VU Research Portal

(1)

VU Research Portal

Inter-individual variation in hepatic drug metabolism

den Braver-Sewradj, S.P.

2018

document version

Publisher's PDF, also known as Version of record

Link to publication in VU Research Portal

citation for published version (APA)

den Braver-Sewradj, S. P. (2018). Inter-individual variation in hepatic drug metabolism: The potential of in vitro assays in unraveling the role of metabolism in drug induced liver toxicity.

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal ?

Take down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

E-mail address:

vuresearchportal.ub@vu.nl

(2)

Inter-individual variation in hepatic drug metabolism

The potential of in vitro assays in unraveling the

role of metabolism in drug induced liver toxicity

(3)

The research described in this thesis was part of the MIP-DILI consortium, supported by the Innovative Medicines Initiative Grant Agreement number 115336, and was carried out at the AIMMS-Division of Molecular Toxicology, Department of Chemistry and Pharmaceutical Sciences, Vrije Universiteit Amsterdam, De Boelelaan 1108, 1081 HZ Amsterdam, The Netherlands.

Printed by: ProefschriftMaken || www.proefschriftmaken.nl Cover design: Patricia Brouwer Design

ISBN: 978-94-6295-835-7

(4)

VRIJE UNIVERSITEIT

Inter-individual variation in hepatic drug metabolism. The potential

of in vitro assays in unraveling the role of metabolism in drug

induced liver toxicity

ACADEMISCH PROEFSCHRIFT ter verkrijging van de graad Doctor aan

de Vrije Universiteit Amsterdam, op gezag van de rector magnificus

prof.dr. V. Subramanian, in het openbaar te verdedigen ten overstaan van de promotiecommissie

van de Faculteit der Bètawetenschappen op dinsdag 20 februari 2018 om 11.45 uur

in de aula van de universiteit, De Boelelaan 1105

door

(5)

promotor: prof.dr. N.P.E. Vermeulen copromotoren: dr. J.C. Vos

(6)

(7)

Reading committee: prof.dr. P. Jennings prof.dr. G.M.M. Groothuis prof.dr. J. de Boer

(8)

Chapter 1

(11)

Page | 10

Together with drug excretion, drug metabolism is an important mechanism of drug elimination from the human body. This is achieved by biotransformation of (lipophilic) drug entities into more polar metabolites. Although drug metabolizing enzymes are expressed in many organs and tissues, the principle site of drug metabolism is the human liver (Figure 1). As most drugs are administered orally, the majority of drugs have to pass the gastrointestinal (GI)-tract before being taken up into the circulation via the liver. Drugs may thereby undergo a so-called first-pass effect which may significantly lower bioavailability (1). The clearance, which describes the elimination of a drug from the human body and which is the most important pharmacokinetic parameter, strongly depends on the metabolic capacity of the liver (2). Metabolism of new drug candidates is consequently thoroughly investigated during development of new drug entities to properly estimate the effective systemic concentration in humans and to prevent adverse drug reactions resulting from pharmaco- or toxico-active metabolites (Figure 2).

Figure 1. Drug uptake, excretion and possible first-pass effect following oral administration of drugs. Metabolism takes place in many organs and tissues, including the gut wall, kidney, skin and lungs, but drug metabolizing enzymes are mostly expressed in the liver. The liver is therefore a major determinant of bioavailability. Adapted from van de Waterbeemd and Gifford, 2003 (3).

(12)

Safety-related attrition is however not significantly/completely reduced. This is partially because information regarding efficacious exposures is limited and many validated and well understood toxicology assays consequently still have a low positive predictive value (5). As a consequence many potential leads are discontinued early in the process, and it has been hypothesized that a better approach would encompass “testing the right things at the right time to make a diligent decision instead of testing everything early” (6). The validation and characterization of in vitro assays in terms of factors affecting the experimental outcome, such as metabolic competence, is needed in this paradigm shift towards choosing the right test system and identification of “alerts” in nonclinical safety testing instead of decision making for discontinuation of a compound.

Figure 2. Investigation of drug metabolism in all phases of drug discovery. Adapted from Zhang et al., 2012 (7).

1.1. Drug metabolism and adverse drug reactions

(13)

Page | 12

(9). Drug induced liver injury (DILI) is the most important contributor to ADR-related drug withdrawals (Figure 3) (8,10), which can be rationalized by the central role of the liver in drug metabolism (10,11).

Figure 3. Toxicity types associated with ADR-related drug withdrawals from one or more countries. Adapted from Siramshetty et al., 2016 (10).

