- Email: [email protected]
Reactive oxygen as modulator of TNF and Fas receptor-mediated apoptosis in vivo: studies with glutathione peroxidase-deficient mice
274A
AASLD ABSTRACTS
HEPATOLOGY October 2001
407
408
INCREASED RESISTANCE OF RAT ItEPATOCYTES AGAINST APOPTOSIS IN BILIARY FIBROSIS IS DUE TO CYTOKINE A N D BILE ACID IND U C E D EXPRESSION OF ANTI-APOPTOTIC GENES. Marieke H Schoe-
REACTIVE OXYGEN AS MODULATOR OF TNF A N D FAS RECEPTORMEDIATED APOPTOSIS IN VIVO: STUDIES W I T H GLUTATHIONE PEROXIDASE-DEFICIENT MICE. Mary Lynn Bajt, University of Arkansas for
maker, Manon Homan, Alexandra Beuving, Harry Van Goor, University Hospital Groningen, Groningen Netherlands; Klaas Poelstra, University of Groningen; Dept Pharmacy, Groningen Netherlands; Peter L Jansen, Han Moshage, University Hospital Groningen, Groningen Netherlands
Medical Sciences, Little Rock, AR; Ye-Shih Ho, Wayne State University, Detroit, MI; Hartmut Jaeschke, University of Arkansas for Medical Sciences, Little Rock, AR
Background: In biliary fibrosis, bepatocytes are exposed to increased levels of bile acids. Bile acids are known to induce apoptosis in hepatocytes in vitro. However, little is known about the extent of apoptosis in hepatocytes and their adaptation to become resistant to apoptosis in biliary fibrosis. Since, cholestasis is accompanied by increased portal blood levels of endotoxin this may result in increased intrahepatic generation of cytokines, leading to the synthesis of NF-~B-regulated survival genes. Questions: 1) Is apoptotic cell death of hepatocytes important in biliary fibrosis? 2) Are there changes in the expression of apoptosis-related genes, in particular Bel-2 family members and members of the IAP (lnhibitors of Apoptosis) family. 3) Do bile acids and cytokines change the expression of these genes? Methods: Bile duct ligatiou (BDL) was used as model of biliary fibrosis. Sham operated animals served as controls. Apoptosis was determined by measuring caspase-3 activity in liver tissue and by immunohistochemical staining for activated caspase-3 and caspase-cleaved cytokeratin-18, mRNA levels of the Bcl-2 family members Bcl-2, Bcl-xl, A1 (anti-apoptutic), Bax and Bak (pro-apoptotie) were determined by RT-PCR. Bcl-2 expression was also anaIyzed by immunohistochemistry and Western blotting. In addition, mRNA expression of the IAP family members cIAP-1, cIAP-2 and XIAP and the inflammatory eytokine TNFa was determined. Primary cultures of rat hepatocytes were exposed for 12 and 24 hrs to a cytokine mixture and to glycochenodeoxycholic acid (GCDCA; 25, 50#M) and tauroursodeoxycholic acid (TUDCA; 50, 100/*M). mRNA levels of Bd-2 and IAP family members were analyzed by RT-PCR. In some experiments, the NF-KBpathway was blocked by transfection with an adenovims expressing dominant negative IKB (Ad5IKB). Results: In vivo: Hepatic caspase-3 activity was transiently increased, peaking at day 4 and almost returned to normal at day 7 after BDL. Immunohistochemistry for activated caspase-3 and easpase-3-deaved cytokeratin-18 revealed very few apoptotic hepatocytes at day 4 and day 7 after BDL. The mRNA expression of Bcl-2, A1, Bak, clAP-2 and TNF~ was significantly increased 1 week after BDL. The mRNA expression of Bcl-xl, Bax, cIAP-1 and XIAP was not changed. Immunohistochemistry revealed strong expression of Bd-2 in bile duct epithelialcells and weakly increased expression in hepatocytes. In vitro: In cultured rat bepatocytes, the mRNA expression of A1, Bak and cIAP-2 was induced by cytokines in an NF-KB dependent manner. Bcl-2 expression was induced by GCDCA, but not by TUDCA. Neither cytokines nor bile acids changed the expression of Bcl-xl, Bax, clAP-1 and XlAP. Conclusion: In the early stages of biliary fibrosis in the rat there is only a transient rise in easpase activity, and apoptosis of hepatocytes is limited. The increased resistance against apoptosis may be due m increased expression of anti-apoptofic AI, cIAP-2 and Bcl-2, despite increased expression of pro-apoptotic Bak. Inflammatory cytokines are induced in bifiary fibrosis and regulate the expression of A1, cIAP-2 and Bak, whereas Bcl-2 is regulated by the toxic bile acid GCDCA. Our results suggest that hepatocytes adapt to apoptosis in biliaryfibrosis and that NF-KBactivation may be involved in this adaptation.
