Potential conflict of interest: Nothing to report.
Kupffer cells, resident tissue macrophages of the liver, play a key role in the regulation of hepatic inflammation, hepatocyte death, and fibrosis that characterize liver diseases. However, it is controversial whether Kupffer cells promote or protect from liver injury. To explore this issue we examined the role of Kupffer cells in liver injury, cell death, regeneration, and fibrosis on cholestatic liver injury in C57BL/6 mice using a model of partial bile duct ligation (BDL), in which animals do not die and the effects of BDL can be compared between injured ligated lobes and nonligated lobes. In cholestatic liver injury, the remaining viable cells represented tolerance for tumor necrosis factor alpha (TNF-α)-induced hepatocyte apoptosis and regenerative features along with AKT activation. Inhibition of AKT by adenovirus expressing dominant-negative AKT abolished the survival and regenerative properties in hepatocytes. Moreover, Kupffer cell depletion by alendronate liposomes increased hepatocyte damage and the sensitivity of TNF-α-induced hepatocyte apoptosis in ligated lobes. Kupffer cell depletion decreased hepatocyte regeneration and liver fibrosis with reduced AKT activation. To investigate the impact of acid sphingomyelinase (ASMase) in Kupffer cells, we generated chimeric mice that contained ASMase-deficient Kupffer cells and -sufficient hepatocytes using a combination of Kupffer cell depletion, irradiation, and the transplantation of ASMase-deficient bone marrow cells. In these mice, AKT activation, the tolerance for TNF-α-induced apoptosis, and the regenerative responses were attenuated in hepatocytes after BDL. Conclusion: Kupffer cells have a protective role for hepatocyte damage and promote cell survival, liver regeneration, and fibrosis in cholestatic liver disease. Kupffer cell-derived ASMase is crucial for AKT activation of hepatocytes that is required for the survival and regenerative responses. (HEPATOLOGY 2009.)
Chronic liver disease is associated with inflammatory cell infiltration, cytokine production, and liver cell death. Persistent hepatocyte death impairs hepatocyte regeneration accompanied with excessive production of extracellular matrix proteins causing liver fibrosis. Kupffer cells, resident tissue macrophages of the liver, function as both a promoter and a protector against liver injury. Lipopolysaccharide activates Kupffer cells and induces liver injury and inhibition of Kupffer cells prevents liver injury.1 In addition, inhibition of Kupffer cell activation prevents liver injury induced by melphalan2 and fumonisin B1.3 In contrast, reduced Kupffer cell activity augments some kinds of liver injuries, such as hepatectomy- or acetaminophen-induced liver injury.4, 5 Activated Kupffer cells release various types of inflammatory cytokines and growth factors,6 and these mediators are thought to regulate liver injury and regeneration. Especially, tumor necrosis factor alpha (TNF-α) from activated Kupffer cells plays a major role in the pathogenesis of various liver injuries.7, 8 Cholestasis is associated with many liver diseases. Bile duct ligation (BDL) causes hepatocyte damage, hepatic stellate cell (HSC) activation, and liver fibrosis accompanied by Kupffer cell activation leading to the production of a variety of cytokines and chemokines that are involved in liver damage and fibrosis.9–11 Because these features are similar to human cholestatic diseases, common BDL has been used as an animal model of chronic liver disease. However, in this model, common bile duct ligation causes total bile acid reflux to damage whole liver, and the animals show high mortality due to liver failure. We have previously established a partial BDL (PBDL) model, in which animals showed a typical liver injury only in the BDL lobes but no damage in the nonligated lobes with viable liver functions. In this study we examined the role of Kupffer cells in chronic liver injury using the PBDL model.
