Biochemical and Cellular Mechanisms of Toxic Liver Injury

• ABSTRACT
• APOPTOSIS VERSUS NECROSIS
• INITIATION OF CELL DEATH AND THE ROLE OF MITOCHONDRIA
• SURVIVAL MECHANISMS
• SENSITIZATION TO TUMOR NECROSIS FACTOR
• OXIDATIVE STRESS
• THERAPEUTIC POSSIBILITIES AND PITFALLS
• FINANCIAL DISCLOSURE
• ABBREVIATIONS
• REFERENCES

ABSTRACT

The pathogenesis of drug- or toxin-induced liver injury usually involves the participation of toxic metabolites that either elicit an immune response or directly affect the biochemistry of the cell. The clinical appearance of hepatitis is then a consequence of cell death mediated by either the extrinsic immune system (e.g., cytotoxic T cells) or intracellular stress. Intracellular stress can lead to apoptotic or necrotic cell death, depending on the extent of mitochondrial involvement and the balance of factors that activate and inhibit the Bcl₂ family of proteins and the caspases. Drug metabolites can undergo or promote a variety of chemical reactions, including covalent binding, depletion of reduced glutathione, or oxidative stress with consequent effects on proteins, lipids, and DNA. These chemical consequences can directly affect organelles such as mitochondria, cytoskeleton, endoplasmic reticulum, microtubules, or nucleus or indirectly influence these organelles through activation or inhibition of signaling kinases, transcription factors, and gene expression profiles. The outcome may be either triggering of the necrotic or apoptotic process or sensitization to the lethal action of cytokines of the immune system intrinsic to the liver.

Cell death is the crucial event leading to the clinical manifestations of drug-induced hepatotoxicity. In this review, the current concepts of cell death will be reviewed and placed in the context of drug hepatotoxicity. At the outset, it should be emphasized that the pathogenesis of clinical drug hepatitis reflects either an immune-mediated attack on the liver or a biochemical effect of toxic metabolites leading to a loss of cell viability.[1] [2] Immune attack involves the participation of death receptors such as Fas or the porin-mediated introduction of granzyme to activate the death cascade distal to the death receptor.[3] [4] The outcome is apoptosis of hepatocytes (or perhaps in some circumstances bile duct or endothelial cells).

The liver is usually the target of the immune system because of the production of drug metabolites that bind to proteins which are degraded to peptides presented on the cell surface with major histocompatibility complex class I. This leads to sensitization, which allows T cells to recognize the drug metabolite bound to liver protein (e.g., CYP) targeted to the hepatocyte surface.[5] In contrast, some drugs produce metabolites that cause cell death independent of the extrinsic immune system.[6] There are two ways in which this can happen: the reactive metabolites can disrupt the balance of factors that favor survival, leading to direct loss of viability, or they can modify this balance so as to render liver cells susceptible to the lethal effects of the intrinsic immune system, that is, cytokines such as tumor necrosis factor (TNF), produced by activation of resident inflammatory cells in the liver. The outcome can range from maintenance of viability to apoptosis or necrosis depending on the drug, the extent of exposure to reactive metabolites, and a variety of environmental and genetic factors that modulate drug metabolism, transport, defense, regeneration, cytokines, and prodeath and prosurvival genes.[2] [4] Because most examples of drug-induced liver disease occur in a very small proportion of patients using any given agent, it is likely that the individual risk is determined by various combinations of the environmental and genetic variations in multiple factors.

APOPTOSIS VERSUS NECROSIS

Two types of cell death can be distinguished by morphological features, although it is likely that these are two ends of a spectrum with possible intermediate forms.[3] [4] [7] Apoptosis involves shrinkage, nuclear disassembly, and fragmentation of the cell into discrete bodies with intact plasma membranes. These are rapidly phagocytosed by neighboring cells. The process involves the participation of specialized proteolytic enzymes, caspases, that contain cysteine at the active sites and cleave at aspartate.[8] Thus, specific proteins are cleaved, leading to both activation and inactivation of specific targets with resultant dismantling of the cell and inactivation of repair.

An important feature of apoptosis is the requirement for adenosine triphosphate (ATP) to initiate the execution phase.[9] In contrast, necrotic cell death is characterized by cell swelling and lysis. This is usually a consequence of profound loss of mitochondrial function and resultant ATP depletion, leading to loss of ion homeostasis, including volume regulation, and increased Ca_i ²⁺; the latter activates a number of nonspecific hydrolases (i.e., proteases, nucleases, and phospholipases). A fundamental distinction between necrosis and apoptosis is the loss of plasma membrane integrity; this is integral to the former but not the latter.[4] As a consequence, lytic release of cellular constituents promotes a local inflammatory reaction, whereas the rapid removal of apoptotic bodies minimizes such a reaction.

The distinction between the two modes of death is easily accomplished in vitro but not in vivo. Thus, although claims that certain drugs induce apoptosis in the liver have been made, these are relatively unconvincing. DNA fragmentation can occur in necrosis, leading to positive TUNEL staining of liver sections.[10] [11] Conversely, when apoptosis is massive, it can exceed the capacity for rapid phagocytosis, resulting in the eventual appearance of secondary necrosis. In drug-induced hepatitis it appears likely that direct toxicity of drug metabolites and profound oxidative stress induce necrotic cell death (e.g., acetaminophen), whereas in circumstances in which the drug sensitizes the liver to cytokines, apoptotic cell death may predominate. These distinctions are not simply of theoretical interest because strategies designed to protect the liver need to be directed at the relevant pathway for specific drugs.

INITIATION OF CELL DEATH AND THE ROLE OF MITOCHONDRIA

The initiation of cell death begins either at the plasma membrane with the binding of TNF or FasL to their cognate receptors (i.e., death receptors[12] [13]) or within the cell. The latter is due to the occurrence of intracellular stress in the form of biochemical events such as oxidative stress, redox changes, covalent binding, lipid peroxidation, and consequent functional effects on mitochondria, endoplasmic reticulum (ER), microtubules, cytoskeleton, or DNA (Fig. [1]). Intracellular stress either directly affects mitochondria or can lead to effects on other organelles, which then send signals to the mitochondria to recruit participation in the death process. The components of this cell death machinery are listed in Table [1].

