Portal fibroblasts with mesenchymal stem cell features form a reservoir of proliferative myofibroblasts in liver fibrosis.

BACKGROUND AND AIMS: In liver fibrosis, myofibroblasts derive from HSCs and as yet undefined mesenchymal cells. We aimed to identify portal mesenchymal progenitors of myofibroblasts. APPROACH AND RESULTS: Portal mesenchymal cells were isolated from mouse bilio-vascular tree and analyzed by single-cell RNA-sequencing. Thereby, we uncovered the landscape of portal mesenchymal cells in homeostatic mouse liver. Trajectory analysis enabled inferring a small cell population further defined by surface markers used to isolate it. This population consisted of portal fibroblasts with mesenchymal stem cell features (PMSCs), i.e., high clonogenicity and trilineage differentiation potential, that generated proliferative myofibroblasts, contrasting with nonproliferative HSC-derived myofibroblasts (-MF). Using bulk RNA-sequencing, we built oligogene signatures of the two cell populations that remained discriminant across myofibroblastic differentiation. SLIT2, a prototypical gene of PMSC/PMSC-MF signature, mediated profibrotic and angiogenic effects of these cells, which conditioned medium promoted HSC survival and endothelial cell tubulogenesis. Using PMSC/PMSC-MF 7-gene signature and slit guidance ligand 2 fluorescent in situ hybridization, we showed that PMSCs display a perivascular portal distribution in homeostatic liver and largely expand with fibrosis progression, contributing to the myofibroblast populations that form fibrotic septa, preferentially along neovessels, in murine and human liver disorders, irrespective of etiology. We also unraveled a 6-gene expression signature of HSCs/HSC-MFs that did not vary in these disorders, consistent with their low proliferation rate. CONCLUSIONS: PMSCs form a small reservoir of expansive myofibroblasts, which, in interaction with neovessels and HSC-MFs that mainly arise through differentiation from a preexisting pool, underlie the formation of fibrotic septa in all types of liver diseases.

Role of Angiogenesis in the Pathogenesis of NAFLD.

Non-alcoholic fatty liver disease (NAFLD) has become the leading cause of chronic liver disease, exposing to the risk of liver fibrosis, cirrhosis, and hepatocellular carcinoma (HCC). Angio-genesis is a complex process leading to the development of new vessels from pre-existing vessels. Angiogenesis is triggered by hypoxia and inflammation and is driven by the action of proangiogenic cytokines, mainly vascular endothelial growth factor (VEGF). In this review, we focus on liver angiogenesis associated with NAFLD and analyze the evidence of liver angiogenesis in animal models of NAFLD and in NAFLD patients. We also report the data explaining the role of angiogenesis in the progression of NAFLD and discuss the potential of targeting angiogenesis, notably VEGF, to treat NAFLD.

Cholangiopathy aggravation is caused by VDR ablation and alleviated by VDR-independent vitamin D signaling in ABCB4 knockout mice.

BACKGROUND & AIMS: Cholangiopathies are chronic liver diseases in which damaged cholangiocytes trigger a proinflammatory and profibrotic reaction. The nuclear vitamin D receptor (VDR) is highly expressed in cholangiocytes and exerts immune-regulatory functions in these cells. In the present study, we examined the protective function of VDR and other vitamin D signaling pathways in chronic cholangiopathy and cholangiocytes. METHODS: Vdr was invalidated in Abcb4 knockout mice, a widely used animal model of chronic cholangiopathy. The impact of vitamin D signaling on cholangiopathy features was examined in vivo and in cholangiocytes (primary and cell lines). RESULTS: Cholangiopathy features (i.e, cholestasis, ductular reaction and fibrosis) were aggravated in Vdr;Abcb4 double knockout mice compared to the Abcb4 simple knockout, and associated with an overexpression of proinflammatory factors. The proinflammatory phenotype of cholangiocytes was also exacerbated following VDR silencing in vitro. The expression of proinflammatory factors and the severity of cholangiopathy were reduced in the double knockout mice treated with the vitamin D analog calcipotriol or with vitamin D. In vitro, the inflammatory response to TNFα was significantly reduced by calcipotriol in biliary cells silenced for VDR, and this effect was abolished by co-silencing the plasma membrane receptor of vitamin D, protein disulfide-isomerase A3 (PDIA3). CONCLUSIONS: Our results demonstrate an anti-inflammatory role of VDR signaling in cholangiocytes and cholangiopathy. They also provide evidence for PDIA3-mediated anti-inflammatory effects of vitamin D and vitamin D analog in these settings.