Acid sphingomyelinase (ASMase) hydrolyses sphingomyelin into ceramide and phosphorylcholine and is involved in various cell functions. Ceramide has been identified as a bioactive mediator of various cellular functions.12 In addition, roles for sphingomyelin and ceramide in membrane lipid rafts have been reported,13 which is related with transmitting signals across the plasma membrane. In macrophages, ASMase contributes to cytokine and chemokine release. Its inhibitor, sphingomyeline difluoromethylene analogue-7 (SMA-7), suppressed lipopolysaccharide-induced releases of TNF-α, interleukin (IL)-1β, and IL-6 from macrophages, and it reduces the severity of inflammatory bowel disease induced by dextran sodium sulfate.14 In contrast, production of macrophage inflammatory protein-1α and -2 is increased in ASMase-deficient macrophages.15 In addition, ASMase-deficient macrophage is impaired in killing bacteria.16 Thus, ASMase contributes to various immunoresponses. In liver damage, although deficiency of ASMase leads to resistance to hepatocyte cell death induced by TNF-α,17, 18 the role of ASMase in Kupffer cells remains unclear.
In this study we assessed the roles of Kupffer cells and ASMase during chronic liver injury using PBDL mice. We found that Kupffer cells reduce liver damage, and induce hepatocyte survival and regeneration, and fibrosis. The protective and regenerative effects require AKT in hepatocytes by way of ASMase in Kupffer cells.
ASMase knockout mice (ASMase−/−) (C57Bl/6 background)18 were bred for studies. Eight-week-old male wildtype C57Bl/6J mice were obtained from Japan SLC (Japan). The left hepatic duct was ligated for PBDL as reported.19 The animals were fasted for 12 hours before sacrifice at 10 days after the surgery. As necessary, hepatocyte apoptosis was induced by mouse TNF-α (R&D Systems, Minneapolis, MN) (0.5 μg/mouse intravenously) with D-galactosamine (GalN) (Nacalai Tesque, Japan) (20 mg/mouse intraperitoneally) 10 days after the PBDL20 and the animals were killed 6 hours after TNF-α administration. All procedures were approved by the Institutional Animal Care Committee of Gifu University.
Depletion of Kupffer Cells.
Alendronate was reported to deplete Kupffer cells.1 A single injection of liposome-encapsulated alendronate (Ale-lip) depleted F4/80-positive cells in the liver at 2-3 days after injection and the cells started to restore at 6 days (Supporting Fig. 1A). Ale-lip had no effect on hepatocytes with hematoxylin and eosin (H&E) (Supporting Fig. 1B) and alanine transaminase (ALT) (data not shown). The vitamin A autofluorescence and desmin-positive cells, characteristic features of HSCs, were not decreased by Ale-lip (Supporting Fig. 1CD). Ale-lip was injected to the operated mice 3 times at 1 day before surgery and 3 and 6 days after the surgery. Phosphate-buffered saline (PBS) encapsulated liposomes (PBS-lip) were used for control.
Bone Marrow Transplantation.
Bone marrow transplantation was performed as reported.11 The wildtype mice received Ale-lip injection twice at 1 and 4 days prior to lethal irradiation (11 Gy). Total bone marrow cells were collected from wildtype or ASMase−/− mice and injected to the irradiated recipient mice (107 cells). PBDL was performed 10 weeks after the transplantation.
Other Experimental Procedures.
Other experimental procedures are described in the Supporting experimental procedures. These include preparation of liposome-encapsulated alendronate, adenovirus infection, histological analysis, western blot, quantitative real-time reverse-transcription polymerase chain reaction (RT-PCR), hydroxyproline measurement, and statistical analysis.
Kupffer Cell Depletion Increases Hepatocellular Damage Induced by BDL.