The initiation of cell death by both death receptors (extrinsic trigger) and intracellular stress (intrinsic trigger) usually involves the participation of mitochondria with the release of proapoptotic proteins from the intermembrane space (Fig. [2]). When death receptors are engaged, adaptor proteins (FADD, TRADD) bind to the cytoplasmic tail of the receptors. Procaspase 8 is then recruited and self-cleaves to release caspase 8, which in turn cleaves Bid, a proapoptotic Bcl-2 family member.[3] [4] tBid then causes the translocation of Bax to the mitochondria and the aggregation of Bax and Bak.[14] Bax and Bak promote the permeabilization of the mitochondria to release cytochrome c, Smac, AIF, and some procaspases sequestered in the intermembrane space. Cytoskeletal stress (e.g., anoikis) releases Bmf (sequestered with myosin motors), which translocates to mitochondria and promotes permeabilization.[15] Microtubular stress releases Bim (sequestered with dynein motors) in a similar fashion[15] (Fig. [3]). Bcl₂ and Bcl-XL prevent permeabilization.[14] [15]

Thus, the interaction of proapoptotic Bcl₂ family members, Bax, Bak, Bmf, and Bim, and antiapoptotic members of this family, Bcl-2 and Bcl-XL, serves as a gateway that regulates the integrity of mitochondria.[14] [15] The extrinsic initiation pathway signals via caspase 8 and tBid to Bax and Bak, whereas the intrinsic initiation of apoptosis involves the direct participation of Bax, Bak, Bmf, and Bim (Fig. [2]). The precise details of what activates these proapoptotic Bcl₂ members as a consequence of intracellular stress are uncertain at present. It is also unclear how proapoptotic and antiapoptotic Bcl2 members function and antagonize each other. Furthermore, in addition to activation and translocation of proapoptotic Bcl₂ members, inactivation of antiapoptotic members may be of importance in disrupting the balance at this gateway. The JNK (stress-activated kinase)-mediated phosphorylation of Bcl-XL is an example.[16]

A critical issue is the mechanism of mitochondrial permeabilization. One school of thought is that the proapoptotic Bcl₂ members selectively permeabilize the outer membrane, allowing the release of intermembrane proteins. The other school of thought is that opening of a megachannel composed of inner and outer membrane proteins, referred to as the membrane permeability transition (MPT) pore, determines release. The pore consists of the outer membrane voltage-dependent anion channel (VDAC) and the peripheral benzodiazepine receptor, the inner membrane adenine nucleotide translocase (ANT), matrix cyclophin (binds cyclosporin A, which inhibits channel opening), and cytosol hexokinase and creatine kinase. Some of the Bcl-2 family directly bind to pore constituents, such as ANT, and thus may regulate the pore-open versus pore-closed configuration. The pore opening is promoted by Ca²⁺ and oxidative stress. The pore contains functional dithiols, whose oxidation in response to thiol-disulfide redox changes promotes pore opening.[17] Extensive opening of the pore depolarizes mitochondria and results in profound ATP depletion. Thus, pore opening is a critical and perhaps universal event in necrotic cell death. To reconcile pore opening with propagation of the apoptotic cascade, it has been proposed that not all mitochondria in a cell participate or that the pore fluctuates from open to closed status. The availability of ATP is required for continued propagation of the apoptosis cascade.[9] MPT opening in the presence of sustained ATP will lead to apoptosis but in the absence of a critical threshold level of ATP will lead to necrosis.

The ultimate evolution of apoptosis is dependent on the participation of caspases. Caspases can be divided into initiators and executioners; caspases exist as zymogens, which are activated either by self-cleavage or cleavage by other caspases.[8] The initiator caspases include caspase 8, as described earlier, and caspase 9, which is activated (self-cleavage) by interaction with cytochrome c/apaf-1 adaptor complex (so-called apoptosome). Thus, the release of cytochrome c is of major importance in initiating the formation of the activating complex. ATP is required for the proper confirmation of the apoptosome (Fig. [3]). Caspase 9 then initiates apoptosis by cleaving caspase 3. This executes the final dismantling of the cell with the participation of other caspases, such as 6 and 7, which are cleaved by caspase 3.[18] Thus, intracellular stress promotes apoptosis at the level of the mitochondria to cause the release of cytochrome c and the activation of caspase 9 to initiate the cascade. There is evidence for self-amplification so that caspase 3 can cleave caspase 8 later in the process and recruit the upstream components of the cascade.

Aside from cytochrome c, other intermembrane proteins are released from mitochondria and may have considerable importance. Among these are apoptosis-inducing factor,[19] [20] which activates DNA fragmentation, and Smac, which binds caspase inhibitory proteins (i.e., inhibitor of apoptosis proteins [IAPs]), removing these inhibitors from interaction with caspases.

Another type of intracellular stress that may not involve mitochondria is ER stress. Malfolding or protein overloading induced by toxins, oxidative stress, viral protein, or mutant protein accumulation in the ER up-regulates production of caspase 12; this is sequestered on the cytoplasmic face of the ER, where it binds and activates cytoplasmic caspase 7, resulting in cleavage of procaspase 12.[21] Caspase 12 is then released into the cytosol, where it binds and cleaves procaspase 9 to initiate apoptosis independent of mitochondria or the apoptosome (Fig. [3]).

SURVIVAL MECHANISMS

An important means of protecting against apoptosis is through inhibition of caspases (Table [1]). This can be accomplished chemically by attack on the cysteine thiol in the active site by NO,[22] reactive oxygen species,[23] or the redox effect of profound depletion of reduced glutathione (GSH).[24] Alternatively, specific protein interactions can inhibit caspases. Inhibitors include IAP family members[25] and heat shock proteins (HSPs).[26] In addition, FLIP inhibits activation of caspase 8 by acting as a decoy for binding to FADD.[27] Bcl-2 and Bcl-X_L inhibit mitochondrial permeabilization,[28] [29] and phosphatidylinositol 3-kinase (PI-3-kinase)/Akt promotes survival by phosphorylation of caspase 9 and activation of nuclear factor κB (NF-κB).[30]

NF-κB is an important transcription factor that promotes the transcription of a number of survival genes including IAPs, inducible nitric oxide synthase (iNOS), and Bcl-X_L.[31] [32] NF-κB is sequestered in the cytoplasm bound to IκBα. Phosphorylation of IκBαby IκBα kinase (IKK) leads to proteosomal degradation of IκBα, releasing NF-κB for translocation to the nucleus. IKK is activated by tumor necrosis factor receptor (TNF-R) signaling, oxidative stress, and PI-3-kinase.