Inhibition of receptor-interacting protein kinase 1 improves experimental non-alcoholic fatty liver disease.

BACKGROUND & AIMS: In non-alcoholic fatty liver disease (NAFLD), hepatocytes can undergo necroptosis: a regulated form of necrotic cell death mediated by the receptor-interacting protein kinase (RIPK) 1. Herein, we assessed the potential for RIPK1 and its downstream effector mixed lineage kinase domain-like protein (MLKL) to act as therapeutic targets and markers of activity in NAFLD. METHODS: C57/BL6J-mice were fed a normal chow diet or a high-fat diet (HFD). The effect of RIPA-56, a highly specific inhibitor of RIPK1, was evaluated in HFD-fed mice and in primary human steatotic hepatocytes. RIPK1 and MLKL concentrations were measured in the serum of patients with NAFLD. RESULTS: When used as either a prophylactic or curative treatment for HFD-fed mice, RIPA-56 caused a downregulation of MLKL and a reduction of liver injury, inflammation and fibrosis, characteristic of non-alcoholic steatohepatitis (NASH), as well as of steatosis. This latter effect was reproduced by treating primary human steatotic hepatocytes with RIPA-56 or necrosulfonamide, a specific inhibitor of human MLKL, and by knockout (KO) of Mlkl in fat-loaded AML-12 mouse hepatocytes. Mlkl-KO led to activation of mitochondrial respiration and an increase in ß-oxidation in steatotic hepatocytes. Along with decreased MLKL activation, Ripk3-KO mice exhibited increased activities of the liver mitochondrial respiratory chain complexes in experimental NASH. In patients with NAFLD, serum concentrations of RIPK1 and MLKL increased in correlation with activity. CONCLUSION: The inhibition of RIPK1 improves NASH features in HFD-fed mice and reverses steatosis via an MLKL-dependent mechanism that, at least partly, involves an increase in mitochondrial respiration. RIPK1 and MLKL are potential serum markers of activity and promising therapeutic targets in NAFLD. LAY SUMMARY: There are currently no pharmacological treatment options for non-alcoholic fatty liver disease (NAFLD), which is now the most frequent liver disease. Necroptosis is a regulated process of cell death that can occur in hepatocytes during NAFLD. Herein, we show that RIPK1, a gatekeeper of the necroptosis pathway that is activated in NAFLD, can be inhibited by RIPA-56 to reduce not only liver injury, inflammation and fibrosis, but also steatosis in experimental models. These results highlight the potential of RIPK1 as a therapeutic target in NAFLD.

Endoplasmic reticulum stress induces inverse regulations of major functions in portal myofibroblasts during liver fibrosis progression.

Portal myofibroblasts (PMF) form a sub-population of highly proliferative and proangiogenic liver myofibroblasts that derive from portal mesenchymal progenitors. Endoplasmic reticulum (ER) stress was previously shown to modulate fibrogenesis, notably in the liver. Our aim was to determine if ER stress occurred in PMF and affected their functions. PMF were obtained after their expansion in vivo from bile duct-ligated (BDL) rats and referred to as BDL PMF. Compared to standard PMF obtained from normal rats, BDL PMF were more myofibroblastic, as assessed by higher alpha-smooth muscle actin expression and collagen 1 production. Their proangiogenic properties were also higher, whereas their proliferative and migratory capacities were lower. CHOP expression was detected in the liver of BDL rats, at the leading edge of portal fibrosis where PMF accumulate. BDL PMF displayed ER dilatation and an overexpression of the PERK pathway downstream targets, Chop, Gadd34 and Trb3, in comparison with standard PMF. In vitro, the induction of ER stress by tunicamycin in standard PMF, caused a decrease in their proliferative and migratory activity, and an increase in their proangiogenic activity, without affecting their myofibroblastic differentiation. Conversely, the treatment of BDL PMF with the PERK inhibitor GSK2656157 reduced ER stress, which caused a decrease in their angiogenic properties, and restored their proliferative and migratory capacity. In conclusion, PMF develop ER stress as they expand with the progression of fibrosis, which further increases their proangiogenic activity, but also inhibits their proliferation and migration. This phenotypic switch may restrict PMF expansion while they support angiogenesis.

