus, understanding the molecular mechanisms for inflammationinduced ADM in human pancreatic tissue would pave the way to engineer human PDAC tumorigenesis model. Methods: To understand the underlying mechanism of pancreatitis-induced ADM in human tissues, an Ulex europaeus lectin (UEA) and CD133 staining-based method was developed for lineage tracing of human primary pancreatic acinar cells in vitro. BD-matrigel-based 3D culture was established to identify pancreatitis-associated cytokines and growth factors induced ADM of human primary pancreatic acinar cells. Results: The human primary pancreas acinar cells and duct cells were defined by the UEA+ CD133and UEA- CD133+ surface staining pattern, respectively. Under 3D culture condition, TGF-β1, a growth factor significantly up-regulated during chronic pancreatitis, efficiently promoted ADM-associated alterations in human primary pancreatic tissues, characterized by the generation of acinar-derived duct-like cells with UEA+ CD133+ surface staining and sphere formation ability. Blocking TGFBR1/ALK5 completely abolished TGF-β1 induced ADM. Conclusion: TGF-β1 might play important roles in pancreatitis-induced ADM via activating canonical SMAD pathway and non-canonic pathways. The model established in current study will setup a platform to develop human PDAC tumorigenesis model and screen targets for early diagnosis and intervention. Funded by CPRIT first time faculty award, CPRIT training grant (RP 140105) 77 Trainee Symposium on Cancer Research in Texas January 30, 2015 Mancha, Anna - UT San Antonio Poster #71 The Novel Role of 11β-Hydroxysteroid Dehydrogenases in Non-Melanoma Skin Cancer Mancha-Ramirez A1, Liang H1, Junco J2 and Slaga T3 1 Department of Cellular and Structural Biology, Graduate School of Biomedical Sciences, University of Texas Health Science Center at San Antonio 2 Department of Pediatrics, Baylor College of Medicine 3 Department of Pharmacology, University of Texas Health Science Center at San Antonio Corresponding author: Mancha-Ramirez A, Department of Cellular and Structural Biology, Graduate School of Biomedical Sciences, University of Texas Health Science Center, 7703 Floyd Curl Drive, San Antonio, Texas, 78229-3900, Email: [email protected] In addition to the many biological roles of endogenous glucocorticoids (GCs), they have also been a mainstay in clinical medicine as the most effective anti-inflammatory and immunosuppressive drugs. Moreover, GCs are used as a treatment for an array of cancers, with their main anti-cancer mechanism of action being GR-mediated mechanisms triggering cell death. Isozymes 11β-hydroxysteroid dehydrogenase 1 and 2 (11βHSD1 and 11βHSD2) elicit their effect on GCs at the preceptor level by controlling the activation or deactivation of GCs in a tissue-specific manner. Significant evidence suggests that 11βHSD2 may play a role in tumorigenesis, as many studies have shown that it is expressed in tumors and malignant cells, although it is not expressed in the respective normal counterparts. Despite the growing body of evidence supporting a role for 11βHSDs in various different cancers, the role of 11βHSD2 in skin cancer development and progression as well as GC-resistance has not yet been evaluated. As topically administered GCs are a mainstay in the treatment of non-melanoma skin cancer (NMSC), we aim to determine the roles of 11βHSD1 and 11βHSD2 in regulating GCs during nonmelanoma skin cancer development and progression as well as in GC resistance. Our in vitro studies thus far show considerable evidence that 11βHSD2 is upregulated in TPA-treated pre-neoplastic P+ cells as well as in basal levels of transformed RT101 cells. In addition to in vitro work, results from two pilot studies, one utilizing chemically-induced carcinogenesis and the other utilizing UVinduced, suggest the same trend as seen in in vitro studies, whereby 11βHSD2 is upregulated with TPA or UVB treatment vs. untreated counterparts. Although the animal studies are pilot studies, the data gleaned from these experiments coupled with the in vitro data suggest that the role of 11βHSDs in nonmelanoma skin cancer may be of significant importance. Currently, 11βHSD2 knockdown experiments are underway to determine the integral role of these enzymes in NMSC. Moreover, we are also performing enzymatic assays to determine if there is also a corresponding increase in 11βHSD2 activity in cancer tissues vs. normal counterparts. Modulation of cell-specific 11βHSD2 expression and activity can be extremely efficacious because it may diminish GC resistance and even serve as a potential target to aid in prevention of skin cancer development and progression. Moreover, if a link can be established between these enzymes and GC resistance in NMSC, this could potentially have strong clinical impacts as well. This project was supported by pre-doctoral training grant CPRIT RP140105. 78 Trainee Symposium on Cancer Research in Texas January 30, 2015 Onyeagucha, Benjamin - UT San Antonio Poster #72 Targeting MiR-2x6 Improves the Efficacy of Paclitaxel in Human Triple Negative Breast Cancer Onyeagucha BC1,2, Rajamanickam S1,2, Bansal S2, Subbarayalu P1,2, Bansal H2, Chang Y1,2, Timilsina S1,2 and Rao MK1,2 1 Department of Cellular and Structural Biology, University of Texas Health Science Center at San Antonio 2 Greehey Children’s Cancer Research Institute, University of Texas Health Science Center at San Antonio Corresponding author: Rao MK. GCCRI, UT Heath Science Center at San Antonio, San Antonio, TX 78229 Triple negative breast cancer (TNBC) is an aggressive subtype type of breast cancer that lacks the expression of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER-2) genes. It accounts for most of the death due to breast cancer in the United States. Currently, there is no standard form of treatment for this subset of breast cancer. Increasing evidence has demonstrated the role of microRNAs (miRNAs) in cancer progression. Therefore, the identification of key miRNA that de-pacifier TNBC to paclitaxel treatment is crucial for the development of effective treatment strategy for TNBC. In this study, we identified miR-2x6 among other miRNAs using miRNA inhibitor library screen, as a potent paclitaxel de-pacifier in TNBC. Inhibition of miR-2x6 expression decreased migration, invasion, and colony formation in TNBC compared to control. We also found that attenuation of miR-2x6 expression led to G2/M cell arrest and apoptosis. Combine treatment with paclitaxel and miR-2x6 inhibitor or inhibitor alone, attenuated growth of human breast cancer xenograft compared to control. Next using microarray approach, we identified Bcl2-related ovarian cancer killer (BOK), a pro-apoptotic gene as one of the targets of miR-2x6 in TNBC. BOK expression was dramatically reduced in ectopic miR-2x6 cells. Down-regulation of BOK promoted colony formation and migration in TNBC compared to control. Our data suggests that miR-2x6 has oncogenic effects in TNBC and that therapeutic strategies with inhibitor of miR-2x6 may be efficacious in TNBC treatment. This work was supported by CPRIT Training Grant (RP 140105), Voelcker Foundation, Owens Foundation, and ThriveWell Foundation 79 Trainee Symposium on Cancer Research in Texas January 30, 2015 Sareddy, Gangadhara - UT San Antonio Poster #73 Selective Estrogen Receptor β Agonists as Novel Therapeutic Agents to Treat Glioblastoma Sareddy GR, Gruslova A, Cavazos DA, Tekmal RR, Brenner AJ, and Vadlamudi RK Correspondence: Ratna K Vadlamudi, Department of OB-GYN, 7703, Floyd Curl Drive, University of Texas Health Science Center at San Antonio, San Antonio, TX, Email: [email protected]; Phone: (210) 5674930 Glioblastoma (GBM) is the most common type of primary brain tumors with poor prognosis. A gender bias exists in the development of GBM with incidence of developing GBM is greater in males compared to females. Epidemiological studies and several lines of evidence suggest tumor suppressive role of female sexual hormone estrogen on brain tumors. However, estrogen as potential therapy for GBM has limited therapeutic application due to safety concerns including breast cancer, uterine cancer, heart disease, and feminization in men. Therefore alternative agents that mimic estrogen effects will have potent utility in treating GBM. Estrogen effects are mediated though its cognate receptors ERα and ERβ and ERβ functions as a tissue-specific tumor suppressor. Recent studies identified a plant derived Liquiritigenin and a synthetic LY500307 as the selective ERβ agonists with