From genes to function


67

Fibrinogen-like protein 1 is a hepatoprotectant

Chinweike Ukomadu1, Anal Desai1, Valeriy Demchev1, Hamed Nayeb-Hashemi1, Agoston Agoston1, Xintong Chen2, Joana F. Neves1, Richard S. Blumberg1, Yujin Hoshida2

1Brigham and Women's Hospital, Boston, MA; 2Medicine, Icahn School of Medicine at MT. Sinai, New York, NY

Background: Fibrinogen like protein 1(Fgl1) is a hepatocyte secreted protein whose expression increases following acute liver injury. Fgl1 stimulates uptake of tritiated thymidine into hepatocytes and activates pathways involved in hepatocyte proliferation. As such, it is believed that Fgl1 functions as a hepatocyte mitogen. We previously showed that Fgl1 is an acute phase protein and have recently generated mice with targeted loss of Fgl1.Aim: To determine the role of Fgl1 during liver injury. Methods: We used two acute injury models. 1)We evaluated mice wild type (Fgl1+/+) or null (Fgl1-/-) for Fgl1 at 24 hour intervals (0 h-120 h) following a one-time administration of CCL4 (0.4 mg/kg). 2) We administered alpha-galactosylceramide (α-GalCer) (2μg/mouse), an inducer of natural killer T cell mediated liver injury to Fgl1 +/+ and Fgl1-/- mice. We determined ALT levels and collected liver tissue at 24 h post injection. For chronic liver injury, we injected mice twice weekly with CCL4 (0.2 mg/kg) for four weeks and then performed gene-expression microarray analyses on liver tissues. We used gene set enrichment analysis to identify biologic pathways affected by Fgl1.Results: Following acute CCL4 administration we found a lag in the time to repair of liver injury in the Fgl 1-/mice, with injury in Fgl 1 +/+ largely resolved by 72 h post injection while it persisted until 120 h in the Fgl1-/- mice. We found no difference in hepatocyte proliferation by PCNA staining, suggesting that this effect is not due to loss of Fgl1 mitogenic activity. We found no differences in the number, morphology and distribution of bile ducts, stellate cells and endothelial cells by immunohistochemistry. Mice treated with αGalCer also showed a 4 fold increase in ALT and marked hepatocellular necrosis in Fgl1-/- when compared to the Fgl1+/+ mice. Consistent with a role in protection from inflammatory liver injury, chronic administration of CCL4 results in a more sustained inflammatory injury. Gene-expression microarray data implicate pathways involved in inflammation and hepatocyte injury (for example Toll, NFKB, IL22 soluble receptor and IL1 receptor pathways) as upregulated in the Fgl1 null mice. Conclusions: Mice with targeted deletion of Fgl1 have more pronounced acute and chronic liver injury when compared to wild type mice. Gene expression data suggest that pathways that protect hepatocytes from injury are perturbed in the Fgl1 null mice. Given its role as an acute phase reactant, we speculate that Fgl1 is enhanced following injury to protect hepatocytes from damage. Future studies will delineate specific immunologic targets of Fgl1.

Disclosures:

Chinweike Ukomadu - Consulting: Gilead Sciences

The following people have nothing to disclose: Anal Desai, Valeriy Demchev, Hamed Nayeb-Hashemi, Agoston Agoston, Xintong Chen, Joana F. Neves, Richard S. Blumberg, Yujin Hoshida

68

Integrative genomic analyses of human fibrolamellar hepatocellular carcinoma

Helena Cornella1, Clara Alsinet1, Ke Hao2, Laia Cabellos2, Alberto Quaglia3, Zhongyang Zhang2, Xintong Chen2, Augusto Villanueva1, Yujin Hoshida2, Nasra H. Giama4, David M. Nagorney4, Swan N. Thung2, Stephen C. Ward2, Leonardo Rodriguez-Carunchio1, Anja Lachenmayer2, Beatriz Minguez5, Lewis R. Roberts4, Vincenzo Mazzaferro6, Myron Schwartz2, Nigel Heaton3, Josep M. Llovet1, 2