To examine the effect of Kupffer cell depletion on chronic liver damage induced by BDL, we initially injected Ale-lip three times to mice operated on with common BDL. Although the treatment with Ale-lip alone did not induce liver injury, the mortality of mice treated with common BDL and Ale-lip was extremely high; 40% 10 days after the surgery. In our established model, PBDL showed liver injury and fibrosis only in BDL lobes, which improved the survival rate up to 100% in Ale-lip-treated mice. Therefore, we decided to use PBDL for this study. In BDL lobes, F4/80-positive cells were increased. The Ale-lip treatment succeeded in deleting F4/80-positive cells (Fig. 1A). Thus, Ale-lip injection can be utilized as a new tool for Kupffer cell depletion. Inflammatory cytokines mainly produced from Kupffer cells were up-regulated in BDL lobes, whereas the Ale-lip treatment markedly inhibited the production of TNF-α and IL-1β (Fig. 1B). Kupffer cell-depleted mice showed an increase of injured lesion in BDL lobes and serum ALT level after the surgery (Fig. 1C). Interestingly, 24 hours after common BDL (Supporting Fig. 2) as well as PBDL (Fig. 1C), there were no significant differences in histological liver injury and elevated ALT activities between control and Kupffer cell-depleted mice. These findings indicate that Kupffer cells were not involved in the early stage of liver damage that occurs by BDL, but in the late stage.
Kupffer Cells Mediate Survival and Regeneration of Hepatocyte by BDL.
As previously reported,20 treatment with TNF-α plus GalN strongly induced hepatocyte destruction and massive hemorrhage with apoptotic cells in nonligated lobes of PBDL animals, whereas hemorrhagic damage and hepatocyte apoptosis were blunted in BDL lobes (Supporting Fig. 3A-C). Kupffer cell depletion itself did not induce hepatocyte apoptosis (Supporting Fig. 3D). In Kupffer cell-depleted livers, GalN plus TNF-α treatment induced hemorrhagic liver damage and hepatocyte apoptosis with the cleavage of poly (ADP-ribose) polymerase (PARP), which is the downstream target of caspase-3, both in nonligated and BDL lobes (Fig. 2A-C).
In the BDL lobes, proliferation cell nuclear antigen (PCNA) or Ki67-positive hepatocytes were increased with up-regulation of cyclin E expression (Fig. 2D-F), indicating that BDL induces hepatocyte regeneration. In Kupffer cell-depleted livers the expressions of PCNA, Ki67, and cyclin E were decreased (Fig. 2D-F). Thus, Kupffer cells are important for survival and regeneration of hepatocytes after BDL.
Kupffer Cells Are Required for Liver Fibrosis.
Fibrosis was induced in BDL lobes as demonstrated by Sirius red staining, hydroxyproline content, expression of α-smooth muscle actin (α-SMA) and desmin, and messenger RNA (mRNA) expression of collagen-α1(I) and transforming growth factor (TGF)-β1 (Fig. 3). Kupffer cell-depleted mice showed reduced fibrosis in BDL lobes (Fig. 3). The number and the activation of HSCs were decreased by Kupffer cell depletion as assessed by desmin and α-SMA expression, respectively. These results suggest that the decrease in the fibrogenic response by Kupffer cell depletion is due to a lack of signal from Kupffer cells to activate and proliferate HSCs.
ASMase Deficiency in Kupffer Cells Diminishes the BDL-Induced Survival and Proliferative Effect.
To further elucidate the mechanisms by which Kupffer cells contribute to BDL-mediated functional changes in liver injury, survival of hepatocyte, regeneration, and fibrosis, we focused on ASMase. The protein level of ASMase (Supporting Fig. 4) and ceramides (Supporting Table 1), the metabolite of ASMase, were increased in BDL lobes compared with those in nonligated lobes, suggesting the contribution of ASMase in the liver. To explore the involvement of ASMase in bone marrow-derived cells we generated ASMase-chimeric mice using a combination of alendronate-induced Kupffer cell depletion, irradiation, and bone marrow transplantation. To confirm the substitution of Kupffer cells in chimeric mice we initially generated the mice transplanted with bone marrow isolated from β-actin promoter-driven green fluorescent protein (GFP)-transgenic mice. In the GFP-chimeric mouse liver all F4/80-positive cells were GFP-positive (Supporting Fig. 5A), suggesting that this protocol achieved full reconstitution of Kupffer cells to bone marrow-derived cells. The chimeric mice containing ASMase−/− bone marrow cells showed an increase of F4/80-positive cells, and TNF-α and IL-1β production after BDL, which were comparable to ASMase+/+ bone marrow-transplanted mice (Supporting Fig. 5BC).