an excellent preclinical profile and are currently in clinical trials for treating hot flashes and Schizophrenia. We recently showed that ERβ is the major ER subtype expressed in GBM. The objective of this study is to determine the therapeutic effects of novel ERβ agonists (Liquiritigenin and LY500307) using both in vitro and orthotopic preclinical models and to determine its mechanism(s) of action. In this study, we have tested the therapeutic efficacy of ERβ agonists using several established as well as patient derived GBM cells. Treatment of ERβ agonists significantly reduced the proliferation of GBM cells with no activity on normal astrocytes suggesting its tumor specific effects. Overexpression of ERβ reduced proliferation of GBM cells and knockdown of ERβ compromised the treatment effect of ERβ agonists further suggesting the specificity. Mechanistic studies revealed that ERβ agonists mediated apoptosis of GBM involve genes activated by both ERβ- classical as well as AP1 mediated non-classical pathways and also involve p38MAPK and JNK pathways. Since Glioma Stem Cells (GSCs) are implicated in tumor initiation, invasion and therapy resistance of GBM, we also tested the effect of ERβ agonists on GSCs. ERβ agonists treatment significantly inhibited the proliferation, neurosphere formation and self-renewal of GSCs and also resulted in the loss of stemness of GSCs and induction of differentiation and apoptosis. Further, Liquiritigenin and LY500307 treatment significantly reduced the tumor growth in in vivo orthotopic GBM models and promoted apoptosis of tumors. Additionally, we validated the isoform expression of ERβ and found that ERβ 1 and ERβ 5 are the major isoforms expressed in GBM and ERβ agonists significantly increased their expression and enhancing their tumor suppressive functions. Together, our results demonstrate that ERβ agonists as a potential therapeutic agent for treatment of GBM. Since ERβ agonists have good blood-brain barrier permeability and less neuronal toxicity, they can be readily transferred to clinical use with current radiation and chemotherapies, thereby providing an additional tool for enhancing survival in GBM patients. Grant Support: CPRIT training grant (RP 140105) 80 Trainee Symposium on Cancer Research in Texas Stanley, Dennis - UT San Antonio January 30, 2015 Poster #74 Dosimetric Effect of Photon Beam Energy on VMAT Treatment Plan Quality Due to Body Habitus for Advanced Prostate Cancer Stanley DN1, Popp T, 1, Ha CS1, Swanson GP2, Eng TY1 and Gutiérrez AN1. 1 University of Texas Health Science Center at San Antonio 2 Department of Radiation Oncology, Scott & White Healthcare In present day, advanced radiation delivery techniques have enabled radiation therapy to become an alternative option to treating high-risk prostate patients despite their larger, complex target volumes due to the inclusion of the prostate, seminal vesicles, and pelvic lymph nodes. These larger target volumes present a challenge for radiation therapy because of their three-dimensional (3D) shape complexity and their relative location to radiosensitive normal tissues. An approach in addressing the complexity has been the implementation of volumetric modulated arc therapy (VMAT). With the understanding that treatment plan quality may also be dependent upon target and patient volume sizes in addition to photon beam energy, a potential study of highrisk prostate patients of varying body habitus using a VMAT delivery technique and varying photon beam energy may help quantify a transition criterion between 6MV and 10MV photon beam energy-a novel investigation to date. In light of this, the purpose of this study was to dosimetrically compare the use of 6MV and 10MV photon beam energies in high-risk prostate cancer patients of varying body habitus using a VMAT radiation delivery technique. The objectives of the study are to evaluate if dosimetric differences exist between 6MV and 10MV and to investigate if differences are dependent on patient body habitus. This study involved forty (n=40) selected patients previously treated to the prostate and pelvic lymph nodes with a VMAT technique using photon beam energies of either 6 or 10 MV. The patients were selected such that they possessed varying body sizes.. To characterize the body habitus of each patient the target to body volume ratio (TBV), the lateral thickness of the patient along the axis of the final isocenter dLAT and the anterior-posterior thickness of the patient along the axis of the final isocenter dAP were evaluated. All patients were planned in the Pinnacle3 treatment planning system (TPS) version 9.6 (Philips Medical, Fitchburg WI) using full (~360°) single or double VMAT arcs for both 6MV and 10 MV photon energies. The plan settings were equivalent for each patient but varied among patients. All patients were optimized with the same planning objectives and normalized such that 95% of the PTV received 100% of the prescription dose (54 Gy delivered 1.8 Gy in 30 fraction). Patients were evaluated for PTV and organ at risk (OAR) parameters for the Rectum, Sigmoid, Bladder, Small Bowel and penile Bulb. For the PTV, there were no statistically significant differences in Dmean, D2cc ,CN and HI values, but the R50 showed an improvement of 6.7% (p<0.01) with 10MV over 6MV for all patients. Concerning the OARs, The statistically significant findings were found in the small bowel, penile bulb, and bladder. For the small bowel, median values of the Dmean, V60%, V40%, and V20% reduced by 1.7% (p <0.01), 2.2% (p <0.01), 3.6% (p <0.01) and 0.9% (p <0.01), respectively, when using 10MV. For the penile bulb, the median value of the D2cc was shown to increase by 3.5% (p <0.01) with 10MV. The bladder median value for the V40%, reduced by 1.8% (p <0.01) using 10MV. A correlation was noted between dAP and R50 that shows that an improvement in R50 of 0.436%/cm is achieved when using 10MV over 6 MV plans. A retrospective, treatment planning study involving forty patients previously treated to the prostate and pelvic lymph nodes with a VMAT technique using photon beam energies of either 6 or 10 MV was analyzed for a correlation between photon energy and body habitus. A correlation between the anterior-posterior thickness and the dose falloff was established. Additionally 10 MV seems to provide a better overall falloff in comparison to 6MV for patients treated with prostate and pelvic lymph nodes using VMAT as demonstrated by the R50. The evaluated patient metrics showed either comparable or preferential results for a photon energy of 10 MV. 10 MV can be considered to be an effective energy for the clinical treatment of high-risk prostate patients with whole pelvic lymph node irradiation utilizing VMAT. This project was supported by the CPRIT training Grant RP 140105 81 Trainee Symposium on Cancer Research in Texas January 30, 2015 Yu, Xiaojie - UT San Antonio Poster #75 Preclinical Evaluation of MiR-195 as a Therapeutic Agent in NSCLC Yu X1, Zhao Z2, Ma X2, Du L2,3 and Pertsemlidis A2,3,4 1 Graduate School of Biomedical Sciences, UT Health Science Center at San Antonio 2 Greehey Children’s Cancer Research Institute, UT Health Science Center at San Antonio 3 Department of Cellular and Structural Biology, UT Health Science Center at San Antonio 4 Department of Pediatrics, UT Health Science Center at San Antonio Corresponding author: Pertsemlidis A, Greehey Children’s Cancer Research Institute, Department of Pediatrics, UT Health Science Center at San Antonio, 8403 Floyd Curl Drive, Texas 78229, USA, Email: [email protected] MicroRNAs (miRNAs) are endogenous small non-coding RNAs that regulate gene expression by binding to the mRNAs of target genes to direct their posttranscriptional repression. A large number of miRNAs have been shown to play important roles of in regulating various cellular physiological and pathological pathways, including cell proliferation and even tumorigenesis. Lung cancer, which is the leading cause of adult cancer deaths, exhibits significant dysregulation of miRNAs, several of which have been shown to regulate cell viability in vitro and in vivo. Intriguingly, some miRNAs also sensitize lung cancer cells to microtubule-targeting agents, such as paclitaxel. These findings highlight the potential application of miRNA mimics as the next generation of therapeutic agents, either alone or in combination with other chemotherapeutic drugs. In order to identify miRNAs with therapeutic relevance to lung cancer, we systematically screen libraries of chemically synthesized