292 Current , 2003, 4, 292-312 Human Hepatocytes as a Tool for Studying and Drug Metabolism

M.J. Gómez-Lechón*, M.T. Donato, J.V. Castell and R. Jover

Centro de Investigación, Hospital La Fe, Avda Campanar 21, 46009-Valencia, Spain

Abstract: Drugs are usually biotransformed into new chemical species that may have either toxic or therapeutic effects. Drug metabolism studies are routinely performed in laboratory animals but, due to metabolic interspecies differences when compared to man, they are not accurate enough to anticipate the metabolic profile of a drug in humans. Human hepatocytes in primary culture provide the closest in vitro model to human liver and the only model that can produce a metabolic profile of a given drug that is very similar to that found in vivo. However their availability is limited due to the restricted access to suitable tissue samples. The scarcity of human liver has led to optimising the cryopreservation of adult hepatocytes for long-term storage and regular supply. Human hepatocytes in primary culture express typical hepatic functions and express drug metabolising enzymes. Moreover, qualitative and quantitative similarities between in vitro and in vivo metabolism of drugs were observed. Different strategies have been envisaged to prolong cell survival and delay the spontaneous decay of the differentiated phenotype during culture. Thus, hepatocytes represent the most appropriate model for the evaluation of integrated drug metabolism, toxicity/metabolism correlations, mechanisms of , and the interactions (inhibition and induction) of xenobiotics and drug-metabolising enzymes. However, in view of limitations of primary hepatocytes, efforts are made to develop alternative cellular models (i.e. metabolic competent CYP-engineered cells stably expressing individual CYPs and transient expression of CYPs by transduction of hepatoma cells with recombinant adenoviruses). In summary, several cellular tools are available to address key issues at the earliest stages of drug development for a better candidate selection and hepatotoxicity risk assessment. Key Words: Cryopreservation, , CYP-engineered cells, drug-drug interaction, drug metabolism, hepatocytes, hepatotoxicity, drug induction.

INTRODUCTION Nowadays, the process of selecting drug candidates is becoming much more rational as studies on metabolism and After intake, drugs are usually biotransformed into new kinetics of drug candidates are implemented earlier by using chemical species that may have either toxic or therapeutic human in vitro hepatic models for these studies at the early effects. This process occurs at several sites in the organism, pre-clinical stages [7]. but the liver is the most active organ in the fate of foreign Different human-derived models are currently being used compounds [1, 2]. Although biotransformation generally is a to investigate human metabolism of drugs in vitro. On the detoxification process, drugs can also be transformed into one hand are human liver preparations like microsomes and new chemical species displaying greater pharmacological or on the other, more complex cellular systems like human toxic potential. CYP-engineered cells, hepatoma cell lines and isolated and Drug metabolism is a major determinant of drug cultured hepatocytes, which could give a more complete clearance, interindividual pharmacokinetic differences and, picture of the metabolism of a drug in man. Human CYP- indirectly, of the clinical and toxicity of drugs [3-5]. engineered cells are at present the most efficient tool for Altered can result in inadequate determining whether a given CYP can or cannot give rise to concentration of the drug at the site of action. This, in turn, a particular drug metabolite, however, these cells do not results in inappropriate pharmacodynamic action and/or allow anticipating the metabolic profile of a drug in man [8]. great variations in clinical response. Indeed, variability Several metabolic-competent cell lines have recently been reduces the safety margin [6]. Analysis of the metabolic developed by using expression vectors encoding full-length profile and hepatic clearance is important for drugs human CYP genes. Several groups succeeded in expressing eliminated by hepatic metabolism and, hence, development fully active CYP isozymes in cells that constitutively express of a new drug demands exhaustive characterisation of its such activities under the control of a strong promoter [9, 10]. metabolic profile. Therefore, the development of a new drug Nevertheless, human hepatocytes are recognized to be the requires an exhaustive characterisation not only of its closest model to human liver [11, 12]. Hepatocytes in pharmacological activity, but also knowledge of major chemically defined culture conditions express most typical enzymes involved in metabolite formation, and the potential hepatic biochemical functions, among which is the ability to enzyme inhibiting or enzyme inducing properties of the drug. metabolise drugs [1, 12, 13]. Primary hepatocytes are differentiated cells able to reproduce in vitro the response of *Address correspondence to this author at the Centro de Investigación, human liver to pathophysiological factors and are currently Hospital La Fe, Avda de Campanar 21, 46009-Valencia (Spain); Tel: +34 96 considered a valuable in vitro tool for determining drug 1973040; Fax: +34 96 1973018; E-mail: [email protected] metabolism and for assessing the risk of drug hepatotoxicity

1389-2002/03 $41.00+.00 © 2003 Bentham Science Publishers Ltd. Human Hepatocytes as a Tool for Studying Toxicity Current Drug Metabolism, 2003, Vol. 4, No. 4 293 in man [11, 12]. The restricted accessibility to suitable liver obtaining high yields of viable human hepatocytes have been samples has greatly hindered the widespread use of primary extensively studied and improved. Livers for transplantation cultures of human hepatocytes. Moreover, cell cultures need are perfused in situ with cold University of Wisconsin (UW) to be prepared each time from liver tissue, what makes it solution to avoid warm ischemia and organ is usually even more difficult to use human hepatocytes for routine maintained under these conditions for several hours until testing. The irregular supply of human liver for cell hepatocyte isolation. There are some concerns about liver harvesting purposes has led to a need for optimised protocols samples stored in the UW solution, as cold ischemia could be for the long-term storage of hepatocytes so that available considered as a possible factor involved in the efficiency of tissue can be used more efficiently. the isolation procedure and the metabolic competence of As a consequence of the restricted access to human liver cultured cells. It was reported that hepatocytes isolated from samples, alternatives to replace or reduce the need of human fresh liver maintain liver-specific functions better than hepatocytes have been explored. The use of hepatic cell hepatocytes obtained from organ donor liver after hypo- lines, either from naturally occurring hepatomas, or from thermic storage for several hours in UW solution [29-31]. laboratory-immortalized hepatocytes have been proposed as However other studies reported that human liver, isolated an alternative model to primary cultured hepatocytes but hepatocytes or liver slices can be efficiently stored in cold hepatoma cells do not constitute a real alternative, as these solutions for several hours with minor losses of functionality cells poorly express biotransformation activities [14, 15]. [25, 33, 37]. However, promising advances are being made with these A large number of viable adult human hepatocytes can be cells by expressing transcription factors that control CYP isolated from whole livers and from a single liver lobe [34, expression and are lacking in these cells [16-19]. 35], but perfusion requires large amounts of collagenase and Consequently, researchers are exploring several strategies in during the digestion process the perfusate is often irregularly order to achieve hepatic cell lines re-expressing the whole distributed through all the tissue giving low yield. Other spectrum of human xenobiotic-metabolising enzymes, as an proteolytic enzymes, such as liberase, have been successfully alternative to primary cultures [19-21]. used to isolate hepatocytes [23, 36]. Small tissue samples obtained in the course of liver surgery or from tissue ISOLATION OF HEPATOCYTES FROM HUMAN resections for pathological examination constitute a reliable LIVER SAMPLES source of viable hepatocytes. The two-step collagenase perfusion has been also successfully employed to isolate Human liver tissue has become increasingly available for hepatocytes from these samples [12, 29, 37-39]. Whenever research purposes due, in part, to the expansion of the liver possible, a wedge-shaped biopsy of liver with a single cut transplantation programmes. Livers from organ donors, surface and readily accessible blood vessels should be which are not implanted for some reason (i.e. steatosis or obtained. An adequate cannulation of the small vessels to non-identification of an adequate recipient), can be used for ensure a complete perfusion of the piece and an efficient hepatocyte isolation. A major limitation of using human liver oxygenation of the perfusion buffer during cell isolation are for xenobiotic metabolism studies is ensuring a regular key factors to obtain high yields of viable, minimally supply of adequate amounts of the tissue required to sustain membrane-damaged cells. Reduced/excessive flow rates and a research program and, related to this, the quality of the pressure levels should be avoided [12, 38]. Alternatively, tissue. The availability of human liver for research purposes methods without perfusion enable isolation of viable is still too scarce to support the high demand derived from hepatocytes from small non-wedge liver biopsies. Gentle the increasingly use of in vitro assays during drug dissociation of the liver either with collagenase [30] or with development and, in particular, there is a limited number of collagenase followed by dispase [40] provides a yield cell sources of healthy liver to be used for hepatocyte isolation. viability and attachment similar to those obtained by the Different types of hepatic tissue, including whole or split classical perfusion procedure. livers from organ donors or from cadavers, waste liver from Emphasis must be placed on the characteristics of therapeutic hepatectomies or small sized surgical biopsies, collagenase. The specific activity of the enzyme is a critical are used to prepare human hepatocyte cultures [22, 23]. The factor for assuring rapid and complete digestion of the knowledge of the suitability of liver samples from different biopsy without extensive cellular damage. Both under- and origins as sources of viable and metabolically competent over-digestion will compromise the yield and viability of human hepatocytes for drug metabolism research becomes of hepatocytes [35]. Collagenases with higher specific activity great interest, however, up to now, information about the contribute to reduce perfusion time and, in general, render influence of the liver source on functionality of cultured high yields of viable cells [37], although a previous selection hepatocytes is scarce [24-26]. Factors directly derived from of the best collagenase lot is recommended. The viability of differences in the procurement of the liver sample (warm and hepatocytes is usually over 85% and typical yields range 20- cold ischemia) and aspects related to intrinsic characteristics 6 25 x 10 viable cells/g tissue, although this value depends of liver tissue (sex, age, liver pathology, xenobiotic treatment) greatly on the size of the biopsy and the source of the must be considered [25-27]. The healthy/pathological status samples [12, 37]. of the liver is probably a key factor [22].

Experimental Procedures for Hepatocytes Isolation CULTURE OF HUMAN HEPATOCYTES Since the introduction of two-step collagenase perfusion Once isolated, human hepatocytes cultured in monolayer to isolate hepatocytes from human liver [28], the methods for in chemically defined conditions, survive for limited time 294 Current Drug Metabolism, 2003, Vol. 4, No. 4 Gómez-Lechón et al. periods and lose the differentiated phenotype, especially of phenotype. Although the precise mechanisms by which non- drug-metabolising enzymes. However, conditions for short- parenchymal cells modulate the hepatocyte phenotype term monolayer culture, allowing the maintenance of liver remain unelucidated, some new insights on the modes of cell specific functions for approximately one week, are now well signalling, the extent of cell-cell interaction, and the ratio of established [12, 39, 41]. cell populations-are noted. Human hepatocytes co-cultured Considerable advances to prolong cell survival and with an epithelial cell line derived from rat liver survived for preserve liver-specific function in cultured hepatocytes have more than 2 months and secreted high levels of albumin even been gained by sophistication of culture media composition. in a serum-free medium. This long-term survival appeared to Supplementation of chemically defined media with growth correlate with production of an extracellular material, which is rich in collagen Type III [57]. Non-parechymal cells factors and hormones [42-45], specific additives such as DMSO [43], aminoacids [46] or inhibitors of nitric oxide contribute to the metabolism of xenobiotics and may protect synthesis [47] have been used in attempts to preserve the liver from damage by potentially toxic metabolites. The hepatocyte functionality. Other improvements in monolayer beneficial effect for growth and stability of hepatocytes co- cultured with a human hepatic stellate cell line LI90 has also cultures include the use of special substrata to coat the culture plates, such as MatrigelTM [48], collagen [43], fibronectin been reported [58]. These observations suggest that when [12] or a mixture of collagen and fibronectin [12, 37]. adult human hepatocytes are maintained in a culture, which closely resembles their in vivo environment, they are capable Other strategies have been used, including three- of continuing to actively express specific cell functions. dimensional cultures, co-culture of hepatocytes with other cells, hepatocyte spheroids or encapsulated hepatocytes. Spheroid Cultures

