2.2. Human xenobiotic-metabolizing CYP enzymes

2.2.1.1. CYP1A1

CYP1A1 is a major extrahepatic CYP enzyme (Raunio et al. 1995a). It contributes notably to the toxicity of many carcinogens, especially PAHs, since it is the principal enzyme activating them into DNA-binding forms (Shimada et al. 1996a). Its level of expression in human liver is very low (Edwards et al. 1998). CYP1A1 constitutive expression is also very low in extrahepatic tissues, but it is inducible by AHR ligands in almost every tissue studied, including lung, lymphocytes, mammary gland, and placenta (Raunio et al. 1995a). CYP1A1 is highly inducible by PAHs and also by cigarette smoke (Anttila et al. 1991). In human primary hepatocytes, CYP1A1 is induced by the AHR agonists 3-methylcholanthrene and omeprazole (Rodríguez-Antona et al. 2000, Bowen et al. 2000). Because of the significance of CYP1A1 in the activation of procarcinogens, there have been active efforts to link the polymorphisms of the CYP1A1 gene with the individual susceptibility to chemically induced cancers, especially lung cancer (Raunio et al. 1995b). Seven variant alleles have been described, but none of them have been unequivocally shown to correlate with altered catalytic activity of the CYP1A1 protein (CYP allele nomenclature committee homepage: http://www.imm.ki.se/cypalleles/). The T3801C (MspI) and I462V polymorphisms have been studied most. It is concluded that these polymorphisms are associated with a higher risk of lung cancer in the Japanese population, but not in Caucasians (Kawajiri 1999). Recently, CYP1A1 conditional knockout mice were produced. The lack of CYP1A1 protein had no effect on the expression of any other AHR-regulated genes (Dalton et al. 2000).

2.2.1.2. CYP1A2

The expression of CYP1A2 is quite strictly liver-specific, since no CYP1A2 protein has been detected in any other tissue (Raunio et al. 1995a). CYP1A2 constitutes about 13% of the total hepatic CYP content (Shimada et al. 1994, Imaoka et al. 1996). It activates PAHs, nitrosamines, aflatoxin B₁, and especially aryl amines into forms that can bind to DNA and produce mutations (Aoyama et al. 1990, Shimada et al. 1996a, Macé et al. 1997, Hammons et al. 1997, Hecht 1998). The regulation of CYP1A2 is both AHR-ARNT–dependent and independent (Landi et al. 1999). It is induced in vivo by cigarette smoke, charbroiled meat, cruciferous vegetables containing indole-3-carbinol, phenytoin, rifampicin, and omeprazole (Landi et al. 1999). AHR agonists induce CYP1A2 in human hepatocytes (Morel et al. 1990, Xu et al. 2000). There are 6 variant alleles of the CYP1A2 gene, and two of these correlate with increased and decreased induction by smoking (Sachse et al. 1999, Nakajima et al. 1999). These variants could partially explain the observed high level of interindividual variation in the enzymatic activity measured in vivo with caffeine as a probe (Kalow & Tang 1991, Butler et al. 1992). CYP1A2 is the main CYP catalyzing the metabolism of several drugs, including clozapine, theophylline, and tacrine (Landi et al. 1999). Two strains of CYP1A2-null mice have been produced, which develop normally but have altered drug metabolism (Pineau et al. 1995, Liang et al. 1996, Gonzalez & Kimura 1999).

2.2.1.3. CYP1B1

Similarly to CYP1A1, CYP1B1 is also mainly an extrahepatic CYP form expressed in almost every tissue, including kidney, prostate, mammary gland, and ovary (Sutter et al. 1994, Shimada et al. 1996a, Tang et al. 1999). In general, CYP1B1 basal expression is higher compared to CYP1A1 (Shimada et al. 1996a, Eltom et al. 1998). The expression of CYP1B1 in human liver is nonexistent (Sutter et al. 1994, Shimada et al. 1996a, Hakkola et al. 1997, Edwards et al. 1998, Tang et al. 1999). It has been suggested to be overexpressed in tumors (Murray et al. 1997). The induction of CYP1B1 is regulated by the AHR-ARNT pathway (Sutter et al. 1994, Savas & Jefcoate 1994), although the responses of CYP1B1 to AHR ligands differ from those of CYP1A1 (Hakkola et al. 1997). CYP1B1 catalyzes the metabolism of PAHs and aryl amines (Shimada et al. 1996a). The CYP1B1 gene has several alleles, and interestingly, the functionally impaired alleles have been shown to be linked with human primary congenital glaucoma (Stoilov et al. 1997, Stoilov et al. 1998). This finding demonstrates that even the CYPs classified as “xenobiotic-metabolizing” enzymes may have important functions in modulating growth and differentiation. This is in striking contrast to CYP1B1-null mice, which develop normally and have no observable phenotype, demonstrating that CYP1B1 is not required for mouse development (Buters et al. 1999). CYP1B1-null mice are protected against 7,12-dimethylbenz[a];anthracene-induced lymphomas (Buters et al. 1999). CYP1B1 allelic variants that affect the rate of conversion of estradiol into carcinogenic 4-hydroxyestradiol have been described (Shimada et al. 1999, Li et al. 2000).