Culture Model of Rat Portal Myofibroblasts.

Myofibroblasts are matrix-producing cells with contractile properties, usually characterized by de novo expression of alpha-smooth muscle actin, that arise in fibrotic diseases. Hepatic stellate cells (HSCs), known as perisinusoidal cells containing auto-fluorescent vitamin A, are the major although not exclusive source of myofibroblasts in the injured liver. Portal myofibroblasts (PMFs) have been defined as liver myofibroblasts derived from cells that are distinct from HSCs and located in the portal tract. Here, we describe the protocol we have established to obtain rat PMFs in culture. In this method, the biliary tree is (i) separated from the liver parenchyma by in situ enzymatic perfusion of the liver, (ii) minced and further digested in vitro, until bile duct segments are isolated by sequential filtration. Bile duct isolates free of HSC contaminants, form small cell clusters, which initially comprise a large majority of epithelial cells. In culture conditions (fetal bovine serum) that provide a growth advantage to mesenchymal cells over epithelial cells, the epithelial cells die and detach from the substrate, while spindle-shaped cells outgrow from the periphery of the cell clusters, as shown by video-microscopy. These cells are highly proliferative and after 4-5 days, the culture is composed exclusively of fully differentiated myofibroblasts, which express alpha-smooth muscle actin and collagen 1, and secrete abundant collagen. We found no evidence for epithelial-mesenchymal transition, i.e., no co-expression of alpha-smooth muscle actin and cytokeratin at any stage, while cytokeratin becomes undetectable in the confluent cells. PMFs obtained by this method express the genes that were previously reported to be overexpressed in non-HSC or portal fibroblast-derived liver myofibroblasts as compared to HSC-derived myofibroblasts, including the most discriminant, collagen 15, fibulin 2, and Thy-1. After one passage, PMFs retain the same phenotypic features as in primary culture. In conclusion, this straightforward and reproducible method of PMF culture, can be used to identify new markers of PMFs at different stages of differentiation, to compare their phenotype with those of HSC-MFs and ultimately determine their progenitors and specific functions in liver wound-healing.

Portal myofibroblasts promote vascular remodeling underlying cirrhosis formation through the release of microparticles.

UNLABELLED: Liver fibrosis expanding from portal tracts and vascular remodeling are determinant factors in the progression of liver diseases to cirrhosis. In the present study, we examined the potential contribution of portal myofibroblasts (PMFs) to the vascular changes leading to cirrhosis. The analyses of liver cells based on the transcriptome of rat PMFs, compared to hepatic stellate cell HSC-derived myofibroblasts in culture, identified collagen, type XV, alpha 1 (COL15A1) as a marker of PMFs. Normal liver contained rare COL15A1-immunoreactive cells adjacent to the bile ducts and canals of Hering in the portal area. A marked increase in COL15A1 expression occurred together with that of the endothelial marker, von Willebrand factor, in human and rat liver tissue, at advanced stages of fibrosis caused by either biliary or hepatocellular injury. In cirrhotic liver, COL15A1-expressing PMFs adopted a perivascular distribution outlining vascular capillaries proximal to reactive ductules, within large fibrotic septa. The effect of PMFs on endothelial cells (ECs) was evaluated by in vitro and in vivo angiogenesis assays. PMF-conditioned medium increased the migration and tubulogenesis of liver ECs as well as human umbilical vein ECs and triggered angiogenesis within Matrigel plugs in mice. In coculture, PMFs developed intercellular junctions with ECs and enhanced the formation of vascular structures. PMFs released vascular endothelial growth factor (VEGF)A-containing microparticles, which activated VEGF receptor 2 in ECs and largely mediated their proangiogenic effect. Cholangiocytes potentiated the angiogenic properties of PMFs by increasing VEGFA expression and microparticle shedding in these cells. CONCLUSION: PMFs are key cells in hepatic vascular remodeling. They signal to ECs through VEGFA-laden microparticles and act as mural cells for newly formed vessels, driving scar progression from portal tracts into the parenchyma.

