38

166

ASSOCIATION FOR ACADEMIC SURGERY AND SOCIETY OF UNIVERSITY SURGEONS—ABSTRACTS

PC and PCCL with qRT-PCR. MIA PaCa2 cells were transfected with mouse PDX-1 and then co-transfected with either mouse PDX-1 siRNA plasmid or control vector and MTS assay of proliferation was performed. Statistical analyses were performed via unpaired T test (p⬍ 0.05 was considered significant). Results: A single nucleotide polymorphism (SNP) was observed for locus 169068 with C and T alleles resulting in a proline to leucine substitution at the receptor’s intracellular c-terminal (amino acid position 335) in all PCs. CT to T LOH was also observed within the same locus in 33% of PCs. Genomic DNA qPCR verified reduction in copy number of the SSTR5 gene in 44% of PCs. Quantitative RT-PCR revealed reduction in SSTR5 mRNA in 43% of PCs, which was associated with PDX-1 overexpression (33% of the PCs). PCCLs also had the C to T SNP. PDX-1 transfection resulted in significant MIA PaCa2 cell proliferation (P⬍0.05), which was blocked by co-transfection of PDX-1 siRNA, but not by the control vector(Fig. 1A). SSTR5 agonist treatment resulted in significant reduction in proliferation of wild-type SSTR5 transfected cells (89.07% of control p⬍0.01 ), but not mutated SSTR5 transfected cells (105.2% of control p⫽NS)(Fig. 1 B). Conclusion: We found SSTR5 LOH in PC and C to T SNP in PC and PCCL. Reduced SSTR5 expression in both PC and PCCL is associated with PDX-1 overexpression. PDX-1 transfection resulted in enhanced proliferation of PCCL, which was blocked by siRNA PDX-1. SSTR5 agonist with Wild-type SSTR5 transfection resulted in suppression of PCCL proliferation, which was ablated by transfection with the mutated SSTR5. This study suggests that SSTR5 gene is an anti-proliferative gene in PC, which is associated with PDX-1.

tional significance of the observed poly (T) 8 mutations remained unknown. The purpose of our present study was to test the following hypotheses: H-1. CDK2-AP1 3’UTR del T poly T (8) mutations are functionally significant resulting in decreased CDK2AP1 expression in MSI CRC. H-2. The mechanism of the decreased CDK2-AP1 expression observed in association with the presence of the del T poly T (8) mutation is reduced CDK2-AP1 mRNA stability. Methods: To test the functional significance of the 3’ UTR del T poly (T) 8 mutation on CDK2-AP1 expression (H1) we employed a site-directed mutagenesis functional assay using a Lentiviral Vector mediated transfer system (LV-pRRl.sin. PPT.hPGK.EGFP) to introduce both the mutant and wild type poly (T) 8 3’ UTR sequences into separate human CRC SW48 cell lines. Expression of a green florescent protein (GFP) marker gene was used to measure transduction efficiency and surrogate CDK2AP1 expression. CDK2-AP1 mRNA stability (H-2) was measured using a standard actinomycin D assay and the mRNA structure folding software mfold 3.2. Results: H-1. Mutant (del T poly T 8) GFP-3’-UTR samples demonstrated significantly reduced GFP expression compared to wild type GFP-3’-UTR as measured by both FACS and real-time PCR. H-2. Both the actinomycin D assay and mfold software demonstrated significantly reduced mRNA stability for the del T poly (T) 8 product compared to the wild type. Conclusions: In summary, these novel results support our hypotheses that the del T poly (T) 8 observed in the 3’ UTR of the CDK2-AP1 gene in human MSI CRC is functionally significant and results in decreased CDK2-AP1 expression. The results also indicate the mechanism of this decreased expression is due at least in part to decreased mRNA stability. We conclude these results significantly advance our understanding of the role of the growth suppressor gene CDK2-AP1 in human MSI CRC and support its continued investigation. Currently, experiments are underway in our laboratory to determine if the 3’ UTR del T mutation is a somatic or germline event. For the future, we plan investigation of the mechanism by which deficient mismatch repair results in the CDK2-AP1 3’ UTR del T poly T(8) mutation observed in human MSI CRC.