1HCC Translational Research Laboratory, Barcelona Clinic Liver Cancer Group (BCLC), Liver Unit, Hospital Clinic, IDIBAPS, CIBEREHD, University of Barcelona, Barcelona, Spain; 2Mount Sinai Liver Cancer Program, Department of Medicine; Department of Surgery; Department of Pathology; Department of Genetics and Genomic Sciences, Icahn School of Medicine af Mount Sinai, New York, NY; 3Institute of Liver Studies, King's College Hospital, London, United Kingdom; 4Division of Gastroenterologic and General Surgery; Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN; 5Liver Unit, Hospital Vall d'Hebron, Barcelona, Spain; 6Gastrointestinal Surgery and Liver Transplantation Unit, Istituto Nazionale dei Tumori, Milan, Italy

Introduction: Fibrolamellar hepatocellular carcinoma (FLC) arises in non-cirrhotic livers of children/young adults with no etiologic factor or gender bias, representing <1% of all liver cancers. Surgical resection is the main treatment option achieving a 70% 5yr survival, hampered by tumor recurrence. Understanding of the molecular alterations and potential drivers of FLC is limited. Aims: 1)Identify novel drivers as therapeutical targets and 2)provide a molecular classification of FLC. Methods: Formalin-fixed paraffin-embedded tissue from 40 FLC patients from 6 referral hospitals(US, Europe) were included, and results compared to genomic data from 164 hepatocellular carcinomas (HCC) and 149 intrahepatic cholangiocarcinomas (ICC). DASL gene expression was analysed using NMF and CMS(GenePattern) for class discovery and differential expression. GSEA and IPA were used for functional annotation. SNP array(Human〇 mniExpress) and GISTIC2 analyses were used to evaluate copy number variations(CNV). Whole exome-sequencing (HiSeq2000; 50X) was run on 4 FLC-normal liver pairs and mutations were annotated by GATK, SIFT and PolyPhen2.Results: FLC patients (median 25yr-old) were surgically-treated (resected 89%, transplanted 11%), female (58%), non-cirrhotic (98%), without viral hepatitis (95%), 11cm median tumor size (7-13cm), 24% metastasized and presented a median survival of 58mo. Unsupervised NMF clustering revealed 3 molecular classes: FLC-proliferation (FLC-P) (18/35, 51%) enriched with liver proliferation signatures (HCC-G2, FDR=0.04, ICCP, FDR<0.07) and mTOR signaling activation (pRPS6IHC, p=0.03); FLC-inflammation (FLC-I) (9/35, 26%) enrichedwith pro-inflammatory cytokines signaling (AcutePhaseResponse, p<0.01), ICC-I class signature (FDR<0.01) and less aggressive phenotype (lack of vascular invasion, p=0.015); FLC-unannotated (8/35, 23%) enriched in non-liver related cancer signatures (MolecularMechanismsCancer, p<0.006). Class associations were confirmed by unsupervised clustering of 348 liver cancers; FLC-P samples co-clustered with ICC-P and HCCP samples, while FLC-I remained near ICC-I samples. FLC showed low number of CNVs vs HCC and ICC: 6p27 amplification (12%) and deletions at 1p36.32 (12%) and 19p13.3 (24%). FLCs showed a mean of 136 somatic mutations, 4 nonsynonymous, involving a total of 12 genes. Prevalent mutations in HCC (TP53, CTNNB1, ARID1A) were not found. Two mutations (ZNF607, SSTR5) were scored as damaging, being ZNF607 mutation validated by Sanger sequencing. Conclusions: Genomic profiling of FLC reveals 3 molecular classes and heterogenic CNVs. Exome sequencing pilot study identifies a distinctive mutation portrait compared with HCC.

Disclosures:

Lewis R. Roberts - Advisory Committees or Review Panels: Inova; Grant/Research Support: Bristol Myers Squibb, Bayer, Nordion; Speaking and Teaching: Nordion

Vincenzo Mazzaferro - Advisory Committees or Review Panels: Bayer; Grant/Research Support: Nordion; Speaking and Teaching: Merck Serono S. p. A.