ASMase−/− bone marrow-transplanted mice showed an increase of liver injury in BDL lobes at the same degree as ASMase+/+ bone marrow-transplanted mice (Fig. 4), suggesting that ASMase of Kupffer cell is not implicated in the liver injury. Hemorrhagic liver damage and apoptosis with PARP cleavage by GalN plus TNF-α were observed in BDL lobes of ASMase−/− bone marrow-transplanted mice but not in that of ASMase+/+ bone marrow-transplanted mice (Fig. 5A-C). An increase of PCNA or Ki67-positive cells with cyclin E expression were blunted in BDL lobes of ASMase−/− bone marrow-transplanted mice (Fig. 5D-F). These results suggest that ASMase in Kupffer cells contribute to the protection against apoptosis and regeneration in BDL lobes. However, there was no difference between ASMase−/− bone marrow and ASMase+/+ bone marrow-transplanted mice in mRNA expression of fibrogenic markers, Sirius red staining, and hydroxyproline content (Fig. 6). Thus, ASMase of Kupffer cell was not associated with liver fibrosis.
AKT Activation of Hepatocytes Is Required for Survival and Regeneration in BDL Lobes.
Our previous study demonstrated that AKT was up-regulated in BDL lobes and was involved in hepatocyte survival from TNF-α-induced cell death.20 In BDL lobes, phosphorylated-AKT and its downstream target, phosphorylated-glycogen synthase kinase (GSK)3β were increased (Supporting Fig. 6A). Immunohistochemical analysis identified that AKT in hepatocytes was phosphorylated (Supporting Fig. 6B). The AKT activation in BDL lobes was abrogated by the infection of Ad5 dominant negative (DN)-AKT (Supporting Fig. 6C). The inhibition of AKT abolished the survival effect (Supporting Fig. 6D) as reported,20 and eliminated the induction of Ki67-positive cells and cyclin E (Fig. 7A,B) induced by BDL. These findings suggest that AKT activation in hepatocytes is essential for hepatocyte survival and regeneration observed in BDL lobes. In Kupffer cell-depleted mice (Fig. 7C) or ASMase−/− bone marrow-transplanted mice (Fig. 7D), the phosphorylation of AKT and GSK3β by BDL was inhibited. Thus, Kupffer cells and their ASMase are required for AKT-mediated survival and regeneration induced by BDL. Mcl-1 induction, which is a Bcl-2 family member and was regulated by AKT in hepatocytes (Supporting Fig. 6C,E),21 was diminished in Kupffer cell-depleted mice or ASMase−/− bone marrow-transplanted mice, whereas Bcl-XL or Bfl-1 were not affected. These results suggest that survival may be mediated by Mcl-1 at the downstream of AKT.
The present study specifically addressed the role of Kupffer cells and of ASMase in the cholestatic liver injury. Our results demonstrate that depletion of Kupffer cells increased liver injury and susceptibility to TNF-α-induced hepatocyte apoptosis, and decreased hepatocyte regeneration and liver fibrosis with reduced AKT activation. Kupffer cell-derived ASMase was crucial for the AKT activation. The results raise novel therapeutic possibilities for treating liver injury.