miRNA mimics and select those that inhibit lung cancer cell growth and/or sensitize lung cancer cells to paclitaxel. The mechanisms, by which candidate miRNAs inhibit cancer cell growth and metastasis, are elucidated through cell cycle analysis and apoptosis assays, and migration and invasion assays, respectively. The targets of candidate miRNAs are identified by expression profiling in conjunction with miRNA target prediction, and the interactions are validated by luciferase assay. We have shown that miR-195 significantly inhibits the growth of lung cancer cells in vitro, and that transfection of miR-195 into lung cancer cells results in G1 phase arrest through down-regulation of CCND3 and decreased cell migration and invasion through down-regulation of BIRC5. Knockdown of CCND3 and BIRC5 recapitulates the effects of miR-195 on lung cancer cells. We also identified miR-195 as sensitizing lung cancer cells to paclitaxel, which may define a potential regulatory pathway modulating paclitaxel response. Additionally, we found that miR-195 is significantly down-regulated in lung cancer tumors compared to adjacent normal tissues and that its down-regulation in lung cancer patients is associated with worse survival. These results indicate that miR-195 is a tumor suppressor that not only regulates the growth of lung cancer cells but also modulates their response to paclitaxel. Intriguingly, miR-195 does not significantly inhibit the viability of immortalized human bronchial epithelial cells (HBECs), suggesting that its effects may be specific to cancer cells. Further studies are needed to clarify how miR-195 modulates the response of lung cancer cells to paclitaxel and how miR-195 is regulated. This project was supported by NIH grant R01 CA129632 and CPRIT Training Grant RP140105. 82 Trainee Symposium on Cancer Research in Texas January 30, 2015 Yuan, Bin - UT San Antonio Poster #76 Mobilizing Erβ Antitumor Activity Through a Phosphotyrosine Switch Yuan B, Hu YF and Li R1 1 Department of Molecular Medicine, University of Texas Health Science Center at San Antonio, San Antonio Corresponding author: Li R, Dept. of Molecular Medicine, Institute of Biotechnology, University of Texas Health Science Center at San Antonio, STRF, MC 8257, 7703 Floyd Curl Drive, San Antonio, TX, 78229-3900, Email: [email protected] Estrogen receptors ERα and ERβ share considerable sequence homology yet exert opposite effects on breast cancer cell proliferation. While the proliferative role of ERα in breast tumors is well characterized, it is not clear whether the antitumor activity of ERβ can be mobilized in breast cancer cells. Here, we have shown that phosphorylation of a tyrosine residue (Y36) present in ERβ, but not in ERα, dictates ERβ-specific activation of transcription and is required for ERβ-dependent inhibition of cancer cell growth in culture and in murine xenografts. Additionally, the c-ABL tyrosine kinase and EYA2 phosphatase directly and diametrically controlled the phosphorylation status of Y36 and subsequent ERβ function. A nonphosphorylatable, transcriptionally active ERβ mutant retained antitumor activity but circumvented control by upstream regulators. Phosphorylation of Y36 was required for ERβ-mediated coactivator recruitment to ERβ target promoters. In human breast cancer samples, elevated phosphorylation of Y36 in ERβ correlated with high levels of c-ABL but low EYA2 levels. Furthermore, compared with total ERβ, the presence of phosphorylated Y36–specific ERβ was strongly associated with both disease-free and overall survival in patients with stage II and III disease. Together, these data identify a signaling circuitry that regulates ERβ-specific antitumor activity and has potential as both a prognostic tool and a molecular target for cancer therapy. This project was supported by grants Cancer Prevention Research Institute of Texas (RP 110524 and RP 140105), Institute for Integration of Medicine and Science (IIMS) and NCI (CA170306 and CA161349). 83 Trainee Symposium on Cancer Research in Texas January 30, 2015 Zavadil, Jessica - UT San Antonio Poster #77 Improving Therapy for Hepatocellular Carcinoma by Combining a HDAC Inhibitor with an Alkylating Agent Zavadil JA1*, White H2 and Walter CA1,3 1 Cellular and Structural Biology 2 Department of Radiology UT Health Science Center at San Antonio 3 Research and Development, Audie Murphy Hospital Corresponding Author: Christi A. Walter, Ph.D., Department of Cellular & Structural Biology, Mail Code 7762, The University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, TX 78229-3900, Email: [email protected] Within the US, approximately 80% of hepatocellular carcinoma (HCC) is diagnosed in late stages when chemotherapy is the only treatment option. Chemotherapy at late stages provides a median survival of only ~10 months, demonstrating the need for improved therapeutic options. We propose to combine temozolomide (TMZ), a monofunctional DNA alkylating agent, with suberoylanilide hydroxamic acid (SAHA), a histone deacetylase inhibitor, to treat HCC. We hypothesize that these agents will work together by two potential mechanisms: (1) SAHA alters the expression of genes, especially in the DNA repair, apoptosis, and cell cycle regulatory pathways, to impair the cells’ ability to respond to DNA damage induced by TMZ, thus lowering the cells’ capacity for repair and (2) SAHA increases the amount of DNA that is only loosely bound to histones, allowing TMZ to have greater access to DNA and to induce higher levels of damage. Together, these mechanisms lower the capacity for DNA repair and increase the amount of DNA damage incurred, thus increasing the number of cells experiencing damage that exceeds repair capacity and thus will undergo cell death. To test the efficacy of this therapy combination, we have used the C3HeB/FeJ mouse strain, which has a 50% prevalence of HCC in males once they reach 10-12 months old. Surgery and ultrasound were used to identify tumor bearing mice and measure tumor volumes prior to initiation of therapy. Mice are currently undergoing 6 therapy cycles consisting of solvent as a control, SAHA or TMZ alone, or SAHA and TMZ combined. Response to therapy will be measured during treatment in a subset of mice using ultrasound. Experiments to test the proposed mechanisms of action of the combined TMZ and SAHA treatment are being initiated. Funding: CPRIT Predoctoral Fellowship (RP 140105) 84 Trainee Symposium on Cancer Research in Texas January 30, 2015 Baek, Guem Hee – UT Southwestern Poster #78 MCT4 Defines a Glycolytic Subtype of Pancreatic Cancer with Poor Prognosis and Unique Metabolic Dependencies Baek G1, Tse YF1, Hu Z3, Cox D1, Buboltz N4, McCue P4, Yeo C5, White M6, DeBerardinis RJ2,3, Knudsen ES1,2* and Witkiewicz AK1,2* 1 Department of Pathology, UT Southwestern Simmons Cancer Center, UT Southwestern 3 Children’s Medical Center Research Institute, UT Southwestern 4 Department of Surgery, Thomas Jefferson University 5 Department of Pathology, UT Southwestern 6 Department of Cell Biology, UT Southwestern 2 Corresponding author: Agnieszka K. Witkiewicz, MD, Department of Pathology, UT Southwestern Dallas, TX, Email: [email protected] Recent studies have suggested a plethora of metabolic features of pancreatic ductal adenocarcinoma (PDA) that represent adaptations to the tissue architecture and genetic underpinnings of the disease. Here, we demonstrate that Slc16A3 gene encoding MCT4 is highly upregulated in PDA and associated with a glycolytic gene expression program. Immunohistochemical staining of MCT4 in PDA cases demonstrated an association with poor survival that remained significant in multivariate analyses. The attenuation of MCT4 compromised glycolytic flux with compensatory induction of oxidative phosphorylation and increase in macropinocytosis and autophagy. In spite of these adaptations, MCT4-depletion induced pronounced cell death in models harboring high-endogenous expression that was associated with elevated reactive oxygen species and metabolic crisis. Correspondingly, cell death induced by MCT4-depletion was further augmented by inhibition of metabolic and scavenging pathways. In xenograft models, MCT4 had a significant impact on tumor metabolism and was required for rapid tumor growth. Together, these findings indicate that MCT4 is a critical determinant of PDA biology and represents a target for therapeutic intervention. This project was supported by CPRIT grants RP140110. 