38. GENE SILENCING IN PANCREATIC CANCER: IDENTIFICATION OF NOVEL TRANSCRIPTIONAL REPRESSORS. M. J. Truty, S. Tsuji, G. A. Lomberk, M. FernandezZapico, R. Urrutia; Mayo Clinic, Rochester, MN

37. A DEL T POLY T (8) MUTATION IN THE 3’ UNTRANSLATED REGION (UTR) OF THE CDK2-AP1 GENE IS FUNCTIONALLY SIGNIFICANT CAUSING DECREASED MRNA STABILITY RESULTING IN DECREASED CDK2AP1 EXPRESSION IN HUMAN MICROSATELLITE UNSTABLE (MSI) COLORECTAL CANCER (CRC). J. Shin 1, Z. Yuan 1, K. Fordyce 2, P. Sreeramoju 1, T. Kent 3, J. Kim 1, V. Wang 1, D. Schneyer 4, T. K. Weber 1; 1Departments of Surgery & Molecular Genetics, Albert Einstein College of Medicine, New York, NY, 2Computational Decision Science Group, IBM USA, Burlington, VT, 3Department of Surgery, University of Pittsburgh Medical Center, Pittsburgh, PA, 4Honors Biology Research Program, Horace Greeley High School, Chappaqua, NY Introduction: In our most recent presentation to this forum (San Diego, 2006) we presented our results demonstrating cyclin dependent kinase-2-associated protein (CDK2-AP1) expression is significantly decreased in microsatellite unstable human colorectal cancer cell lines and tissues. We have published results indicating the decreased expression of CDK2-AP1 is associated with mutations in the poly (T) 8 repeat sequence within the 3’ UTR of the gene (Oncogene 2005 24:3657-3668). However, the true func-

Introduction: Alterations leading to the inhibition of the TGF-␤ signaling pathway are frequently observed in cancer. Several recent studies suggest that silencing of the type II TGF-␤ receptor is associated with pancreatic cancer development and progression. The mechanism underlying this transcriptional control remains elusive. Here we present data on a novel TGF-␤inducible mediated mechanism of transcriptional repression through the Sp/KLF family of transcription factors. Methods/ Results: Using a combination of bioinformatics and gene expression analyses we identified several Sp/KLF’s that are inducible by TGF-␤ 1 which have not been previously reported. Using laboratory constructed cloned expression vectors for each of our candidate proteins and luciferase reporters for the type II TGF-␤ receptor promoter; we identified KLF14, a Sp/KLF transcription factor, as a novel TGF-␤1 inducible repressor of the type II TGF-␤ receptor. KLF14 showed marked transcriptional repression of both the full-length TGF␤RII promoter as well as a truncated promoter containing several putative “Sp1-like” binding sites at a wide range of concentrations. RT-PCR was used to validate this effect on endogenous type II TGF-␤ receptor levels. Using chromatin immunoprecipitation assays we confirmed that KLF14 directly binds to the type II TGF-␤ receptor promoter and this binding is enhanced by TGF-␤1. In order to identify the specific physical sites of repression on the promoter sequence, we created

ASSOCIATION FOR ACADEMIC SURGERY AND SOCIETY OF UNIVERSITY SURGEONS—ABSTRACTS deletion and site-directed mutant constructs of the TGF␤RII promoter in addition to performing gel-shift assays which confirmed that KLF14 represses TGF␤RII activity through ‘Sp1-like’ sites located on the proximal promoter. Moreover, KLF14 antagonizes the usual Sp1 transcriptional activation by competition for this site to determine the overall relative transcriptional activity. Analysis of this repressive transcriptional mechanism show HDAC and mSin3a, known co-repressors, are recruited by KLF14 to this promoter. Further studies showed that KLF14 is able to induce specific repressive chromatin modifications. Looking at the chromatin landscape we found an increase in methylation markers and a decrease in acetylation markers. Furthermore using luciferase reporter assays, we found that KLF14 is able repress the promoter of other downstream target genes of the TGF-␤ cascade such as p21, which is an important inhibitor of cell cycle progression. This was confirmed using RT-PCR on endogenous p21 levels. Conclusion: In summary, KLF14 is a novel TGF-␤inducible transcriptional repressor of the type II TGF-␤ receptor in pancreatic cancer that leads to repressive chromatin modifications consistent with gene silencing. This is the first known description of an alternative mechanism of inhibition of the TGF-␤ pathway. Such findings support the role of the type II TGF-␤ receptor as a tumor suppressor which may be silenced in pancreatic cancer, and KLF14 may play a role as an important therapeutic target for this devastating disease.