Vitamin D nuclear receptor deficiency promotes cholestatic liver injury by disruption of biliary epithelial cell junctions in mice.

UNLABELLED: Alterations in apical junctional complexes (AJCs) have been reported in genetic or acquired biliary diseases. The vitamin D nuclear receptor (VDR), predominantly expressed in biliary epithelial cells in the liver, has been shown to regulate AJCs. The aim of our study was thus to investigate the role of VDR in the maintenance of bile duct integrity in mice challenged with biliary-type liver injury. Vdr(-/-) mice subjected to bile duct ligation (BDL) displayed increased liver damage compared to wildtype BDL mice. Adaptation to cholestasis, ascertained by expression of genes involved in bile acid metabolism and tissue repair, was limited in Vdr(-/-) BDL mice. Furthermore, evaluation of Vdr(-/-) BDL mouse liver tissue sections indicated altered E-cadherin staining associated with increased bile duct rupture. Total liver protein analysis revealed that a truncated form of E-cadherin was present in higher amounts in Vdr(-/-) mice subjected to BDL compared to wildtype BDL mice. Truncated E-cadherin was also associated with loss of cell adhesion in biliary epithelial cells silenced for VDR. In these cells, E-cadherin cleavage occurred together with calpain 1 activation and was prevented by the silencing of calpain 1. Furthermore, VDR deficiency led to the activation of the epidermal growth factor receptor (EGFR) pathway, while EGFR activation by EGF induced both calpain 1 activation and E-cadherin cleavage in these cells. Finally, truncation of E-cadherin was blunted when EGFR signaling was inhibited in VDR-silenced cells. CONCLUSION: Biliary-type liver injury is exacerbated in Vdr(-/-) mice by limited adaptive response and increased bile duct rupture. These results indicate that loss of VDR restricts the adaptation to cholestasis and diminishes bile duct integrity in the setting of biliary-type liver injury.

Origins and functions of liver myofibroblasts.

Myofibroblasts combine the matrix-producing functions of fibroblasts and the contractile properties of smooth muscle cells. They are the main effectors of fibrosis in all tissues and make a major contribution to other aspects of the wound healing response, including regeneration and angiogenesis. They display the de novo expression of α-smooth muscle actin. Myofibroblasts, which are absent from the normal liver, are derived from two major sources: hepatic stellate cells (HSCs) and portal mesenchymal cells in the injured liver. Reliable markers for distinguishing between the two subpopulations at the myofibroblast stage are currently lacking, but there is evidence to suggest that both myofibroblast cell types, each exposed to a particular microenvironment (e.g. hypoxia for HSC-MFs, ductular reaction for portal mesenchymal cell-derived myofibroblasts (PMFs)), expand and exert specialist functions, in scarring and inflammation for PMFs, and in vasoregulation and hepatocellular healing for HSC-MFs. Angiogenesis is a major mechanism by which myofibroblasts contribute to the progression of fibrosis in liver disease. It has been clearly demonstrated that liver fibrosis can regress, and this process involves a deactivation of myofibroblasts, although probably not to a fully quiescent phenotype. This article is part of a Special Issue entitled: Fibrosis: Translation of basic research to human disease.

Distinct proteomic features of two fibrogenic liver cell populations: hepatic stellate cells and portal myofibroblasts.