85 Trainee Symposium on Cancer Research in Texas January 30, 2015 Bozeman, Ronald – UT Southwestern No Poster Biological Effects on Lung Cancer Susceptible K-Rasla1 Mice Irradiated with an Acute Dose of Protons Compared to Protracted Protons Provided as a Solar Particle Event Simulation Bozeman RG, Luitel K, Kasani A, Delgado O, Barron S and Shay JW1 1 Department of Cell Biology University of Texas Southwestern Medical Center Corresponding author: R.G. Bozeman. Department of Cell Biology, UT Southwestern Medical Center 5323 South Harry Hines Blvd., Dallas, Texas 75390, Email: [email protected] During space exploration astronauts are exposed to varying types of radiation. These include Galactic Cosmic Radiation (GCR), which include protons, alpha particles and High Atomic Number (Z) and Energy (HZE) particles. Protons account for the vast majority of radiation in the GCR occasionally in the form of a Solar Particle Event (SPE). SPEs are comprised of varying energies and doses, and occurrences are difficult to predict. The aim of this project is to investigate the mechanistic details that contribute to lung cancer progression following exposure to low LET acute doses of energetic protons using various energies as compared to equivalent doses of protons provided as a SPE simulation in a protracted manner (over 2 hours). The biological effect will be compared to equivalent terrestrial (X-ray) radiation doses. Biologically significant data collected and compared to radiation type, dose, dose-rates, and energies will provide valuable information to more effectively assess acceptable risk limits for astronauts on long-term space missions. Utilizing the K-rasLA1 mouse model of lung carcinogenesis, which mimics lung cancer progression in humans, we investigated the effects of low LET exposure in these mice. K-rasLA1 transgenic mice exposed to a simulated SPE event with varying energies (50-150 MeV/n) using a total dose of 2 Gy (100 cGY/hour). We also exposed mice to a single acute dose of 2 Gy (50 MeV/n or 150 MeV/n) proton (20 cGy/min) or X-ray (250 kVp) at a dose rate of 20 cGy/min as a reference radiation exposure. Acute single dose x-rays, protons, and protracted SPE irradiation of K-rasLA1 mice led to a significant decrease in life span compared to unirradiated control K-rasLA1 mice. In addition, K-rasLA1 mice irradiated with acute and SPE protracted radiation exhibited a significant increase in the incidence of invasive carcinoma compared to unirradiated K-rasLA1 controls. There was a slight increase in carcinoma incidence in mice exposed to X-ray but it was not statistically significant in comparison to the unirradiated controls. K-rasLA1 mice also had a significant increase in the size and number of hyperplastic lesions following 100 days of SPE irradiation compared to unirradiated K-rasLA1 controls. We previously reported that persistent inflammatory responses may play a role in radiation-induced lung carcinogenesis in K-rasLA1 mice. Preliminary data suggest that acute, SPE, and X-ray irradiation may lead to increased tumor formation by activating various mitogenic signaling pathways. This is supported by increased activation of various growth factor signaling pathways such as mTOR, MAPK/ERK, and Stat3. Studies comparing these mitogen signaling pathways to X-ray treated mutant and WT mice are in progress. Collectively, these preliminary findings suggest that both acute and SPE exposure leads to activation of various oncogenic signaling pathways that may contribute to lung tumor development and progression in this mouse model. Studies are currently underway to more fully elucidate underlying mechanisms. This project was supported by a CPRIT Postdoctoral Fellowship. Cancer Intervention and Prevention Discoveries Program RP140110 86 Trainee Symposium on Cancer Research in Texas January 30, 2015 Dutchak, Paul – UT Southwestern Poster #79 Regulation of Amino Acid Metabolism and Hematopoiesis by mTORC1 Negative Regulator NPRL2 Dutchak PA1, Laxman S1, Estill SJ1, Wang C2, Bulut GB3, Gao J2, Huang LJ3 and Tu BP1 1 Department of Biochemistry, University of Texas Southwestern Medical Center 2 Department of Pharmacology, University of Texas Southwestern Medical Center 3 Department of Cell Biology, University of Texas Southwestern Medical Center Corresponding author: Tu B.P., Department of Biochemistry, UT Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, TX 75390-9038, USA Nitrogen permease regulator-like 2 (NPRL2) is a negative regulator of mTORC1 (mammalian Target Of Rapamycin Complex 1). To determine the physiological role of NPRL2 we have created a novel global Nprl2 knockout mouse. Here we show that NPRL2 is essential for fetal liver hematopoiesis and amino acid homeostasis. Loss of NPRL2 produces an apparent “folate-trap”, as determined by targeted mass spectrometry, implicating mTORC1 as a new factor in cobalamin (Vitamin B12)-deficiency. NPRL2 KO liver and mouse embryonic fibroblasts show defective processing of the cobalamin-transport protein called transcobalamin 2 (TCN2), concurrent with impaired lysosomal acidification and lysosomal gene expression. Mechanistically, we show NPRL2 KO cells have impaired metabolic flux from homocysteine to methionine, resulting in methionine-dependent cell proliferation. Treatment with an exogenous cobalamin analog, cyanocobalamin, can rescue NPRL2 KO cell growth when homocysteine is provided as the precursor for methionine synthesis. Collectively, these observations reveal that NPRL2 controls amino acid homeostasis by regulating mTORC1, and controls cobalamin-availability for the methionine and folate cycle, an essential metabolic pathway in hematopoiesis. This work was funded by a Damon Runyon-Rachleff Innovation Award, a Packard Fellowship, a Cancer Prevention Research Institute of Texas (CPRIT) individual investigator award, (B.P.T.), and a research training grant from CPRIT (P.A.D.). 87 Trainee Symposium on Cancer Research in Texas January 30, 2015 Kendall, Genevieve – UT Southwestern No Poster Zebrafish Modeling of PAX3-FOXO1 Driven Rhabdomyosarcoma Kendall GC1 and Amatruda JF1-3 1 Department of Pediatrics, University of Texas Southwestern Medical Center 2 Department of Internal Medicine, University of Texas Southwestern Medical Center 3 Department of Molecular Biology, University of Texas Southwestern Medical Center Corresponding author: Kendall GC, Department of Pediatrics, University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 75390, USA, Email: [email protected] Rhabdomyosarcoma (RMS) accounts for 40% of pediatric soft tissue sarcomas in patients under 20, and is associated with a misregulation of skeletal muscle developmental pathways. Of the two histological subtypes: embryonal (ERMS) and alveolar (ARMS), ARMS has a poorer prognosis, increased metastases, and a higher mortality rate. The majority of ARMS cases have a defining t(1;13) or t(2;13) chromosomal translocation and fusion of either PAX3 or PAX7 with FOXO1, creating a transcriptionally active chimeric protein in which the PAX DNA binding domain is fused to the transactivation domain of FOXO1. PAX3FOXO1 is the most common ARMS chromosomal translocation and is associated with reduced overall survival; however, the impact of this PAX3-FOXO1 fusion oncogene and its gain of function activities remains to be fully understood in the context of a vertebrate developmental system. Remaining questions include defining PAX3-FOXO1 mechanisms of pathogenesis such as targeted cell types, affected transcriptional pathways, and characterizing the interplay and regulation of vertebrate development (specifically, skeletal muscle terminal differentiation). I have shown that PAX3-FOXO1 expression in zebrafish in specific cell subsets alters normal development, including inhibiting somitogenesis or inducing cyclopia. I propose to expand on these findings and model human PAX3-FOXO1 dependent ARMS in a transgenic zebrafish system, thus providing a unique opportunity to: 1) characterize the basic biology and relevant targeted cell type in ARMS in a vertebrate system 2) identify PAX3-FOXO1 targets and associated pathogenic regulation of transcriptional networks, and 3) perform a chemical modifier screen to identify PAX3-FOXO1 suppressors and activators. Delineating how this ARMS-specific PAX3FOXO1 fusion oncogene promotes tumorigenesis will identify associated molecular pathways and likely therapeutic targets. The applications of such a PAX3-FOXO1 transgenic zebrafish model range from studying basic ARMS biology to more high-throughput studies, such as genetic modifier and drug discovery efforts, which are uniquely suited for zebrafish systems. This project is supported by grants from the Cancer Prevention and Research Institute of Texas (RP120685), the National Institute of Genetics Collaborative Research Program, Mishima, Japan (2014A11, in collaboration with K. Kawakami), and a CPRIT postdoctoral fellowship through the UTSW Cancer Intervention and Prevention Discoveries training program. 