Myron Schwartz - Consulting: Gilead, Inova

Nigel Heaton - Advisory Committees or Review Panels: Novartis, Roche; Speaking and Teaching: Astellas

Josep M. Llovet - Consulting: Bayer Pharmaceutical, Bristol Myers Squibb, Imclone, Biocompatibles, Novartis; Grant/Research Support: Bayer Pharmaceutical, Bristol Myers Squibb, Boehringer-Ingelheim

The following people have nothing to disclose: Helena Cornella, Clara Alsinet, Ke Hao, Laia Cabellos, Alberto Quaglia, Zhongyang Zhang, Xintong Chen, Augusto Villanueva, Yujin Hoshida, Nasra H. Giama, David M. Nagorney, Swan N. Thung, Stephen C. Ward, Leonardo Rodriguez-Carunchio, Anja Lachenmayer, Beatriz Minguez

69

Targeting PRAJA1 to govern TERT gene regulation via TGF-β signaling pathway in hepatocellular cancer

Jian Chen, Jiun-Sheng Chen, Vivek Shukla, Zhixing Yao, Wilma Jogunoori, Bibhuti Mishra, Lopa Mishra

MD Andrson Cancer Center, Houston, TX

Purpose: Hepatocellular Cancer (HCC) is a rising and lethal disease, that is difficult to treat due to late diagnosis with few viable targeted therapeutics. Recent studies demonstrate a high frequency of TERT promoter mutations in early stage HCCs, suggesting that these promoter mutations may function as driver events that contribute to oncogenesis through TERT dysregulation in HCCs. However, telomerase remains a challenge to target effectively. We have previously found that deletion of Smad3/4 adaptor β2SP results in spontaneous HCC with loss of TGF-β signaling in mouse model. CCCTC-Binding Factor CTCF is a highly conserved zinc finger protein that has diverse regulatory functions, including transcriptional activation/repression/imprinting of molecules such as IGF-2, c-Myc and TERT. More importantly, our data reveal that Smad3/β2SP/ the substrates of PRAJA1, forms a complex with CTCF regulating TERT on its promoter region. Therefore, our hypothesis is that inhibiting PRAJA1 may suppress TERT by rescuing TGF-β pathway via stabilizing the p2SP/Smad/ CTCF complex. Moreover, triterpenoids targeting PRAJA1 successfully reduce tumor burden with inhibition of telomerase. Materials & Methods: Database of HCC Genomics (COSMIC) and transcriptomics (TCGA) were analyzed. Whole mount in situ hybridization histochemistry assay was used to determine knock down PRAJA1 in zebrafish embryos. Soft agar assay and colony formation assay were performed to elucidate PRAJA1 oncogenic activity in HCC cells.

Results: (1)Genomics and transcriptomics analyses revealed aberrant TGF-β signaling in 70% of HCCs. (2) p2SP+/-Smad3+/- mice develop visceromegaly, multiple cancers, spontaneously and increased the levels of TERT, phenocopy the hereditary human cancer syndrome Beckwith-Wiedemann. (3) TGF-β promotes the complex of p2SP/Smad3/CTCF at TERT promoter region. (4) PRAJA1 expression is dramatically raised in human HCCs with loss of TGF-β signaling. (5) PRAJA1 interacts with p2SP/Smad3 and downregulates CTCF in a TGF-βdependent manner. (6) Inhibition of PRAJA1 in developing Zebrafish embryos and HCCs leads to high levels of apoptosis. (7) RTA402 and RTA405 inhibit PRAJA1 and restore TGF-β tumor suppressor function in HCC cells. Conclusions: PRAJA1 upregulates TERT gene expression via disrupting TGF-β pathway in HCC. Small molecule inhibitors such as triterpenoids that specifically target PRAJA1 could be very useful in HCC therapy, through targeting TERT, restoring TGF-β tumor suppressor function. This study may lead to new therapeutics targeting this lethal cancer and potentially to a Phase I clinical trial in HCC.

Disclosures:

The following people have nothing to disclose: Jian Chen, Jiun-Sheng Chen, Vivek Shukla, Zhixing Yao, Wilma Jogunoori, Bibhuti Mishra, Lopa Mishra

70

KLF14 transcriptionally activates sphingosine kinase 1 and sphingosine-1-phosphate production in response to FGF2 in liver endothelial cells linking metabolic syndrome with hepatic stellate cell activation

Thiago de Assuncao, Sheng Cao, Gwen Lomberk, Usman Yaqoob, Yan Bi, Angela Mathison, Raul A. Urrutia, Vijay Shah