Three-Dimensional Cultures Aggregates of liver cells have been successfully established in an attempt to promote in vitro the kind of Extracellular matrix geometry is an important modulator cellular interactions, which are likely to be found in vivo. of cell polarity and function, and affects the phenotype of Amongst the cell types incorporated into the culture both hepatocytes and non-parenchymal cells in the liver. The aggregates are parenchymal and non-parenchymal liver cells, importance of the reconstruction of matrix geometry for including Kupffer cells that allow modulation of metabolic primary hepatocyte culture was firstly recognized by Dunn profiles by parenchymal cells. Several reports indicate that [49]. Sandwiching primary hepatocytes as monolayers within culturing hepatocytes as multicellular aggregates maintain two layers of extracellular matrix is aimed at imitating its prolonged expression of liver-specific genes and achieve bilateral presence with respect to the sinusoidal surfaces of polarity and cell-to-cell contact resulting in up-regulation of the hepatocytes (space of Disse). Extracellular matrix within function [59, 60]. Spheroid culture systems favour the three- the space of Disse next to the central vein is predominantly dimensional cellular organisation and avoid the constraints composed of collagen type I. It has been shown that collagen of cell attachment support [59]. It has been suggested that the type I and other extracellular matrix proteins (ie. Laminin, addition of liver-derived proteoglycans to culture medium fibronectin) modulate hepatocyte growth, gene expression promotes the formation of spheroids. These spheroids of and stability of liver specific functions [50-54]. A general human hepatocytes were found to be viable up to at least one effect is the modulation of gene expression in human hepato- month in culture where they express a high cell functional cytes. The establishment of a three-dimensional co-culture of hepatic activity [48, 61]. liver cells, mimicking more closely the original microen- vironment of the liver, was therefore the next logical step. FUNCTIONALITY AND SURVIVAL OF HUMAN This includes the opportunity to investigate interac-tions, to HEPATOCYTES IN CULTURE examine means of maintaining differentiation of all cell types, and to study the effects of modification of the microenviron- A major drawback of primary culture of hepatocytes is ment [48, 53, 54]. For instance, biliary epithelial cells establish that the expression of many hepatic specific functions three-dimensional structures that subsequently developed into declines during culture probably as a consequence of well differentiated polarized luminal ducts when they are co- hepatocyte adaptation to the new in vitro environment [12, cultured with human hepatocytes in a three-dimensional 13, 62]. It has been shown that complex systems (three- system [55]. Microencapsulation in alginate, a rela-tively dimensional models, co-cultures, etc) are able to support inert biocompatible matrix, also mimics the biological more efficiently than monolayer cultures the long-term extracellular matrix allowing three-dimensional configuration survival and differentiation of primary human liver cells culture to cultivate successfully human hepatocytes [53, 56]. [54]. These culture systems have been shown to maintain the continued transcription of liver specific genes preventing the Co-Cultures dedifferentiation of hepatocytes in primary culture [44, 63, 64]. Heterotypic interactions between cells and non- parenchymal neighbours have been reported to modulate cell Carbohydrate Metabolism growth, migration, and/or differentiation. In both the developing and adult liver, cell-cell interactions are The liver plays a central role in the maintenance of blood imperative for coordinated organ function. In vitro, co- glucose homeostasis and one of the major liver functions is cultivation of hepatocytes and non-parenchymal cells has the ability of hepatocytes to store glucose in the form of been used to preserve and modulate the hepatocyte glycogen. The study of the carbohydrate metabolism of Human Hepatocytes as a Tool for Studying Toxicity Current Drug Metabolism, 2003, Vol. 4, No. 4 295 cultured human hepatocytes reveals interesting features. micrograms albumin/ml per day) [71]. Hepatocytes cultured Human hepatocytes in culture maintain high levels of in three-dimensional systems, either collagen sandwich intracellular glycogen, similar to those reported for fed cultures [48, 52, 54] or spheroids [59], as well as with human liver in vivo (1980 nmol glucose/ mg cell protein) complex culture media [42] synthesize and secrete albumin [37]. Basal glycogenolysis in unstimulated human consistently for more than 3 months. Plasma proteins hepatocytes is low, but insensitive to low levels of external involved in the typical acute-phase response of human liver glucose, and can be totally suppressed by insulin [12, 37, during inflammation are synthesized and secreted by human 65]. The depletion of glycogen stores in human hepatocytes hepatocytes in culture. Their synthesis is regulated by requires hormonal activation. Both glycogenolysis and inflammatory cytokines and its production is additionally glycogen output could be significantly stimulated by stimulated when dexamethasone 10-9 M is present in the physiological concentrations of glucagon, and more than culture medium [72]. 90% of glycogen can be mobilized within 1-2 h by adding -8 Apoproteins are of primary importance for assembly, 10 M glucagon to the culture medium [65]. In cells that secretion, stability and metabolic fate of lipoproteins. Several were previously glycogen-depleted with glucagon, synthesis isoforms of apoproteins, ApoA1 and ApoB100 are also and accumulation of glycogen can be stimulated by produced in significant amounts by hepatocytes in physiological concentrations of insulin [12, 37, 65], but it is monolayer culture [34, 37] and long-term synthesis of both strongly dependent on the presence of lactate in the apoproteins is produced by hepatocytes cultured in a incubation media at postprandial portal concentrations [66, complex culture medium [42]. 67]. Hepatocytes efficiently synthesized glucose from lactate, and other physiological gluconeogenic substrates (fructose, pyruvate, glycerol, aminoacids), at similar rates to Lipid Metabolism and Transport those estimated in fasted human liver (150-240 g glucose/day) [37], and secreted glucose with a negligible Fatty acids and monoacylglycerols are converted to contribution of glycogen [12, 37]. Beside glucose production triglycerids and are transported as lipoprotein complexes. and secretion, the synthesis of glucuronides and ascorbate, The composition of hepatocyte culture medium significantly which is dependent on the extent of the glycogen stores and affects lipid metabolism. Insulin's anabolic effects, such as on the rate of glycogenolysis, has been shown in culture [68]. stimulation of triglyceride storage, have been reported [73]. Basal glycolysis in human cultured hepatocytes is somewhat Hepatocyte growth factor (HGF) stimulates de novo higher than that estimated for liver cells in man (100-200 g synthesis of lipids by affecting activities of lipid metabolic lactate/day) [37]. Gluconeogenesis and glycolysis by cultured gene through its tyrosine kinase-associated receptor, c-met human hepatocytes could be stimulated, as expected, by [74]. concentrations close to the physiological value of glucagon Lipids are transported in the blood in four major classes -9 -8 (10 M) and insulin (10 M), respectively [12, 37]. Many of lipoproteins. One of the principal functions of hepatocytes regulatory mechanisms of carbohydrate metabolism in is the synthesis and secretion of lipoproteins, which are human liver have been investigated in human cultured classified depending on the amount of lipids and proteins hepatocytes [69, 70]. they carry: triacylglycerol-rich ones (chylomichrons and very low density,VLDL), cholesterol-rich ones (low density, Ureogenesis LDL and high density HDL). Hepatic VLDL is the major precursor for LDL, which in humans is the major vehicle Urea cycle occurs exclusively in the liver, therefore urea responsible for transporting cholesterol to peripheral tissues. synthesis from ammonia is a typical hepatic function and a Human hepatocytes in culture secrete triglyceride-rich good indicator of the degree of mitochondrial preservation. lipoproteins in the VLDL density range, apolipoprotein B, Under basal conditions, human hepatocytes in monolayer and apolipoprotein A-I [71]. Binding and degradation of culture synthesize urea at a rate of 2.5-3.5 nmol/mg cell lipoprotein(a) and LDL have also been reported in human protein per min [12, 37]. For human liver a maximal rate of hepatocytes [75]. urea synthesis of 1.2 nmol/mg cell protein per mim was measured in liver, while 24-hour cultured hepatocytes could Bile Acids Synthesis and Canalicular Transport be maximally stimulated to synthesize 4 to 130 nmol urea /mg cell protein per mim [37]. Urea secretion, as well as the Bile acids, which are synthesized in the liver from ammonia removal rate, were largely unaffected by insulin cholesterol, are important to the production of bile flow, but increased with amino acid supplementation [37]. excretion of cholesterol, and intestinal digestion and absorption of fats. Mature hepatocytes in primary culture Plasma Protein Synthesis and Secretion maintain the capability to synthesize bile acids, uptake by basolateral membrane intracellular transport and canalicular Primary human hepatocytes cultured in nutritionally secretion [76]. Moreover the composition of bile acids defined medium synthesize and secrete plasma proteins to secreted in culture closely resembles that found in vivo, and the culture medium, namely, albumin, a 1-antitrypsin, a 1- it is interesting to note that the capacity of human antichymotrypsin, a 1-acid glycoprotein, C-reactive protein, hepatocytes to produce bile acids is higher in culture than in serum amyloid A, fibronectin, fibrinogen, a 1-macroglobulin vivo [76, 77]. A highly differentiated feature of cultured and the major plasma lipoproteins [4, 12, 39]. They display primary hepatocytes is the active uptake of the bile salt albumin secretion at a level comparable to human liver [27, taurocholate [78]. It has been shown in human hepatocyte in 296 Current Drug Metabolism, 2003, Vol. 4, No. 4 Gómez-Lechón et al. culture that the protective effect of ursodeoxycholic acid or The conjugates are much more soluble thus facilitating the one of its conjugates in cholestatic human liver diseases elimination of lipophillic substances. occurs through its effect on hepatocyte transport systems [79]. Are In Vivo Drug-Metabolising Enzymes Maintained by Hepatocytes are polarized, forming bile canaliculi-like Human Hepatocytes? vacuoles where exogeneous organic anions accumulate. Expression of the membrane transporter multi-drug CYP superfamily catalyses the oxidation of many resistance protein (P-glycoprotein, MDR-1) and multidrug endogenous and exogenous compounds. Drugs are resistance-associated protein (MRP-1), known to be predominantly catalysed in the human liver by CYPs 1A2, localized at the canalicular membrane, are maintained in 2A6, 2B6, 2Cs, 2D6, 2E1, and 3A4 [2]. CYP3A4 is the most culture [44, 80]. In addition, model inducers like rifampicin, abundantly expressed isozyme and represents ca. 30-40% of phenobarbital, were able to increase MDR-1 and MRP-1 the total CYP protein in human adult liver [83]. Variability protein levels significantly [44]. Bile canaliculi and in CYP expression is a well-known fact determined by desmosomes are long-term preserved in hepatocytes three- hormonal status, diet, smoking, age, genotype, and drug dimensional cultured in collagen [51] and alginate [81], as exposure [5, 27]. Variability is the norm rather than the well as cultured as spheroids [82]. exception, making the definition of a normal liver complicated. Similarly to human liver, high interindividual variability in CYP activities is found in cultured human DRUG METABOLISING CAPABILITY OF HUMAN hepatocytes prepared from different donors [1, 13, 35, 84]. HEPATOCYTES IN CULTURE Liver microsomes contain numerous CYP enzymes and Biotransformation of xenobiotics enables the elimination they are the most widely used in vitro model for of lipophillic substances that otherwise might accumulate in characterizing CYP isozyme pattern in human liver. tissues, thereby causing toxic effects. This process involves Comparison of CYP expression in hepatic tissue and in chemical modifications of the compounds mostly through hepatocytes obtained from the same livers helps to clarify redox reactions catalysed by CYP activities (phase I whether the metabolic activities of cultured cells are reactions). The result is the formation of new metabolites representative of hepatic metabolism. As seen in Fig. (1) that usually are more polar and can be further conjugated by high similarities can be observed between both hepatic hepatocytes with endogenous molecules (phase II reactions). models when activity and mRNA levels of multiple CYPs

Fig. (1). CYP activities and mRNA levels in human liver and in human hepatocytes. CYP activities were assayed in microsomes prepared from human livers or in intact 24-h cultured monolayers of human hepatocytes by means of specific substrates [87, 143]. Samples of human livers or cultured hepatocytes were processed to extract total RNA and specific CYP mRNA were measured by quantitative RT- PCR [143, 145]. Human Hepatocytes as a Tool for Studying Toxicity Current Drug Metabolism, 2003, Vol. 4, No. 4 297 are comparatively quantified, suggesting that cultured mRNA levels precedes decreases in CYP protein content and hepatocytes provide a good reflection of hepatic CYP activity (Fig. (3)). These results suggest alteration in CYP isozyme pattern. This can be confirmed by the use of mRNA content as the cause of the decline of metabolic testosterone, a multienzymatic substrate stereoselectively capacity of cultured human hepatocytes during culture [15, hydroxylated by individual CYPs that provides a rapid 89]. Although early reduction in CYP expression during method for assessing the activity of several CYPs hepatocyte culture is accepted for all CYPs, there is some simultaneously [85, 86]. Comparison of the profile of discrepancy in the literature regarding the fate of individual testosterone hydroxylated metabolites formed in cultured isozymes. The general consensus remains that the mRNA hepatocytes with those produced by liver microsomes reveals levels of some enzymes (i.e. CYP3A4, CYP2C9 and the high metabolic similarities existing between both hepatic CYP2D6) recovers the initial levels after 72-96 hr of culture, preparations [87]. Human hepatocytes not only reproduce whereas that of others (i.e. CYP1A2 and CYP2E1) does not hepatic CYP isozyme patterns, but also express conjugating [62, 84, 89, 90]. Differential expression of the individual (phase II) enzymes similarly to human liver [12, 35, 87]. In isoforms of the major CYP enzymes over time in culture may contrast to liver microsomes, hepatocytes in culture contain limit its application for drug metabolism studies. However, the entire hepatic drug-metabolising enzyme system in an over a 24-72 hr period of culture, primary hepatocytes consti- integrated form similarly to human liver in vivo and they tute a good model for drug metabolism [11, 12, 35, 91, 92]. provide a useful in vitro model to investigate the metabolic profile of a drug [25, 35, 88, 89]. The mechanism that leads to decreased CYP expression in cultured cells is still largely unknown. It has been sugges- ted that hepatocyte isolation procedure with collagenase Stability of CYP Expression in Human Cultured triggers alteration in liver gene expression, although CYP Hepatocytes content is lost to the same extent in hepatocytes isolated non- enzymically [93]. A lower transcription of CYP genes seems The expression of many hepatic specific proteins, to be the cause of decreased mRNA levels [15]. A recent including CYPs, declines during culture and hepatocytes study shows that the loss of CYP mRNA levels correlates dedifferentiate becoming more similar to hepatic cell lines. with an alteration in the expression of key transcription The study of the evolution during culture of mRNA levels of factors in cultured hepatic cells [15]. These results are in CYPs 1A1, 1A2, 2A6, 2B6, 2C9, 2C19, 2D6, 2E1, 3A4 and agreement with series of previous results on the possible role 3A5, revealed a time-dependent reduction (Fig. (2)). The of inhibitory and activatory liver-enriched transcription CYP mRNA contents decline rapidly after hepatocyte factors in the control of CYP expression in rat hepatocytes isolation and after 4 hours of culture CYP mRNA levels and hepatoma cells [19, 62, 64]. represent 10-30% of that of human liver. This rapid loss in Primary Human Hepatocytes Is the Model of Choice for Drug Metabolism Studies.

The metabolic profile of a drug (i.e. identification of stable and/or reactive metabolites), knowledge of major metabolic routes involved in metabolite formation and the human enzymes so far involved, and potential enzyme-inhibiting or enzyme-inducing properties of the drug are key issues in drug development. To speed up the identification of new drug candidates, pharmaceutical companies are increasingly making use of in vitro drug metabolism models (Table 1). Metabolism of a new chemical entity can be investigated, in a first approach, by incubating the drug with hepatic microsomes, which are easily prepared from liver tissue and can be stored for years with little or no loss of CYP enzyme activities. The major limitations of microsomes are that they have very low phase II activities and incubations can be performed only for a short time (poorly metabolised drugs as well as secondary metabolism are hardly seen). As a consequence of this, it may happen that the results obtained in vitro are noticeably different from those obtained in vivo. Hepatocytes have some advantages that make them the closest model to in vivo [87, 88]. As intact cells are used, the entire plasma membrane is preserved and hepatocytes retain Fig. (2). CYP mRNA levels of human hepatocytes in primary active uptake/excretion mechanisms similarly to hepatocytes cultures of human hepatocytes. mRNA contents from different in the liver in vivo. Integrated metabolic pathways (both CYPs (assayed by RT-PCR) were measured in a small piece of liver phase I and phase II enzymes) and physiological cofactor- and at 2, 4, 16 and 24 h of culture. Results are expressed as enzyme levels are reasonably well maintained for several percentage of the corresponding CYP mRNA content in the liver. hours/days in culture. Nowadays this hepatic model is 298 Current Drug Metabolism, 2003, Vol. 4, No. 4 Gómez-Lechón et al.