88 Trainee Symposium on Cancer Research in Texas January 30, 2015 Mgbemena, Victoria – UT Southwestern No Poster Identification and Characterization of BRCA1 Modifiers Mgbemena V1*, Wijayatunge R1*, Kolski L1, Soyombo A1, Foley SB1 and Ross TS1 1 Departments of Internal Medicine and Cancer Genetics, UT Southwestern Medical Center Corresponding author: Theodora S. Ross, Department of Internal Medicine, UT Southwestern Medical Center, 5323 Harry Hines Blvd. Dallas, TX, 75390 USA, Email: [email protected] BRCA1 associates with several proteins to repair double strand breaks, regulate the cell cycle in response to DNA damage and modify transcription of certain genes. Deleterious mutations in BRCA1 increase the chance of developing cancer, especially breast and ovarian cancers. We recently sequenced the genomes of 200 BRCA- mutant patients and observed the presence of a CREB binding protein (CREBBP) mutation (p. N1978S) in six patients with early onset cancers. We hypothesized that CREBBP mutations may be BRCA1 modifiers. CREBBP is a transcriptional regulator that physically interacts with BRCA1. The effect of this interaction on BRCA1 function has not been investigated. Our short-term goal has been to study how CREBBP expression affects BRCA1 and whether it contributes to the function of BRCA1 as a tumor suppressor. We have found that CREBBP increases the levels of BRCA1 in cultured cells. We have also demonstrated that expression of CREBBP variants associated with Rubenstein-Taybi Syndrome (RTS), a pediatric dysmorphic syndrome associated with cognitive challenges and neural crest derived tumors, do not stabilize BRCA1. We conclude that CREBBP may contribute to BRCA1 stability and affects BRCA1 function. This project was supported by grants R01 CA82363-03, R01 CA098730-01, Burroughs Wellcome Fund Clinical Scientist Award in Translational Research and the Cancer Intervention and Prevention Discoveries Program Grant RP140110 89 Trainee Symposium on Cancer Research in Texas January 30, 2015 Min, Jaewon – UT Southwestern No Poster Human Telomere Overhangs in Telomerase-negative Cells are Generated by Multi-step Processes Min J1, Wright WE1 and Shay JW1 1 Department of Cell Biology, University of Texas Southwestern Medical Center Corresponding author: Shay JW, Dept. of Cell Biology, University of Texas Southwestern Medical Center, Dallas, Texas 75390, USA, E-mail: [email protected] Telomeres are composed of TTAGGG repeated sequences located at the end of chromosomes. In human somatic cells, cell divisions are accompanied with the telomere length shortening due to lack of or insufficient telomerase activity (termed the end replication problem). In combination with other alterations, telomere shortening can accelerate genetic instability in the early stage of tumorigenesis driving neoplastic transformation. Thus, dissecting the mechanism underlying progressive telomere shortening may provide a deeper understanding of how normal cells become cancer cells as well as the molecular mechanism of telomere maintenance. It is generally believed that telomere length shortening results from the removal of the RNA primer binding site at the Okazaki fragment located in lagging DNA strand synthesis. If it is correct, the rate of telomere shortening should be around 8-10 bp per cell division (RNA primer size / 2 = (16 ~ 20 nt) / 2). Nevertheless, telomerase negative human cells showed telomere shortening ~70 bp per cell division. From previous studies, we found that telomere overhang generation processes may be important in determining telomere shortening rates. We found that the lagging strand has ~110 nt overhang length and the leading strand possesses ~30 nt overhang length. These results can potentially explain the telomere shortening rate in telomerase negative cells (lagging strand overhang 110 nt + leading strand overhang 30 nt /2 = 70nt). However, the molecular mechanisms for overhang generation processes remain to be determined. We have recently found that human telomere overhangs in telomerase-negative cells are processed by multi-steps and leading/lagging strands have distinct mechanisms. Leading strand overhangs have an early process that occurs throughout S phase, and a late process that occurs at the late S/G2 phase. In contrast, lagging strand overhang processes are completed with DNA replication. At present we are determining the molecular factors involved in overhang generation processes using biochemical approaches as well as a new system for introducing telomerase negative condition in various human cell lines using advanced techniques of genome editing (ex. CRISPR/Cas9). This project was supported by Cancer Intervention and Prevention Discovery Training Program. 90 Trainee Symposium on Cancer Research in Texas January 30, 2015 Padanad, Mahesh – UT Southwestern Poster #80 Acyl-CoA Synthetase Long-chain Family Member 3 is Required for Mutant KRAS-Driven Lung Cancer Padanad MS1, Konstantinidou G1, Yang C2, Melegari M1, Venkateswaran N1, Batten K3, Huffman KE4, Shay J W3, Minna JD4, DeBerardinis RJ2 and Scaglioni PP1 1 Department of Internal Medicine, UT Southwestern Medical Center 2 Children’s Medical Center Research Institute, UT Southwestern Medical Center 3 Department of Cell Biology, UT Southwestern Medical Center 4 Hamon Center for Therapeutic Oncology Research, UT Southwestern Medical Center Corresponding author: Scaglioni PP, Department of Internal Medicine, Division of Hematology/Oncology, The University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 753908852, Email: [email protected] The goal of this study was to identify cellular networks that mediate the maintenance of mutant KRAS lung cancer. To this end, we functionally analyzed the transcriptome of transgenic mouse lung tumors manipulated in vivo to undergo mutant Kras extinction, providing isogenic comparisons between mutant KRAS extinguished versus non-extinguished tumors for the discovery of new therapeutic targets. We determined that mutant KRAS controls tumor metabolism by regulating not only glycolysis but also lipid homeostasis. Inhibition of rate-limiting glycolytic enzymes caused only modest antiproliferative effects in human lung cancer cells in culture. By contrast, we found that Acyl-CoA synthetase long-chain family member 3 (ACSL3), which converts fatty acids into fatty Acyl-CoA esters, the substrate for lipid synthesis and β-oxidation, is required for the survival of mutant KRAS lung cancer cells. Further, we confirmed that ACSL3 is a mutant KRAS responsive gene expressed in the respiratory epithelium, in lung cancer cells and in primary human cancers. As predicted by these experiments, ACSL3 is essential for the ability of mutant KRAS human lung cancer cells to form colonies in soft agar or to establish xenografts in immunocompromised mice. With mechanistic experiments we determined that mutant KRAS stimulates the uptake and retention of fatty acids as well as their β-oxidation by lung cancer cells, in an ACSL3-dependent manner. ACSL3 silencing causes significantly up to 80% reduction in β-oxidation. Furthermore, pharmacologic inhibition of β-oxidation with etomoxir causes the apoptotic death of mutant KRAS NSCLC cells. This data support the hypothesis that mutant KRAS NSCLC cells rely on fatty acid β-oxidation for energy production. Acsl3 null mice are developmentally normal and their characterization is in progress. Our data demonstrate that mutant KRAS reprograms lipid homeostasis in lung cancer, establishing a cancer specific metabolic vulnerability that could be exploited for therapeutic gain. This project was supported by Cancer Intervention and Prevention Discoveries Program Training Grant from CPRIT, American Cancer Society Scholar Award ACS 13-068-01-TBG, Texas 4000 Fighting Cancer Every Mile, UT Southwestern Friends of the Comprehensive Cancer Center. 