Gastroenterology Research Unit, Mayo Clinic, Rochester, MN

Background/Aims: Lipids contribute to the development of liver disease by inducing hepatocyte cell death and stimulating fibrosis. Sphingosine kinase 1 (SK1) is an endothelial cell (EC) enzyme responsible for generation of sphingosine-1-phosphate (S1P), a lipid molecule implicated in the activation of hepatic stellate cells (HSC). Since binding of fibroblast growth factor (FGF2) to its cognate receptor FGF receptor 1 (FGFR1) leads to EC activation, we hypothesized that that this pathway may stimulate EC to produce SK1 as part of a lipid signaling cascade that regulates HSC activation. Methods/Results: S1P (0.5 μm) increased HSC chemotaxis in Boyden cell migration assay (vehicle: 45.8±26 vs S1P: 1 74.67±68; p<0.05) and stimulated HSC contractility as assessed by increase in phalloidin staining of actin stress fibers. In vivo, mRNA levels of SK1 and FGFR1 were upregulated in human cirrhotic liver tissue compared to normal liver by qPCR. In isolated liver EC, FGF2 upregulated SK1 based on qPCR (2-fold; p<0.05), Western blot, and ELISA (2-fold; p<0.05). Studies using the 1.9-Kb SK1 promoter and several deletion mutants revealed that the FGF2/FGFR1 pathway regulated the expression of SK1 at the level of transcription. Highest basal and FGF2 stimulated-promoter activity was mapped to two GC-rich regions located within 633 bp from the transcription start site (p<0.05). Sitedirected mutagenesis demonstrated that disruption of these GCrich sites resulted in a 5-7 fold decrease in basal and FGF2 stimulated promoter activity. Screening for GC-rich binding transcription factors that could activate this site demonstrated that KLF14, a gene implicated in metabolic syndrome, binds to this region. Congruently, overexpression of KLF14 increased basal and FGF2 stimulated WT SK1 promoter activity by 3-fold (p<0.05), but not upon mutation of the GC-rich sites. In addition, KLF14 siRNA transfection decreased SK1 mRNA levels by 3-fold (p<0.05) and SK1 protein levels in presence andabsence of FGF2 stimulation. Finally, SK1 mRNA and protein levels were decreased in livers from KLF14 knockout (ko) mice compared to wild-type mice (WT: 2.9±0.28 vs KLF14ko: 1.17±0.32 p<0.05). Conclusions: These results show the importance of FGF2 and KLF14 in the activation of the SK1 gene in liver EC and potentially link metabolic syndrome with HSC activation through EC derived S1P.

Disclosures:

The following people have nothing to disclose: Thiago de Assuncao, Sheng Cao, Gwen Lomberk, Usman Yaqoob, Yan Bi, Angela Mathison, Raul A. Urrutia, Vijay Shah

71

Activity of NMDA Receptors and Roles in Hepatic Injury Offer Novel Mechanisms for Liver Diseases and Therapeutic Development

Nicole Pattamanuch2,1, Preeti Viswanathan2,1, Sylvia O. Suadicani1, David C. Spray1, Sanjeev Gupta1

1Marion Bessin Liver Research Center, Albert Einstein College of Medicine, Bronx, NY; 2Pediatrics, Children's Hospital at Montefiore, Bronx, NY

Expression of N-methyl-D-aspartate receptors (NMDARs) is classically associated with excitoxic injury in neuronal tissues, e. g., ischemic or traumatic insults, Alzheimer's, Parkinson's, schizophrenia, etc. This spurred wide interest in drugs to suppress NMDAR activity. However, NMDARs display multiple subunits and ligand-binding sites, which add complexities in identifying their activity. Few studies have been devoted to NMDARs in nonneural tissues, and presence of NMDAR activity in liver has not been defined. Nonetheless, in liver failure, plasma ammonia may induce neurotoxicity via NMDARs in brain, and previously NMDAR antagonists, e. g., MK801 and memantine, improved survival in animals with liver failure. Recently, neurotropic receptor expression was unexpectedly identified in liver after a period of anoxia, which led us to hypothesize that NMDARs may directly contribute in hepatotoxicity. To develop this possibility, we cultured HuH-7 cells and primary mouse hepatocytes with or without NMDA and acetaminophen (APAP), MK801, and memantine. MTT assays were performed to assess cytotoxicity. Intracellular Ca++ fluxes were measured in hepatocytes with NMDA and NMDAR blockers. Brain and liver tissues were examined for multiple NMDARs by RT-PCR, western, and immunostaining. C57BL/6 mice were used for studies with 500 mg/kg APAP, 2 mg/kg MK801 or 30 mg/kg memantine, Besides mortality, liver injury was evaluated by histology and liver tests. We found NMDAR were expressed in liver at RNA and protein levels. Moreover, HuH-7 cells and mouse hepatocytes were sensitive to NMDA with cytotoxicity as shown by MTT assays. Although APAP-induced cytotoxicity in HuH-7 cells or mouse hepatocytes was not potentiated by simultaneous presence of NMDA, it was abolished when cells were cultured with APAP plus either MK801 or memantine. In mouse hepatocytes, NMDA dose-dependently induced intracellular Ca++ fluxes. APAP alone did not directly stimulate intracellular Ca++ fluxes, as was expected. By contrast, MK801 and memantine blocked intracellular Ca++ oscillations. In APAPtreated mice, we observed significant mortality, liver necrosis and liver test abnormalities. When mice treated with APAP were given MK801 or memantine, survival of animals was prolonged and liver histology improved. Conclusions: The NMDARs were expressed in hepatocytes, especially after liver injury, and contributed to APAP-induced hepatotoxicity. Decreases in hepatic injury after blockade of NMDARs by MK801 or memantine indicated further studies of hepatic NMDARs will be helpful. In particular, pathophysiological studies of NMDARs in liver diseases should be relevant for their therapeutic implications.