2.2.2.1. CYP2A

The human CYP2A subfamily contains three genes i.e. CYP2A6, CYP2A7, and CYP2A13, and two pseudogenes located in the CYP2A-2B-2F gene cluster on chromosome 19 (Hoffman et al. 1995). The CYP2A6 protein has been detected in liver (Yun et al. 1991), where it constitutes about 4% of total CYP content (Shimada et al. 1994, Imaoka et al. 1996). There is also evidence of the expression of a CYP2A-related protein in adult and fetal nasal mucosa (Getchell et al. 1993, Su et al. 1996, Gu et al. 2000). There has been a growing interest towards CYP2A6, due to its major role in the metabolism of nicotine in vitro (Nakajima et al. 1996, Messina et al. 1997, Yamazaki et al. 1999) and in vivo (Kitagawa et al. 1999) and in the activation of tobacco-specific nitrosamine NNK (4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (Yamazaki et al. 1992, Hecht 1998). Surprisingly, a recent study indicated that CYP2A6 is also involved in the formation of NNK from nicotine (Hecht et al. 2000). CYP2A6 catalyzes the metabolism of the gasoline additive MTBE (Hong et al. 1999) and some pharmaceuticals, including halothane and losigamone (Raunio et al. 1999b).

The genetic polymorphisms of CYP2A6 have been associated with interindividual differences in smoking behavior (Pianezza et al. 1998), although this report has methodological shortcomings (Oscarson et al. 1998). Deletions of the CYP2A6 gene have also been connected to a reduced risk of lung cancer (Miyamoto et al. 1999). Coumarin has been used as a probe drug to assess the activity of CYP2A6 in vivo (Cholerton et al. 1992, Rautio et al. 1992, Iscan et al. 1994). Very little is known about the mechanisms of regulation of CYP2A6 (Raunio et al. 1999b). It is induced in vivo by phenobarbital and other antiepileptic drugs (Sotaniemi et al. 1995). In human hepatocytes, CYP2A6 is induced by phenobarbital and rifampicin (Dalet-Beluche et al. 1992, Rodríguez-Antona et al. 2000). The CYP2A7 protein is non-functional due to its inability to incorporate heme (Yamano et al. 1990). Relatively high levels of CYP2A13 mRNA have been detected in human lung and adult and fetal nasal mucosa (Koskela et al. 1999, Gu et al. 2000, Su et al. 2000), and recent results indicate that heterologously expressed CYP2A13 is highly active in the activation of NNK (Su et al. 2000).

2.2.2.2. CYP2B6

CYP2B6 is a minor CYP form in human liver, accounting for only 1-2% of total hepatic CYP (Mimura et al. 1993, Shimada et al. 1994, Imaoka et al. 1996). Its expression appears to be regulated tissue-specifically, since in lung and kidney it is expressed as a splicing variant (Czerwinski et al. 1994, Nelson et al. 1996, Gervot et al. 1999). This splicing variant was previously called CYP2B7. The substrates for CYP2B6 include 6-aminochrysene (Mimura et al. 1993), methoxychlor (Dehal & Kupfer 1994), NNK (Code et al. 1997), and cyclophosphamide (Chang et al. 1993). The CYP2B forms in rodents are typically induced by phenobarbital, the classic CYP inducer (Honkakoski & Negishi 1998). Human CYP2B6 is also induced by phenobarbital and rifampicin in primary hepatocytes (Chang et al. 1997, Gervot et al. 1999, Rodríguez-Antona et al. 2000, Pascussi et al. 2000). This induction is mediated by nuclear receptor CAR (constitutively active receptor)(Sueyoshi et al. 1999) and probably also by PXR (pregnane X receptor) (Pascussi et al. 2000) (see chapters 2.3.2. and 2.3.3.). The human CYP2B subfamily also contains the CYP2B7P pseudogene (Nelson et al. 1996). The CYP2B genes are located in the CYP2A-2B-2F gene cluster on chromosome 19 (Hoffman et al. 1995).