91 Trainee Symposium on Cancer Research in Texas January 30, 2015 Colosimo, Dominic – UT Southwestern Poster #81 Discoipyrroles: A Multifaceted Approach to Understanding Novel Marine Natural Products Colosimo DA1 and MacMillan JB1 1 Department of Biochemistry, Division of Basic Sciences Graduate Program, UT Southwestern Medical Center Corresponding author: Colosimo DA, Department of Biochemistry, Division of Basic Sciences Graduate Program, UT Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390-9038 Email: [email protected] Non-small cell lung cancer (NSCLC) contributes to 80% of lung cancer in the United States, which is estimated to kill more than 159,000 Americans in 2014. Around 8% of patients with squamous cell carcinoma, the most prevalent subtype of NSCLC, carry genetic mutations in discoidin domain containing receptor 2, or DDR2. This transmembrane protein is a receptor tyrosine kinase that binds collagen in the extracellular matrix and initiates signaling cascades vital for cell migration. A novel family of natural products, the discoipyrroles, produced by the marine bacteria Bacillus hunanensis have been hypothesized to inhibit the DDR2 signaling pathway of particular NSCLC cell lines (Hu Y., JACS 2013). We have shown that the discoipyrroles, particularly discoipyrrole A (DPA), elicit cytotoxicity of DDR2 mutated, but not DDR2 wild-type NSCLC cell lines demonstrating a potential use in targeted therapy (Figure 1). Currently, we are investigating multiple aspects of discoipyrrole drug development. Using chemistry inspired by the bacterial biosynthesis, over 90 analogs of discoipyrrole A have been generated and resubmitted to a previously established NSCLC cell line cytotoxicity assay. Concordantly, pharmacokinetic studies characterizing the drug-like characteristics of these analogs are being completed. Phase I Figure 2 – Discoipyrroles Induce Cytotoxicty of DDR2 Mutated Cell and II in vitro studies in S9 cell extracts and Lines. (a) The discoipyrrole family of natural products (b) Cytotoxicity of discoipyrrole compounds to NSCLC cell lines. hepatocytes, respectively, have been completed. Additionally, in vivo absorption, distribution, metabolism, and excretion (ADME) analysis have been performed. To identify the molecular mechanism behind the biological effect of discoipyrrole A, we are using discoipyrrole analogs fused with affinity or fluorescent tags to identify protein binding partners or cell localization, respectively. Recently, using a streptavidin enrichment system we have identified a putative list of proteins that show binding to biotinylated DPA analogs. After validation of this interaction using competition assays with non-biotinylated DPA, we are progressing with RNAi based methods to investigate the role of identified binding partners in DDR2 mutated NSCLC. Apart from the clinical investigation of DPA as modulator of the DDR2 signaling pathway we are exploring the biosynthetic pathway of the discoipyrroles. These compounds have a unique biosynthetic mechanism that incorporates both enzyme driven and enzyme free chemical reactions. Utilizing nuclear magnetic resonance (NMR) and selective isotope labeling we are capable of following the non-enzymatic mechanism of this biosynthetic pathway in real time. Preliminary results have identified previously unknown intermediates in the formation of DPA. Understanding this pathway will uncover novel biomimetic synthetic routes with application in drug design. This work is funded by a CPRIT training grant (RP140110), the National Institute of Health (5R01CA149833-03), and the Chilton Foundation. 92 Trainee Symposium on Cancer Research in Texas January 30, 2015 Gallolu, Sachith – UT Southwestern No Poster The Effect of with no Lysine (WNK) Pathway on Autophagy Perera GKS, Lee AY and Cobb MH1 1 Department of Pharmacology, University of Texas Southwestern Medical Center Corresponding author: Melanie H. Cobb, Department of Pharmacology, University of Texas Southwestern Medical Center, 6001 Forest Park Road, Dallas TX 75390 The With No Lysine kinases (WNKs) are an atypical family of Ser/Thr kinases which have a conserved lysine residue in their catalytic domains shifted from the conventional position. This arrangement gives them unique structural and functional properties. In mammals, there are four isoforms of WNKs, including WNK1 which is ubiquitously expressed. WNKs are activated by osmotic stress and once activated; they phosphorylate and activate downstream target kinases oxidative-stress responsive 1 (OSR1) and SPS1/Ste20-related proline-alanine-rich kinase (SPAK). These in turn phosphorylate SLC12 family of cation chloride cotransporters, including the sodium potassium cotransporter 1 (NKCC1), either activating or inhibiting them. Thus, the WNK pathway mediates the cellular response to changing osmotic conditions. This pathway governs essential physiological activities such as blood pressure, secretion and reabsorption in kidney, inhibitory neurotransmission and muscle contraction. In addition to acting as the master regulator of cation chloride cotransporters via the WNK pathway, WNK1 is shown to be involved in other cellular processes such as cell proliferation, migration, endocytosis and angiogenesis. Recently, unpublished work from the Cobb lab has shown that WNK1 regulates lysosomal localization and autophagy. Autophagy is a cellular degradation pathway that protects the cells under stress conditions and supplies them with nutrients. It starts with the formation of structures called phagophores in mammalian cells. The phagophores then engulf the cellular material and extend to generate double membrane organelles called autophagosomes. Finally, they fuse with cellular lysosomes to form autolysosomes where the degradation of cellular materials occurs. Autophagy is altered in various human diseases and genes involved in autophagy are implicated in different types of cancers. The objective of this project is to understand the effect of WNK1 on autophagy and elucidate the molecular mechanisms involved in that process. The knockdown of WNK1 by siRNA in HeLa cells increases autophagy under both fed and starved conditions. In addition, overexpression of an N-terminal fragment of WNK1 in cells inhibits autophagy. The siRNA-mediated knockdown of the WNK1 downstream target kinase SPAK also increases autophagy. Current work is underway to determine whether the effect of WNK1 on autophagy is dependent on SPAK and its downstream target NKCC1. In conclusion, the kinase WNK1 acts as an inhibitor of autophagy and it is likely mediated by its downstream target kinase SPAK. Funding: CPRIT - Cancer Intervention and Prevention Discoveries Program RP140110 NIGMS R01 GM53032 93 Trainee Symposium on Cancer Research in Texas January 30, 2015 Manjunath, Hema – UT Southwestern Poster #82 Evaluating the Tumor Suppressor Activity and Mechanism of the let-7 miRNA Family in vivo Manjunath H1 and Mendell JT1, 2, 3 1 Department of Molecular Biology, 2Simmons Cancer Center, and 3Center for Regenerative Science and Medicine, University of Texas Southwestern Medical Center Corresponding author: Manjunath H, Department of Molecular Biology, University of Texas Southwestern Medical Center, NA6.200, 6000 Harry Hines Blvd, Dallas, Texas 75390, Email: [email protected] Since its identification as one of the first recognized miRNAs over 15 years ago, let-7 has emerged as a master regulator of proliferation and differentiation in multiple model organisms. The related variants of let7 comprise a large miRNA family that post-transcriptionally regulate genes such as cell cycle regulators, apoptotic genes and several important oncogenes, including the RAS genes and MYC. Accordingly, numerous studies have implicated let-7 as a tumor suppressor in human cancers. Nevertheless, the consequences of loss of let-7 function on tumorigenesis in vivo remain largely unexplored – mainly due to redundancy within this large miRNA family. The 9 murine let-7 family members, for instance, arise from 12 precursor transcripts located across 8 genomic loci, and are predicted to share significant target specificity. I aim to study let-7 function by employing a published ‘miRNA tough decoy’ (TuD) strategy to inhibit function of this miRNA family. TuDs are RNA decoys that are designed to bind and inhibit a miRNA of interest. TuDs are expressed from strong RNA PolIII promoters and possess an optimized secondary structure that allows them to be efficiently exported from the nucleus and resist rapid degradation, thus facilitating potent inhibition of miRNA function. I have demonstrated that a TuD designed to target let-7 (TuD-let-7) effectively inhibits the members of this miRNA family in vitro. In order to examine function of the let-7 family in vivo, I am generating a transgenic mouse that expresses the TuD-let-7 