ONCOLOGY II: APOPTOSIS 39. THE ESSENTIAL ROLE OF REACTIVE OXYGEN SPECIES IN CISPLATIN-INDUCED SENSITIZATION OF MALIGNANT PLEURAL MESOTHELIOMA CELLS TO FAS LIGAND MEDIATED APOPTOSIS. J. H. Stewart, IV 1, W. Tsai 2, D. S. Schrump 3, D. M. Nguyen 3; 1Wake Forest University School of Medicine, Winston-Salem, NC, 2The University of Pittsburgh, Pittsburgh, PA, 3National Cancer Institute, Bethesda, MD Background: Despite adequately expressing Fas, the majority of cultured malignant pleural mesothelioma (MPM) cells are refractory to the cytotoxic effect of soluble Fas ligand (sFasL). Pre-treating FasL-resistant cells with sublethal concentrations of cisplatin (CDDP) sensitizes them to subsequent exposure to FasL. The goal of this study is to evaluate the role of reactive oxygen species (ROS) in CDDP-mediated enhancement of sFasL cytotoxicity. Methods: Nine Fas-positive MPM cells were treated with sFasL (5 to 100 ng/ml x 48 hours) with or without pretreatment with CDDP (0.5 to 4 microg/ml x 24 hours). Cell viability and apoptosis were determined by MTT and TUNEL-based ApoBrdU assays. Generation of ROS in control and treated cells were quantified using H2DCFDA dye and flow cytometry. Mitochondrial inner membrane potential was determined by JC-1 staining and flow cytometry. Activated Bax was detected by immunoprecipitation and immunoblotting. Specific caspase proteolytic activity was quantified by colorimetric assays. Results: Three of nine MPM cells were sensitive to sFasL (IC50 values ⬍100 ng/ml). CDDP significantly sensitized MPM cells, regardless of their intrinsic susceptibility to sFasL cytotoxicity, to this apoptosis-inducing ligand with 2- to ⬎ 10-fold reduction of sFasL IC50’s. While CDDP (0.5 or 1.0 microg/ml) or sFasL (50 ng/ml) mediated ⬍20% apoptosis, 80% to 95% of combination-treated cells were apoptotic. ROS generation in CDDP/sFasL-treated MPM cells (detected by 2-fold increase of DHDFA fluorescence) played an important role in combination treatment-induced cytotoxicity as the anti-oxidant N-acetylcysteine not only eliminated ROS production but also completely abrogated the cytotoxic effect of this drug combination. Moreover, cleavage of BID and release of cytochrome c from mitochondria, collapse of mitochondria inner membrane potential as well as activation of caspase 8, 9 and 3 in CDDP/sFasL-treated MPM cells were all suppressed by N-acetylcysteine. Moreover, treating MPM cells with