2.2.2.3. CYP2C

The human CYP2C subfamily contains four highly homologous genes: 2C8, 2C9, 2C18 and 2C19, which are located in a cluster on chromosome 10 (Gray et al. 1995, Nelson et al. 1996). Interestingly, the splicing of CYP2C mRNA transcripts has been shown to produce chimeric mRNAs containing exons from several CYP2C genes (Finta & Zaphiropoulos 2000). The biological function of these mRNAs is unknown. CYP2C accounts for about 20% of the human total liver CYP content (Shimada et al. 1994, Imaoka et al. 1996). CYP2C9 is the main CYP2C in human liver, followed by CYP2C8 and CYP2C19 (Edwards et al. 1998). The CYP2C18 protein is not expressed in liver (Richardson et al. 1997). CYP2C mRNA and protein are induced in primary hepatocytes by phenobarbital and rifampicin (Morel et al. 1990) (Chang et al. 1997). Recent results indicate that PXR and CAR mediate CYP2C8 and CYP2C9 induction in human hepatocytes (Pascussi et al. 2000). However, there are also reports showing no induction of CYP2C9 and CYP2C19 mRNAs (Rodríguez-Antona et al. 2000) and proteins (Runge et al. 2000). Rifampicin and barbiturates can induce CYP2C proteins and related activities in vivo (Zilly et al. 1977, Perrot et al. 1989, Treluyer et al. 1997). Recently, it was postulated that CYP2C8 has important physiological functions in the production of endothelium-derived hyperpolarizing factor (EDHF) (Fisslthaler et al. 1999). Pharmaceutical substrates for CYP2C include diazepam, omeprazole, mephenytoin, tolbutamide, and warfarin (Guengerich 1995) as well as many non-steroidal anti-inflammatory drugs (Pelkonen et al. 1998). Selective substrates include taxol for CYP2C8, tolbutamide for CYP2C9, and mephenytoin for CYP2C19 (Pelkonen et al. 1998). The CYP2C19 poor metabolizer phenotype is detected in 2-4 % of Caucasians and in about 20% of Asians (Ingelman-Sundberg et al. 1999). CYP2C9 also has two variants (CYP2C9*2 and CYP2C9*3) with decreased metabolism (Miners & Birkett 1998, Ingelman-Sundberg et al. 1999).

2.2.2.4. CYP2D6

The CYP2D subfamily has one gene and four pseudogenes (Nelson et al. 1996). The CYP2D6 polymorphism was the first defect in drug metabolism to be specifically associated with altered expression of CYP enzyme (Gonzalez et al. 1988). The CYP2D6 poor metabolizer (PM) phenotype is detected in about 6% of Caucasians (Ingelman-Sundberg et al. 1999), and it has profound effects on the metabolism of several commonly used pharmaceuticals, including several tricyclic antidepressants, haloperidol, metoprolol, propranolol, codeine, and dextromethorphan (Pelkonen et al. 1998). There are at least 30 different defective CYP2D6 alleles, six of which contribute to 95-99% of PM phenotypes (Ingelman-Sundberg et al. 1999). Interestingly, duplications of the CYP2D6 gene up to 13 gene copies have been reported (Johansson et al. 1993), giving rise to the ultra-rapid metabolizer phenotype. Ultra-rapid metabolizers show increased metabolism and decreased drug effects of CYP2D6 substrates, such as tricyclic antidepressants (Dalen et al. 1998). About 4% of Caucasians have multiple CYP2D6 genes (Ingelman-Sundberg et al. 1999). It has been speculated that since CYP2D6 is not inducible, the duplications are a way to adapt to environmental chemical pressures, most likely to alkaloids in the diet (Ingelman-Sundberg et al. 1999). CYP2D6 polymorphisms have been linked to altered susceptibility to Parkinson’s disease and lung cancer. A meta-analysis did not reveal a link to Parkinson’s disease, but PM individuals seem to be somewhat protected (OR = 0.69) against lung cancer (Rostami-Hodjegan et al. 1998). CYP2D6 has a minor, but not crucial, role in the activation of tobacco-derived nitrosamine NNK (Crespi et al. 1991, Hecht 1998). It has been speculated that the CYP2D6 polymorphism might affect the risk of lung cancer through modulating smoking behavior, since CYP2D6 might be involved in the signal transduction of the dopaminergic pathway in brain (Saarikoski et al. 2000). CYP2D6 constitutes about 2% of total hepatic CYP (Shimada et al. 1994, Imaoka et al. 1996), and the protein is also expressed in duodenum and brain (Pelkonen & Raunio 1997).