Fig. (3). Stability of CYPs in culture. CYP mRNA levels (assayed by RT-PCR) (circles), protein content (measured by immunoblot using specific antibodies) (triangles) and activities (assayed with specific substrates) (squares) were measured in human hepatocytes at different times in culture. Results are expressed as percentage of the corresponding value at 4 h of culture.

Table 1. Advantages and Limitations on In Vitro Models

Issue Advantages Limitations Major applications Future improvements

Liver . Nearly ilimited availability . Lack of cytosolic phase II . Metabolic profilea microsomes enzymes . Easy handling . Kinetic studies . Uncoupled metabolic . Great metabolic capability pathways . CYPs involved in metabolism . High levels of most CYPs . No membrane integrity . Interspecies comparison . Impossibility of induction studies . Inhibition . Only short-term studies Human . Integrated metabolism . Limited availability . Metabolic profile . Long-term cultures with stable hepatocytes phenotype . Membrane integrity . Technically demanding . Interspecies comparison (mechanistic transport studies) . Cryopreservation . Phenotype instability . Drug transport studies . Induction-based drug-drug . Funtional immortalised cells interactions . High interindividual variations . Inhibition . High in vitro/in vivo . Induction correlations . Possibility of long-term studies CYP . Nearly ilimited availabitiy . Lack of phase II enzymes . Kinetic studies . Co-expression of several engineered cells CYPs . High levels of individual . Uncoupled metabolic . CYP enzymes involved CYPs pathways in metabolism . Expression of phase II enzymes . Mechanistic studies on single . No physiological levels of . Drug transport studies CYPs enzymes . Development of cells . Inhibition responsive to induction . Stable phenotype . Impossibility of induction studies . Membrane integrity . No in vitro/in vivo correlations . Long-term studies Incomplete profile in many cases. Human Hepatocytes as a Tool for Studying Toxicity Current Drug Metabolism, 2003, Vol. 4, No. 4 299 considered more appropriate than any other in vitro approach 106]. Such an integrated approach enables the strengths of for the prediction of in vivo drug metabolism [11, 24, 25, 35, each model system to be exploited. Thus, by comparing the 87, 91, 94]. metabolic kinetics in CYP cell lines and human hepatocytes, and the activity of each CYP isozyme both in the cell line Metabolic stability of the compound is one of the criteria and in hepatocytes, it is possible to determine the degree of for selection of new drug candidates. Because of the known participation of a given CYP isoform in the overall species differences in drug metabolism, experiments should metabolism of a drug [106]. be performed in in vitro human models. This parameter can easily be evaluated by incubating the drug with human hepa- tocytes and later quantification of the rates of disappearance CRYOPRESERVED HUMAN HEPATOCYTES AS A from the incubation medium. Hepatocytes are also well SOURCE FOR CELL CULTURES. suited for investigating the metabolic profile of a drug across species, helping to better select the most appropriate animal Differentiated hepatocytes do not proliferate in vitro and species for further pharmacokinetic/bioavailability studies cell cultures need to be prepared each time from liver tissue. [87, 95]. This has greatly hindered the widespread use of human hepatocytes because of the restricted, unpredictable and Different throughput metabolic in vitro assays have been irregular accessibility to suitable liver samples, further developed to accurately extrapolate to humans [7, 96-99]. reduced by legal and ethical considerations in certain Prediction of in vivo metabolic clearance from in vitro data is countries [107]. The limited supply of human liver for cell still difficult and controversial. In vivo extrapolation of in harvesting purposes has led to a need for optimised protocols vitro data is firstly based on the determination of in vitro for the long-term storage of hepatocytes so that available clearance (calculated from Km and Vmax in vitro values) tissue from larger species and rare human material can be and, secondly, on in vivo parameters such as the hepatic used more efficiently. Moreover, human liver samples often mass, hepatic blood flow rate, unbound drug fraction in allow the isolation of a large number of viable cells, which blood [100, 101]. The in vivo clearance of drugs is better are not necessarily required for immediate use for research predicted in studies with hepatocytes than with subcellular purposes. Cryopreservation is the only method presently models such as microsomes [98, 100]. The overall rate of available for long-term storage of freshly isolated hepato- metabolism of a drug can easily be determined in vitro by cytes. Therefore, two major considerations have stimulated measuring the disappearance as well as the formation of the search for successful freezing techniques for human specific metabolites. Metabolites formed in vivo can be hepatocytes. First, the desire to create a batch-controlled cell predicted by the analysis of metabolite profiles obtained after bank, and second, the need to supply functional hepatocytes hepatocyte incubation with the drug [95, 102]. With the use on demand and ready to use for pharmacotoxicology and of appropriate controls, primary cultured human hepatocytes hepatic cell therapy research purposes. can be a good model to quantitatively anticipate the in vivo metabolic profile of a drug. In a recent study, the metabolism of aceclofenac could be examined both in vitro and in vivo Critical Parameters Influencing Cryopreservation [102]. Hepatocytes were isolated and cultured from liver samples obtained in the course of programmed surgery. The Several protocols have been proposed over the last 20 metabolic profile of the drug and its rate of metabolism was years, mostly for rat hepatocytes but also for monkey, porcine investigated and compared to that observed in vivo in the and human hepatocytes [23, 98, 108]. Their effectiveness greatly varies as a function of the characteristics of the same individual. Variations in drug metabolism were observed in the various cell preparations, which were actually reflecting method used. There is a general consensus on the various the inter-individual in vivo variation among donors [102]. aspects of hepatocyte cryopreservation as well as the applications of the cryopreserved hepatocytes for xenobiotic CYP isozymes involved in the metabolism of a drug are metabolism, toxicity evaluation, and transplantation in acute achieved in liver microsomes with specific chemical inhibi- liver failure. Key parameters to successful cryopreservation tors or inhibiting CYP antibodies [86] or, more recently, by include a) the source of the liver, b) the choice of the the use of genetically manipulated cells expressing single cryoprotectant, c) the composition of the cryopreservation CYP genes [86, 96, 97, 103, 104]. CYP engineered cells, medium, d) the cellular density and e) the cooling and however, do not provide quantitative input about the partici- thawing rates. pation of a given isozyme into the overall metabolism of the drug [86, 105]. Relative concentrations of other proteins The recovery as well as the viability of thawed hepato- (NADPH cytochrome P450 reductase and cytochrome b5) or cytes is influenced by the source of the tissue sample used membrane lipid composition may differ in a heterologous for cell harvesting. Liver samples are usually obtained from expressing system compared to human hepatocytes/human livers discarded for transplantation and surgical waste tissue liver and, hence, influence the results [96, 103, 104]. after hepatic tumour resection or from partial hepatectomy. Moreover, in cDNA-expressing systems a single CYP Few results obtained from liver resections [92, 109-111] are interacts with an electron-carrier/supplier protein, while in available compared to the data on hepatocytes obtained from liver hepatocytes many CYPs can interact with them. This discarded livers for transplantation after washing with UW can lead to incorrect predictions of the relative contributions solution [31, 112-114]. Viability of different human liver of individual CYPs to the metabolism of a drug. Taking into samples varies substantially, affecting the recovery account the limitations inherent to each experimental model efficiency and functionality after cryopreservation [23, 111]. (Table 1), it is best to use them in a complementary way [97, High variability among liver samples in the rate of cell 300 Current Drug Metabolism, 2003, Vol. 4, No. 4 Gómez-Lechón et al. attachment to culture dishes and plating efficiency (more Hepatic Functionality and Drug Metabolising stringent measures of cell viability since only undamaged Competence of Hepatocytes in Primary Culture After hepatocytes are likely to attach), has been reported [111, Deep-Freezing Storage 1114, 1115]. Although in most studies, evaluation of hepatic functionality and metabolic competence of cryopreserved Satisfactory cryopreservation of adult hepatocytes has hepatocytes is carried out only in cells from a few hepatocyte been achieved using several protocols and frozen cells have donors, it is generally shown that differentiated hepatic shown, in general, to retain after deep freeze storage the functions (namely: albumin synthesis, glutathione content, hepatic functionality (synthesis of plasma proteins, ATP, the pattern of CYPs and conjugating enzyme activities) show lipids and GSH, carbohydrate and urea metabolism, and great differences among cell preparations, suggesting great synthesis and conjugation and secretion of biliary acids), variability in the resistance to cryopreservation either of maintenance of steroid metabolism and hormone hepatocytes obtained from different liver donors, or of the responsiveness, and the expression of most detoxifying different hepatocyte subpopulations in the liver [92, 111]. enzymes (Phase I and Phase II activities) essential for Interindividual functional variations due not only to the xenobiotic metabolism, at levels close to those expressed in donor (genetic polymorphism, liver diseases, premedication, corresponding unfrozen hepatocyte monolayers [98, 108, etc.), but also to the conditions of hepatocyte preparation (the 109, 111, 112, 117, 124]. The metabolic patterns of several time between liver resection and dissociation, duration of drugs were qualitatively and quantitatively similar before warm ischemia etc.) make interpretation of the data even and after cryopreservation [98, 126]. However, survival, more complicated. functionality and metabolic competence of hepatocytes in culture has been shown to be improved when they are Several cryopreservation agents (DMSO, glycerol, cryopreserved microencapsulated in calcium-alginate gels propylene glycol, acetamide or polyethylene glycol, glycerol, [31, 119], or after being cultured on gelatin or collagen polyvinylpyrrolidone, dextrans), and combinations of these [120]. Secondly, hepatocytes with membrane alterations agents, have been used [23, 1116]. When compared to other arising from enzymatic digestion during and following the cryoprotective agents, DMSO was found to be the most isolation process are unlikely to survive osmotic changes that potent cryoprotectant ranging between 10 and 20% on the usually occur during cryoprotection freezing and thawing, basis of cell viability, cell attachment and metabolic however, the use of a Percoll gradient allows clearing dead activities after thawing [31, 108, 111, 117]. DMSO delays cells from the suspension after thawing [92, 108, 117]. and reduces ice formation during the freezing process and Finally, placing hepatocytes in complex and favorable interacts electrostatically with phospholipid membranes, culture conditions after thawing allows better expression of which seem to be critical for membrane stability during hepatic phenotype [92, 108, 117, 119]. freezing and thawing [31]. Cryopreserved hepatocytes maintain their responsiveness Concerning the composition of the cryopreservation to specific inducers requiring prolonged exposure in a medium, it has been reported that the concentration of serum fashion, which for the investigated inducers, is similar to that in the freezing medium did not affect critically cryopreserva- in cultures from freshly isolated hepatocytes. In fact, CYP tion results, although it is usually included in the composition activities have been shown to be induced in thawed of most cryopreservation media [92, 117]. Protocols using hepatocytes immobilized in calcium-alginate gels [121], and UW solution and fetal bovine serum have been described even in monolayer cultures [109, 110, 122]. [113, 115]. The most reproducible results appear to be obtained with culture media supplemented with l0% or 20% FCS and a cryoprotectant, preferentially DMSO [31, 92, 108, Suitability of Cryopreserved Hepatocytes for Drug 117, 118]. Metabolism and Toxicity Evaluation Cell density has been found to be a critical parameter for Higher throughput screening assays have been developed successful cryopreservation of hepatocytes. It has been using cryopreserved human hepatocytes for hepatotoxicity, reported that tightly packing of hepatocytes during metabolic stability, and inhibitory drug-drug interactions cryopreservation reduces both the viability and the metabolic [116, 125], as well as CYP induction of drugs [109, 110, activity of viable hepatocytes after thawing [118]. 121, 122]. Dose-dependent cytotoxicity was observed for The cooling and thawing rate is a very critical point for known model hepatotoxins, thus suggesting that the cell recovery after cryopreservation. Necrosis does occur in freezing/thawing steps did not change cell sensitivity to toxic primary hepatocytes following isolation and cryopreserva- compounds [107, 117]. In addition, the suitability of tion, but the contribution of apoptosis, another form of cell cryopreserved hepatocytes as indicator cells for the study of death, in primary hepatocytes banked for transplantation has DNA repair to discover possible carcinogenicity in also been shown [124, 125]. Although much of the cellular chemicals has been described [123]. damage (necrosis) in freezing is caused by formation of ice In conclusion, successful freezing protocols offer a crystals within the cells, this is largely prevented by the slow practical means of establishing a cell bank of viable addition of cryoprotectants and microcomputer-controlled hepatocytes for biomedical and toxicological research. Also, slow stepwise rate of freezing with adjustment for the heat it allows the researchers to perform experiments not of crystallization, storage at -150 degrees C, and rapid plausible with freshly isolated hepatocytes, such as the direct thawing and immediate dilution of the cryoprotectant. [108, comparison of xenobiotic toxicity and metabolism in 111, 112, 117]. hepatocytes from multiple human donors in a single Human Hepatocytes as a Tool for Studying Toxicity Current Drug Metabolism, 2003, Vol. 4, No. 4 301 experiment. However, further modifications to the process potency of related drugs. Alternatively, inhibition assays are still needed to optimise the long-term survival and based on individual cDNA-expressed CYPs have also been maintenance of liver-specific functions in the thawed cells in used [103]. The major limitation of these later assays is that culture. inhibition potential could not be identified in the case that the active inhibiting metabolite is generated by a different APPLICATION OF PRIMARY HUMAN HEPATO- CYP. Primary culture of human hepatocytes can extend CYTES IN THE EVALUATION OF POTENTIAL findings obtained with microsomes or CYP engineered cells. DRUG-DRUG INTERACTIONS Although adding more complexity to the assays, a number of studies demonstrated that this in vitro model represents a Multi-drug therapy is a common clinical practice. Simu- valuable tool for the prediction of drug-drug interactions ltaneous administration of two or more drugs may result in based on inhibitory mechanisms [127-129]. metabolic drug-drug interactions having pharmacological and/or toxicological implications. The pharmaceutical CYP Induction in Human Hepatocytes industry is committed to market safer drugs with predictable pharmacokinetic properties and minor drug-drug interactions The phenomenon of CYP induction was first discovered [99]. A pharmacokinetics drug interaction is produced by the and studied in experimental animals, but it was soon alteration of the disposition of one drug by the presence of recognised to occur in humans. Most CYPs can be another compound. Metabolic drug-drug interactions occur influenced to a certain extent, some of them being clearly when a drug affects the metabolism of another drug, as a induced by xenobiotics [130, 131]. Notably, there appears to consequence of an inhibition/induction of a metabolising be more variation in response to enzyme inducers among enzyme. As this is an undesirable feature for a drug humans than in animals, probably due to genetics, lifestyles candidate, information about CYP inhibition or induction by and dietary habits. Some compounds do show CYP inductive the compound should be obtained before a drug candidate is potential across species (i.e. inducers of CYP1A enzymes) considered for the clinical stages of development. There is an whilst for other substances there exist significant differences increasing need to develop high-throughput in vitro assays to in their inducing abilities in animals. Rifampicin, for anticipate drug-drug interactions. The major purpose of these example, is a potent inducer in man and rabbit but it is a poor studies is minimising the potential for drug-drug interactions inducer in rat [132-134]. In contrast, pregnenolone 16a - associated with marked inhibition or induction in CYP carbonitrile, a potent inducer of CYP3A in rat, is not an activities. inducer in either rabbit or man [132, 133]. Inducers are usually classified on the basis of their action Inhibition of Drug Metabolism on individual CYP isozymes (Table 2). From studies in experimental animals, five major groups of CYP inducers As drugs are metabolised by a limited number of have been classically considered: polycyclic aromatic enzymes, they can compete with each other as substrates for hydrocarbons (inducers of CYP1A enzymes); phenobarbital- the same enzyme. Thus, inhibition of a CYP by one drug like compounds (CYP2B inducers); ethanol (CYP2E1 could result in elevations in plasma/tissue concentrations of inducer); glucocorticoids and macrolid antibiotics (CYP3A other drugs. For compounds with a narrow , inducers); and peroxisome proliferators (CYP4A inducers) this can lead to overdosage symptoms and/or toxicity. [130, 131]. Liver microsomes are tools frequently used to investigate The availability of stable cultures of human hepatocytes the potential of a drug candidate as an enzyme inhibitor [7, would allow a direct study of the mechanisms involved in 105]. The inhibitory effect is readily assessed by incubation the induction of drug metabolism in a human model. of several concentrations of the compound with human However, a prerequisite is the characterisation of the microsomes in the presence of CYP selective substrates. This inducibility of CYP isozymes in this in vitro model. The is followed by kinetic measurements to determine whether response of CYPs to a series of classical inducers, which the tested drug acts as an inhibitor on a given CYP isoform, were chosen as representative of the major classes of as well as the type of inhibition caused [7]. These inducers, was examined in human hepatocytes in primary experiments allow an easy comparison of the inhibitory culture (Fig. (4)). The results show that, as observed in