Reactive oxygen species (ROS) can directly induce or enhance TNF-mediated apoptosis in a number of different cell lines. Exposure to TNF causes intracellular ROS formation (Goossens et al., Proc Natl Acad Sci USA 92: 8115-9, 1995). To test the relevance of intracellular ROS in modulating apoptotic signaling in vivo, we evaluated hepatocellular apoptosis mediated by the TNF or Fas receptor in wildtype and glutathione peroxidase-1 (Gpx-1-/-) deficient mice (129SV/B6 background). Previous studies with these animals demonstrated the redox-sensitivity of endotoxin-induced NF-KB activation and TNF formation in Kupffer cells and the enhanced susceptibility of hepatocytes to neutrophil-induced oxidant stress in vivo (Jaeschke et al., Hepatology 29: 443-450, 1999). RESULTS: Apoptosis developed in livers of wildtype animals 4-6 h after ip administration of 700 mg/kg galactosamine/0.1 mg/kg endotoxin (G/ET). Apoptosis was indicated by processing of procaspases-3, -8, and - 9 (assessed by Western blotting), a 5-fold increase in caspase-3 activity (DEVDAMC as substrate), and a 44-fold increase in DNA fragmentation (ELISA). The time course and magnitude of apoptosis was the same in Gpx-/- mice. In contrast, plasma ALT values were significantly higher in Gpx-/- mice (5940 -+ 1300 U/L) compared to wildtype animals (1600 -+ 240 U/L) at 6 h. While liver injury was further enhanced in wildtype mice at 7 h (3950 + 1450 U/L), all Gpx-/- animals died. W h e n the dose of ET was reduced to 0.01 mg/kg, neither wildtype nor Gpx-/- mice had increased plasma ALT levels (25 -+ 5 U/L in both groups) at 6 h. Under these conditions, the increase in caspase-3 activity and DNA fragmentation was again the same in both strains of mice. Treatment of wildtype mice with the anti Fas antibody Jo-2 (0.6 mg/kg iv) resulted in processing of procaspase-3, -8, and -9, a 7-fold increase in caspase-3 activity and a 28-fold increase in DNA fragmentation at 3 h. Similar results were obtained with Gpx-/- mice. SUMMARY AND CONCLUSIONS: Our data demonstrated that Gpx-l-deficient mice developed hepatocelullar apoptosis in vivo with the same time course and to the same extent in response to TNF or Fas receptor stimulation. These data suggest that under our experimental conditions, intracellular ROS did not modulate the death receptor-initiated apoptotic signaling cascade in hepatocytes. Since Gpx-1 is located in the cytosol and in mitochondria, which are the main cellular compartments involved in apoptotic signaling, our findings indicate that the oxidant stress in vivo was insufficient to modulate these signaling pathways. (Supported in part by NIH grant ES-06091).