2.2.2.5. CYP2E1

CYP2E1 is the only gene in this subfamily (Nelson et al. 1996). The CYP2E1 enzyme has been studied extensively due to its role in the metabolism of ethanol and also as an activator of chemical carcinogens (Lieber 1997). CYP2E1 also actives some tobacco-specific nitrosamines, but not NNK (Yamazaki et al. 1992, Kushida et al. 2000), the most important nitrosamine in tobacco (Hecht 1999). Most of the over 70 substrates demonstrated are small and hydrophobic compounds (Ronis et al. 1996), including only a few pharmaceuticals, such as paracetamol, chlorzoxazone, enflurane, and halothane (Guengerich 1995). Disulfiram is a clinically used inhibitor of CYP2E1 (Guengerich & Shimada 1991). About 7% of the liver CYP content consists of CYP2E1 (Shimada et al. 1994, Imaoka et al. 1996). It is also expressed in lung and brain (Raunio et al. 1995a). The regulation of CYP2E1 is complex, since it is regulated transcriptionally, pretranslationally, translationally, and posttranslationally (Song 1995). Transcriptional regulation seems to play a minor role, in contrast to many other CYP forms. Many substrates of CYP2E1 are also CYP2E1-inducing agents, including acetone, ethanol, pyridine, pyrazole, and isoniazid (Ronis et al. 1996). Ethanol intake increases the human CYP2E1 content in liver in vivo (Perrot et al. 1989), and it is also induced in lymphocytes of poorly controlled insulin-dependent diabetics (Song et al. 1990). The regulation of CYP2E1 will be discussed in more detail in chapter 2.3.5. In addition to activating procarcinogens, CYP2E1 also produces free radicals causing tissue injury. These radicals are formed both in the absence and in the presence of substrate (Lieber 1997). Several allelic variants of the CYP2E1 gene have been detected. Three of these cause amino acid substitutions (Hu et al. 1997, Fairbrother et al. 1998). One of these alleles (CYP2E1*2) produces a protein with reduced stability (Hu et al. 1997), while another (CYP2E1*1D) is associated with increased activity after alcohol exposure and in obese subjects (McCarver et al. 1998). CYP2E1 knockout mice develop normally and are protected against paracetamol-induced hepatotoxicity (Lee et al. 1996).

2.2.2.6. CYP2F1

The CYP2F1 gene is located in the CYP2A-2B-2F gene cluster in chromosome 19 (Hoffman et al. 1995). Interestingly, it has two full-length copies in addition to one pseudogene, but it is not known whether both of them produce functional transcripts (Hoffman et al. 1995). CYP2F1 mRNA has been identified in human lung (Nhamburo et al. 1990) and placenta (Hakkola et al. 1996), but not in liver (Hakkola et al. 1994). No expression of CYP2F1 protein has been demonstrated in any human tissue. Recombinant CYP2F1 enzyme is capable of activating pulmonary toxicants 3-methylindole and naphthalene (Lanza et al. 1999).

2.2.2.7. CYP2J2

CYP2J2 is expressed mainly extrahepatically in heart, kidney (Wu et al. 1996), lung (Zeldin et al. 1996), pancreas (Zeldin et al. 1997b), and gastrointestinal tract (Zeldin et al. 1997a). CYP2J2 is involved in the metabolism of arachidonic acid into epoxyeicosatrienoic acids (EETs), which have physiological functions (Wu et al. 1996, Zeldin et al. 1997a, Capdevila et al. 2000). CYP2J2 has not been demonstrated to exhibit activity towards xenobiotics.

