Liver biology and pathobiology


  • Potential conflict of interest: Nothing to report.

During the past 4 years, many interesting and important advances in liver biology and pathobiology have been identified in the Highlights section. However, several recurrent themes stand out reflecting foci of research for which Hepatology has provided a major forum in rapidly advancing fields. These include apoptosis, drug-induced liver injury, and alcoholic liver disease (ALD)/nonalcoholic fatty liver disease (NAFLD). Some of the previously highlighted articles in these three areas of research will be put in the context of these rapidly evolving fields.

Apoptosis/Cell Death

The area of apoptosis has been a major focus of research in all fields of medicine, and a number of publications in Hepatology reflect this. As shown in Fig. 1, cell death is triggered by either an extrinsic pathway involving ligation of Fas and TNF-R1 death receptors or an intrinsic pathway involving activation of stress-induced mechanisms in specific subcellular compartments. In either case, usually the triggering and upstream signaling mechanisms feed into a common pathway after recruitment of mitochondria. Here the commitment to die is made although the mode of death will depend on the status of preservation of mitochondrial function (e.g., adenosine triphosphate [ATP] depletion and oxidative stress inactivate caspases and thus favor a necrotic demise). Several important advances have been reported in Hepatology . Among the most notable has been the identification of the importance of sustained JNK activation in tumor necrosis factor (TNF)-induced cell death as a consequence of inhibition of nuclear factor kappa B (NF-κB).1, 2 Normally, JNK activation is transient because NF-κB responsive genes activated by TNF shut off JNK by promoting antioxidant processes; candidate NF-κB regulated genes include Mn SOD and ferritin heavy chain. Recent research suggests that TNF-induced mitochondrial oxidative stress1 as well as other sources of reactive oxygen species (ROS)2 activate JNK. Current evidence points to oxidative stress-mediated inactivation of JNK phosphatase or release of JNK and upstream kinase (ASK-1) from sequestration by redox-sensitive binding-partners (e.g., thioredoxin and monomeric GSH S-transferases).

Figure 1.

Extrinsic and Intrinsic Pathways of Apoptosis

Some controversy has concerned what are targets of JNK, and this may be context specific, but in primary hepatocytes and intact liver, the evidence favors targets upstream of mitochondria (possibly Bcl2 family regulators of mitochondrial permeability) rather than c-jun/ AP-transcriptional regulation.3 One very important Hepatology paper in this field4 demonstrated that tBid promotes mitochondrial ROS production, which then leads to proteosomal degradation of FLIP, an NF-κB–responsive inhibitor of caspase 8 activation. In Bid null mice, this amplification of Bid cleavage through destruction of caspase 8 inhibitor did not occur.4 Thus, sufficient caspase 8 needs to be cleaved (sufficient tBid formation) to fully recruit mitochondria. The initial ligation of death receptors may only activate a small percentage of caspase 8 and lead to enough Bid cleavage to induce increased mitochondrial ROS release. ROS then play a key role in initiating feed-forward, self-amplification of the caspase 8-Bid-mitochondria pathway.

Another important concept emerged recently, namely, that stress within various organelles can recruit the participation of mitochondria and lead to apoptosis. This occurs with DNA damage (p53), as well as cytoskeleton (Bmf and Bim) and endoplasmic reticulum (ER) (JNK, CHOP) stress. ER stress can be induced by prolonged protein malfolding or calcium depletion. ER stress can also be caused by homocysteinylation and impaired glycosylation or Golgi processing. ER stress–induced apoptosis has been implicated in chronic hepatitis C, alcoholic liver disease, and α1-antitrypsin deficiency. Interestingly, ursodeoxycholate has broadened its protective umbrella to include prevention of ER stress in hepatocytes.5 It is important to recognize that ER stress also activates SREBPs leading to fatty liver.