167

100 microM of H2O2 for 2 hours increased the sensitivity of MPM cells to sFasL. Finally, over-expression of Bcl2 or concurrent exposure to selective caspase 9 inhibitor also abrogated CDDP/sFasL- or H2O2/sFasL-mediated cytotoxicity. Conclusion: Cisplatin sensitizes MPM cells to sFasL-mediated cytotoxicity and apoptosis via molecular pathway(s) that depend(s) on ROS and mitochondriaassociated death signaling cascade. 40. TRA-8 (TRAIL-R2 ANTIBODY) BASED COMBINATION CHEMOTHERAPY PRODUCES A SURVIVAL BENEFIT IN A PANCREATIC CANCER ORTHOTOPIC MODEL. J. W. Long1, L. C. Derosier1, S. M. Vickers1, J. Sellers2, W. Wang3, J. P. Arnoletti1, D. J. Buchsbaum2; 1UAB Medical CenterDepartment of Surgery, Birmingham, AL, 2UAB Medical CenterDepartment of Radiation Oncology, Birmingham, AL, 3UAB Medical Center-Department of Medicine, Birmingham, AL Background: TRA-8 is an agonistic monoclonal antibody that binds selectively to the Tumor Necrosis Factor-Related ApoptosisInducing Ligand Receptor 2, TRAIL-R2 (DR5). TRA-8 has been shown to increase the in vitro cytotoxicity of standard chemotherapy drugs and inhibit tumor growth in vivo in sensitive cell lines. Recent meta-analysis has shown that adding platinum drugs to standard Gemcitabine (Gem) chemotherapy increases survival time. The purpose of this study was to evaluate the survival benefit of TRA-8 and Oxaliplatin (Ox) based combination therapy in vivo against a TRA-8 resistant pancreatic cancer model. Methods: Screening cell proliferation assays were performed to compare the sensitivity of S2013 human pancreatic cancer cells to TRA-8 therapy alone versus combination therapy of TRA-8 with Oxaliplatin and Gemcitabine (TOG) or Irinotecan (TOI). S2013 cells were injected into the pancreas of SCID mice. In the first study, animals were injected I.P. with 4 doses of 200 ␮g TRA-8 (days 19, 22, 26, and 29) and 2 doses of 120 mg/kg Gem (days 20 and 27) or with 4 doses of TRA-8 and 4 doses of 25 mg/kg Irinotecan (days 20, 23, 27 and 30). In the second study, mice received 4 doses of 6 mg/kg Oxaliplatin (Ox) (days 20, 23, 27, and 30) in addition to 4 doses of TRA-8 and 2 doses of Gem 500 mg/kg. A final study investigated the effects of TOG versus TOI in which animals were treated with either 6 doses of TRA-8, 6 doses of Ox and 3 doses of Gem or TRA-8 and Ox (same dosing schedule) and 6 doses of Irinotecan. Survival results were compared against non-treated controls using a two-sided log-rank test. The significance level was adjusted for the multiple comparisons using Bonferroni’s method. Results: In vitro cellular proliferation assays showed no inhibition of proliferation in S2013 cells treated with only TRA-8. When cells were treated with TRA-8 and Gem or Irinotecan, mean cell counts fell 38.5% and 60.3% compared to untreated cells, respectively. Combination treatment with TOG or TOI produced further reductions in cell counts of 57.2% and 64.4%, respectively. In vivo, the TRA-8/Gem group had a mean survival time of 51.1 ⫾ 2.0 days (7.5 day increase over control; p ⫽ 0.0493) and the TRA-8/Irinotecan group had a mean survival of 53.8 ⫾ 2.1 days (10.2 day increase over control, p ⫽ 0.0026). When Ox was added to the TRA-8/Gem combination, mean survival was 61.6 ⫾ 3.6 days (16 day increase, p ⫽ 0.0001). Mice treated with the higher dose regimen of TOG showed further improvement with a 17.5 day increase in mean survival over untreated controls (63.2 ⫾ 1.9 days, p ⬍ .0001). Those treated with TOI had the greatest mean survival time of 70.4 ⫾ 1.9 days (24.7 day increase over control; p ⬍ .0001). Conclusions: TRA-8 combination therapy with Oxaliplatin and Irinotecan produced the most significant survival benefit against a TRA-8 resistant pancreatic cancer cell line in an orthotopic model. This approach may be useful for treatment of localized or metastatic pancreatic cancer. 41. BOTH IKK␣ AND IKK␤ ARE INVOLVED IN THE ACTIVATION OF NF⌲B IN THE CHEMORESISTANCE OF SARCOMAS. B. Bednarski, A. Baldwin, X. Ding, H. J. Kim; University of North Carolina, Chapel Hill, NC