2.2.3.1. CYP3A4

The CYP3A4 enzyme is the most important drug-metabolizing CYP in human liver. About 30-40% of the total hepatic CYP content consists of CYP3A4 (Shimada et al. 1994, Imaoka et al. 1996) and it is also present in small intestine (Kolars et al. 1992). It has been estimated that about 50% of the drugs metabolized by CYPs are metabolized by CYP3A4 (Bertz & Granneman 1997). The substrates for this enzyme include drugs, such as quinidine, nifedipine, diltiazem, lidocaine, lovastatin, erythromycin, cyclosporin, triazolam, and midazolam, and endogenous substances, including testosterone, progesterone, and androstenedione (Pelkonen et al. 1998, Guengerich 1999). Midazolam and erythromycin have been used as in vivo probes for CYP3A4 activity (Thummel & Wilkinson 1998). CYP3A4 also activates procarcinogens, including aflatoxin B₁ (Aoyama et al. 1990), PAHs, NNK (Hecht 1999), and 6-aminochrysene (Yamazaki et al. 1995). CYP3A4 is induced in human hepatocytes by rifampicin (Morel et al. 1990, Schuetz et al. 1993), dexamethasone (Pichard et al. 1992, Schuetz et al. 1993, Kocarek et al. 1995), and phenobarbital (Schuetz et al. 1993, Kocarek et al. 1995) among others. CYP3A4 is induced in vivo by rifampicin and barbiturates in liver (Perrot et al. 1989, Ged et al. 1989) and by rifampicin in small intestine (Kolars et al. 1992). The induction of CYP3A4 is mainly regulated by the novel orphan receptor PXR (Lehmann et al. 1998, Moore et al. 2000a), but also other receptors, including CAR and, indirectly, the glucocorticoid receptor are involved (Sueyoshi et al. 1999, Pascussi et al. 2000, Moore et al. 2000a)(See chapters 2.3.2. and 2.3.3.). Three variant alleles have been detected for the CYP3A4 gene. A 5’-flanking allelic variant has been associated with prostate cancer (Rebbeck et al. 1998) and leukemia (Felix et al. 1998). However, the functional significance of this alteration is uncertain (Westlind et al. 1999). Recently, two other amino acid alterations have been discovered, one of which encodes an enzyme with reduced activity (Satav 2000).

2.2.3.2. CYP3A5

CYP3A5 is expressed polymorphically in human liver (Wrighton et al. 1989), but consistently in lung (Kivistö et al. 1996), colon (Gervot et al. 1996), kidney (Schuetz et al. 1992, Haehner et al. 1996), oesophagus (Lechevrel et al. 1999), and anterior pituitary gland (Murray et al. 1995), demonstrating CYP3A5 to be a more extrahepatic CYP3A form. About 20-25% of livers have substantial levels of CYP3A5 protein (Aoyama et al. 1989, Wrighton et al. 1989). However, more sensitive protein or mRNA detection methods reveal CYP3A5 in almost every liver sample (Jounaidi et al. 1996, Edwards et al. 1998). There are some variant alleles of the CYP3A5 gene (Jounaidi et al. 1996, Paulussen et al. 2000). An allelic variant with linked –45A>G and –369T>G sequence mutations has been shown to lead to increased expression of CYP3A5 protein in liver (Paulussen et al. 2000). There are also two CYP3A5 pseudogenes (Nelson et al. 1996). In comparison to CYP3A4, CYP3A5 shows roughly the same substrate preference pattern, but the turnover rates are usually lower (Aoyama et al. 1989, Wrighton et al. 1990, Yamazaki et al. 1995). However, CYP3A5 is unable to metabolize some CYP3A4 substrates, including erythromycin and quinidine (Wrighton et al. 1990). CYP3A5 is generally thought to be uninducible in human liver and primary hepatocytes (Wrighton et al. 1989, Schuetz et al. 1993, Chang et al. 1997), but there is some evidence for the induction of CYP3A5 in colon carcinoma cell lines by reserpine and clotrimazole (Schuetz et al. 1996). Surprisingly, in a recent study, CYP3A5 mRNA was induced in human primary hepatocytes by rifampicin and phenobarbital (Rodríguez-Antona et al. 2000). The promoter region of the CYP3A5 gene has been shown to contain functional glucocorticoid-responsive element half-sites that mediate induction by dexamethasone in the HepG2 cell line (Schuetz et al. 1996). Unlike the other CYP3A enzymes, CYP3A5 is probably not regulated by PXR (Blumberg et al. 1998).

2.2.3.3. CYP3A7

CYP3A7 is mainly expressed in human fetal liver, where it is the major CYP form (Kitada & Kamataki 1994). Low levels of CYP3A7 mRNA have also been detected in adult liver (Hakkola et al. 1994, Schuetz et al. 1994). CYP3A7 has similar catalytic properties compared with other CYP3A enzymes, including testosterone 6β -hydroxylation (Kitada et al. 1985, Kitada et al. 1987, Kitada et al. 1991). CYP3A7 is induced in adult human primary hepatocytes by rifampicin (Greuet et al. 1996), and this induction is mediated by the PXR pathway (Pascussi et al. 1999). Whether or not CYP3A7 is induced in fetal liver is unknown.