Disclosures:

The following people have nothing to disclose: Nicole Pattamanuch, Preeti Viswanathan, Sylvia O. Suadicani, David C. Spray, Sanjeev Gupta

72

Hepatic Lipid Droplet Metabolism is Dependent upon Autophagic Lysosomal Dynamics that are Regulated by the Large GTPase Dynamin 2

Ryan Schulze1, Shaun Weller1, Barbara Schroeder1, Eugene W. Krueger1, Susan Chi1, Carol A. C asey 2, Mark A. McNiven1

1Biochemistry ond Molecular Biology, Moyo Clinic, Rochester, MN; 2Internal Medicine, University of Nebraska Medical Center, Omaha, NE

Lipid droplets (LDs) are the major cellular storage sites of esterified fatty acids and are the central organelle contributing to hepatic steatosis. The specific machinery orchestrating the breakdown of these structures remains unclear. The goal of this study was to further define the hepatocellular machinery that supports LD metabolism. In hepatocytes, a role for autophagy has been proposed as one mechanism for LD turnover in response to starvation. We hypothesized that the large GTPase Dynamin 2 (Dyn2), well known to support membrane remodeling and trafficking events throughout the cell, might participate in either the vesiculation, or the autophagic breakdown, of LDs. Results: Indeed, either depletion or pharmacologic inhibition of Dyn2 results in a substantial accumulation of LDs in hepatocytes. Surprisingly, co-localization and biochemical experiments suggest that Dyn2 does not associate directly on LDs. Instead, we observe by electron and immunofluorescence microscopy that the targeted disruption of Dyn2 function induces a dramatic 4- to 5-fold increase in the size of autophagic autolysosomal compartments. Moreover, Dyn2 inhibition results in the extensive tubulation of the autolysosomal membrane. These tubules exhibit numerous varicosities and constrictions, as if a scission process has been halted. Importantly, upon restoration of enzymatic function, Dyn2 associates along the length of these tubules, resulting in the vesiculation and fragmentation of the autolysosomal membranes. Rescue of Dyn2 function results in the restoration of LD breakdown. Conclusion: We predict that Dyn2 participates in autophagic lysosomal reformation, a poorly-studied process of lysosomal regeneration from autolysosomal membranes during starvation conditions. The inhibition of Dyn2 therefore results in an inability to repopulate the cellular lysosome pool, preventing further LD degradation by autophagy. This data provides new evidence for the participation of the autolysosome in hepatic LD catabolism and implicates a novel role for Dyn2 in mediating the function and biogenesis of autophagic compartments. This study was supported by grants 5R37DK044650 (MAM), 5R01AA020735 (MAM and CAC), 5T32DK007352 (RJS), NIH Challenge Grant AA19032 (mAm and CAC), and funding from the Robert and Arlene Kogod Center on Aging.

Disclosures:

The following people have nothing to disclose: Ryan Schulze, Shaun Weller, Barbara Schroeder, Eugene W. Krueger, Susan Chi, Carol A. Casey, Mark A. McNiven

Ancillary