409
410
TRANSLOCATION OF TISSUE TRANSGLUTAMINASE FROM CYTOPLASM TO NUCLEUS PLAYS A ROLE IN ETHANOL-INDUCED APOPTOSIS IN RAT HEPATOCYIES. Liang-Wen Song, Jian Wu, UC Davis Medical
INHIBITORY ACTION OF T U M O R NECROSIS FACTOR-o~ ON HEPATOCYTE APOPTOSIS THROUGH INHIBITOR OF APOPTOSIS PROTEINS (IAPS) IN MICE. Sumiko Nagoshi, 3rd Dept of Internal Medicine, Saitama
Center, Sacramento, CA; Soichi Kojima, Tsukuba Institute, Tsukaba Japan; Mark A Zern, UC Davis Medical Center, Sacramento, CA Background: Tissue transglutaminase (tTG) is a ubiquitous protein that catalyzes the posttranscriptional modification of nuclear, cytosolic, and extracellular proteins by cross-linking via e ('y-glutamyl lysine) bonds. Recent evidence indicates that this cross-linking of specific nuclear proteins induces apoptosis. Our previous study showed that ethanol treatment causes apoptosis, inhibition of cell proliferation, and an increase in tTG cross-linking activity in rat hepatocytes (J Biol Chem 2000; 275:22213-22219). However, the mechanism by which the increased tTG activity leads to ethanol-induced hepatocyte apoptosis remains unclear. The aim of the present study is to investigate whether nuclear tTG activity increases in ethanol-treated rat hepatocytes. Methods: Rat hepatocytes were isolated by two-step collagenase digestion and incubated in Williams Medium E. tTG activity in the nuclear protein extracts was measured by [l÷C]-putrescine incorporation. Localization of tTG in the nucleus was determined by immunohistochemistry and visualized by confocal microscopy. Results: Treatment of rat hepatocytes with ethanol at 100 mM overnight led to a significant increase in [>C]-putrescine incorporation in the nuclear protein extracts compared to the untreated cells (6188-+304.9 vs. 3807-+ 155.5cpm, p<0.01). Confocal microscopy examination showed that rat hepatocytes stained with specific anti-tTG antibodies were positive for tTG in the cytoplasm and contained a very faint staining in the nucleus. Ethanol treatment caused markedly enhanced tTG signals in the nucleus. At the same time, apoptosis was observed in the cells as indicated by fragments of dense nuclei. Conclusion: Our findings indicate that ethanol-induced apoptosis was associated with enhanced nuclear tTG activity, and that this apparently resulted from cytoplasmic translocation of the enzyme.
Medical Sch, Saitama Japan; Takayuki Yoshimoto, Intractable Disease Research Center, Tokyo Medical University, Tokyo Japan; Satoshi Mochida, Kenji Fujiwara, 3rd Dept of Internal Medicine, Saitama Medical Sch, Saitama Japan Tumor Necrosis Factor-~¢ (TNF-a) produced by activated macrophages in the hepatic sinusoids regulates the development of massive liver necrosis, and also acts as a cytokine to induce liver regeneration. It is well known that TNF-a binding to its receptor activates the caspase cascade resulting in apoptosis, whereas NFKB, an endogenous transcriptional factor activated by TNF-a, can inhibit this apoptosis. When mice received a small amount of TNF-a, liver injury is not induced, but massive liver necrosis following apoptosis occurs after TNF-c¢ administration in mice pretreated with d-galactosamine (GalN), a transcription inhibitor. On the other hand, hepatocyte apoptosis after Fasligand injection was retarded in partially hepatectomised mice compared to sham operated mice, and TNF-~e pretreatment suppressed Fas-mediated hepatocyte apoptosis in normal mice. In the present investigation, the mechanisms of this inhibitory action of TNF-o¢ on apoptosis was studied in relation to inhibitor of apoptosis proteins (IAPs) which block apoptosis by binding to caspases. METHODS and RESULTS As IAPs, IAP-1, IAP-2, XIAP and survivin were used, and the expressions of these IAPs in the liver were evaluated by RT-PCR. 1) When mice received TNF-oq IAP-1 and IAP-2 were expressed in the liver from 1 to 5 hr with the peaks at 1 hr, but the expressions of XIAP and survivin showed the control levels until 8 hr. W h e n mice were injected with GalN and with TNF-cz 30 min later, the peak levels of IAP-1 and IAP-2 expressions showed the control levels, and apoptosis of hepatocytes developed with high serum ALT values. 2) IAP-1, IAP-2 and XIAP expressions were not changed from 10 mill to 5 days after 70% partial hepatectomy in mice, while survivin mRNA level was increased from 36 to 48 hr after the operation. DISCUSSION The inhibitory action of TNF-~e on apoptosis was suggested to be produced via IAP-1 and IAP-2. However, during liver regeneration, hepatocytes may be protected from apoptosis through different mechanisms. The role of survivin would be important. CONCLUSION IAP-1 and IAP-2 induced by T N F - a may act as survival factors for massive liver necrosis.