DILI can be subdivided into two categories: intrinsic liver injury and idiosyncratic liver injury. Intrinsic liver injury is dose-dependent and has a predictable cascade of symptoms (11). In contrast, idiosyncratic drug induced liver injury (IDILI) often occurs with drugs which are prescribed at the doses higher than 50-100 mg (12,13) and generally does not show a dose-dependent risk within the therapeutic range (14). It is characteristic for IDILI that liver injury has a delayed onset, usually 1-3 months (14). Upon re-challenge with the drug, a rapid onset of liver injury is typically observed (14). Diagnosis of IDILI is challenging since the symptoms are very divers and no selective biomarkers have been identified so far (15,16). Although its incidence is relatively low (1-10,000 to 1-1000,000) (15), patient morbidity and mortality is significant. The U.S. acute liver failure study group found that spontaneous patient survival following IDILI is only 29%, while 41% of the patients needed a liver transplantation (17).

(14)

consistent clinical characteristics suggests the involvement of multiple mechanisms (18). Given the complexity, the low incidence and the limited understanding of the underlying mechanisms, current preclinical models are non-predictive and clinical cohorts are too small to detect IDILI (15,19). Therefore the manifestation of IDILI is mostly not recognized until the drug is launched to the market and administered to millions of patients (20).

Figure 4. IDILI is considered to be a multifactorial process. Drug specific factors increase likeliness of IDILI, however not every patient is susceptible. Patient-specific susceptibility is caused by factors associated with host and environment. No significant associations have been found between drug- or environmental factors and IDILI, however, they are likely to contribute to the risk for IDILI. Associations at the host level have been found, e.g. polymorphisms and susceptibility to IDILI; polymorphisms of UGT1A6 and APAP-induced DILI (19); and polymorphisms of UGT2B7 and CYP2C8 to diclofenac-induced IDILI (21). Adapted from Licata, 2016 (22).

(15)

Page | 14

can also result in a therapeutically active metabolite or a chemically reactive metabolite (CRM) (Figure 5).

Figure 5. Effects of drug metabolism on activity and/or toxicity of drugs and metabolites. Adjusted from Gunaratna, 2000 (25).

1.1.1. Phase I metabolism

Phase I metabolism often encompasses the introduction of a polar functional group on a molecule. The most important family of phase I enzymes is the cytochrome P450 (CYP) superfamily, as human clearance of approximately 75% of drugs is facilitated by CYPs (26). The majority of these reactions is covered by five isoforms, CYP3A4 (27%), CYP2D6 (13%), CYP2C9 (10%), CYP1A2 (9%) and CYP2C19 (9%) (26). To date, 57 human CYPs have been identified. CYPs are heme-containing oxygenases catalyzing many reactions including carbon hydroxylation, heteroatom oxygenation, heteroatom dealkylation and epoxidation, but also less frequent reactions such as reduction, dehydration and desaturation (27,28). Human CYPs are membrane bound and are mainly expressed in the membrane of the endoplasmic reticulum. It is generally acknowledged that CYPs are anchored in the membrane by their hydrophobic NH2-terminus with their active side facing the cytoplasm

(Figure 6) (29,30), although there are indications that particular CYPs face the lumen (31). Drug-metabolizing CYPs are predominantly expressed in the liver, but are also present in extrahepatic tissue including the respiratory and gastrointestinal tracts (32,33). Substrate specificity is mainly determined by the positioning of the substrate in the active site and the active site topology, which varies greatly between CYP isoforms (27,34).

(16)

or DT diapohorase) and NRH: quinone reductase 2 (NQO2 or QR2). NQO1 and NQO2 are FAD-bound proteins and are 49% similar at protein level, which a main difference being the shorter C-terminal tail in NQO2 (35–37). These missing 43 amino acids results in one of the major differences between the enzymes, as NQO1 uses NADPH as a cofactor and

Figure 6. Cellular location of phase I and phase II enzymes (upper figure) and protein abundances in human liver (lower figures, expressed per mg microsomal or cytosolic protein for CYP and UGT or SULT, respectively). Abundance data of UGT and SULT protein is based on data from 76-154 and 28 human livers, respectively. Abbreviations: AADAC, arylacetamide deacetylase; ABHD10, α/β hydrolase domain containing 10 (prefix c and m stands for cytosolic and microsomal, respectively); BLQ, below limit of quantification; CES, carboxylesterase; NPR, NADPH-P450 reductase; P450, cytochrome P450; SULT, sulfotransferase; UGT, UDP-glucuronosyltransferase. Adjusted from Oda et al., 2015; Achour, Rostami-Hodjegan and Barber, 2014; Achour, Rostami-Hodjegan, 2014; Riches et al., 2009 (45–48).

(17)

Page | 16

very low levels in mice (42). NQO1 and NQO2 are expressed in many tissues, including the liver, kidney, lung and skeletal muscle (36,43).