SENSITIZATION TO TUMOR NECROSIS FACTOR

A role for nonparenchymal cells in hepatotoxicity has been demonstrated in various experimental systems, such as animal models of CCl₄ and acetaminophen hepatotoxicity. In the case of CCl4 , TNF appears to be a key mediator of toxicity as its neutralization[33] or the use of TNF-R1 knockout mice prevents toxicity despite a lack of change in hepatocellular metabolism of CCl₄.[34] This suggests that CCl₄ sensitizes hepatocytes to the lethal action of TNF. In the case of acetaminophen, macrophage inhibitors protect,[35] but the role of TNF is less clear. No change in susceptibility to acetaminophen was observed in TNF knockout mice.[36] However, enhanced susceptibility was observed in C-C chemokine receptor 2 (CCR2) knockouts.[37] CCR2 is a receptor for MCP-1. Acetaminophen stimulates MCP-1 production in hepatocytes. MCP-1, via CCR2, down-regulates nonparenchymal cell production of TNF. In CCR2 knockouts, increased TNF production was seen in response to acetaminophen and immunoneutralization of TNF abrogated the toxicity of acetaminophen.

How can toxins sensitize the hepatocyte to the lethal action of TNF? To address this question it is important to recognize that hepatocytes, although expressing abundant TNF-R1, are resistant to the lethal effects of TNF. When TNF engages TNF-R1, multiple signaling systems are activated-several promote cell death and involve activation of caspase 8 or stress-activated kinases, whereas another signaling pathway involves activation of NF-κB and protection. The NF-κB transcriptional up-regulation of survival genes then protects against apoptosis while enhancing inflammation through the up-regulation of cytokines, chemokines, and adhesion molecules. Inhibition of NF-κB unmasks the lethal effects of TNF through the apoptotic cascade.[38] Certain toxins, such as galactosamine and α-amanitin (mushroom poison) as well as actinomycin D, sensitize to TNF by inducing a transcriptional arrest that blocks the production of survival genes.[39] The extent to which inhibition of NF-κB activation or transcription occurs with CCl₄, acetaminophen, or clinically important toxins is uncertain.

Another mechanism of sensitization to TNF is GSH depletion. Although there is some controversy about whether GSH depletion sensitizes or inhibits TNF action, the timing of depletion and its duration may be important.[40] An important factor is compartmentation of GSH in mitochondria and cytosol. Depletion of mitochondrial GSH in fibrosarcoma cells and hepatocytes sensitizes to TNF-induced necrosis.[41] [42] [43] TNF increases oxidative stress in mitochondria, which is believed to be a result of the translocation of GD3 sphingolipids to mitochondria and blockage of electron transport.[44] Selective mitochondrial GSH depletion by 3-hydroxy-4-pentenoate or chronic alcohol feeding appears to render hepatic mitochondria incapable of detoxifying the TNF-induced burden of reactive oxygen species (ROS).[42] [43] Presumably as a consequence, MPT opening occurs with resultant collapse of the mitochondrial membrane potential, ATP depletion, and necrosis. It is also conceivable that the redox effects of mitochondrial GSH depletion directly sensitize the MPT pore to opening. Antioxidants protect against this TNF-induced necrosis occurring with selective mitochondrial GSH depletion, underscoring the role of enhanced mitochondrial oxidative stress.[42] [43]

Cytoplasmic GSH depletion may sensitize to TNF by additional mechanisms. We have observed sensitization of mouse hepatocytes to TNF-induced apoptosis by acute, moderate GSH depletion induced by either diethyl maleate or acetaminophen (N. Kaplowitz et al, unpublished observations). This was not inhibited by antioxidants, indicating that sensitization was due to the redox changes in thiol-disulfide status as a consequence of GSH depletion. Candidates for redox-sensitive factors that could sensitize to TNF include stress kinases and NF-κB binding and transactivation. GSH depletion is known to activate JNK and p38 kinase through redox inhibition of thioredoxin and GSH S-transferase Pi, which normally binds to and inhibits stress kinases.[45] [46] Furthermore, GSH depletion is known to inhibit NF-κB- dependent transcription in Jurkat cells.[47]

OXIDATIVE STRESS

Oxidative stress occurs when cells are exposed to increased ROS. This occurs when there is enhancement of reactive oxygen production or inhibition of its removal.[48] The mitochondria are the most important source of ROS, mainly because of auto-oxidation of ubisemiquinone at the level of complex III. This occurs to some extent under physiological, aerobic conditions and explains why profound depletion of GSH alone is lethal for aerobic cells. It is enhanced by blockage of electron transport, as occurs with TNF action or as consequence of loss of cytochrome c, or because of increased electron flow, as occurs with the uncoupling action of some drugs. Other important cellular sources of ROS include CYP2E1, peroxisomal oxidases, iron and copper overload, and NADPH oxidase of phagocytic cells.

The biochemical consequences of oxidative stress include lipid peroxidation, protein thiol oxidation, DNA oxidation, and changes in the GSH/GSSG redox buffer. The consequences of these chemical changes may include MPT pore opening, activation of redox-sensitive kinases and transcription factors, such as AP-1, NF-κB, and p53. This may lead to alterations of gene expression and either promote or protect against cell death.

The role of oxidative stress in cell death is complex. Massive oxidative stress, particularly in mitochondria, induces necrosis. Lesser exposure may be sufficient to trigger apoptosis-possible upstream targets include JNK, other kinases, p53, ER protein accumulation, microtubule assembly, and so forth. Aside from initiating apoptosis, oxidative stress may occur as a downstream secondary event due to uncoupling of mitochondria, cytochrome c release, or the nearly universal massive export of cell GSH that accompanies apoptosis. In addition, oxidative stress may protect against apoptosis in some circumstances as a consequence of the inhibition of caspases by ROS or the activation of NF-κB and increased expression of survival genes. Obviously, the effects of oxidative stress are complex and contradictory.

In liver injury, profound, directly lethal oxidative stress is rarely encountered and is typified by acetaminophen-induced necrosis.[49] [50] Otherwise, it is encountered in response to exogenous sources, such as infiltrating inflammatory cells, TNF action, or sometimes CYP2E1 induction or storage of unsaturated fatty acids. Thus, fatty liver predisposes to oxidative stress presumably by amplifying the capacity for free radical chain reactions.[51] Perhaps the most important consequence of oxidative stress in the pathogenesis of liver disease is the promotion of inflammation through the activation of transcription of cytokines, chemokines, and adhesion molecules.