Table 2. Inducers of Human CYP Isozymes

CYP Model inducer Others

1A2 3-methylcholanthrene polycyclics aromatic hidrocarbons, 3-methyl-indole (cruciferous vegetables), omeprazole

2B6 phenobarbital antiepileptic drugs, pesticides

2E1 ethanol isoniazide, organic solvents (acetone, benzene, pyridine)

3A4 dexamethasone, rifampicin glucocorticoids, macrolide antibiotics, phenobarbital, lovastatin, simvastatin, omeprazole?

4A clofibrate peroxisome proliferators (fibrate,…) 302 Current Drug Metabolism, 2003, Vol. 4, No. 4 Gómez-Lechón et al.

Fig. (4). CYP activities in primary cultures of human exposed to model inducers. After 24 h of culture, human hepatocytes were exposed to 2 mM 3 methylcholanthrene, 100 mM ethanol, 1 mM phenobarbital, 1 mM dexamethasone, 50 mM rifampicin, or 1 mM clofibric acid and CYP activities were measured 48 h later. Individual CYP activities were measured by using specific substrates: 7-methoxyresorufin (CYP1A2), coumarin (CYP2A6), 7-benzoxyresorufin (CYP2B6), diclofenac (CYP2C9), dextromethorphan (CYP2D6), chlorzoxazone (CYP2E1), and testosterone (CYP3A4). Results are expressed as fold induction over control (untreated hepatocytes). animals both in vivo or in cultured hepatocytes, human hepa- In addition to measurements of catalytic activities, tocytes respond to model inducers by selectively increasing information on enzyme inducibility can also be obtained by the activity of individual CYP isozymes. Upon incubation measuring total CYP content [13], by immunodetection of hepatocytes for 24-72 h with the inducers, typically a 10-20- individual CYPs using Western blot analysis [127, 133, 135, fold induction can be observed for CYP1A2 activity, and 2- 136, 139] or by quantitative determination of CYP mRNA 10 fold for the rest of inducible CYPs. 3-Methylcholanthrene levels [127, 139-141]. As the regulation of most CYPs is (a polyciclic aromatic hydrocarbon compound) and ethanol essentially pretranslational, mRNA levels allow a good highly increased CYP1A1/2 and CYP2E1 activities, estimation of CYP activity [142, 143]. As seen in Fig. (5), respectively, and no effects on other CYP activities were the inductive pattern of 3-methylcholanthrene, phenobarbital observed (Fig. (4)) [1, 13, 90, 135, 136]. In contrast, after and rifampicin on CYP isozymes determined by quantitative treatment of human hepatocytes with phenobarbital, a RT-PCR measurement is similar to that shown by activity classical CYP2B1 inducer in rats, a moderate increase in determinations (Fig. (4)). Enzyme induction is generally CYP2B6 activity was observed but this effect is not selective preceded by an increase of CYP transcription and the use of and CYP2A6, CYP3A4 and CYP1A1/2 activities are also PCR-RT quantitative techniques for CYP mRNA measure- induced (Fig. (4)) [1, 13, 135, 136]. Dexamethasone, a ments presents certain advantages with respect to catalytic synthetic glucocorticoid, produced increases in CYP3A4, assays. The overlapping substrate specificity of CYPs makes CYP2A6 and CYP2B6 activities (Fig. (5)) [1, 135, 136]. selection of a compound metabolised exclusively by a single Clofibric acid, a typical CYP4A inducer, also increases CYP extremely difficult, and in many cases specificity has CYP2B6 activity in human hepatocytes and, to a lesser only a relative meaning. By the use of mRNA quantification extent, CYP2E1 and CYP3A4 (Fig. (4)) [1]. However, the techniques it is possible to study the response to induction of effects on CYP3A4 activity in human cells are very low certain CYPs (i.e. CYP3A5) for which specific substrates are when compared with those seen in rat hepatocytes [133], not available. Moreover, except for activity measurements showing the notable differences in CYP inducibility existing based on fluorimetric techniques that can be determined in between man and experimental animals. Rifampicin action hepatocytes cultured on 96-well plates [12, 144], catalytic on CYP3A and CYP2A isoforms constitute another example analysis require a large amount of cells. This is a serious of interspecies differences: it induced CYP3A4 in human drawback for human studies, where sample availability is hepatocytes but not CYP3A1 in rat hepatocytes [132, 133, limited. PCR-RT quantitative techniques have been proposed 137]; on the other hand, it strongly induces CYP2A5 activity as reliable and sensitive tools when only a reduced amount in mouse hepatocytes but seems to have no effects on of cells or tissue is available [140, 142, 145]. CYP2A6 in human cells [138]. These differences advise The high variability in the response to inducers found caution before extrapolating to man the results found in between different hepatocyte preparations is noteworthy [1, animal models. 84, 133]. As indicative of this, the inductive effect of 3- Human Hepatocytes as a Tool for Studying Toxicity Current Drug Metabolism, 2003, Vol. 4, No. 4 303 methylcholantrene and dexamethasone on cultured subsequently administered therapeutic agents, by changing hepatocytes obtained from different human livers ranged the expression of drug metabolising enzymes. Exposure to from 6- to 21-fold over control CYP1A2 activity and from certain chemicals (pollutants, cigarette smoke, alcohol, and none to 3-fold over control CYP3A4 activity, [1]. There is dietary constituents) can also induce drug metabolism probably a threshold above which the CYP cannot be enzymes. Induction may result in a rapid fate of the drug and induced and, then, the inducibility depends on the basal level lower plasma concentration to a level which is no longer of expression of the enzyme in each particular liver donor efficacious. Enzyme induction also accounts for the onset of [84]. An inverse correlation between CYP levels in control tolerance to some therapeutic agents and risk of appearance cells and the fold-induction observed after treatment with an of metabolic drug-drug interactions. On the other hand, a inducer has been suggested [84]. possible consequence of enzyme induction is the increased formation of pharmacologically or toxicologically active metabolites. Thus, enzyme induction significantly contri- butes to interindividual differences in drug metabolism and toxicity [131]. CYP induction was classically investigated in vivo in experimental animals but because of the known interspecies differences the convenience of examining CYP induction in man was soon recognised [133]. Human studies are limited, for obvious ethical reasons, to compounds which are at a late stage of clinical development, and not to drug candidates at the pre-clinical stage. In this context, the use of human in vitro models will contribute to the detection of inductive effects of new drugs during early development stages. Cultured hepatocytes respond to CYP inducers [1, 13, 92, 134, 146], and hence represent a suitable in vitro model for scrutinising chemicals of unknown CYP induction potential. Enzyme induction can be easily monitored in vitro as an increase in catalytic activity of cells in response to a stimulus [1, 13, 84, 92, 147]. Hepatocytes are incubated for 24-72 h with non-cytotoxic concentrations of the compound and CYP activities are assessed using specific substrates. The activity measured in induced cells is then compared to untreated cells. As enzyme induction is preceded by an increase of CYP transcription, the use of PCR-RT assays can be effi- ciently used for screening new chemical entities as potential CYP induces. By selecting appropriate primers, the 10 most relevant human CYPs can be quantitatively measured in human hepatocytes incubated with the compounds. Differences vs. control in CYP mRNA are best observed after 48 h of continuous incubation with the inducer [145]. Human hepatocytes have been successfully applied for the screening of the inductive potential of drugs of wide clinical use, providing a rationale for the clinically observed drug-drug interactions [95, 137, 146, 147]. The use of human hepatocytes allows easy analysis of CYP induction produced by structurally related chemicals. An example of this application is the study of CYP1A2 and CYP3A4 induction by a series of proton pump inhibitors widely used in the treatment of peptic ulcers [147]. This study showed that Fig. (5). CYP mRNA induction in human hepatocytes. After 24 modification of chemical structures results in significant h of culture, human hepatocytes were exposed to 2 mM 3 reductions of CYP induction potential without similar methylcholanthrene, 1 mM phenobarbital, and 50 mM rifampicin reduction in therapeutic efficacy. So human hepatocytes in and CYP mRNA contents were evaluated 48 h later by RT-PCR. culture constitute a valuable tool to guide the selection of Results are expressed as fold induction over control (untreated chemical structures with the least drug-drug interaction hepatocytes). potential via the mechanism of CYP induction. Nevertheless the use of human hepatocytes for induction studies presents Primary Human Hepatocytes Is the Model of Choice for certain limitations derived from their low availability and Induction Studies. from the high interindividual differences in CYP levels, which makes the interpretation of results very difficult, Upon repeated administration certain drugs can alter their particularly when a limited number of individual cell own metabolism, or that of other simultaneously or preparations are used. These drawbacks suggest the 304 Current Drug Metabolism, 2003, Vol. 4, No. 4 Gómez-Lechón et al. development of high-throughput experimental models (i.e. xenobiotics and/or pathophysiological factors occurs at inducer-receptor interaction assays in genetically manipulated transcriptional, translational, and posttranslational levels systems) for preliminary induction screening [148]. These [131]. At low ethanol concentrations protein levels are assays are promising tools for CYP induction screening, but increased without changes in mRNA. This involves mainly are still in a developmental state. Inducers that require protein stabilisation after binding of the inducer to the active previous metabolism or that act via alternative, non-receptor site of the enzyme [152]. One proposed mechanism for mediated, mechanisms cannot be detected. The generalized CYP2E1 stabilisation is protection of the protein from cAMP use of these systems requires a previous validation by degradation by the enzyme-bound substrate. At higher comparison with in vivo data or human hepatocyte results. concentrations, ethanol produces additional induction by increases in transcription [150]. Mechanisms Involved in CYP Induction Induction processes are more complex than initially assumed and their study requires living cells. Although most Expression of constitutive and inducible CYP genes is mechanistic studies have been performed in other controlled at multiple levels, which determine tissue-specific experimental models, hepatocytes in primary culture have expression, regulation by physiopathological factors and contributed to the identification of mechanisms of induction inducibility by xenobiotics. Most, but not all, CYPs involved at the molecular level [84, 139, 141]. Through the use of in drug metabolism are inducible by xenobiotics [2]. cultured hepatocytes from different species, variations in Although the phenomenon of CYP induction has long been CYP3A response after exposure to inducers have been known, only in recent years has the use molecular biological explained by interspecies differences in the activation of techniques enabled the identification of mechanisms nuclear receptors [153]. Again, human hepatocytes are more responsible of CYP induction. Most inducible CYP genes likely to reflect regulation of CYP in human liver than other can be transcriptionally activated by a receptor-dependent more simple in vitro models. The relevance of information mechanism resulting in up-regulated CYP expression [149], on CYP induction process obtained from other models have although other mechanisms of CYP induction have been to be confirmed in human in vivo studies or, alternatively, in described [131, 150] (Table 3). Ah receptor, a helix-loop- human hepatocytes. helix family transcription factor, activates transcription of CYP1A genes. Upon binding the inducer to cytosolic Ah receptor, the complex is translocated to the nucleus, where THE USE OF CULTURED HUMAN HEPATOCYTES FOR HEPATOTOXICITY EVALUATION. heterodimerizes with the nuclear factor Arnt, and binds to an enhancer/promotor DNA region of CYP1A [151]. Drug-Induced Hepatic Injury Significant advances in our understanding of CYP regulation have been made with the identification of orphan nuclear Substances capable of producing liver damage and, more receptors mediating the induction of different CYPs. specifically, hepatocyte damage are known as hepatotoxins. Constitutive androstane receptor (CAR), preganane X Intrinsic hepatotoxins are substances that exert their effects receptor (PXR) and peroxisome proliferators-activated in all individuals, in a dose-dependent and hence predictable receptor (PPAR) are recognised players in the inductive manner. These toxins can interfere directly with cell mechanisms of CYP2B, CYP3A and CYP4A enzymes metabolism (active hepatotoxins) [154]. or become toxic [149]. These three nuclear receptors share a common once they have been biotransformed (latent hepatotoxins). heterodimerization partner, the retinoid X-receptor (RXR), Idiosyncratic hepatotoxicity, on the other hand, may be the and are subject to cross-talk interaction with other receptors. consequence of an abnormal metabolism of the drug by The regulatory mechanisms involved in the induction susceptible individuals (metabolic idiosincrasy) or be CYP2A6 and CYP2C genes remain unknown. It has been elicited by an immune-mediated hepatocyte injury (allergic suggested that nuclear receptors, some of them currently hepatitis). The former has a geno- or phenotypic basis that unknown, could mediate the regulation of the expression of results in the over/under expression of drug metabolizing these inducible CYP genes [148]. enzymes, a different drug metabolism pattern and eventually Ethanol-type induction of CYP2E1 is a non-receptor- the abnormal production of a toxic metabolite. This type of mediated mechanism. Regulation of CYP2E1 expression by idiosyncratic toxicity is dose-dependent in susceptible

