Hepatology highlights


  • Potential conflict of interest: Nothing to report.

Alcohol Takes Many Tolls

The livers of alcohol fed animals seem to be conditioned to be more susceptible to a variety of second hits. A variety of evidence has suggested that the second hit may be mediated by the innate immune response, including cytokines. Ethanol promotes lipopolysaccharide (LPS) exposure via gut permeabilization, primes Kupffer cells to respond more strongly to LPS, and/or sensitizes target cells to the pro-inflammatory or lethal actions of cytokines. Gustot et al. examined the effect of oral ethanol feeding in mice on one aspect of this putative cascade, namely the expression and action of toll-like receptors (TLR) 1-9, the gatekeepers of the innate immune response. Previous alcohol work of others examined only the role of TLR4, the LPS receptor, but did not consider the other TLRs which respond to other bacterial and viral molecules. After 10 days, the oral ethanol feeding induced hepatic steatosis but no inflammation. Real-time PCR identified substantial increases in mRNA for TLR1,2,4,6,7,8,9 but not TLR 3 and 5 in response to ethanol. The mice were then challenged with ligands specific for individual receptors and exhibited increased TNF mRNA only in response to the ligands for receptors which were upregulated by ethanol, accompanied by marked increase in ALT and liver inflammatory foci. Antibiotics decreased fecal flora and moderately but significantly attenuated the hepatic steatosis (27.5% decrease) but did not decrease the effect of ethanol on expression of TLRs. Alcohol feeding was accompanied by increased lipid peroxidation and decreased GSH. Inhibition of NADPH oxidase with DPI suppressed oxidative stress and attenuated the ethanol induced steatosis while reducing the upregulation of TLR 2,4,6, and 9 but not TLR 1,7, and 8. The study demonstrates that ethanol upregulates a repertoire of TLRs which sensitize the liver to injury induced by specific ligands for these receptors. Bacterial products including, but not limited to LPS, and viral molecules thus evoke a myriad of injurious responses through their specific TLR targets. (See HEPATOLOGY 2006;43:989-1000.)

The Liver Loses by KO

The liver responds to cholestatic insults by upregulating basolateral efflux of toxic bile acids. Candidates for the mediator of this self-protective response are Ostα-Ostβ, Mrp3, or Mrp4. Mennone et al. studied the adaptive response to common bile duct ligation (CBDL) in wild type and Mrp4 null mice. Mrp4 null mice exhibited more severe necrosis and ALT elevation 7 days following CBDL (see Fig.). Concomitantly serum bile acid levels were four fold lower in null mice but serum bilirubin levels were comparably increased in both genotypes after CBDL. Mrp4 was upregulated in the wild type mice. Mrp3 protein was upregulated in both wild type and KO mice after CBDL. Ost protein increased in both genotypes after CBDL (confirming that Mrp3 governs bilirubin glucuronide efflux into plasma). The downregulation of uptake transporters was similar in both genotypes. The findings in this report indicate that Mrp4 plays a critical role in adaptive protection of the mouse liver against toxic bile acid accumulation. Upregulation of Mrp3 and Ost are unable to compensate for the lack of Mrp4. As noted by the authors, human liver expresses much greater levels of Ost so its role could be more important in modulating the severity of cholestatic liver disease in humans. (See HEPATOLOGY 2006;43:1013-1021.)

Illustration 1.

Transpicuous Transfer of Transport in a Transplant

Non-anastomotic strictures of the bile ducts occasionally complicate OLT and correlate with cold ischemia time. A role for bile salts in mediating this injury has been suggested. To test this hypothesis, Hoekstra et al. performed OLT using donor wild type and Mdr2+/− livers transplanted into wild type recipient mice. Mdr2 governs bile phospholipids secretion which buffers the toxic bile acids. Heterozygous Mdr2 +/− animals normally do not develop the bile duct injury seen in homozygous Mdr2 null mice. Following one hour cold storage the livers were implanted with portal vein and hepatic artery anastomoses. 14 days post-OLT, the wild-type liver grafts were normal, whereas the Mdr2+/− grafts showed histological abnormalities: portal inflammation, ductular proliferation and purulent cholangitis in the intermediate and large ducts (see Fig.). These morphological changes were accompanied by increased serum AST, ALT, alkaline phosphatase and bile acid levels in the Mdr2+/− transplanted mice. Biliary phospholipid was decreased pre- and post-transplant in the Mdr2+/− mice. Post-transplant Ntcp and Bsep expression decreased in heterozygotes, as expected in response to cholestasis. Mdr2 mRNA doubled post-transplant in the +/− mice and protein increased but only to <50% of wild-type. This elegant study demonstrates that non-anastomotic bile duct injury initiated by and following cold ischemia closely correlates with an increased bile salt-phospholipid ratio in bile. It will be of interest to determine what are the long-term consequences of this early bile ductular injury (compensation or strictures) and to determine if depletion of bile salts (e.g., cholestyramine) or replacement with non-toxic bile salts (e.g., ursodeoxycholate) will protect the recipient, which would prove that the bile salt-phospholipid ratio is directly responsible. Once could envisage implementing these measures in potential liver graft recipients in the period before transplantation to prevent ductular injury. (See HEPATOLOGY 2006;43:1022-1031.)