Finally, an important study demonstrated the critical role of FADD in death receptor apoptosis in hepatocytes using transgenic animals expressing a dominant negative FADD under the control of the albumin promoter. Aside from showing the expected finding that inhibition of FADD action on the cytoplasmic side of the death receptor interrupted the apoptotic signal cascade (caspase 8 activation) in hepatocytes, the in vivo model showed that hemorrhage and microvascular injury also did not occur (despite dominant negative FADD expression only in hepatocytes).6 Thus, the extensive hepatic hemorrhage and microvascular disruption in FasL and TNF/galactosamine models is not due to direct action on nonparenchymal cells as previously suspected but occurs as a secondary response to hepatocellular apoptosis.


ALD, alcoholic disease; NAFLD, nonalcoholic fatty liver disease; ATP, adenosine triphosphate; TNF, tumor necrosis factor; NFκB, nuclear factor kappa B; ROS, reactive oxygen species; ER, endoplasmic reticulum; APAP, acetaminophen; GSH, glutathione; NO, nitric oxide.

Drug-Induced Liver Injury

Drug-induced liver injury has become a major focus of research, and acetaminophen (APAP) toxicity has occupied center stage, both because it is the major cause of acute liver failure and because it is the most reliable and widely studied animal model of drug-induced liver injury. APAP is activated by CYP2E1 to a toxic metabolite, N-acetyl-p-quinoneimine, which is detoxified by glutathione (GSH) (Fig. 2). The mitochondrion is a key target as toxicity requires depletion of GSH and covalent binding in the mitochondrial compartment. Although some controversy exists regarding the role of chronic alcoholism in increasing susceptibility to APAP in humans, this has been well documented in animal models. Ethanol feeding for 10 days increased Cyp2e1, whereas after 6 weeks of ethanol the mitochondrial GSH selectively decreased, and susceptibility to APAP was heightened.7 Thus, a combination of enhanced toxification and impaired detoxification contribute to the mechanism for ethanol's sensitization to APAP in animal models.

Figure 2.

Pathogenesis of Acetaminophen (APAP) Toxicity

APAP toxicity depends on collapse of mitochondrial function, which was found to be mediated by permeability transition (MPT), leading to loss of ATP and oncotic necrosis.8 Sustaining some ATP through anaerobic glycolysis could switch necrosis to an apoptotic mode of cell death because of release of cytochrome c and ATP-dependent activation of caspases.8 Fasting-induced glycogen depletion therefore favors necrotic cell death. Although the contribution of apoptosis is controversial, a tendency for apoptosis may be suppressed by fasting and influence the severity of APAP-induced liver injury because necrosis may promote greater inflammation and collateral damage.

N-acetyl-P -quinoneimine–dependent consumption of GSH leads to an altered redox state and oxidative stress. This causes the nuclear translocation of Nrf-2, a transcription factor normally sequestered in the cytosol bound to redox sensitive proteins. After oxidative stress releases Nrf-2 from its tether, it translocates to the nucleus where it binds to the antioxidant response element, triggering transcription of protective genes such as glutamyl cysteine ligase (rate-limiting enzyme of GSH synthesis) and heme oxygenase.9 Even minimally toxic doses of APAP trigger this pathway.9 Thus, a dynamic counterbalance occurs in hepatocytes exposed to APAP in which toxic and protective mechanisms are simultaneously activated, modulating the threshold for injury.

The role of nitric oxide (NO) in APAP toxicity is controversial. V-pyrro/NO, a prodrug that delivers NO selectively to hepatocytes, protects against APAP, possibly by decreasing FasL expression,10 whereas subtoxic dose of Fas agonist worsens APAP toxicity, possibly by pre-induction of iNOS.11 NO can improve blood flow, scavenge ROS, and inhibit caspases; conversely, it can combine with ROS to generate more toxic radicals. This appears to be a tenuous balance influenced by compartmental factors (intra-versus extra-cellular and subcellular) and extent of exposure.