Table 3. Mechanisms of Induction of CYP

Inducers group Regulatory mechanism Responsive receptor Heterodimer patner CYP

Polyciclic aromatic hydrocarbons transcriptional activation AhR Arnt 1A, 1B

Phenobarbital transcriptional activation CAR RXR 2B (3A, 2Cs ?)

Glucocorticoids, rifampicin transcriptional activation PXR RXR 3A

Peroxisome proliferator transcriptional activation PPAR RXR 4A

Ethanol enzyme stabilization - - 2E1 Human Hepatocytes as a Tool for Studying Toxicity Current Drug Metabolism, 2003, Vol. 4, No. 4 305 individuals. Idiosyncratic drug toxicity is a rare human- cell injury [159]. 3) Lipid peroxidation. It is a free radical specific event and therefore not detectable in experimental process leading to the oxidative degradation of lipids that animals and impossible to be studied in clinical trials [155]. finally may disrupt the structure and functionality of the cell Immune-mediated hepatitis appears all of a sudden in membranes [160]. 4) Alteration of intracellular Ca2+ individuals who have had previous asymptomatic contacts concentration. Intracellular calcium participates in many with the drug, regardless of the administered dose. Some cellular functions and its levels should be perfectly regulated xenobiotics are electrophilic in nature, and others are to obtain a proper cell function. Many substances can bioactivated by the liver to highly reactive metabolites interfere in intracellular calcium homeostasis control, thus generally more toxic than the parent compound, which is the leading to cell malfunction and death [154, 161]. 5) key to many toxic phenomena [154, 156, 157]. To minimize Oxidative stress is produced by compounds able to undergo these effects, hepatocytes have effective defence repeated oxidation and reduction cycles within the cell [163]. mechanisms, and ultimately it is the balance between This redox cycling causes the continuous production of bioactivation, detoxification and defence/repair mechanisms reactive oxygen species (e.g. superoxide anion) and that determines whether a compound will or will not elicit a depletion of GSH and nicotinamide nucleotide pools, with a toxic effect. concomitant increase in lipid peroxidation and intracellular Ca2+ accumulation. The ability of cells to reduce all radicals Molecular Mechanisms Involved in the Hepatotoxicity of with GSH to avoid modification of basic macromolecules Xenobiotics (proteins, lipids, ...) protects the cell against the xenobiotic. When GSH is depleted, cellular malfunction and death can Fig. (6) summarizes the molecular events that can be be promoted. 6) Finally, biotransformation of xenobiotics involved in hepatocyte toxicity: 1) The impairment of the can also result in the formation of intermediates capable of biochemical functions of hepatocytes by the drug or by any covalently binding to cell macromolecules (proteins, DNA of its stable metabolites is the first possible mechanism of and RNA) to form stable drug adducts [163]. hepatotoxicity [156-158]. 2) The mitochondrion is a frequent target of hepatotoxic drugs and the alteration of its function Screening of Potential Hepatotoxicity of Drugs has immediate effects on the energetic balance of cells [158]. Depletion of ATP is, in fact, an early event in the course of Despite considerable progress in the understanding of the drug-induced toxicity that precedes the irreversible stages of mechanism of liver toxicity we are not yet able to design

Fig. (6). Mechanisms of drug hepatotoxicity. Different mechanisms can be involved in hepatocyte toxicity: Mitochondrion is a frequent target of hepatotoxic drugs and the alteration of its function has immediate effects on the energetic balance of cells (depletion of ATP). Lipid peroxidation, oxidative stress, alteration of Ca2+ homeostasis and covalent binding to cell macromolecules are the molecular mechanisms more frequently involved in the toxicity of xenobiotics. 306 Current Drug Metabolism, 2003, Vol. 4, No. 4 Gómez-Lechón et al. non-hepatotoxic molecules rationally. In most cases interpretation of results are essential. Even simple parame- hepatotoxicity is detected at later stages of drug development ters for assessing cell toxicity have yielded promising results in animal toxicity studies or clinical trials [164]. when comparing in vitro effects with human toxicity as false negatives are infrequent in compounds that are toxic without In vitro liver preparations are increasingly used for the study of hepatotoxicity of chemicals. In recent years their biotransformation. However, the in vitro models may lack actual advantages and limitations have been better defined. sensitivity for xenobiotics that require a high degree of The primary hepatocyte cultures, appear to be the most biotransformation or prolonged exposures to exert their toxic effect. powerful in vitro systems, as liver-specific functions and responsiveness to inducers are retained either for a few days A second point is the metabolic relevance of the observed or several weeks depending on culture conditions. In vitro alteration and its reversion upon withdrawal of the xenobio- liver cell models have various applications in : tic from incubation media [94]. Certain cell functions can be screening of cytotoxic and genotoxic compounds, evaluation transiently altered by a xenobiotic, yet this might lack in vivo of chemoprotective agents, and determination of character- significance if the cell rapidly recovers upon elimination of istic liver lesions and associated biochemical mechanisms the xenobiotic. induced by toxic compounds [94, 165, 166]. Extrapolation of A final point is the influence of in vivo pharmacokinetics the results to the in vivo situation remains a matter of debate. on the toxicity of the compound [155] In in vitro experiments, Currently, the most convincing applications of liver cell xenobiotics are kept in culture plates at a constant concen- models are the studies on different aspects of metabolism tration over the incubation time. This contrasts with what and mechanisms of toxicity. occurs in vivo where the concentration of the drug reaches a A first point is to ascertain the hepatic-specific or basal maximum and decreases thereafter in a characteristic concen- toxic effects and whether toxicity is associated with the tration/time curve. Whenever possible, this circumstance metabolism of a given compound. This should be assessed should be taken into account in the experimental design of in by combining the use of metabolically competent hepato- vitro experiments. This is, however, technically complex in cytes with non-differentiated cell lines. For example, by vitro. A simplified procedure to bring the in vivo situation comparing the concentration-toxicity curves of the compound closer to in vitro experimental conditions is to incubate cells 1) in fully competent primary cultured hepatocytes; 2) in for a time and concentration equivalent to the in vivo AUC non-hepatic cells; and 3) in hepatoma cell lines that lack (area under the concentration/time curve), experimentally cytochrome P-450, it is possible to ascertain whether the determined or estimated by PBPK models [167]. There is no compound elicits toxic effects preferentially on hepatocytes doubt that ADME/Tox drug properties, absorption, distribu- or whether bioactivation (metabolisation) of the xenobiotic is tion, metabolism, elimination and toxicity, are properties required in order to produce cellular damage [154, 156, 157]. crucial to the final clinical success of a drug candidate. It has been estimated that nearly 50% of drugs fail because of Cytotoxicity end-points (cell viability assays: MTT, unacceptable efficacy, which includes poor bioavailability as neutral red uptake tests, etc; cell membrane permeability a result of ineffective intestinal absorption and undesirable alteration: enzyme leakage; etc.) represent a first approach to metabolic stability and up to 40% of drug candidates have assess hepatotoxicity, but evaluation of these parameters failed in the past because of safety issues. The methodo- alone may leave out of consideration xenobiotics that impair logies that are available for use in drug development as in cell function without causing cell death. This may not be vitro human-based screens for ADME/Tox drug properties critical for the hepatocyte itself, but can be of toxicological are very promising tools [155, 165, 166]. significance for the whole organism [94]. By examining the effects on hepatocyte-specific metabolism, it is possible to A simple way to rank the relative potential toxicity of a find out whether relevant hepatic specific functions become drug within a homologous series of compounds is to altered by the presence of a xenobiotic. Currently, several compare the plasma concentration of the drug in vivo with metabolic parameters, representative of the liver’s most the concentration causing toxic effects in vitro. The toxicity characteristic functions, should be evaluated, namely, risk (TR) is thus defined as the quotient of both magnitudes. gluconeogenesis, glycogen metabolism, ureogenesis, plasma The larger the values of TR (closer to 1 or even greater), the protein synthesis, synthesis of VLDL, etc. In general, greater the toxicity risk will be for a given drug. It can be metabolic parameters are more sensitive to the toxic effect of reasonably assumed that if a drug reaches a plasma hepatotoxins than cytotoxicity indicators. To be fully concentration at a concentration that is toxic in vitro and confident of the possible implications that a compound might stays there for a period of time that is also conducive to have for hepatocytes in vivo, it may be necessary to examine toxicity in vitro, the compound will show toxic effects in several relevant metabolic functions in order to estimate the vivo [154]. potential effects of the drug before going into a more precise investigation of the mechanisms involved in the observed Improvement in Knowledge on Toxicity and Toxico- toxic effects. Concentrations to which cells are exposed for genomics cell metabolism studies should not cause perceptible cell death (they should be up to the MNTC). Understanding complex toxicological processes requires better knowledge of the early cellular events, and the The ultimate goal of in vitro experiments is to generate relationships between cell lesion, gene induction (and/or the type of scientific information needed to identify repression), protein translation and toxic events. Toxic compounds that are potentially toxic to man for which responses in vitro models are limited in the sense they can purpose not only the design of experiments but also the only show part of these complex processes, but can easily Human Hepatocytes as a Tool for Studying Toxicity Current Drug Metabolism, 2003, Vol. 4, No. 4 307 show immediate and/or early responses affecting essential regulatory factors (LETFs) [16]. Comparative analysis of cellular processes. Phenomena such as genotoxicity, embryo- four major transcription factors (HNF1, HNF3, HNF4 and toxicity or organ toxicity are ultimately the result of different C/EBPa ) in HepG2 cells and in human hepatocytes showed cellular responses, which occur in target cells. Our present that some of them are poorly expressed in hepatoma cells. In knowledge in molecular biology, biochemistry and cell contrast, HNF4 mRNA is found at similar levels in both cell biology tell us that the majority of toxic responses in vivo types [19]. CYP genes appear to be largely regulated at the depend on the individual cellular responses and that the final level of transcription initiation. Thus, it is conceivable that toxic effect depends of how cell responds. Some lesions may the lack of CYP gene expression in hepatomas is the be phenotypically silent and, thus, they may escape the consequence of an altered expression of key transcription detection system in the experimental model. If a certain factors. A candidate to verify the feasibility of this compound possesses a very limited DNA target sequence for hypothesis is the CCAAT/enhancer binding protein alpha its mutagenic action it may cause mutations in non-coding or (C/EBP-a ), which plays key roles in differentiation of a in non-expressed DNA. Hence, these lesions may appear as number of cell types, including hepatocytes. Upon non-toxic because they do not affect the cell phenotype. transfection of HepG2 cells with C/EBP-a , it was found a Cells respond to minimise the harmful effect of the significant increase in three different CYPs from family 2, toxicant (detoxification processes, DNA repair, homeostatic suggesting that lack of C/EBP- a expression must be regulation), minimising or recovering from the initial injury relevant in the transcriptional control of the CYP2 family [19]. However, it seems reasonable to postulate that the re- without major consequences. Thus, although cells may be affected to a certain extent corrective responses are induced expression of all CYP isoforms in hepatoma cells is likely to and the final toxic effect is not relevant. Conversely, homeo- require the concerted action of several liver-enriched static/adaptative changes, which may not be relevant in short transcription factors. A promising experimental approach is times, may give rise to secondary responses that, upon the use of adenoviral expression vectors that allow simultaneous expression of multiple genes as well as the accumulation with time, may have negative consequences in the overall control of an organ function resulting in toxicity. modulation of the level of expression of the transgenes [171]. Adenoviruses encoding two of the most relevant liver- Finally, cells may die by necrosis and/or apoptosis to an enriched transcription factors (HNF3-g and C/EBPa ) have extent sufficient to result in compromised organ functioning been successfully generated [172]. HepG2 cells have been or loss of homeostasis and, then, in toxic response. This is transduced with combinations of adenovirus expressing the “classical” outcome of the toxic phenomena, which is HNF3-g and C/EBP-a and assessed the impact on CYP observed as loss of survival signals followed by cell death. In levels (Fig. (7)). The preliminary results have evidenced addition, these changes correlate with alterations of cell’s synergistic effects on CYP3A4 gene, the major isoform of specific functionality. The balance between necrosis/apopto- human liver. The combined effects of both factors, acting on sis is very likely linked to the type of primary cellular lesion. endogenous CYP3A4 gene results in more than 60-fold The monitoring of gene expression of cells exposed to increase CYP3A4 mRNA expression [172]. It is likely that toxic xenobiotics (toxicogenomics), has gained great HNF3-g, by alleviating chromatin structure, and C/EBP- popularity among researchers [168-170]. The ease with a, by direct transactivation, result in a substantial increase of which thousands of genes can be measured has led most of CYP3A4 gene transcription. the scientists to register changes of as many genes as HNF4 a is another key activator of liver-specific gene possible with the hope that an in-depth biostatistical analysis expression [173]. When an adenoviral vector encoding anti- would reveal which of that gene were clearly linked to toxic sense HNF4a was transfected to human hepatocytes [20] events. analysis of CYP gene expression in transfected human In summary, it is hoped that through a better understan- hepatocytes indicated that CYP3A4, CYP3A5 and CYP2A6 ding of cellular mechanisms of toxicity, combined with showed a clear, dose-dependent down-regulation upon blockage of HNF4 translation, while a moderated inhibition mechanistically directed toxicogenomic analysis, the accuracy and predictivity of in vitro screening for toxicants could be was observed on CYP2B6, CYP2C9 and CYP2D6 greatly improved. Ultimately, the medium term-goals are the expression (35-40% reduction). The levels of CYP2E1 were design of a more mechanistic approach in the in vitro unaffected. cytotoxicity models, which may increase the performance It has been measured the expression level of several and applicability of such assays in the screening of HNF4a splicing variants in HepG2 cells, and compared with compounds both of pharmacological interest as well as non-hepatic cell lines (HeLa) and primary cultured human potentially toxic to man and the environment. hepatocytes. Preliminary results showed a deregulated expression of the HNF4a isoforms pattern in HepG2 cells. ALTERNATIVE HEPATIC MODELS TO HUMAN Despite the adult isoforms being more abundant the foetal- HEPATOCYTES FOR METABOLISM AND HEPATO- isoform HNF4a 7 is also present in hepatoma cells. Notably, TOXICITY STUDIES this foetal splicing variant has demonstrated lower transactivation potential on promoters of hepatic genes There is a need for hepatic cell lines expressing the whole associated with the adult phenotype (e.g. ApoCIII) [174]. spectrum of human xenobiotic-metabolizing enzymes, as an Similarly, AldoB mRNA, a target gene of HNF4a , is alternative to primary cultures. The hepatic-specific expressed in hepatocytes while it is lacking in the hepatoma expression of a given gene is accomplished by the concerted cell line, despite the higher abundance of HNF4 in HepG2. action of a number of liver-enriched and ubiquitous Therefore, deregulated expression of HNF4a isoforms could 308 Current Drug Metabolism, 2003, Vol. 4, No. 4 Gómez-Lechón et al.