THERAPEUTIC POSSIBILITIES AND PITFALLS

A number of strategies can be envisioned to protect the liver against toxins. If cytokines and death receptors participate, neutralization of cytokines with monoclonal antibodies or soluble receptors or inhibition of their synthesis or release (e.g., with pentoxifylline[52]) may be effective. Alternatively, we have shown that inhibition of microtubule-dependent targeting of death receptors to the cell surface protects hepatocytes against TNF and FasL.[53] A variety of nonselective and selective caspase inhibitors are available that can inhibit initiator and/or executioner caspases. Cyclosporin A and ursodeoxycholate inhibit the MPT opening.[54] [55] If oxidative stress or GSH depletion is involved in toxicity, the use of antioxidants such as tocopherol and silymarin or GSH precursors such as S-adenosylmethionine, N-acetylcysteine, and GSH esters may be effective. In addition, the use of inducers of gene expression or gene therapy may be employed to boost expression of survival genes such as IAP, FLIP, HSP, iNOS, Bcl-X_L or protective genes such as δ-glutamylcysteine synthetase, superoxide dismutase, GSH S-transferases, and peroxidases. Protective cytokines such as interleukin-10 (IL-10), MCP-1, or IL-6 may be of value in certain circumstances by down-regulating toxic cytokines, promoting survival gene expression or liver regeneration.[56]

There are a number of pitfalls that must be taken into consideration in designing an effective therapeutic strategy. Although inhibition of caspases may inhibit apoptosis, intracellular stress may persist in leading to MPT opening, resulting in necrosis or a switch from apoptotic to necrotic cell death.[57] [58] In addition, interference with TNF production or receptor signaling may have the adverse consequence of inhibiting hepatic regeneration[38] or altering susceptibility to and outcome of complicating bacterial infections.

Designing strategies to treat toxin exposure is best accomplished by a detailed understanding of the mechanisms involved in individual circumstances. It is crucial to have knowledge of the role of the extrinsic and intrinsic pathways, the intracellular triggers of stress, the participation of the MPT, and the contribution of apoptosis and necrosis as well as the potential for switching from one mode of cell death to the other.

FINANCIAL DISCLOSURE

The author is supported by the USC Research Center for Liver Diseases (NIH P30DK48522-07) and the USC-UCLA Research Center for Alcoholic Liver and Pancreatic Diseases (NIH P50AA11999-03). The author is a consultant for GlaxoSmithKline, Pfizer, Metabolex, Bristol Myers Squibb, Serono, Praecis Pharmaceutical, Socratech, and Actelion and is on the speaker bureau of Axcan Scandipharm.