After BDL, hepatocytes are exposed to elevated concentrations of bile acid, and hydrophobic bile acids lead to hepatocyte cell death22 through various factors such as reactive oxygen species (ROS) generation from mitochondria23 and activation of Fas signaling in a ligand-independent manner by altering cellular trafficking of Fas.24 Indeed, expression of 4-hydroxy-2-nonenal (HNE), which is produced by lipid peroxidation, was increased on 1 day after the surgery of BDL (data not shown). Because Kupffer cell depletion did not increase the initial liver damage by BDL (1 day after the surgery), it is likely that this damage is induced by a direct toxic effect of bile acid rather than subsequent immune responses because Kupffer cells are not activated in this early stage. The initial hepatocyte cell death stimulates subsequent inflammatory responses leading to further liver injury and fibrosis.25, 26 In BDL liver, the engulfment of apoptotic or necrotic body in Kupffer cells is observed,27 which leads to production of cytokines including TNF-α and TGF-β.9 Either a promotive9 or protective10 effect of Kupffer cells on BDL-induced liver injury have been reported. In the present study, alendronate treatment, which depleted Kupffer cells in the livers, increased liver injury and reduced fibrosis 10 days after BDL, suggesting that Kupffer cells have a protective effect on the subsequent damage of hepatocytes and a promotive effect on fibrosis in the late stage. The increase of liver injury is probably explained by the diminished Kupffer cell functions, including the phagocytosis of injured tissue and the production of protective factors for hepatocytes. The reduced fibrosis is most likely due to decreased fibrogenic cytokines from Kupffer cells. Cytokines including TGF-β and TGF-α are released from Kupffer cells,28 and HSCs are stimulated to induce collagen I α1 transcription by TGF-β.29
In the liver chronically injured by BDL, hepatocytes represented the survival and regenerative properties, and AKT was a critical factor for the survival and regeneration of the remaining viable hepatocytes. Indeed, overexpression of constitutive active-AKT led hepatocytes to be resistant against TNF-α-induced apoptosis in primary cultured hepatocytes (data not shown) and promoted hepatocyte proliferation by cyclin E.30 Because depletion of Kupffer cells diminished the survival and regeneration of hepatocytes with reduced AKT activation, Kupffer cells could produce factors that activate AKT in hepatocytes. In our study, the survival and regenerative effects of AKT activation were abrogated in ASMase−/− bone marrow-transplanted mice, suggesting that ASMase in Kupffer cells requires the production of unknown factors that lead to the activation of AKT in hepatocytes. mRNA expression of TNF-α, IL-1β, and IL-6 in ASMase−/− bone marrow-transplanted mice were similar to those in ASMase+/+ bone marrow-transplanted mice (Supporting Fig. 5 and data not shown) after BDL. mRNA levels of hepatocyte growth factor (HGF) and heparin-binding epithelial growth factor (HB-EGF), which induce hepatocyte proliferation,31, 32 were not changed in ASMase−/− bone marrow-transplanted mice (Supporting Fig. 7). Accumulation of CD3-positive T cells in BDL lobes in ASMase−/− bone marrow-transplanted mice was also similar to those in ASMase+/+ bone marrow-transplanted mice (data not shown). The factors that lead to AKT-dependent hepatocyte protection and regeneration are currently unknown. Further studies are needed to determine these factors.
ASMase has various roles in both parenchymal and nonparenchymal cells. ASMase in hepatocytes modulates hepatocyte apoptosis.18 Although ASMase in Kupffer cells did not contribute to liver fibrosis, ASMase in HSCs promotes collagen production. Administration of ASMase to human HSCs increased collagen expression. ASMase plus TGF-β treatment further increased collagen production in HSCs (Supporting Fig. 8A). The collagen expression by ASMase is, at least in part, stimulated by way of the modulation of intracellular signals, Smad2/3, downstream targets of TGF-β receptor, and p38, which increases collagen α1(I) mRNA stability in HSCs.33 The administration of ASMase also phosphorylated p38 (Supporting Fig. 8B). Moreover, exogenous membrane permeable ceramide exerts a stimulatory effect of basal and TGF-β-induced collagen promoter activity in foreskin fibroblast.34
In conclusion, Kupffer cells regulate liver injury, hepatocyte survival, regeneration, and fibrosis after chronic liver damage by BDL. AKT activation in hepatocytes, which is induced by way of ASMase of Kupffer cells, is required for the survival and regeneration of hepatocytes. The hypothetical roles of Kupffer cells are schematically summarized in Fig. 8.