Fig. (7). Effects of Ad-C/EBPa and Ad-HNF-3g on CYP3A4 expression in HepG2 cells. HepG2 cells were infected for 90 min with Ad- C/EBPa , Ad-HNF-3g and Ad-pAC (control insert-less adenovirus). 48 h after infection total RNA was isolated and specific CYP3A4 and b- actin mRNAs were measured by semi quantitative RT-PCR. The CYP3A4 mRNA content was quantified and normalized with b-actin mRNA content and expressed as fold induction over control cells. Bars are the mean of four independent experiments ± SD. Western blot: Total protein extracts (20 µg) were obtained from cells and the expression of C/EBPa and HNF-3g was analysed by immunoblotting. A representative western blot is depicted after detection with the specific antibodies. Multiplicity of Infection (MOI) of Adenoviruses: Infection doses of Ad-C/EBPa and Ad-HNF-3g used. Control cells were infected with 15 MOI of Ad-pAC. be a possible mechanism to explain why HNF4 is non- P/CAF and the p160 family of proteins (SRC-1, -2 and -3) operative in HepG2. interact with a variety of transcription factors and cooperate CYP transcription in hepatoma cells could also be to stimulate transcription. Acting as bridging proteins between activators and the general transcriptional machinery inhibited as a consequence of an increased activity of transcription repressors. Several C/EBP protein isoforms as well as by acetylation of the surrounding nucleosomes, corresponding to full length as well to truncated proteins are coactivators facilitate the accessibility of transcription differentially expressed in hepatomas and hepatocytes [185, factors to the promoter [178, 179]. Thus, deregulation of co- activators and histone deacetylases could also be at the origin 186]. The C/EBPb mRNA directs production of two isoforms: of CYP down-regulation in hepatoma cells. a 35-kDa LAP (liver-enriched transcriptional activating protein) and a 20-kDa LIP (liver-enriched transcriptional The results combining adenoviral vectors encoding inhibitory protein). An example of this inhibitory action can transcription factors along with chemical inhibitors of be observed in the case of CYP3A4 expression where histone deacetylases (i.e. tricostatin A) suggest that truncated C/EBPb protein, by competing with other chromatin condensation in hepatoma cells is likely to be an constitutive C/EBP activating factors, particularly C/EBPa , important factor responsible for the low transcription of down-regulates the expression of human CYP3A4 [21]. some CYP genes [172]. CYP transcription could also be inhibited in hepatoma In summary, hepatoma cell lines show very limited cells as a result of DNA methylation / histone deacetylation metabolic capacity due to a very low expression of CYP or a decrease in acetyltransferases co-activators. [177]. genes. The altered expression of key activating liver- Last but not least, the communication between transcrip- enriched transcription factors (e.g. C/EBPa and HNF3g) in tion factors and the basal transcriptional machinery consti- the hepatoma cells is a demonstrated reason explaining tutes one of the key steps in the regulation of eukaryotic gene decreased CYP transcription and functionality. Indeed, expression. Transcriptional co-activators such as p300, CBP, tailored re-expression of these missing activators in Human Hepatocytes as a Tool for Studying Toxicity Current Drug Metabolism, 2003, Vol. 4, No. 4 309

Fig. (8). Possible mechanisms involved in CYP repression in hepatomas. The expression of CYP genes is under the stringent control of several hepatic-specific transcription factors, which act cooperatively in a complex cross-regulatory network that has been postulated to determine the hepatic phenotype. Key liver-enriched transcription factors could be down-regulated, or even lacking, in poorly differentiated hepatomas (1). It is also conceivable that CYP transcription could be inhibited as a consequence of increased activity of transcription repressors (2), inaccessibility of gene promoters due to chromatin compactation (3) or a decrease in acetyltransferases co-activators. hepatoma cells causes significant re-expression of relevant [10] Macé, K.; Offord, E.A. and Pfeifer, A.M.A. (1997) in In vitro CYPs, opening a new experimental strategy to generate Methods in Pharmaceutical Research, (Castell, J.V. and Gómez- Lechón, M.J. Eds.), Academic Press, London, pp. 433-456. competent cell lines for human drug metabolism studies. [11] Maurel, P. (1996) Adv. Drug Deliv. Rev., 22(3), 105-132. Other mechanisms, certainly influencing gene regulation [12] Gómez-Lechón, M.J.; Donato, T.; Ponsoda, X.; Fabra, R.; play a significant role. Thus, the existence in hepatoma cells Trullenque, R. and Castell, J.V. (1997) in Isolation, culture ad use of non-functional CYP regulators such as HNF4a , the of human hepatocytes in drug research, In Vitro Methods in Pharmaceutical Research, (Castell, J.V. and Gómez-Lechón, M.J. existence of specific CYP repressors and chromatin Eds.), Academic Press, London, pp. 129-153. condensation around CYP genes in hepatoma cells, need to [13] Donato, M.T.; Gómez-Lechón, M.J. and Castell, J.V. (1990) be taken into consideration when attempting to develop a Biochem. Pharmacol., 39(8), 1321-1326. drug-metabolism competent cell line (Fig. (8)). [14] Gómez-Lechón, M.J.; Donato, T.; Jover, R.; Rodriguez, C.; Ponsoda, X.; Glaise, D.; Castel1, J.V. and Guguen-Guillouzo, C. (2001) Eur. J. Biochem., 268(5), 1448-1459. REFERENCES [15] Rodríguez-Antona, C.; Donato, M.T.; Boobis, A., Edwards, R.J.; Watts, P.S.; Castell, J.V. and Gómez-Lechón, M.J- (2002) 1] Donato, M.T.; Castell, J.V. and Gómez-Lechón, M.J. (1995) Drug [ Xenobiotica, 32 (6), 505-520. Metab. Dispos., 23 (5), 553-558. [16] Cereghini, S. (1996) FASEB J., 10(2), 267-282. [2] Gonzalez, F.J. (1990) Pharmacol. Ther., 45(1), 1-38. [17] Ourlin, J.E.; Jounaïdi, Y.; Maurel, P. and Vilarém, M. J. (1997) J. [3] Meyer, U.A. and Zanger, U.M. (1997) Ann. Rev. Pharmaco. Hepatol., 26(Suppl. 2), 54-62. , (3), 269-296. Toxicol. 37 [18] Jover, R.; Bort, R.; Gómez-Lechón, M.J. and Castell, J.V. (2000) in [4] Pelkonen, O. and Raunio, H. (1997) Environ. Health Perspt., Progress in the Reduction, Refinement and Replacement of Animal 105(Suppl. 4), 767-774. Experimentation. (Balls, M.; van Zeller, A.M.; Halder, M.E. Eds.) [5] Lu, A.Y.H. (1998) , (12), 1217-1222. Drug Metab. Disp. 26 pp. 127-137. [6] Kennedy, T. (1997) , (2), 436-444. Drug Discovery Today 2 [19] Jover, R.; Bort, R.; Gomez-Lechón, M.J. and Castell, J.V. (1998) [7] Rodrigues, A.D.; Wong, S.L. (1997) Adv. Pharmacol., 43, 65-101. FEBS Lett., 431(2), 227-230. [8] Chen, Q.; Galleano, M. And Cederbaum, A.I. (1998b) Alcohol. [20] Jover, R.; Bort, R.; Gomez-Lechón, M.J. and Castell, J.V. (2001) Clin. Exp. Res., 22(4), 782-784. Hepatology, 33(3), 668-675. [9] Crespi, C.L.; Gonzalez, F.J.; Steimel, D.T.; Turner, T.R.; Gelboin, [21] Jover, R.; Bort, R.; Gomez-Lechón, M.J. and Castell, J.V. (2002) H.V.; Penman, B.W. and Langenbach, R. (1991) Chem. Res. FASEB J., 10.1096/fj.02-0195fje. Toxicol., 4(5), 566-572. 310 Current Drug Metabolism, 2003, Vol. 4, No. 4 Gómez-Lechón et al.