ABBREVIATIONS

ANT adenine nucleotide translocase

ATP adenosine triphosphate

CCR2 C-C chemokine receptor 2

ER endoplasmic reticulum

GSH reduced glutathione

HSP heat shock protein

IAP inhibitor of apoptosis protein

IL-10 interleukin-10

INOS inducible nitric oxide synthase

MPT membrane permeability transition

NF-κB nuclear factor κB

PI-3-kinase phosphatidylinositol 3-kinase

TNF-R tumor necrosis factor receptor

VDAC voltage-dependent anion channel

REFERENCES

• 1 Zimmerman H. Hepatotoxicity: The Adverse Effects of Drugs and Other Chemicals on the Liver, 2nd ed. Philadelphia. Lippincott Williams & Wilkins 1999
  Search in Google Scholar
• 2 Kaplowitz N. Drug-induced liver disorders: implications for drug development and regulation.  Drug Saf . 2001;  24 483-490
  PubMedSearch in Google Scholar
• 3 Kaplowitz N. Cell death at the millennium: implications for liver diseases.  Clin Liver Dis . 2000;  4 1-23
  PubMedSearch in Google Scholar
• 4 Kaplowitz N. Mechanisms of liver cell injury.  J Hepatol . 2000;  32 39-47
  PubMedSearch in Google Scholar
• 5 Robin M, LeRoy M, Descatoire V, Pessayre D. Plasma membrane cytochromes P450 as neoantigens and autoimmune targets in drug-induced hepatitis.  J Hepatol . 1997;  26 23-30
  CrossrefPubMedSearch in Google Scholar
• 6 Liu Z-X, Kaplowitz N. Immune-mediated drug-induced liver disease (in press).  Clin Liver Dis.
  PubMed
• 7 Kaplowitz N. Mechanisms of cell death and relevance to drug hepatotoxicity. In: Kaplowitz N, DeLeve L, eds. Drug Hepatotoxicity New York: Marcel Dekker (in press)
• 8 Thornberry N A, Lazebnik Y. Caspase: enemies within.  Science . 1998;  281 1312-1316
  CrossrefPubMedSearch in Google Scholar
• 9 Leist M, Single B, Castoldi A F. Intracellular adenosine triphosphate (ATP) concentration: a switch in the decision between apoptosis and necrosis.  J Exp Med . 1997;  185 1481-1486
  CrossrefPubMedSearch in Google Scholar
• 10 Grasl-Kraupp B, Ruttkay-Nedecky B, Koudelka H. In situ detection of fragmented DNA (TUNEL assay) fails to discriminate among apoptosis, necrosis and autolytic cell death.  Hepatology . 1995;  21 1465-1468
  CrossrefPubMedSearch in Google Scholar
• 11 Dong Z, Saikumar P, Weinberg J, Venkatachalam M. Internucleosomal DNA cleavage triggered by plasma membrane damage during necrotic cell death.  Am J Pathol . 1997;  151 1205-1213
  PubMedSearch in Google Scholar
• 12 Ashkenazi A, Dixit V. Death receptors: signaling and modulation.  Science . 1998;  281 1305-1308
  CrossrefPubMedSearch in Google Scholar
• 13 Faubion W A, Gores G. Death receptors in liver biology and pathology.  Hepatology . 1999;  29 1-4
  CrossrefPubMedSearch in Google Scholar
• 14 Wei M C, Zong W X, Cheng E. Pro-apoptotic BAX and BAK: a requisite gateway to mitochondrial dysfunction and death.  Science . 2001;  292 727-730
  CrossrefPubMedSearch in Google Scholar
• 15 Puthalakath H, Villunger A, O'Reilly L A. Bmf: a proapoptotic BH3-only protein regulated by interaction with the myosin V actin motor complex, activated by anoikis.  Science . 2001;  293 1829-1832
  CrossrefPubMedSearch in Google Scholar
• 16 Kyriakis J, Banerjee P, Nilolakaki E. The stress-activated protein kinase subfamily of c-Jun kinases.  Nature . 1994;  369 156-160
  CrossrefPubMedSearch in Google Scholar
• 17 Zamzani N, Marzo I, Susin S. Thiol crosslinking agent diamide overcomes the apoptosis-inhibitory effect of Bcl-2 by enforcing mitochondrial permeability transition.  Oncogene . 1998;  16 1055-1063
  CrossrefPubMedSearch in Google Scholar
• 18 Slee E A, Adrain C, Martin S J. Executioner caspase-3, -6, and -7 perform distinct, non-redundant roles during the demolition phase of apoptosis.  J Biol Chem . 2001;  276 7320-7326
  CrossrefPubMedSearch in Google Scholar
• 19 Susin S, Lorenzo H, Zamzani N. Molecular characterization of mitochondrial apoptosis-inducing factor.  Nature . 1999;  397 441-446
  CrossrefPubMedSearch in Google Scholar
• 20 Lorenzo H K, Susin S A, Penninger J, Kroemer G. Apoptosis inducing factor (AIF): a phylogenetically old, caspase-independent effector of cell death.  Cell Death Differ . 1999;  6 516-524
  CrossrefPubMedSearch in Google Scholar
• 21 Rao R V, Hermel E, Castro-Obregon S. Coupling endoplasmic reticulum stress to the cell death program: mechanism of caspase activation.  J Biol Chem . 2001;  276 33869-33874
  CrossrefPubMedSearch in Google Scholar
• 22 Kim Y-M, Talanian R V, Billiar T R. Nitric oxide inhibits apoptosis by preventing increases in caspase-3-like activity via two distinct mechanisms.  J Biol Chem . 1997;  272 31138-31148
  CrossrefPubMedSearch in Google Scholar
• 23 Hampton M B, Orrenius S. Dual regulation of caspase activity by hydrogen peroxide: implications for apoptosis.  FEBS Lett . 1997;  414 552-556
  CrossrefPubMedSearch in Google Scholar
• 24 Hentze H, Kunstle G, Volbracht C. CD95-mediated murine hepatic apoptosis requires an intact glutathione status.  Hepatology . 1999;  30 177-185
  CrossrefPubMedSearch in Google Scholar
• 25 Goyal L. Cell death inhibition: keeping caspases in check.  Cell . 2001;  104 805-808
  CrossrefPubMedSearch in Google Scholar
• 26 Masser D D, Antoine C, Bourget L. Role of the human heat shock protein hsp 70 in protection against stress-induced apoptosis.  Mol Cell Biol . 1997;  17 5317-5327
  PubMedSearch in Google Scholar
• 27 Hu S, Vincenz C, Ni J. I-FLICE, a novel inhibitor of tumor necrosis factor receptor-1- and CD-95-induced apoptosis.  J Biol Chem . 1997;  272 17255-17257
  CrossrefPubMedSearch in Google Scholar
• 28 Adams J M, Cory S. The Bcl-2 protein family: arbiters of cell survival.  Science . 1998;  281 1322-1326
  CrossrefPubMedSearch in Google Scholar
• 29 Kroemer G. The proto-oncogene Bcl-2 and its role in regulating apoptosis.  Nat Med . 1997;  3 614-620
  CrossrefPubMedSearch in Google Scholar
• 30 Rust C, Karnitz L M, Paya C V. The bile acid tauroursodeoxycholate activates a phosphatidylinositol 3-kinase dependent survival signaling cascade.  J Biol Chem . 2000;  275 20210-20216
  CrossrefPubMedSearch in Google Scholar