In chronic liver diseases, the accumulation of extracellular matrix leading to fibrosis is caused by myofibroblasts, the origins of which are debatable. We performed a comparative proteomic study to identify markers and gain insight into distinct functions of myofibroblasts derived either from hepatic stellate cells (HSCs) or from portal mesenchymal cells. After isolation from normal liver and culture in similar conditions, myofibroblastic HSCs (MF-HSCs) presented enlarged cytoplasms whereas portal myofibroblasts (PMFs) were more proliferative, and formed more stress fibers. The two cell types were subjected to comparative analyses by 2-D MS/MS. Six proteins were overexpressed in PMFs, with myofibroblast-related typical functions. Among them, cofilin-1 showed the greatest difference in expression and a lower pI than expected. Immunoblot demonstrated higher levels of phosphorylation, a modification of the protein implicated in stress fiber formation. Eleven proteins, mostly involved in stress response, were overexpressed in MF-HSCs. Cytoglobin had the highest level of overexpression, as confirmed by reverse transcription quantitative real-time PCR, immunoblot and immunocytochemical analyses. These results identify cytoglobin as the best marker for distinguishing MF-HSCs from PMFs and suggest different functions for the two cell populations in the liver wound healing response, with a prominent role for PMFs in scar formation.

IGF-1R contributes to stress-induced hepatocellular damage in experimental cholestasis.

The insulin-like growth factor type 1 receptor (IGF-1R) controls aging and cellular stress, both of which play major roles in liver disease. Stimulation of insulin-like growth factor signaling can generate cell death in vitro. Here, we tested whether IGF-1R contributes to stress insult in the liver. Cholestatic liver injury was induced by bile duct ligation in control and liver-specific IGF-1R knockout (LIGFREKO) mice. LIGFREKO mice displayed less bile duct ligation-induced hepatocyte damage than controls, while no differences in bile acid serum levels or better adaptation to cholestasis by efflux transporters were found. We therefore tested whether stress pathways contributed to this phenomenon; oxidative stress, ascertained by both malondialdehyde content and heme oxygenase-1 expression, was similar in knockout and control animals. However, together with a lower level of eukaryotic initiation factor-2 alpha phosphorylation, the endoplasmic reticulum stress protein CHOP and its downstream pro-apoptotic target Bax were induced to lesser extents in LIGFREKO mice than in controls. Expression levels of cytokeratin 19, transforming growth factor-beta1, alpha-smooth muscle actin, and collagen alpha1(I) in LIGFREKO mice were all lower than in controls, indicating reduced ductular and fibrogenic responses and increased cholestasis tolerance in these mutants. This stress resistance phenotype was also evidenced by longer post-bile duct ligation survival in mutants than controls. These results indicate that IGF-1R contributes to cholestatic liver injury, and suggests the involvement of both CHOP and Bax in this process.

Hepatocyte proliferation during liver regeneration is impaired in mice with liver-specific IGF-1R knockout.

Recent evidence indicates that growth hormone (GH) is involved in liver regeneration. To test whether insulin-like growth factor I (IGF-I) mediates this effect, we studied liver regeneration induced by partial hepatectomy in liver-specific IGF type 1 receptor knockout (LIGFREKO) mice. The absence of IGF-1R caused a significant decrease in hepatocyte proliferation in males (-52%), but not in females, as assessed by Ki67 immunohistochemistry. Cyclin D1 and cyclin A protein levels in the livers of LIGFREKO males were only half those in controls, indicating that cyclin induction during liver regeneration is dependent on IGF-1R signaling. Analyzing the signaling cascade initiated by IGF-1R, we observed a lack of IRS-1 induction in LIGFREKO livers. In contrast, the induction of IRS-2 synthesis was similar in LIGFREKO and control groups, suggesting the existence of differential regulation of IRS synthesis during liver regeneration. Regenerating livers from LIGFREKO animals also showed significantly less activated ERKs than controls. Our findings demonstrate that IGF-1R makes a significant contribution to liver regeneration. Using the LIGFREKO model, we provide new evidence that IGF-1R/IRS-1/ERK signaling may be the intracellular pathway controlling the cell cycle via cyclin D1 and cyclin A in the regenerating liver.

GSK-3beta reactivation with LY294002 sensitizes hepatoma cells to chemotherapy-induced apoptosis.