[22] Chapman, D.E.; Christensen, T.A.; Michener, S.R. and Powis, G. [54] Zeilinger, K.; Sauer, I.M.; Pless, G.; Strobel, C.; Rudzitis, J.; (1993) in Human drug metabolism from molecular biology to man, Wang, A.; Nussler, A.K.; Grebe, A.; Mao, L.; Auth, S.H.; Unger, (Jeffrey, E.H. Ed.), CRC Press, London, pp. 53-63. J.; Neuhaus, P.; Gerlach, J.C. (2002) Altern. Lab. Anim., 30(5), [23] Mitry, R.R.; Hughes, R.D. and Dhawan, A. (2002) Semin. Cell 525-538. Dev. Biol., 13(6), 463-467. [55] Auth, M.K.; Joplin, R.E.; Okamoto, M.; Ishida, Y.; McMaster, P.; [24] Olinga, P.; Merema, M.; Hof, I.H.; de Jong, K.P.; Slooff, M.J.H.; Neuberger, J.M.; Blaheta, R.A.; Voit, T. and Strain, A.J. (2001) Meijer, D.K.F.; Groothuis, G.M.M. (1998) Drug Metab. Dispos., Hepatology, 33(3), 519-529. 26(1), 5-11. [56] Selden, C.; Shariat, A.; McCloskey, P.; Ryder, T.; Roberts, E.; [25] Olinga, P.; Merema, M.T.; Hof, I.H.; De Jager, M.H.; De Jong, Hodgson, H. (1999) Ann. N. Y. Acad. Sci., 875, 353-363. K.P.; Slooff, M.J.H.; Meijer, D.K.F. and Groothuis, G.M.M. (1998) [57] Clement, B.; Guguen-Guillouzo, C.; Campion, J.P.; Glaise, D.; Xenobiotica, 28(4), 349-360. Bourel, M. and Guillouzo, A. (1984) Hepatology, 4(3), 373-380. [26] Serralta, A.; Donato, M.T.; Orbis, F.; Castell, J.V.; Mir, J. and [58] Bhatia, S.N.; Balis, U.J.; Yarmush, M.L. and Toner M. (1999) Gómez-Lechón, M.J. (2003) Toxicol. in vitro, (in press). FASEB J., 13(14), 1883-1890. [27] Schmucker, D.L. (2001) Drugs Aging, 18(11), 837-851. [59] Tong, J.Z.; Sarrazin, S.; Cassio, D., Gauthier, F., Alvarez, F. (1994) [28] Reese, J.A.; Byard, J.L. (1981) In Vitro, 17(11), 935-940. Biol. Cell., 81(1),77-81. [29] Vons, C.; Pegorier, J.P.; Ivanov, M.A.; Girard, J.; Melcion, C.; [60] Khalil, M.; Shariat-Panahi, A.; Tootle, R.; Ryder, T.; McCloskey, Cordier, A. and Franco, D. (1990) Toxicol. in vitro, 4(3), 432-436. P.; Roberts, E.; Hodgson, H.; Selden, C. (2001) J. Hepatol., 34 (3), [30] David, P.; Viollon, C.; Alexandre, E; Azimzadeh, A.; Nicod, L.; 68-77. Wolf, P.; Jaeck, D.; Boudjema, K. and Richert, L. (1998) Hum. [61] Ijima, H.; Nakazawa, K.; Mizumoto, H.; Matsushita, T.; Funatsu, Exp. Toxicol., 17(10), 544-553. K. (1998) J. Biomater. Sci. Polym. Ed., 9(7), 765-778. [31] Guillouzo, A.; Guyomard, C.; Fautrel, A. and Chesné, F. (2000) in [62] Padgham, C.R.W.; Boyle, C.C.; Wang, X.J.; Raleigh, S.M.; The Hepatocyte Review (Berry, M.N. and Edwards, A. M., Eds.) Wright, M.C. and Paine, A.J. (1993) Biochem. Biophys. Res. Kluwer Academic Press London, pp. 125-145. Commun., 197 (2), 599-605. [32] Caraceni, P.; Gasbarrini, A.; Nussler, A.; Di Silvio, M.; Bartoli, F.; [63] Nagaki, M.; Shidoji, Y.; Yamada, Y.; Sugiyama, A.; Tanaka, M.; Borle, A.B. and Van Thiel, D.H. (1994) Hepatology, 20(5), 1247- Akaike, T.; Ohnishi, H.; Moriwaki, H.; Muto, Y. (1995) Biochem. 1254. Biophys. Re.s Commun., 210(1),38-43. [33] Spinelli, S.V.; Rodríguez, J.V.; Quintana, A.B.; Mediavilla, M.G. [64] Gómez-Lechón, M.J.; Jover, R.; Donato,T.; Ponsoda, X.; and Guibert, E.E. (2002) Cell Transplant., 11(2), 161-168. Rodriguez, C.; Stenzel, K.G.; Klocke, R.; Paul, D.; Guillen, I.; [34] Guguen-Guillouzo, C.; Bourel, M.; Guillouzo, A. (1986) Prog. Bort, R.; Castell, J.V. (1998) J. Cell. Physiol., 177(4), 553-562. Liver Dis., 8(8), 33-50. [65] López, M. P.; Gómez-Lechón, M.J. and Castell, J. V. (1991) [35] Li, A.P. (2000) in The hepatocyte review (Berry, M.N. and Diabetes, 40 (2), 263-268. Edwards, A. Eds), Kluwer Academic Publishers, Dordrecht, The [66] Lopez, M. P.; Gómez-Lechón, M.J. and Castell, J. V. (1991) Netherlands, pp. 391-410. Biophys. Acta, 1094 (2), 200-206. [36] Donini, A.; Baccarani, U.; Lavaroni, S.; Dialti, V.; Adami, V.; [67] Keppens, S.; Vandekerckhove, A.; Moshage, H.; Yap, S.H.; Aerts, Risaliti, A, Cautero, N.; Degrassi, A. and Bresadola, F. (2001) R.; De Wulf, H. (1993) Hepatology, 17(4), 610-614. Transplant. Proc., 33(1-2), 1972-1973. [68] Braun, L.; Puskas, F.; Csala, M.; Meszaros, G.; Mandl, J. and [37] Gómez-Lechón, M.J.; López, P.; Donato, T.; Montoya, A.; Banhegyi, G. (1997) Free Radic. Biol. Med., 23(5), 804-808. Larrauri, A.; Giménez, P.; Trullenque, R.; Fabra, R. and Castell, [69] An, J.; Li, Y.; van De Werve, G. and Newgard, C.B. (2001) J. Biol. J.V. (1990) In Vitro Cell. Biol. Dev., 26(1), 67-74. Chem., 276(14), 107221-107229. [38] Ryan, C.M.; Carter, E.A.; Jenkins, R.L.; Sterling, L.M.; Yarmush, [70] Iynedjian, P.B.; Marie, S.; Gjinovci, A.; Genin, B.; Deng, S.P.; M.L.; Malt, R.A. and Tompkins, R.G. (1993) Surgery, 113(1), 48- Buhler, L.; Morel, P.; Mentha, G. (1995) J. Clin. Invest., 95(5), 54. 1966-1973. [39] Kim, H.M.; Han, S.B.; Hyun, B.H.; Ahn, C.J.; Cha, Y.N.; Jeong, [71] Schippers, I.J.; Moshage, H.; Roelofsen, H.; Muller, M.; Heymans, K.S.; Oh, G.T. (1995) J. Toxicol. Sci., 20(5), 565-578. H.S.; Ruiters, M.; Kuipers, F. (1997) Cell. Biol. Toxicol., 13(4-5), [40] Miyazaki, K.; Takaki, R.; Nakayama, F.; Yamauchi, S.; Koga, A.; 375-386. Todo, S. (1981) Cell Tissue Res., 218(1), 13-21. [72] Castell, J.V.; Gómez-Lechón, M.J.; David, M.; Fabra R.; [41] Skett, P.; Bayliss, M. (1996) Xenobiotica, 26(1), 1-7. Trullenque, R. and Heinrich P. C. (1990) Hepatology, 12(5), 1179- [42] Ferrini, J.B.; Pichard, L.; Domerge, J. and Maurel, P. (1997) Chem. 1186. Biol. Interact., 107(1-2), 31-45. [73] Chan, C.; Berthiaume, F.; Washizu, J.; Toner, M. and Yarmush, [43] Hino, H.; Tateno, C.; Sato, H.; Yamasaki, C.; Katayama, S.; M.L. (2002) Biotechnol. Bioeng., 78(7), 753-760. Kohashi, T.; Aratani, A.; Asahara, T.; Dohi, K.; Yoshizato, K. [74] Kaibori, M.; Kwon, A.H.; Oda, M.; Kamiyama, Y.; Kitamura, N.; (1999) Biochem. Biophys. Res. Commun., 256(1), 184-191. Okumura, T. (1998) Hepatology, 27(5), 1354-1361. [44] Runge, D.; Runge, D.M.; Jager, D.; Lubecki, K.A.; Beer Stolz, D.; [75] Snyder, M.L.; Hay, R.V.; Whitington, P.F.; Scanu, A.M.; Fless, Karathanasis, S.; Kietzmann, T.; Strom, S.C.; Jungermann, K., G.M. (1994) Arterioscler. Thromb., 14(5),770-779. Fleig, W.E.; Michalopoulos, G.K. (2000) Biochem. Biophys. Res. [76] Ellis, E.; Goodwin, B.; Abrahamsson, A.; Liddle, C.; Mode, A.; Commun., 269(1), 46-53. Rudling, M.; Bjorkhem, I. and Einarsson, C. (1998) Hepatology, [45] Katsura, N.; Ikai, I.; Mitaka, T.; Shiotani, T.; Yamanokuchi, S.; 27(2), 615-620. Sugimoto, S.; Kanazawa, A.; Terajima, H.; Mochizuki, Y. and [77] Axelson, M.; Ellis, E.; Mork, B.; Garmark, K.; Abrahamsson, A.; Yamaoka, Y. (2002) J. Surg. Res., 106(1), 115-123. Bjorkhem, I.; Ericzon, B.G. and Einarsson, C. (2000) Hepatology, [46] Washizu, J.; Chan, C.; Berthiaune, F.; Tompkins, R.G.; Toner, M. 31(6), 1305-1312. and Yarmush, M. L. (2000) Tissue Eng., 6(5), 497-504. [78] Olinga, P.; Merema, M.; Sandker, G.W.; Slooff, M.J.; Meijer, [47] Donato, M.T.; Ponsoda, X.; O´Connor, E.; Castell, J.V. and D.K.; Groothuis, G.M. (1998) J. Pediatr. Gastroenterol. Nutr., Gómez-Lechón, M.J. (2001) Xenobiotica, 31(5), 249-264. 27(3), 366-368. [48] Chen, H.L.; Wu, H.L.; Fon, C.C.; Chen, P.J.; Lai, M.Y. and Chen, [79] Hillaire, S.; Ballet, F.; Franco, D.; Setchell, K.D.; Poupon, R. D.S. (1998) J. Biomed. Sci., 5(6), 435-440. (1995) Hepatology, 22(1), 82-87. [49] Dunn, J.C.; Yarmush, M.L.; Koebe, H.G. and Tompkins, R.G. [80] Fardel, O.; Morel, F. and Guillouzo, A. (1993) Carcinogenesis, (1989) FASEB J., 3(2), 174-177. 14(4), 781-783. [50] Reid, L.M.; Fiorino, A.S.; Sigal, S.H.; Brill, S.; Holst, P.A. (1992) [81] Capuani. G.; Miccheli, A.; Tomassini, A.; Falasca, L.; Aureli, T. Hepatology, 15(6), 1198-1203. and Conti, F. (2000) Artif. Cells Blood Substit. Immobil. [51] Koebe, H. G.; Pahernik, S.; Eyer, P. and Schildberg, F.W. (1994) Biotechnol., 28(4), 293-305. Xenobiotica, 24(2), 95-107. [82] Lian, W.N.; Tsai, J.W.; Yu, P.M.; Wu, T.W.; Yang, S.C.; Chau, [52] Kono, Y.; Yang, S. and Roberts, D. (1997) In vitro Cell. Dev. Biol., Y.P.; Lin, C.H. (1999) Hepatology, 30(3), 748-760. 33(5), 467-472. [83] Shimada, T.; Yamazaki, H.; Mimura, M.; Inui, Y. and Guengerich, [53] Bader, A.; Knop, E. and Haverich, A. (2000) in The Hepatocyte F.P. (1994) J. Pharmacol. Exp. Ther., 270 (1), 414-423. Review, (Berry, M.N. and Edwards, A.M. Eds.), Kluwer Academic [84] LeCluyse, E.L. (2001) Eur. J. Pharmac. Sci., 13 (4), 343-368. Publuishers, London, pp. 97-116. Human Hepatocytes as a Tool for Studying Toxicity Current Drug Metabolism, 2003, Vol. 4, No. 4 311