AASLD ABSTRACTS
HEPATOLOGY October 2001
407
408
INCREASED RESISTANCE OF RAT ItEPATOCYTES AGAINST APOPTOSIS IN BILIARY FIBROSIS IS DUE TO CYTOKINE A N D BILE ACID IND U C E D EXPRESSION OF ANTI-APOPTOTIC GENES. Marieke H Schoe-
REACTIVE OXYGEN AS MODULATOR OF TNF A N D FAS RECEPTORMEDIATED APOPTOSIS IN VIVO: STUDIES W I T H GLUTATHIONE PEROXIDASE-DEFICIENT MICE. Mary Lynn Bajt, University of Arkansas for
maker, Manon Homan, Alexandra Beuving, Harry Van Goor, University Hospital Groningen, Groningen Netherlands; Klaas Poelstra, University of Groningen; Dept Pharmacy, Groningen Netherlands; Peter L Jansen, Han Moshage, University Hospital Groningen, Groningen Netherlands
Medical Sciences, Little Rock, AR; Ye-Shih Ho, Wayne State University, Detroit, MI; Hartmut Jaeschke, University of Arkansas for Medical Sciences, Little Rock, AR
Background: In biliary fibrosis, bepatocytes are exposed to increased levels of bile acids. Bile acids are known to induce apoptosis in hepatocytes in vitro. However, little is known about the extent of apoptosis in hepatocytes and their adaptation to become resistant to apoptosis in biliary fibrosis. Since, cholestasis is accompanied by increased portal blood levels of endotoxin this may result in increased intrahepatic generation of cytokines, leading to the synthesis of NF-~B-regulated survival genes. Questions: 1) Is apoptotic cell death of hepatocytes important in biliary fibrosis? 2) Are there changes in the expression of apoptosis-related genes, in particular Bel-2 family members and members of the IAP (lnhibitors of Apoptosis) family. 3) Do bile acids and cytokines change the expression of these genes? Methods: Bile duct ligatiou (BDL) was used as model of biliary fibrosis. Sham operated animals served as controls. Apoptosis was determined by measuring caspase-3 activity in liver tissue and by immunohistochemical staining for activated caspase-3 and caspase-cleaved cytokeratin-18, mRNA levels of the Bcl-2 family members Bcl-2, Bcl-xl, A1 (anti-apoptutic), Bax and Bak (pro-apoptotie) were determined by RT-PCR. Bcl-2 expression was also anaIyzed by immunohistochemistry and Western blotting. In addition, mRNA expression of the IAP family members cIAP-1, cIAP-2 and XIAP and the inflammatory eytokine TNFa was determined. Primary cultures of rat hepatocytes were exposed for 12 and 24 hrs to a cytokine mixture and to glycochenodeoxycholic acid (GCDCA; 25, 50#M) and tauroursodeoxycholic acid (TUDCA; 50, 100/*M). mRNA levels of Bd-2 and IAP family members were analyzed by RT-PCR. In some experiments, the NF-KBpathway was blocked by transfection with an adenovims expressing dominant negative IKB (Ad5IKB). Results: In vivo: Hepatic caspase-3 activity was transiently increased, peaking at day 4 and almost returned to normal at day 7 after BDL. Immunohistochemistry for activated caspase-3 and easpase-3-deaved cytokeratin-18 revealed very few apoptotic hepatocytes at day 4 and day 7 after BDL. The mRNA expression of Bcl-2, A1, Bak, clAP-2 and TNF~ was significantly increased 1 week after BDL. The mRNA expression of Bcl-xl, Bax, cIAP-1 and XIAP was not changed. Immunohistochemistry revealed strong expression of Bd-2 in bile duct epithelialcells and weakly increased expression in hepatocytes. In vitro: In cultured rat bepatocytes, the mRNA expression of A1, Bak and cIAP-2 was induced by cytokines in an NF-KB dependent manner. Bcl-2 expression was induced by GCDCA, but not by TUDCA. Neither cytokines nor bile acids changed the expression of Bcl-xl, Bax, clAP-1 and XlAP. Conclusion: In the early stages of biliary fibrosis in the rat there is only a transient rise in easpase activity, and apoptosis of hepatocytes is limited. The increased resistance against apoptosis may be due m increased expression of anti-apoptofic AI, cIAP-2 and Bcl-2, despite increased expression of pro-apoptotic Bak. Inflammatory cytokines are induced in bifiary fibrosis and regulate the expression of A1, cIAP-2 and Bak, whereas Bcl-2 is regulated by the toxic bile acid GCDCA. Our results suggest that hepatocytes adapt to apoptosis in biliaryfibrosis and that NF-KBactivation may be involved in this adaptation.