Constitutive activation of phosphatidylinositol 3-kinase (PI3K) confers resistance to apoptotic stimuli induced by chemotherapeutic agents in a variety of cancer cells. Therefore, the comprehension of mechanisms whereby PI3K downregulation interferes with chemotherapy is of major clinical interest for the elaboration of combined anticancer treatment modalities. Here, we examined the molecular mechanisms whereby the PI3K inhibitor LY294002 sensitized p53- and Fas-deficient hepatoma cells to etoposide and camptothecin. LY294002 increased Hep3B cell susceptibility to chemotherapy-induced apoptosis by enhancing the expression of DR4 and DR5 and the activation of caspase-8 and -3. Moreover, LY294002-mediated sensitization to chemotherapy involved mitochondrial Bax translocation and cytosolic cytochrome c accumulation. In Hep3B cells, LY294002 led to the reactivation of glycogen synthase kinase-3beta (GSK-3beta) by promoting its dephosphorylation on the serine 9 residue independently from Akt inhibition. The transient transfection of a constitutively active and non-phosphorylable S9AGSK-3beta mutant sensitized cells to etoposide cytotoxic effects while cell treatment with the small GSK-3beta inhibitor SB-415286 repressed the sensitizing effect of LY294002 on chemotherapy-induced apoptosis and caspase-8 activation. Altogether, our results show that LY294002 sensitizes hepatoma cells to chemotherapy-induced apoptosis via death receptor and mitochondria signalling pathways and that GSK-3beta reactivation is involved in this process. Therefore, PI3K-mediated GSK-3beta inhibition could be a mechanism by which cancer cells escape from chemotherapy-induced apoptosis.

c-myc-induced hepatocarcinogenesis in the absence of IGF-I receptor.

Numerous tumours, including hepatocarcinomas, produce IGFs, and some depend on these growth factors in a paracrine or autocrine fashion. We have shown that c-myc-induced experimental hepatocarcinogenesis is associated with enhanced production of IGF-II. To assess the role of the IGF-I receptor (IGF-IR) in hepatocarcinogenesis, we generated conditional mutant mice that overexpressed c-myc and were knocked out for IGF-IR specifically in the liver. We compared these mice with littermate controls that also overexpressed c-myc but had wild-type IGF-IR alleles. We found that the pretumoral phase, induced by early c-myc expression and characterised by increased cell proliferation, was largely unaffected by the lack of IGF-IR. To our further surprise, hepatocellular carcinomas (HCCs) lacking IGF-IR readily developed and progressed at the same rate as control HCCs. At 9 months, all c-myc transgenic mice displayed well-differentiated multifocal tumours, regardless of whether their livers-and their tumours-were able to produce IGF-IR. Levels of IRS-1 and IRS-2 were elevated in all tumours in the presence or absence of IGF-IR, suggesting that the signalling pathway downstream of IGF-IR is activated via IGF-IR-independent mechanisms in HCC. In conclusion, the deregulation of IGF signalling pathways, which often occurs during liver tumorigenesis, does not necessarily require IGF-IRs, and hepatic IGF-IR alone may not play a determinant role in c-myc-induced hepatocarcinogenesis.

GSK-3beta inhibition by lithium confers resistance to chemotherapy-induced apoptosis through the repression of CD95 (Fas/APO-1) expression.

Lithium exerts neuroprotective actions that involve the inhibition of glycogen synthase kinase-3beta (GSK-3beta). Otherwise, recent studies suggest that sustained GSK-3beta inhibition is a hallmark of tumorigenesis. In this context, the present study was undertaken to examine whether lithium modulated cancer cell sensitivity to apoptosis induced by chemotherapy agents. We observed that, in different human cancer cell lines, lithium significantly reduced etoposide- and camptothecin-induced apoptosis. In HepG2 cells, lithium repressed drug induction of CD95 expression and clustering at the cell surface as well as caspase-8 activation. Lithium acted through deregulation of GSK-3beta signaling since (1) it provoked a rapid and sustained phosphorylation of GSK-3beta on the inhibitory serine 9 residue; (2) the GSK-3beta inhibitor SB-415286 mimicked lithium effects by repressing drug-induced apoptosis and CD95 membrane expression; and (3) lithium promoted the disruption of nuclear GSK-3beta/p53 complexes. Moreover, the overexpression of an inactivated GSK-3beta mutant counteracted the stimulatory effects of etoposide and camptothecin on a luciferase reporter plasmid driven by a p53-responsive sequence from the CD95 gene. In conclusion, we provide the first evidence that lithium confers resistance to apoptosis in cancer cells through GSK-3beta inhibition and subsequent repression of CD95 gene expression. Our study also highlights the concerted action of GSK-3beta and p53 on CD95 gene expression.