• 31 Aggarwal B. Apoptosis and nuclear factor-κB: a tale of association and dissociation.  Biochem Pharmacol . 2000;  60 1033-1039
  CrossrefPubMedSearch in Google Scholar
• 32 Wang C Y, Mayo M W, Korneluk R G. NF-κB antiapoptosis: induction of TRAF1 and C-IAP1 and c-IAP2 to suppress caspase-8 activation.  Science . 1998;  281 1680-1683
  CrossrefPubMedSearch in Google Scholar
• 33 Czaja M, Xu J, Alt E. Prevention of carbon tetrachloride- induced rat liver injury by soluble tumor necrosis factor receptor.  Gastroenterology . 1995;  108 1849-1854
  PubMedSearch in Google Scholar
• 34 Morio L A, Chiu H, Sprowles K A. Distinct roles of tumor necrosis factor-α and nitric oxide in acute liver injury induced by carbon tetrachloride in mice.  Toxicol Appl Pharmacol . 2001;  172 44-51
  CrossrefPubMedSearch in Google Scholar
• 35 Laskin D L, Gardner C R, Price V F, Jollow D J. Modulation of macrophage functioning abrogates the acute hepatotoxicity of acetaminophen.  Hepatology . 1995;  21 1045-1050
  CrossrefPubMedSearch in Google Scholar
• 36 Boess F, Bopst M, Althaus R. Acetaminophen hepatotoxicity in tumor necrosis factor/lymphotoxin-α gene knockout mice.  Hepatology . 1998;  27 1021-1029
  CrossrefPubMedSearch in Google Scholar
• 37 Hogaboam C M, Bone-Lasron C L, Steinhauser M L. Exaggerated hepatic injury due to acetaminophen challenge in mice lacking c-c chemokine receptor 2.  Am J Pathol . 2000;  156 1245-1252
  PubMedSearch in Google Scholar
• 38 Iimuro Y, Nishiura T, Hellerbrand C. NFκB prevents apoptosis and liver dysfunction during liver regeneration.  J Clin Invest . 1998;  101 802-811
  PubMedSearch in Google Scholar
• 39 Leist M, Gantner F, Naumann H. Tumor necrosis factor-induced apoptosis during the poisoning of mice with hepatotoxins.  Gastroenterology . 1997;  112 923-934
  PubMedSearch in Google Scholar
• 40 Haouzi D, Lekehal M, Tinel M. Prolonged, but not acute, glutathione depletion promotes Fas-mediated mitochondrial permeability transition and apoptosis in mice.  Hepatology . 2001;  33 1181-1188
  CrossrefPubMedSearch in Google Scholar
• 41 Goosens V, Grooten J, DeVos K, Fiers W. Direct evidence for tumor necrosis factor-induced mitochondrial reactive oxygen intermediates and their involvement in cytotoxicity.  Proc Natl Acad Sci USA . 1995;  92 8115-8119
  PubMedSearch in Google Scholar
• 42 Colell A, Garcia-Ruiz C, Miranda M. Selective glutathione depletion of mitochondria by ethanol sensitizes hepatocytes to tumor necrosis factor.  Gastroenterology . 1998;  115 1541-1551
  PubMedSearch in Google Scholar
• 43 Colell A, Coll O, Garcia-Ruiz C. Tauroursodeoxycholic acid protects hepatocytes from ethanol-fed rats against TNF-induced cell death by replenishing mitochondrial GSH.  Hepatology . 2001;  34 964-971
  CrossrefPubMedSearch in Google Scholar
• 44 Garcia-Ruiz C, Colell A, Paris R, Fernandez-Checa J C. Direct interaction of GD3 ganglioside with mitochondria generates reactive oxygen species followed by mitochondrial permeability transition, cytochrome c release and caspase activation.  FASEB J . 2000;  14 847-858
  PubMedSearch in Google Scholar
• 45 Sithoh M, Nishitoh H, Fuju M. Mammalian thioredoxin is a direct inhibitor of apoptosis signal-regulating kinase (ASK)-1.  EMBO J . 1998;  17 2596-1206
  CrossrefPubMedSearch in Google Scholar
• 46 Adler V, Yin Z, Fuchs S. Regulation of JNK by GSTpi.  EMBO J . 1999;  18 1321-1334
  CrossrefPubMedSearch in Google Scholar
• 47 Ginne-Pease M E, Whisler R L. Optimal NF-κB mediated transcriptional responses in Jurkat T cells exposed to oxidative stress are dependent on intracellular glutathione and costimulatory signals.  Biochem Biophys Res Commun . 1996;  226 695-702
  CrossrefPubMedSearch in Google Scholar
• 48 Kaplowitz N, Tsukamoto H. Oxidative stress and liver disease.  Prog Liver Dis . 1996;  14 131-159
  PubMedSearch in Google Scholar
• 49 Adamson G, Harman A. Oxidative stress in cultured hepatocytes exposed to acetaminophen.  Biochem Pharmacol . 1993;  45 2289-2294
  CrossrefPubMedSearch in Google Scholar
• 50 Arnaiz S, Llesuy S, Cutrin J, Boveris A. Oxidative stress by acute acetaminophen administration in mouse liver.  Free Radic Biol Med . 1995;  19 303-310
  CrossrefPubMedSearch in Google Scholar
• 51 Sneed R A, Buchweitz J P, Jean P A, Ganey P E. Pentoxifylline attenuates bacterial lipopolysaccharide-induced enhancement of allyl alcohol hepatotoxicity.  Toxicol Sci . 2000;  56 203-210
  CrossrefPubMedSearch in Google Scholar
• 52 Berson A, DeBeco V, Letteron P. Steatohepatitis-inducing drugs cause mitochondrial dysfunction and lipid peroxidation in rat hepatocytes.  Gastroenterology . 1998;  114 764-774
  PubMedSearch in Google Scholar
• 53 Feng G, Kaplowitz N. Colchicine protects mice from the lethal effect of agonistic anti-Fas antibody.  J Clin Invest . 2000;  105 329-339
  PubMedSearch in Google Scholar
• 54 Botla R, Spivey J, Aguilar H. Ursodeoxycholate (UDCA) inhibits the mitochondrial membrane permeability transition induced by glycochenodeoxycholate: a mechanism of UDCA cytoprotection.  J Pharmacol Exp Ther . 1995;  272 930-938
  PubMedSearch in Google Scholar
• 55 Rodrigues C MP, Fan G, Ma X. A novel role for ursodeoxycholic acid in inhibiting apoptosis by modulating mitochondrial membrane perturbation.  J Clin Invest . 1998;  101 2790-2799
  CrossrefPubMedSearch in Google Scholar
• 56 Kovalovich K, Li W, DeAngelis R. Interleukin-6 protects against Fas-mediated death by establishing a critical level of anti-apoptotic hepatic proteins FLIP, Bcl-2, and Bcl-xL.  J Biol Chem . 2001;  276 26605-26613
  CrossrefPubMedSearch in Google Scholar
• 57 Lemasters J J. Mechanisms of hepatic toxicity v.  necrapoptosis and the mitochondrial permeability transition: shared pathways to necrosis and apoptosis. Am J Physiol . 1999;  276 G1-G6
  PubMedSearch in Google Scholar
• 58 Lemaire C, Andreau K, Souvannavong V. Inhibition of caspase induces a switch from apoptosis to necrosis.  FEBS Lett . 1998;  425 266-270
  CrossrefPubMedSearch in Google Scholar