Reactive oxygen species (ROS) can directly induce or enhance TNF-mediated apoptosis in a number of different cell lines. Exposure to TNF causes intracellular ROS formation (Goossens et al., Proc Natl Acad Sci USA 92: 8115-9, 1995). To test the relevance of intracellular ROS in modulating apoptotic signaling in vivo, we evaluated hepatocellular apoptosis mediated by the TNF or Fas receptor in wildtype and glutathione peroxidase-1 (Gpx-1-/-) deficient mice (129SV/B6 background). Previous studies with these animals demonstrated the redox-sensitivity of endotoxin-induced NF-KB activation and TNF formation in Kupffer cells and the enhanced susceptibility of hepatocytes to neutrophil-induced oxidant stress in vivo (Jaeschke et al., Hepatology 29: 443-450, 1999). RESULTS: Apoptosis developed in livers of wildtype animals 4-6 h after ip administration of 700 mg/kg galactosamine/0.1 mg/kg endotoxin (G/ET). Apoptosis was indicated by processing of procaspases-3, -8, and - 9 (assessed by Western blotting), a 5-fold increase in caspase-3 activity (DEVDAMC as substrate), and a 44-fold increase in DNA fragmentation (ELISA). The time course and magnitude of apoptosis was the same in Gpx-/- mice. In contrast, plasma ALT values were significantly higher in Gpx-/- mice (5940 -+ 1300 U/L) compared to wildtype animals (1600 -+ 240 U/L) at 6 h. While liver injury was further enhanced in wildtype mice at 7 h (3950 + 1450 U/L), all Gpx-/- animals died. W h e n the dose of ET was reduced to 0.01 mg/kg, neither wildtype nor Gpx-/- mice had increased plasma ALT levels (25 -+ 5 U/L in both groups) at 6 h. Under these conditions, the increase in caspase-3 activity and DNA fragmentation was again the same in both strains of mice. Treatment of wildtype mice with the anti Fas antibody Jo-2 (0.6 mg/kg iv) resulted in processing of procaspase-3, -8, and -9, a 7-fold increase in caspase-3 activity and a 28-fold increase in DNA fragmentation at 3 h. Similar results were obtained with Gpx-/- mice. SUMMARY AND CONCLUSIONS: Our data demonstrated that Gpx-l-deficient mice developed hepatocelullar apoptosis in vivo with the same time course and to the same extent in response to TNF or Fas receptor stimulation. These data suggest that under our experimental conditions, intracellular ROS did not modulate the death receptor-initiated apoptotic signaling cascade in hepatocytes. Since Gpx-1 is located in the cytosol and in mitochondria, which are the main cellular compartments involved in apoptotic signaling, our findings indicate that the oxidant stress in vivo was insufficient to modulate these signaling pathways. (Supported in part by NIH grant ES-06091).