under inducible control. I will use this unique tool to examine the tumor suppressive activity of let-7, specifically in the context of Wilms tumor, or nephroblastoma. This common renal malignancy accounts for 95% of pediatric kidney cancers and affects nearly 1 in 10,000 children in North America. No targeted therapies have been developed to treat Wilms tumor, due in part, to a lack of basic understanding of its molecular mechanisms. Our interest in studying this malignancy arose from multiple lines of evidence suggesting that loss of let-7 function was tied to Wilms tumorigenesis. For instance, the repression of let-7 has been implicated in the development of ‘embryonal tumors’ such as medulloblastoma and neuroblastoma – a class of malignancies including Wilms tumor, that are thought to originate from progenitor cells that have lost their ability to differentiate properly. Furthermore, exome sequencing of primary Wilms tumors revealed novel mutations in crucial enzymes of the miRNA biogenesis pathway that impair production of a specific subset of miRNAs - including all members of the let-7 family. We therefore hypothesize that loss of let-7 is a key event in the pathogenesis of some Wilms tumors. The TuD-let-7 transgenic mouse will be used to test this hypothesis and dissect mechanisms through which let-7 regulates tumorigenesis in this context. This mouse will also be valuable in elucidating the role of let-7 in other malignancies, including embryonal rhabdomyosarcoma, where similar mutations in the miRNA processing machinery have been observed. This project was supported by the Cancer Intervention and Prevention Discoveries Program Training Grant RP140110 (to H.M), CPRIT grant R1008 (to J.T.M.), and NIH grant R01CA120185 (to J.T.M.). 94 Trainee Symposium on Cancer Research in Texas January 30, 2015 Miller, Jason – UT Southwestern Poster #83 Zwitterionic Aminolipids for siRNA Delivery Miller, JB1 and Siegwart, DJ1 1 Department of Biochemistry, Universiy of Texas Southwestern Medical Center Corresponding author: Siegwart, DJ, Department of Biochemistry, Simmons Cancer Center, University of Texas Southwestern Medical Center, 2201 Inwood Road, Dallas, Texas 75390, Email: [email protected] Since the discovery of RNA interference, the silencing of target genes using small interfering RNA (siRNA) has become an essential tool for biologists to study and develop therapeutics for diseases including cancer. Many different cationic lipid systems have been chemically synthesized to bind and transport highly anionic siRNAs across cell membranes, but delivery to tumors remains elusive due to preferred liver accumulation. Poly(ethylene glycol) (PEG) modifications have been extensively studied as a way to mitigate aggregation in serum and reduce protein absorption that modulates biodistribution, but the lack of success using PEG demands the development of alternative approaches. Zwitterionic chemical groups are net neutral, and introduce a layer of hydration to separate the materials from the surrounding environment. These properties make zwitterions excellent candidates for improving stability in the blood, reducing non-specific protein absorption, and allowing for passive accumulation in tumor tissue by the enhanced permeability and retention (EPR) effect attainable because of the leaky vasculature created by rapidly growing tumor cells. Using a protecting group free, combinatorial approach a library of ~100 amphiphilic zwitterionic amino lipids (ZALs) was rapidly synthesized, containing a zwitterionic sulfobetaine head group, an amine rich linker region, and assorted hydrophobic tails. Sulfobetaine head groups were appended by conjugate addition of amines with a zwitterionic acrylamide, and subsequent reaction of these zwitterionic intermediates with a variety of hydrophobic electrophiles yielded a structurally diverse library of materials. Biophysical characterization has shown that ZALs can be formulated into nanoparticles that encapsulate siRNA in the presence of cholesterol in physiologically relevant buffer. Lead materials are able to deliver siRNA in vitro to human cancer cell lines with up to 80% silencing of a stable firefly luciferase reporter gene while showing minimal material-derived cytotoxicity. The pharmacokinetic properties and delivery efficacy of lead materials will be evaluated in vivo in established tumor models in collaboration with cancer biologists. This project was supported by CPRIT R1212 and RP140110. 95 Trainee Symposium on Cancer Research in Texas January 30, 2015 Nichols, Brandt – UT Southwestern Poster #84 Tumor-Activated Meiotic Regulators Modulate DNA Damage Response Nichols B1, Vaziri C2, Gao Y2 and Whitehurst AW1 1 Simmons Comprehensive Cancer Center, University of Texas Southwestern Medical Center 2 Department of Pharmacology, University of North Carolina, Chapel Hill Corresponding authors: Nichols B & Whitehurst A, Simmons Comprehensive Cancer Center, University of Texas Southwestern Medical Center, Dallas, TX, 75390, E-mail: [email protected] CT-antigens (CTAs) are genes whose expression is restricted to the testes but reactivated in cancer. We have performed extensive functional assays of 138 CTAs to assess their contributions to cancer cell tumorigenic behaviors and signaling modules. This effort uncovered a cohort of testes proteins that are required for tumor cell viability and which are essential for repairing DNA double-strand breaks (DSBs) during meiosis. This evokes the hypothesis that these meiotic regulators may functionally contribute to tumorigenesis by modulating DNA damage repair. Follow-up experiments show that inhibition of these CTAs prevents DNA repair enzymes from being recruited to sites of damage. Furthermore, depleting these CTAs in NSCLC lines decreased cell viability and increased levels of DNA DSBs. For example, HORMAD2 depletion resulted in decreased EdU incorporation and increased H2AX phosphorylation. Also, HORMAD1 depletion made cancer cells more sensitive to irradiation in a dose-dependent manner. Another meiotic regulator, SPO11, initiates DNA DSBs during meiosis and leads to elevated DNA damage when overexpressed in tumor-derived cell lines. We hypothesize that SPO11 may drive a mutator phenotype and contribute to basal levels of DNA damage in cancer. These findings suggest that tumor cells may engage meiotic regulators to modulate DNA damage and promote tumorigenesis. Future endeavors aim to determine the molecular mechanistic basis for dependence on these CTAs, their relevance in vivo, and whether they may serve as drug targets for enhancing response to current cytotoxic therapies. This project is supported by the Cancer Intervention and Prevention Discoveries Program (RP140100) and the Department of Defense (LC130495). 96 Trainee Symposium on Cancer Research in Texas January 30, 2015 Ren, Junyao – UT Southwestern No Poster IKKβ is an IRF5 Kinase that Instigates Inflammation Ren J1, Chen X1, and Chen ZJ1, 2 1 Department of Molecular Biology and 2 Howard Hughes Medical Institute, University of Texas Southwestern Medical Center Corresponding author: Chen ZJ, Department of Molecular Biology, University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard Dallas, Texas, 75390, Email: [email protected] Inflammation affects tumorigenesis by regulating its promotion, progression and immune surveillance. Interferon regulatory factor 5 (IRF5), a critical mediator of immune cell development and inflammatory response, has recently emerged as a potent tumor suppressor. IRF5 can induce apoptosis, arrest cell cycle and also regulate host immunity against cancer cells. However, the signaling pathways by which IRF5 contributes to antitumor immunity remain poorly understood. We report that the kinase IKKβ, which is known to regulate the Rel/nuclear factor kappa B (NF-κB) family of transcription factors, phosphorylates IRF5 at a specific serine residue, and that this phosphorylation is critical for IRF5 activation and cytokine production. Thus, IKKβ regulates two master transcription factors, NF-κB and IRF5, which coordinately control gene expression to mediate inflammatory responses. These results provide insights into the underlying relationship between tumorigenesis and host anti-tumor immunity and suggest new therapeutic targets for cancer treatment. This work was supported by grants from the National Institutes of Health (R01 AI093967) and the Cancer Prevention and Research Institute of Texas (CPRIT; RP120718). J.R. was supported by a CPRIT predoctoral training fellowship (RP140110). 