REFERENCES

• 1 Zimmerman H. Hepatotoxicity: The Adverse Effects of Drugs and Other Chemicals on the Liver, 2nd ed. Philadelphia. Lippincott Williams & Wilkins 1999
  Search in Google Scholar
• 2 Kaplowitz N. Drug-induced liver disorders: implications for drug development and regulation.  Drug Saf . 2001;  24 483-490
  PubMedSearch in Google Scholar
• 3 Kaplowitz N. Cell death at the millennium: implications for liver diseases.  Clin Liver Dis . 2000;  4 1-23
  PubMedSearch in Google Scholar
• 4 Kaplowitz N. Mechanisms of liver cell injury.  J Hepatol . 2000;  32 39-47
  PubMedSearch in Google Scholar
• 5 Robin M, LeRoy M, Descatoire V, Pessayre D. Plasma membrane cytochromes P450 as neoantigens and autoimmune targets in drug-induced hepatitis.  J Hepatol . 1997;  26 23-30
  CrossrefPubMedSearch in Google Scholar
• 6 Liu Z-X, Kaplowitz N. Immune-mediated drug-induced liver disease (in press).  Clin Liver Dis.
  PubMed
• 7 Kaplowitz N. Mechanisms of cell death and relevance to drug hepatotoxicity. In: Kaplowitz N, DeLeve L, eds. Drug Hepatotoxicity New York: Marcel Dekker (in press)
• 8 Thornberry N A, Lazebnik Y. Caspase: enemies within.  Science . 1998;  281 1312-1316
  CrossrefPubMedSearch in Google Scholar
• 9 Leist M, Single B, Castoldi A F. Intracellular adenosine triphosphate (ATP) concentration: a switch in the decision between apoptosis and necrosis.  J Exp Med . 1997;  185 1481-1486
  CrossrefPubMedSearch in Google Scholar
• 10 Grasl-Kraupp B, Ruttkay-Nedecky B, Koudelka H. In situ detection of fragmented DNA (TUNEL assay) fails to discriminate among apoptosis, necrosis and autolytic cell death.  Hepatology . 1995;  21 1465-1468
  CrossrefPubMedSearch in Google Scholar
• 11 Dong Z, Saikumar P, Weinberg J, Venkatachalam M. Internucleosomal DNA cleavage triggered by plasma membrane damage during necrotic cell death.  Am J Pathol . 1997;  151 1205-1213
  PubMedSearch in Google Scholar
• 12 Ashkenazi A, Dixit V. Death receptors: signaling and modulation.  Science . 1998;  281 1305-1308
  CrossrefPubMedSearch in Google Scholar
• 13 Faubion W A, Gores G. Death receptors in liver biology and pathology.  Hepatology . 1999;  29 1-4
  CrossrefPubMedSearch in Google Scholar
• 14 Wei M C, Zong W X, Cheng E. Pro-apoptotic BAX and BAK: a requisite gateway to mitochondrial dysfunction and death.  Science . 2001;  292 727-730
  CrossrefPubMedSearch in Google Scholar
• 15 Puthalakath H, Villunger A, O'Reilly L A. Bmf: a proapoptotic BH3-only protein regulated by interaction with the myosin V actin motor complex, activated by anoikis.  Science . 2001;  293 1829-1832
  CrossrefPubMedSearch in Google Scholar
• 16 Kyriakis J, Banerjee P, Nilolakaki E. The stress-activated protein kinase subfamily of c-Jun kinases.  Nature . 1994;  369 156-160
  CrossrefPubMedSearch in Google Scholar
• 17 Zamzani N, Marzo I, Susin S. Thiol crosslinking agent diamide overcomes the apoptosis-inhibitory effect of Bcl-2 by enforcing mitochondrial permeability transition.  Oncogene . 1998;  16 1055-1063
  CrossrefPubMedSearch in Google Scholar
• 18 Slee E A, Adrain C, Martin S J. Executioner caspase-3, -6, and -7 perform distinct, non-redundant roles during the demolition phase of apoptosis.  J Biol Chem . 2001;  276 7320-7326
  CrossrefPubMedSearch in Google Scholar
• 19 Susin S, Lorenzo H, Zamzani N. Molecular characterization of mitochondrial apoptosis-inducing factor.  Nature . 1999;  397 441-446
  CrossrefPubMedSearch in Google Scholar
• 20 Lorenzo H K, Susin S A, Penninger J, Kroemer G. Apoptosis inducing factor (AIF): a phylogenetically old, caspase-independent effector of cell death.  Cell Death Differ . 1999;  6 516-524
  CrossrefPubMedSearch in Google Scholar
• 21 Rao R V, Hermel E, Castro-Obregon S. Coupling endoplasmic reticulum stress to the cell death program: mechanism of caspase activation.  J Biol Chem . 2001;  276 33869-33874
  CrossrefPubMedSearch in Google Scholar
• 22 Kim Y-M, Talanian R V, Billiar T R. Nitric oxide inhibits apoptosis by preventing increases in caspase-3-like activity via two distinct mechanisms.  J Biol Chem . 1997;  272 31138-31148
  CrossrefPubMedSearch in Google Scholar
• 23 Hampton M B, Orrenius S. Dual regulation of caspase activity by hydrogen peroxide: implications for apoptosis.  FEBS Lett . 1997;  414 552-556
  CrossrefPubMedSearch in Google Scholar
• 24 Hentze H, Kunstle G, Volbracht C. CD95-mediated murine hepatic apoptosis requires an intact glutathione status.  Hepatology . 1999;  30 177-185
  CrossrefPubMedSearch in Google Scholar
• 25 Goyal L. Cell death inhibition: keeping caspases in check.  Cell . 2001;  104 805-808
  CrossrefPubMedSearch in Google Scholar
• 26 Masser D D, Antoine C, Bourget L. Role of the human heat shock protein hsp 70 in protection against stress-induced apoptosis.  Mol Cell Biol . 1997;  17 5317-5327
  PubMedSearch in Google Scholar
• 27 Hu S, Vincenz C, Ni J. I-FLICE, a novel inhibitor of tumor necrosis factor receptor-1- and CD-95-induced apoptosis.  J Biol Chem . 1997;  272 17255-17257
  CrossrefPubMedSearch in Google Scholar
• 28 Adams J M, Cory S. The Bcl-2 protein family: arbiters of cell survival.  Science . 1998;  281 1322-1326
  CrossrefPubMedSearch in Google Scholar
• 29 Kroemer G. The proto-oncogene Bcl-2 and its role in regulating apoptosis.  Nat Med . 1997;  3 614-620
  CrossrefPubMedSearch in Google Scholar
• 30 Rust C, Karnitz L M, Paya C V. The bile acid tauroursodeoxycholate activates a phosphatidylinositol 3-kinase dependent survival signaling cascade.  J Biol Chem . 2000;  275 20210-20216
  CrossrefPubMedSearch in Google Scholar
• 31 Aggarwal B. Apoptosis and nuclear factor-κB: a tale of association and dissociation.  Biochem Pharmacol . 2000;  60 1033-1039
  CrossrefPubMedSearch in Google Scholar