409
410
TRANSLOCATION OF TISSUE TRANSGLUTAMINASE FROM CYTOPLASM TO NUCLEUS PLAYS A ROLE IN ETHANOL-INDUCED APOPTOSIS IN RAT HEPATOCYIES. Liang-Wen Song, Jian Wu, UC Davis Medical
INHIBITORY ACTION OF T U M O R NECROSIS FACTOR-o~ ON HEPATOCYTE APOPTOSIS THROUGH INHIBITOR OF APOPTOSIS PROTEINS (IAPS) IN MICE. Sumiko Nagoshi, 3rd Dept of Internal Medicine, Saitama
Center, Sacramento, CA; Soichi Kojima, Tsukuba Institute, Tsukaba Japan; Mark A Zern, UC Davis Medical Center, Sacramento, CA Background: Tissue transglutaminase (tTG) is a ubiquitous protein that catalyzes the posttranscriptional modification of nuclear, cytosolic, and extracellular proteins by cross-linking via e ('y-glutamyl lysine) bonds. Recent evidence indicates that this cross-linking of specific nuclear proteins induces apoptosis. Our previous study showed that ethanol treatment causes apoptosis, inhibition of cell proliferation, and an increase in tTG cross-linking activity in rat hepatocytes (J Biol Chem 2000; 275:22213-22219). However, the mechanism by which the increased tTG activity leads to ethanol-induced hepatocyte apoptosis remains unclear. The aim of the present study is to investigate whether nuclear tTG activity increases in ethanol-treated rat hepatocytes. Methods: Rat hepatocytes were isolated by two-step collagenase digestion and incubated in Williams Medium E. tTG activity in the nuclear protein extracts was measured by [l÷C]-putrescine incorporation. Localization of tTG in the nucleus was determined by immunohistochemistry and visualized by confocal microscopy. Results: Treatment of rat hepatocytes with ethanol at 100 mM overnight led to a significant increase in [>C]-putrescine incorporation in the nuclear protein extracts compared to the untreated cells (6188-+304.9 vs. 3807-+ 155.5cpm, p<0.01). Confocal microscopy examination showed that rat hepatocytes stained with specific anti-tTG antibodies were positive for tTG in the cytoplasm and contained a very faint staining in the nucleus. Ethanol treatment caused markedly enhanced tTG signals in the nucleus. At the same time, apoptosis was observed in the cells as indicated by fragments of dense nuclei. Conclusion: Our findings indicate that ethanol-induced apoptosis was associated with enhanced nuclear tTG activity, and that this apparently resulted from cytoplasmic translocation of the enzyme.
Medical Sch, Saitama Japan; Takayuki Yoshimoto, Intractable Disease Research Center, Tokyo Medical University, Tokyo Japan; Satoshi Mochida, Kenji Fujiwara, 3rd Dept of Internal Medicine, Saitama Medical Sch, Saitama Japan Tumor Necrosis Factor-~¢ (TNF-a) produced by activated macrophages in the hepatic sinusoids regulates the development of massive liver necrosis, and also acts as a cytokine to induce liver regeneration. It is well known that TNF-a binding to its receptor activates the caspase cascade resulting in apoptosis, whereas NFKB, an endogenous transcriptional factor activated by TNF-a, can inhibit this apoptosis. When mice received a small amount of TNF-a, liver injury is not induced, but massive liver necrosis following apoptosis occurs after TNF-c¢ administration in mice pretreated with d-galactosamine (GalN), a transcription inhibitor. On the other hand, hepatocyte apoptosis after Fasligand injection was retarded in partially hepatectomised mice compared to sham operated mice, and TNF-~e pretreatment suppressed Fas-mediated hepatocyte apoptosis in normal mice. In the present investigation, the mechanisms of this inhibitory action of TNF-o¢ on apoptosis was studied in relation to inhibitor of apoptosis proteins (IAPs) which block apoptosis by binding to caspases. METHODS and RESULTS As IAPs, IAP-1, IAP-2, XIAP and survivin were used, and the expressions of these IAPs in the liver were evaluated by RT-PCR. 1) When mice received TNF-oq IAP-1 and IAP-2 were expressed in the liver from 1 to 5 hr with the peaks at 1 hr, but the expressions of XIAP and survivin showed the control levels until 8 hr. W h e n mice were injected with GalN and with TNF-cz 30 min later, the peak levels of IAP-1 and IAP-2 expressions showed the control levels, and apoptosis of hepatocytes developed with high serum ALT values. 2) IAP-1, IAP-2 and XIAP expressions were not changed from 10 mill to 5 days after 70% partial hepatectomy in mice, while survivin mRNA level was increased from 36 to 48 hr after the operation. DISCUSSION The inhibitory action of TNF-~e on apoptosis was suggested to be produced via IAP-1 and IAP-2. However, during liver regeneration, hepatocytes may be protected from apoptosis through different mechanisms. The role of survivin would be important. CONCLUSION IAP-1 and IAP-2 induced by T N F - a may act as survival factors for massive liver necrosis.