Illustration 2.

The Phox to Nox Redox

Recent evidence has revealed the capacity of hepatic stellate cells (HSC) to phagocytose apoptotic bodies derived from dead hepatocytes and this can lead to induction of pro-collagen α1 (I) and TGF-β expression. Zhan et al. examined the possible role of NADPH oxidase (NOX), which generates superoxide as a putative signaling process linking stellate cell phagocytosis of apoptotic bodies and fibrogenic activation. Apoptotic bodies, generated by UV radiation of HepG2 cells, were incubated with an immortalized human HSC line. This resulted in increased superoxide production, which was blocked by DPI, an NADPH oxidase inhibitor (see Fig.) Fluorescent tagged apoptotic bodies co-localized with increased DCF signal, representing exposure to reactive oxygen species; furthermore the p47 phox subunit of NOX migrated to membranes after exposure of cells to apoptotic bodies. Concomitant increased mRNA of procollagen α1 (I) was inhibited by DPI but upregulation of TGF-β1 was not, indicating that phagocytic signaling of the former is NADPH oxidase dependent but the latter involves a different signaling pathway. Morphological EM studies in various animal and human liver diseases revealed the presence of phagosomes or phagolysosomes containing engulfed apoptotic bodies in activated HSC. Staining for E-cadherin (an epithelial membrane marker) identified apoptotic bodies presumably derived from hepatocytes in α-smooth muscle actin positive HSC in these sections. This is an important study which links HSC phagocytosis of apoptotic bodies derived from epithelial cells (hepatocytes) to activation of fibrogenesis at least partially through an NADPH oxidase/superoxide redox signaling pathway. More details concerning this redox signaling pathway and the mechanism of phagocytosis-induced upregulation of TGF-β are anticipated. At present, targeting therapy to prevent fibrosis through inhibition of NADPH oxidase seems worthy of exploration, based on this work as well as findings in p47 phox −/− mice which do not develop fibrosis after bile duct ligation. (See HEPATOLOGY 2006;43:435-443.)

Illustration 3.

Less Is Better Than More

IL-6 is an important early signal for liver regeneration after acute injury. Most of the work which supports this conclusion has been derived from IL-6 KO mice coupled with rescue by IL-6 administration. However, in human liver diseases, there is chronic, increased IL-6 levels which correlate with disease severity. To address the long-term role of IL-6, Jin et al. examined the effects of sustained high-dose IL-6 on liver regeneration after partial hepatectomy (PH) or apoptosis induced by agonistic anti-Fas (Jo2). To achieve high level exposure, they injected (i.m.) IL-6 producing CHO cells into athymic nude mice. Two days after injection of cells regeneration after PH was accelerated and apoptosis was reduced. Five days after cell injection, baseline liver mass was increased and there was increased mitosis. However, PH 5-7 days after cell injection was lethal in these mice and apoptosis was not reduced. IL-6 2-day treatment followed by PH decreased caspase 7 and 9 cleavage but 7-day treatment resulted in the opposite, i.e., increased caspase 7 and 9 cleavage in response to PH. Interestingly, caspase 3 was not activated. Thus, IL-6 only protected against PH-induced apoptosis after short-term exposure. Similarly, 2 days of IL-6 exposure protected against Jo2 induced lethality and liver apoptosis but 7-day treatment did not protect. Interestingly, in this model of direct induction of apoptosis, caspase-7 was not activated, whereas caspase-3 cleavage was observed which was inhibited by 2-day IL-6 but not 7-day IL-6 exposure. IL-6 treatment lead to persistent STAT3 activation and sustained induction of anti-apoptotic Bcl-2 and Bcl-XL both after 2 or 7 day exposures. However, the pro-apoptotic Bax protein was markedly induced by prolonged IL-6 exposure (See Fig.) Delivery of IL-6 via osmotic mini-pump for 7 days in immunocompetent mice reproduced the changes in nude mice treated with IL-6-CHO cells. In conclusion, this work demonstrates that short-term IL-6 exposure is anti-apoptotic and pro-proliferative. Inhibition of apoptosis is associated with induction of anti-apoptotic Bcl-2 and Bcl-XL, but not FLIP (in contrast to previous reports). However, despite continued induction of these anti-apoptotic proteins, resistance to apoptosis is lost which correlates with progressive increase in pro-apoptotic Bax expression. It is important to note that these findings provide a hypothesis to explain the resistance and loss of resistance to apoptosis but will require direct examination — e.g., will silencing Bax prevent the loss of resistance with prolonged IL-6 treatment? From a practical viewpoint, this work points out that the duration of IL-6 treatment needs to be limited. Finally, it remains to be seen if the effects of sustained increased endogenous IL-6 exposure in chronic liver diseases mimic the responses to high exposure to exogenous IL-6. (See HEPATOLOGY 2006:43;474-484.)

Illustration 4.