• 32 Wang C Y, Mayo M W, Korneluk R G. NF-κB antiapoptosis: induction of TRAF1 and C-IAP1 and c-IAP2 to suppress caspase-8 activation.  Science . 1998;  281 1680-1683
  CrossrefPubMedSearch in Google Scholar
• 33 Czaja M, Xu J, Alt E. Prevention of carbon tetrachloride- induced rat liver injury by soluble tumor necrosis factor receptor.  Gastroenterology . 1995;  108 1849-1854
  PubMedSearch in Google Scholar
• 34 Morio L A, Chiu H, Sprowles K A. Distinct roles of tumor necrosis factor-α and nitric oxide in acute liver injury induced by carbon tetrachloride in mice.  Toxicol Appl Pharmacol . 2001;  172 44-51
  CrossrefPubMedSearch in Google Scholar
• 35 Laskin D L, Gardner C R, Price V F, Jollow D J. Modulation of macrophage functioning abrogates the acute hepatotoxicity of acetaminophen.  Hepatology . 1995;  21 1045-1050
  CrossrefPubMedSearch in Google Scholar
• 36 Boess F, Bopst M, Althaus R. Acetaminophen hepatotoxicity in tumor necrosis factor/lymphotoxin-α gene knockout mice.  Hepatology . 1998;  27 1021-1029
  CrossrefPubMedSearch in Google Scholar
• 37 Hogaboam C M, Bone-Lasron C L, Steinhauser M L. Exaggerated hepatic injury due to acetaminophen challenge in mice lacking c-c chemokine receptor 2.  Am J Pathol . 2000;  156 1245-1252
  PubMedSearch in Google Scholar
• 38 Iimuro Y, Nishiura T, Hellerbrand C. NFκB prevents apoptosis and liver dysfunction during liver regeneration.  J Clin Invest . 1998;  101 802-811
  PubMedSearch in Google Scholar
• 39 Leist M, Gantner F, Naumann H. Tumor necrosis factor-induced apoptosis during the poisoning of mice with hepatotoxins.  Gastroenterology . 1997;  112 923-934
  PubMedSearch in Google Scholar
• 40 Haouzi D, Lekehal M, Tinel M. Prolonged, but not acute, glutathione depletion promotes Fas-mediated mitochondrial permeability transition and apoptosis in mice.  Hepatology . 2001;  33 1181-1188
  CrossrefPubMedSearch in Google Scholar
• 41 Goosens V, Grooten J, DeVos K, Fiers W. Direct evidence for tumor necrosis factor-induced mitochondrial reactive oxygen intermediates and their involvement in cytotoxicity.  Proc Natl Acad Sci USA . 1995;  92 8115-8119
  PubMedSearch in Google Scholar
• 42 Colell A, Garcia-Ruiz C, Miranda M. Selective glutathione depletion of mitochondria by ethanol sensitizes hepatocytes to tumor necrosis factor.  Gastroenterology . 1998;  115 1541-1551
  PubMedSearch in Google Scholar
• 43 Colell A, Coll O, Garcia-Ruiz C. Tauroursodeoxycholic acid protects hepatocytes from ethanol-fed rats against TNF-induced cell death by replenishing mitochondrial GSH.  Hepatology . 2001;  34 964-971
  CrossrefPubMedSearch in Google Scholar
• 44 Garcia-Ruiz C, Colell A, Paris R, Fernandez-Checa J C. Direct interaction of GD3 ganglioside with mitochondria generates reactive oxygen species followed by mitochondrial permeability transition, cytochrome c release and caspase activation.  FASEB J . 2000;  14 847-858
  PubMedSearch in Google Scholar
• 45 Sithoh M, Nishitoh H, Fuju M. Mammalian thioredoxin is a direct inhibitor of apoptosis signal-regulating kinase (ASK)-1.  EMBO J . 1998;  17 2596-1206
  CrossrefPubMedSearch in Google Scholar
• 46 Adler V, Yin Z, Fuchs S. Regulation of JNK by GSTpi.  EMBO J . 1999;  18 1321-1334
  CrossrefPubMedSearch in Google Scholar
• 47 Ginne-Pease M E, Whisler R L. Optimal NF-κB mediated transcriptional responses in Jurkat T cells exposed to oxidative stress are dependent on intracellular glutathione and costimulatory signals.  Biochem Biophys Res Commun . 1996;  226 695-702
  CrossrefPubMedSearch in Google Scholar
• 48 Kaplowitz N, Tsukamoto H. Oxidative stress and liver disease.  Prog Liver Dis . 1996;  14 131-159
  PubMedSearch in Google Scholar
• 49 Adamson G, Harman A. Oxidative stress in cultured hepatocytes exposed to acetaminophen.  Biochem Pharmacol . 1993;  45 2289-2294
  CrossrefPubMedSearch in Google Scholar
• 50 Arnaiz S, Llesuy S, Cutrin J, Boveris A. Oxidative stress by acute acetaminophen administration in mouse liver.  Free Radic Biol Med . 1995;  19 303-310
  CrossrefPubMedSearch in Google Scholar
• 51 Sneed R A, Buchweitz J P, Jean P A, Ganey P E. Pentoxifylline attenuates bacterial lipopolysaccharide-induced enhancement of allyl alcohol hepatotoxicity.  Toxicol Sci . 2000;  56 203-210
  CrossrefPubMedSearch in Google Scholar
• 52 Berson A, DeBeco V, Letteron P. Steatohepatitis-inducing drugs cause mitochondrial dysfunction and lipid peroxidation in rat hepatocytes.  Gastroenterology . 1998;  114 764-774
  PubMedSearch in Google Scholar
• 53 Feng G, Kaplowitz N. Colchicine protects mice from the lethal effect of agonistic anti-Fas antibody.  J Clin Invest . 2000;  105 329-339
  PubMedSearch in Google Scholar
• 54 Botla R, Spivey J, Aguilar H. Ursodeoxycholate (UDCA) inhibits the mitochondrial membrane permeability transition induced by glycochenodeoxycholate: a mechanism of UDCA cytoprotection.  J Pharmacol Exp Ther . 1995;  272 930-938
  PubMedSearch in Google Scholar
• 55 Rodrigues C MP, Fan G, Ma X. A novel role for ursodeoxycholic acid in inhibiting apoptosis by modulating mitochondrial membrane perturbation.  J Clin Invest . 1998;  101 2790-2799
  CrossrefPubMedSearch in Google Scholar
• 56 Kovalovich K, Li W, DeAngelis R. Interleukin-6 protects against Fas-mediated death by establishing a critical level of anti-apoptotic hepatic proteins FLIP, Bcl-2, and Bcl-xL.  J Biol Chem . 2001;  276 26605-26613
  CrossrefPubMedSearch in Google Scholar
• 57 Lemasters J J. Mechanisms of hepatic toxicity v.  necrapoptosis and the mitochondrial permeability transition: shared pathways to necrosis and apoptosis. Am J Physiol . 1999;  276 G1-G6
  PubMedSearch in Google Scholar
• 58 Lemaire C, Andreau K, Souvannavong V. Inhibition of caspase induces a switch from apoptosis to necrosis.  FEBS Lett . 1998;  425 266-270
  CrossrefPubMedSearch in Google Scholar