97 Trainee Symposium on Cancer Research in Texas January 30, 2015 Updegraff, Barrett – UT Southwestern Poster #85 Sleeping Beauty Mutagenesis for the Functional Identification of Human Melanoma Metastasis Genes Updegraff B1 and O’Donnell K1 1 Department of Molecular Biology, UT Southwestern Medical Center Corresponding author: Updegraff B, Department of Molecular Biology, UT Southwestern Medical Center, 6000 Harry Hines NA6.300, Dallas, TX, Email: [email protected] Melanoma is one of the deadliest malignancies when diagnosed at advanced stages, with a five-year survival rate of less than 15% for patients with distant metastasis. Determining how human primary melanoma becomes intrinsically capable of metastasis remains a significant challenge. While exome sequencing has led to the identification of a handful of bona fide melanoma metastasis oncogenes such as MITF, these efforts have been hindered by a lack of valid model systems that are capable of assaying metastasis of primary tumor cells. Therefore, an unbiased methodology to functionally interrogate human primary melanoma genomes is urgently needed to identify critical driver genes that initiate metastasis. Such studies will facilitate the development of novel diagnostic strategies to differentiate metastatic from non-metastatic melanomas, a major clinical need, and reveal new opportunities for therapy. A number of laboratories including ours have utilized Sleeping Beauty (SB) transposon mutagenesis to identify both loss- and gain-of-function mutations important for the development and acceleration of a number of tumor types. Importantly, the majority of SB screens have been performed using transgenic mice. Recently, our laboratory has successfully established an ex vivo mutagenesis system wherein human cells growing in culture are mutagenized with the SB transposon, transplanted into immunocompromised NOD/SCID/IL2R- (NSG) mice, and transposon insertion sites associated with tumorigenesis and metastasis are recovered. The ability to tractably mutagenize human primary melanomas in an unbiased manner would provide a wealth of information regarding the processes of metastasis. We utilized a published assay wherein human primary stage II-IV melanomas labeled with a luciferase-dsRed reporter are xenotransplanted into NSG mice and monitored for metastasis using bioluminescence imaging (BLI). Because the metastatic behavior of these melanomas in NSG mice is predictive of the clinical outcome in patients, this model provides a powerful and clinically relevant platform for discovering new genes that regulate metastasis. To facilitate the identification of genes that drive metastasis, we optimized delivery of Sleeping Beauty vectors to primary human melanomas shown to infrequently metastasize in both mice and patient donors. In a pilot screen, mutagenesis of patient-derived melanoma cells with the SB transposon and transposase resulted in a significant increase in the frequency of metastasis-bearing animals and a broader organ distribution of metastases compared to controls (n=24, p<0.027). Genomic DNA isolated from metastases, primary subcutaneous tumors, and blood was sequenced and common insertion sites (CISs) associated with metastases were identified using Monte-Carlo analysis. To functionally validate the contribution of genes identified in our forward genetic screen, gain- and loss-of-function experiments are currently being performed in human melanoma cells. Lentiviral cDNA expression vectors are introduced to test whether candidate oncogenes are sufficient to promote the metastasis of inefficiently metastasizing melanomas. We are also using the CRISPR/Cas9 system to inhibit candidate metastasis suppressor genes. Finally, in order to regulate the extent of ongoing mutagenesis, we have generated melanomas with stable conditional expression of the SB transposase. A summary of our latest results will be presented. We anticipate these studies will provide a better understanding of the molecular mechanisms that promote melanoma metastasis and lead to the development of novel therapeutic targets for this deadly disease. This work is supported by the Cancer Intervention and Prevention Discoveries Program Training Grant RP140110 (to B.U.) and the CPRIT Scholar in Cancer Research Award R1101 (to K.O.). 98 Trainee Symposium on Cancer Research in Texas January 30, 2015 Vega, Anthony – UT Southwestern Poster #86 Quantitative Analysis of Cortical Actin-Membrane Microcluster Interactions Vega A1, Ditlev J1, Rosen MK1 and Jaqaman K1 1 Department of Biophysics, Division of Basic Sciences Graduate Program, University of Texas Southwestern Medical Center Corresponding author: Vega A, Dept. of Biophysics, Division of Basic Sciences Graduate Program, University of Texas Southwestern Medical Center, 6001 Forest Park Road, Dallas, TX 75235, Email: [email protected] The spatial organization of receptors on the plasma membrane is crucial for their signaling efficacy. Previous studies have implied that the cortical-actin (CA) meshwork regulates receptor mobility and organization in the plasma membrane. To understand the molecular mechanisms underlying this process, we sought to quantitatively study the role of the CA meshwork in regulating the clustering and mobility of receptors in an in vitro reconstitution system. Specifically, we are studying the mobility, organization and clustering of Linker for Activation of T cells (LAT) complexes in an artificial lipid bilayer with an attached actomyosin meshwork. Our quantitative analyses using light microscopy and computational image analysis have led to the very exciting observation that while LAT complexes exhibit predominantly confined diffusion in the absence of actin, they exhibit nearly equal populations undergoing confined and normal diffusion in the presence of a CA meshwork. These results suggest that actin may be playing a larger and more interesting role in influencing cluster dynamics that extends beyond static spatial organization. Future work will focus on determining the proteins that link LAT complexes to the CA meshwork and on extending this work to cells and other receptor systems. This project was supported by CPRIT Training Grant (RP140110). 99 Trainee Symposium on Cancer Research in Texas January 30, 2015 Zhang, Ning – UT Southwestern Poster #87 Analysis of Telomere Position Effect Based on 3D-Fish Zhang N, Robin J, Han S, Kim W, Shay JW and Danuser G Department of Cell Biology, UT Southwestern Medical Center Corresponding author: Shay JW, Danuser G, Dept of Cell Biology, UT Southwestern Medical Center, 6000 Harry Hines Blvd, Dallas TX, 75235, Email: [email protected]; [email protected] Telomeres are repetitive nucleotide sequences (TTAGGG) capping the end of chromosomes, which can prevent the chromosomes ends from fusion and degradation. The telomere length decreases with each cell division until cells undergo replicative senescence and crisis. Very few cells that bypass the checkpoint will become cancer cells by reactivating telomerase and maintaining the short telomeres length. Our lab previously reported telomere position effect which indicates that the shortening of telomere length can regulate certain genes expression. This regulation effect could even be extended to genes up to 10 Mb away from the telomere by chromosome looping. We designed 3D fluorescent in situ hybridization (FISH) to confirm that the loci of telomeres and target genes become separated with telomere shortening. To explore the mechanism, we knocked down shelterin complex subunits RAP1 and POT1 in fibroblast, and maybe more candidate factors in the future, to test the target gene expression with different lengths telomeres. Also we are using 3D-FISH to measure the gene-telomere distance when knock down happens. Instead of checking the distance of fluorescent probes by eye, it is important to employ the high throughput data analysis algorithm and generate statistical results from those images. Based on the mixture model and our large program repository, we are developing automatic fluorescent spots detection tools in 3D by matlab. The spot classification parameter is the product value of intensity and 3D curvature. We apply nuclei size Gaussian filter to identify each single nucleus, crop out those nuclei, and then reset filter size and threshold to detect fluorescent probes. Hopefully such tools will be able to increase our image analysis efficiency and reliability. This project was supported by CPRIT training grant. 100
© Copyright 2024