[85] Waxman, D.J.; Lapenson, D.P.; Aoyama, T.; Gelboin, H.V.; [117] Li, A.P.; Lu, C.; Brent, J.A.; Pham, C.; Fackett, A.; Ruegg, C.E.; Gonzalez, F.J. and Korzekwa, K. (1991) Arch. Biochem. Biophys., Silber, P.M. (1999) Chem. Biol. Interact., 121(1), 17-35. 290(1), 160-166. [118] De Loecker, W.; Koptelov, V.A.; Grischenko, V.I. and De Loecker, [86] Yamazaki, H.; Inoue, K.; Shaw, P.M.; Checovich, W.J.; P. (1998) Cryobiology, 37(2), 103-109. Guengerich, F.P.; Shimada, T. (1997) J. Pharmacol. Exp. Ther., [119] Guillouzo, A.; Rialland, L.; Fautrel, A.; Guyomard, C. (1999) 283(2), 434-442. Chem. Biol. Interact., 121(1), 7-16. [87] Donato, M.T.; Castell, J.V. and Gómez-Lechón, M.J. (1999) J. [120] Strom, S. C.; Dorko, K.; Thompson, M. T. ; Pisarov, L. A. and Hepatol., 31(1), 542-549. Nussler, A. K. (1998) in Large scale isolation and culture of [88] Guillouzo, A. and Chesné, C. (1996) in Cell culture models of human hepatocytes. (Fraco, D.; Boudejema, K. and Vater, B. Eds.) epithelial tissues: a practical approach (Shaw, A.J. Eds.), Oxford Les Editions INSERM, Paris, pp. 196-205. University Press, Oxford, UK. pp. 67-85. [121] Rialland, L.; Guyomard, C.; Scotte, M.; Chesne, C.; Guillouzo, A. [89] Woodcroft, K.J.; Novak, R.F. (1998) Drug Metab. Dispos., 26 (4), (2000) Cell. Biol. Toxicol., 16(2), 105-116. 372-378. [122] Hengstler, J.G.; Utesch, D.; Steinberg, P.; Platt, K.L.; Diener, B.; [90] Morel, F.; Beaune, Ph.; Ratanasavanh, D.; Flinois, J.P.; Yang, C.S.; Ringel, M.; Swales, N.; Fischer, T.; Biefang, K.; Gerl, M.; Bottger, Guengerich, F.P- and Guillouzo, A. (1990) Eur. J. Biochem., T.; Oesch, F. (2000) Drug Metab. Rev., 32(1), 81-118. 191(2), 437-444. [123] Schehrer, L.; Regan, J.D.; Westendorf, J. (2000) Toxicology, [91] Guillouzo, A.; Morel, F.; Fardel, O. and Meunier, B. (1993) 147(3), 177-191. Toxicology, 82(1-5), 209-219. [124] Skett, P.; Roberts, P.; Khan, S. (1999) Chem. Biol. Interact., [92] Li, A.P.; Maurel, P.; Gomez-Lechon, M.J.; Cheng, L.C.; Jurima- 121(1), 65-76. Romet, M. (1997) Chem. Biol. Interact., 107(1-2), 5-16. [125] Fu, T.; Guo, D.; Huang, X.; O'Gorman, MR.; Huang, L.; Crawford, [93] Paine, A.J. (2000) in The hepatocyte review (Berry, M.N. and SE.; Soriano, HE. (2001) Cell. Transplant., 10(1), 59-66. Edwards, A. Eds.) Kluwer Academic Publishers , Dordrecht, The [126] Reinach, B.; de Sousa, G.; Dostert, P.; Ings, R.; Gugenheim, J.; Netherlands, pp. 411-420. Rahmani, R. (1999) Chem. Biol. Interact., 121(1), 37-48. [94] Gómez-Lechón, M.J.; Donato, M.T. and Castell, J.V. (1997) In [127] Pichard, L.; Fabre, I.; Fabre, G.; Domergue, J., Saint Aubert, B.; Vitro Toxicol., 10(1/2), 63-70. Mourad, G. and Maurel, P. (1990) Drug Metab. Dispos., 18 (5), [95] Kistler, A.; Liechti, H.; Pichard, L.; Wolz, E.; Oesterhelt, G.; 595-606. Hayes, A.; Maurel, P. (2002) Arch. Toxicol., 75(11-12), 665-675. [128] Li, A.P. and Jurima-Romet, M. (1997) Chem. Biol. Toxicol., 13 (4- [96] Rodrigues, A.D. (1999) Biochem. Pharmacol., 57(5), 465-480. 5), 365-374. [97] Ohyama, K.; Nakajima, M.; Nakamura, S.; Shimada, N.; [129] Cohen, L.H.; van Leeuwen, R.E.; van Thiel, G.C.; van Pelt, J.F. Yamazaki, H.; Yokoi, T. (2000) Drug Metab. Dispos., 28(11), and Yap, S.H. (2000) Biopharm. Drug Dispos., 21(9), 353-364. 1303-1310. [130] Pelkonen, O.; Maenpaa, J.; Taavitsainen, P.; Rautio, A.; Raunio, H. [98] Hewitt, N.J.; Buhring, K.U.; Dasenbrock, J.; Haunschild, J.; (1998) Xenobiotica, 28(12), 1203-1253. Ladstetter, B.; Utesch, D. (2001) Drug Metab. Dispos., 29(7), [131] Ronis, M.J.J.; Ingelman-Sundberg, M. (1999) in Handbook of drug 1042-1050. metabolism (Woolf, T.F. Eds.) Marcel Dekker, Inc, New York, pp. [99] McGuinnity, D.F. and Riley, R.J. (2001) Biochem. Soc. Trans., 239-262. 29(11), 135-139. [132] Guzelian, P.S. (1994) in Cytochrome P450 8th International [100] Houston, J.B. (1994) Biochem. Pharmacol., 47(9), 1469-79. Conference, (Lechner, M.C.; Libbey, J. Eds.), Eurotext, Paris, [101] Lave, T.; Coassolo, P.; Reigner, B. (1999) Clin. Pharmacokinet., pp.109-113. 36(3), 211-231. [133] Kocarek, T.A.; Schuetz, E.G.; Strom, S.C.; Fisher, R.A.; Guzelian, [102] Ponsoda, X.; Pareja, E.; Gómez-Lechón, M.J.; Fabra, R.; Carrasco, P.S. (1995) Drug Metab. Dispos., 23(3), 415-421. E.; Trullenque, R. and Castell, J.V. (2001) J. Hepatol., 34 (1), 19- [134] Lu, C. and Li, A.P. (2001) Chem. Biol. Interact., 134(3), 271-281. 25. [135] Meunier, V.; Bourrie, M.; Julian, B.; Marti, E.; Guillou, F.; Berger, [103] Remmel, R.P.; Burchell, B. (1993) Biochem. Pharmacol., 46(4), Y. and Fabre, G. (2000) Xenobiotica, 30(6), 589-607. 559-566. [136] Runge, D.; Kohler, C.; Kostrubsky, V.E.; Jager, D.; Lehmann, T.; [104] McGinnity, D.F.; Parker, A.J.; Soars, M.; Riley, R.J. (2000) D rug Runge, D.M.; May, U.; Stolz, D.B.; Strom, S.C.; Fleig, W.E.; Metab. Dispos., 8(11), 1327-1334. Michalopoulos, G.K. (2000) Biochem. Biophys. Res. Commun., [105] Komatsu, K.; Ito, K.; Nakajima, Y.; Kanamitsu, S.; Imaoka, S.; 273(1), 333-341. Funae, Y.; Green, C.E.; Tyson, C.A.; Shimada, N. and Sugiyama, [137] Chang, T.K.H.; Yu, L.; Maurel, P. and Waxman, D.J. (1997) Y. (2000) Drug Metab. Dispos., 28(4), 475-481. Cancer Res., 57(10), 1946-1954. [106] Bort, R.; Macé, K.; Boobis, A.; Gomez-Lechon, M.J.; Pfeifer, A. [138] Donato, M.T.; Viitala, P.; Rodríguez-Antona, C.; Lindfors, A.; and Castell, J. (1999) Biochem. Pharmacol., 58(5), 787-796. Castell, J.V.; Raunio, H.; Gómez-Lechón, M.J. and Pelkonen O. [107] Skett, P.; Tyson, C.; Guillouzo, A.; Maier, P. (1995) Biochem. (2000) J. Pharmacol. Exp. Ther., 11(28), 1321-1326. Pharmacol., 50(2), 280-285. [139] Gerbal-Chaloin, S.; Pascussi, JM.; Pichard-Garcia, L.; Daujat, M.; [108] Li, A.P.; Gorycki, P.D.; Hengstler, J.G.; Kedderis, G.L.; Koebe, Waechte, F.; Fabre, J.M.; Carrere, N. and Maurel, P. (2001) Drug H.G.; Rahmani, R.; de Sousas, G.; Silva, J.M.; Skett, P. (1999) Metab. Dispos., 29(3), 242-251. Chem. Biol. Interact., 121(1), 117-123. [140] Bowen, W.P.; Carey, J.E.; Miah, A.; McMurray, H.F.; Munday, [109] de Sousa, G.; Langouet, S.; Nicolas, F.; Lorenzon, G.; Placidi, M.; P.W.; James, R.S.; Coleman, R.A. and Brown, A.M. (2000) Drug Rahmani, R. and Guillouzo, A. (1996) Cell. Biol. Toxicol., 12(4-6), Metab. Dispos., 28(7), 781-788. 351-358. [141] Raucy, J.L., Mueller, L.; Duan, K.; Allen, S.W.; Strom, S.; Lasker, [110] Silva, J.M.; Day, S.H.; Nicoll-Griffith, D.A. (1999) Chem. Biol. J.M. (2002) J. Pharmacol. Exp. Ther., 302(2), 475-482. Interact., 121(1), 49-63. [142] Sumida, A.; Kinoshita, K.; Fukuda, T.; Matsuda, H.; Yamamoto, I.; [111] Alexandre, E.; Viollon-Abadie, C.; David, P.; Gandillet, A.; Inaba, T.; Azuma, J. (1999) Biochem. Biophys. Res. Común., Coassolo, P.; Heyd, B.; Mantion, G.; Wolf, P.; Bachellier, P.; 262(2), 499-503. Jaeck, D. and Richert, L. (2002) Cryobiology, 44(2), 103-113. [143] Rodríguez-Antona, C.; Donato, M.T.; Pareja, E.; Gómez-Lechón, [112] Dou, M.; de Sousa, G.; Lacarellem, B.; Placidi, M.; Lechene De La M.J. and Castell, J.V. (2001) Arch. Biochem. Biophys., 393 (2), Porte P.; Domingo, M.; Lafont, H. and R. Rahmani. (1992) 308-315. Cryobilogy, 29(4), 454-469. [144] Donato, M.T.; Gómez-Lechón, M.J. and Castell, J.V. (1993) Anal. [113] Adams, R.M.; Wang, M.; Crane, A.M.; Brown, B.; Darlington, Biochem., 213(1), 29-33. G.J.and Ledley, F.D. (1995) Cell Transplant., 4(6), 579-586. [145] Rodríguez-Antona, C.; Jover, R.; Gómez-Lechón, M.J. and Castell, [114] Ruegg, C. E.; Silber, P.M.; Mughal, R.A.; Ismail, J.; Lu, C.; Bode, J.V. (2000) Arch. Biochem. Biophys., 376(1), 109-116. D.C. and Li, A.P. (1997) In vitro Toxicol., 10(3), 217-222. [146] Curi-Pedrosa, R.; Daujat, M.; Pichard, L.; Ourlin, L.C.; Clair, P.; [115] Loretz, L.J.; Li, A.P.; Flye, M.W. and Wilson, A.G.E. (1989) Gervot, L.; Lesca, P.; Domergue, J.; Joyeux, H.; Fourtanier, G. and Xenobiotica, 19(4), 489-498. Maurel, P. (1994) J. Pharmacol. Exp. Ther., 269(1), 384-392. [116] Coundouris, J. A.; Grant, M. H.; Engest J.; Petrie, J. C. and [147] Masubuchi, N.; Li, A.P. and Okazaki, O. (1998) Chem. Biol. Hawksworth, G.M. (1993) Xenobiotica, 23(12), 1399-1409. Interact., 114(1-2), 1-13. 312 Current Drug Metabolism, 2003, Vol. 4, No. 4 Gómez-Lechón et al.

[148] Pelkonen, O.; Hukkanen, J.; Honkakoski, P.; Hakkola, J.; Viitala, [165] Guillouzo, A. (1998) Environ. Health Perspect., 106(Suppl. 2), P. and Raunio, H. (2002) in Ernst Schering Res Found Workshop 511-532. 37, (Pelkonen, O. and Baumann, A. Eds) pp. 105-137. [166] Davila, J.C.; Rodriguez, R.J.; Melchert, R.B. and Acosta, D. (1998) [149] Waxman, D.J. (1999) Arch. Biochem. Biophys., 369(1), 11-23. Annu. Rev. Pharmacol. Toxicol., 38, 63-96. [150] Ronis, M.J.J.; Huang, J.; Crouch, J.; Mercado, C.; Irby, D.; [167] Krishnan, K. and Andersen, M.E. (1994) in Principles and Methods Valentine, C.R.; Lumpkin, C.K.; Ingelman-Sundberg, M.; Badger, of Toxicology, (Hayes, A.W. Eds.) Raven Press, New York, pp. T.M. (1993) J. Pharmacol. Exp. Ther., 264(2), 944-950. 149-188. [151] Whitlock, J.P. (1999) Ann. Rev. Pharmacol. Toxicol., 39, 103-125. [168] Harries, H.M.; Fletcher, S.T.; Duggan, C.M.; Baker, V.A. (2001) [152] McGehee, R.E.; Ronis, M.J.J.; Cowherd, R.M.; Ingelman- Toxicol. In Vitro, 15(4-5), 399-405. Sundberg, M.; Badger, T.M. (1994) Biochem. Pharmacol., 48(9), [169] Waring, J.F.; Ciurlionis, R.; Jolly, R.A.; Heindel, M.; Ulrich, R.G. 1823-1833. (2001) Toxicol. Lett., 120(1-3), 359-368. [153] Jones, S.A.; Moore, L.B.; Shenk, J.L.; Wisely, G.B.; Hamilton, [170] Schuppe-Koistinen, I.; Frisk, A.L.; Janzon, L. (2002) Toxicology, G.A.; McKee, D.D.; Tomkinson, N.C.; LeCluyse, E.L.; Lambert, 179(3), 197-219. M.H.; Willson, T.M.; Kliewer, S.A. and Moore, J.T. (2000) Mol. [171] Castell, J.; Hernandez, D.; Gomez, A.M.; Guillen, I.; Donato, T. Endocrinol., 14 (1), 27-39. and Gomez-L, M.J. (1997) Gene Ther., 4(5), 455-464. [154] Castell, J.V.; Gómez-Lechón, M.J.; Ponsoda, X. and Bort, R. [172] Rodriguez-Antona, C.; Bort, R.; Jover, R.; Tindberg, N.; Ingelman- (1997) in In Vitro Methods in Pharmaceutical Research. (Catell, Sundberg, M.; Gómez-Lechón, M.J. and Castell, J.V. (2003) Mol. J.V. Gómez-Lechón, M.J. Eds.), Academic Press, London, UK, Pharmacol., 63(5), 1180-1189. pp.375-410. [173] Hadzopoulou-Cladaras, M.; Kistanova, E.; Evagelopoulou, C.; [155] Li, A.P. (2001) Drug Discov. Today, 6(7), 357-366. Zeng, S.; Cladaras, C. and Ladias, J. (1997) J. Biol. Chem., 1(3), [156] Jaeschke, H.; Gores, G.J.; Cederbaum, A.I.; Hinson, J.A.; Pessayre, 539-550. D.; Lemasters, J.J. (2002) Toxicol. Sci., 65(2), 166-176. [174] Torres-Padilla, M.; Fougere-Deschatrette, C. and Weiss, M. (2001) [157] Williams, D.P., Kitteringham, N.R.; Naisbitt, D.J.; Pirmohamed, Mech. Dev., 2(2), 183-193. M.; Smith, D.A.; Park, B.K. (2002) Curr. Drug Metab., 3(4), 351- [175] Descombes, P. and Schibler, U. (1991) Cell, 67(3), 569-579. 366. [176] Ossipow, V.; Descombes, P. and Schibler, U. (1993) Proc. Natl. [158] Kedderis, G.L. (1996) Toxicol. Pathol., 24(1), 77-83. Acad. Sci. USA, 90(17), 8219-8223. [159] Nieminen, A. L.; Saylor, A. K.; Tesfai, S. A.; Herman, B. and [177] Jones, P., Veenstra, G., Wade, P., Vermaak, D., Kass, S., Lemasters, J. J. (1995) Biochem. J., 307( Pt 1), 99-106. Landsberger, N., Strouboulis, J. and Wolffe, A. (1998) Nat. Genet., [160] Ross, D. (1989 J. Pharm. Pharmacol., 41(8), 505-511. 19(1), 187-191. [161] Chapman, D.E. (1995) Cell, 80(3), 259-268. [178] Collingwood, T.; Urnov, F. and Wolffe, A. (1999) J. Mol. [162] Sies, H. (1971) in Oxidative Stress. Oxidants and Antioxidants. Endocrinol., 23(3), 255-275. (Sies, H. Eds.) Academic Press. London. [179] Brown, C.; Lechner, T.; Howe, L. and Workman, J. (2000) Trends [163] Boelsterli, U.A. (1993) Drug Met. Rev., 25(4), 395-451. Biochem. Sci., 25(1), 15-19. [164] Ballet, F. (1997) J. Hepatol., 26(Suppl. 2), 26-36.