Once initial injury has occurred, mounting evidence supports a role for the innate immune system and inflammation in determining the progression and severity of APAP-induced liver injury. APAP appears to prime Kupffer cells (?APAP metabolism in Kupffer cells or hepatocytes) to respond to lipopolysaccharide, which may trigger the production of cytokines and initiate an inflammatory cascade. Lipopolysaccharide binding protein null mice exhibited decreased cytokine mRNA, less necrosis, and improved survival in response to APAP.12

Some parent drugs, rather than their metabolites, target mitochondria to exert toxicity. Examples reported recently in Hepatology include diclofenac,13 tacrine,14 and amiodarone.15 Accumulation in mitochondria can promote mitochondrial permeability transition (diclofenac), impair mitochondrial DNA and topoisomerases (tacrine) or inhibit electron transport, uncouple and inhibit β-oxidation of fatty acids (amiodarone). These findings emphasize the importance of screening for direct effects on mitochondria during drug development. Another approach to screening is the identification of patterns of altered gene expression. Microarray studies with trovafloxacin revealed features of oxidative stress and decreased RNA transcription, possibly due to inhibition of topoisomerase.16 Although not conclusively identifying the cause of idiosyncratic hepatotoxicity of the drug, these studies provide important clues as to the characteristic signature of altered gene expression compared with nontoxic drugs in the same class.

Alcoholic and Nonalcoholic Fatty Liver Disease

In both alcoholic and nonalcoholic fatty liver disease, common themes that occupy investigators include the pathogenesis of steatosis and the mechanisms for progression to injury and fibrosis. Hallmarks of oxidative stress have been identified by many investigators in the liver of alcohol-fed animals. Three major sources of ROS have been considered, and their relative contribution remains uncertain: mitochondria, cyp2e1, and NADPH oxidase (phagocytes). Comparison of p47 phox-null (NADPH oxidase deficient) and Cyp2e1-null mice fed intragastric ethanol has suggested a more important role for NADPH oxidase in necroinflammatory response to ethanol, but Cyp2e1 seems to have a more important role in DNA damage17; this may contribute to promotion of hepatocellular carcinoma. Conversely, Cyp2e1 induction by ethanol in hepatocytes could contribute to either lethal or nonlethal oxidative stress. HepG2 cells that express CYP2E1 release H2O2 and lipid peroxidation products that can act on co-cultured stellate cells to cause activation, proliferation, and collagen synthesis,18 thus providing proof of principle for the concept that oxidative stress in hepatocytes can promote fibrosis.

Animal models of alcoholic liver disease are imperfect and at best seem to mimic only the early stages of the disease. The intragastric feeding model has been favored in recent times over the oral Lieber-DiCarli feeding model because the former exhibits much greater steatosis and significant necroinflammation and apoptosis. However, underscoring the potential importance of oxidative stress, sod1−/− mice fed ethanol orally developed significant necrosis, decreased mitochondrial GSH, and features of oxidative stress even with lower doses of ethanol.19 Thus, in considering the role of oxidative stress, one must recognize that factors that contribute to the production of ROS (pro-oxidants) and to the removal of ROS (antioxidants) are important. One human study identified polymorphism of SOD1 as a risk factor for alcoholic cirrhosis.

Both alcoholic and nonalcoholic liver disease are associated with steatohepatitis and fibrosis. A major area of advancement has been the identification of the role of adipokines such as leptin and adiponectin, which are altered in both ALD and NALD. Leptin was shown to be critical in hepatic fibrosis. Leptin-deficient ob/ob mice were found to be extremely resistant to fibrosis from CC14.20 Conversely, adiponectin deficiency has the opposite effect. Fatty liver sensitizes to TNF toxicity. Adiponectin treatment of KK-Ay obese mice protected against lipopolysaccharide/galactosamine injury and suppressed TNF response.21 A very intriguing effect of leptin deficiency is disruption of adrenohypothalamic regulation of norepinephine. Norepinephine deficiency decreased NKT cells (apoptosis) in the liver22; NKT cells have adrenoreceptors. This was confirmed in dopamine β-hydroxylase null mice, which lack norepinephrine. These mice exhibited decreased hepatic NKT cells.22 Norepinephrine supplementation reversed NKT cell depletion in both ob/ob leptin-deficient and dopamine β-hydroxylase–deficient mice. The decrease in NKT cells in norepinephrine deficiency could disturb the innate immune balance favoring increased Th1 and decreased Th2 response, which would promote inflammation and injury.