New targets of beta-catenin signaling in the liver are involved in the glutamine metabolism.

Inappropriate activation of the Wnt/beta-catenin signaling has been implicated in the development of hepatocellular carcinoma (HCC), but exactly how beta-catenin works remains to be elucidated. To identify, in vivo, the target genes of beta-catenin in the liver, we have used the suppression subtractive hybridization technique and transgenic mice expressing an activated beta-catenin in the liver that developed hepatomegaly. We identified three genes involved in glutamine metabolism, encoding glutamine synthetase (GS), ornithine aminotransferase (OAT) and the glutamate transporter GLT-1. By Northern blot and immunohistochemical analysis we demonstrated that these three genes were specifically induced by activation of the beta-catenin pathway in the liver. In different mouse models bearing an activated beta-catenin signaling in the liver known to be associated with hepatocellular proliferation we observed a marked up-regulation of these three genes. The cellular distribution of GS and GLT-1 parallels beta-catenin activity. By contrast no up-regulation of these three genes was observed in the liver in which hepatocyte proliferation was induced by a signal-independent of beta-catenin. In addition, the GS promoter was activated in the liver of GS(+/LacZ) mice by adenovirus vector-mediated beta-catenin overexpression. Strikingly, the overexpression of the GS gene in human HCC samples was strongly correlated with beta-catenin activation. Together, our results indicate that GS is a target of the Wnt/beta-catenin pathway in the liver. Because a linkage of the glutamine pathway to hepatocarcinogenesis has already been demonstrated, we propose that regulation of these three genes of glutamine metabolism by beta-catenin is a contributing factor to liver carcinogenesis.

Dysregulation of glycogen synthase kinase-3beta signaling in hepatocellular carcinoma cells.

It has been reported that upstream components of the insulin-like growth factor (IGF) signaling axis could be overexpressed during hepatocarcinogenesis in humans and rodents. However, the signal transduction pathways activated downstream have been poorly studied. Here, we examined whether glycogen synthase kinase-3beta (GSK-3beta) could be a target in human hepatoma cell lines and transgenic ASV mice with hepatic expression of the SV40 large T antigen. In HuH7, Mahlavu, and Hep3B cells, basal levels of GSK-3beta(Ser9) phosphorylation were strongly elevated, indicating that GSK-3beta was inhibited. GSK-3beta phosphorylation was insensitive to exogenous IGFs and was blocked with an IGF-1 receptor-neutralizing antibody in Mahlavu and Hep3B cells. By using LY294002 and ML-9, which act as phosphatidylinositol 3-kinase (PI3-K) and Akt inhibitors, respectively, we showed that GSK-3beta phosphorylation required PI3-K activation in both cell lines whereas downstream Akt activation was required only in Mahlavu cells. However, in the 2 cell lines, GSK-3beta(Ser9) phosphorylation was controlled by protein kinase C (PKC)zeta because it was blocked by an inhibitory PKCzeta peptide. The blockage of GSK-3beta phosphorylation markedly inhibited glycogen synthesis and decreased beta-catenin expression. In addition, the overexpression of a constitutively active GSK-3beta reduced AP-1-mediated gene transcription in Hep3B cells. Finally, we observed that reexpression of IGF-2 in tumoral livers from ASV mice was associated with a marked phosphorylation of GSK-3beta. In conclusion, our results identify GSK-3beta as a molecular target of the constitutive activation of the IGF axis in in vitro and in vivo models of hepatocarcinogenesis. Persistent phosphorylation of GSK-3beta could be critical for regulation of glycogen metabolism and cell growth in hepatoma cells.