Other relevant phase I enzymes include carboxylesterases, epoxide hydrolases, amidases, alcohol and aldehyde dehydrogenases, xanthine oxidases and flavin-containing monooxygenases (44).

1.1.2. Phase II metabolism

Phase II drug metabolism encompasses the enzymatic conjugation of a drug or a metabolite to an endogenous hydrophilic cofactor, the metabolic product of which is more easily excreted from the human body (49). In terms of drug metabolism, the most common phase II enzymes include UDP-glucuronosyltransferase (UGT), sulfotransferase (SULT), glutathione-S-transferase (GST), N-acetyltransferase (NAT) and methyltransferases (50). Following CYPs, UGTs are the second most important enzymes in clearance of drug, covering 15% of the top 200 most prescribed drugs (51). The contribution of UGTs in drug metabolism is expected to increase in the future, as extensive CYP metabolism is generally avoided for new drug candidates (52). Although exact numbers have not been reported for SULTs and GSTs, it is well established that their contribution to drug metabolism is significant as well (53–55).

(18)

selectivity is greater than for CYPs (60). Consequently, glucuronidation of drug is often catalyzed by multiple UGT isoforms (62). Drug-drug interactions are therefore less pronounced for UGTs compared to CYPs (63).

SULTs catalyze the sulfonation of many endogenous and exogenous compounds, by transferring the sulfonyl moiety from 3’-phosphoadenosine 5’-phosphosulfate (PAPS) to an acceptor group with a N-, O- or S-nucleophilic atom (64). In contrast to the majority of other drug metabolizing enzymes, highest expression is found in the intestinal tract, followed by the liver (65,66). SULTs are expressed in both membranes and cytosol, however only cytosolic SULTs are active in sulfonation of xenobiotics (Figure 6) (67). The superfamily of cytosolic SULTs in human consists of four gene families (SULT1, SULT2, SULT4 and SULT6) comprising a total of 13 isoforms (66,68,69). Of these, the most important isoforms for xenobiotic metabolism are SULT1A1, SULT1A3, SULT1B1, SULT1C2, SULT1E1 and SULT2A1 (65). Like other drug metabolizing enzymes, substrate selectivity is broad and many xenobiotics can be sulfated by multiple SULTs. Moreover, many SULT substrates can be conjugated by UGTs as well. A particular property of SULTs compared to other drug metabolizing enzymes comprises their relatively high affinities for xenobiotics, with Km values in the nM to low µM range (54). As it has been estimated that

the concentration of the major hepatic isoforms SULT1A1 and SULT2A1 in a hepatocyte is around 40 and 20 µM, respectively, sulfonation is consequently often the first metabolic route at low drug doses (51,70).

(19)

Page | 18

major subunits in human liver are GSTA1, GSTA2, GSTM1 and GSTT1 (74,75). Even though data is still limited, enzymatic GSH conjugation of CRMs has been shown for a number of GSTs, including GSTA1-1, GSTA2-2, GSTM1-1, GSTM3-3 and GSTP1-1 (76–80). GSTs are multifunctional proteins and are involved in many cellular processes besides detoxification of cell damaging molecules, such as glutathionylation, regulation of ion channels and regulation of signaling pathways (75). GST activity can therefore be ambiguous. Very illustrative in this context is the high GSTP1-1 activity in GSH conjugation of the acetaminophen (APAP)-derived reactive metabolite N-acetyl-p-benzoquinonimine (NAPQI) found in vitro, while APAP-induced liver injury was reduced in GSTP1 knock out mice (81,82).

1.1.3. Phase III transport

Active cellular uptake and excretion of drugs and/or metabolite is referred to as phase 0 and phase III metabolism. Transporters control the passage of xenobiotics across plasma- and cell membranes and consequently have a central role in the uptake into hepatocytes, excretion into the bile, or reuptake into the systemic circulation of drugs (83,84). Indeed, it is suggested that transporters work side by side with drug metabolizing enzymes to regulate drug absorption and elimination (85). A lot of progress has been made in the past decades in characterization of transporters and identification of substrates (84). Transporters are expressed in varying abundances in the human body and therefore have a key role in drug distribution and absorption (85). Hepatocytes express a wide range of transporters (Figure 7). Although most transporters are considered bidirectional, active hepatic uptake is facilitated by members of the solute carrier (SLC) superfamily which are expressed in the basolateral membrane (sinusoidal) and include organic anion transporting polypeptide 1B1/1B3/2B1 (OAT1B1/1B3/2B1), organic anion transporter 2 and 7 (OAT2/7), Na+_{-taurocholate co-transporting peptide (NTCP) and organic cation transporter 1 (OCT1).}

(20)

transporters, however also many endogenous processes are regulated by (hepatic) transporters (e.g. disposition of bile salts, BSEP; efflux of endogenous metabolites, MRP1; regulation of bile salt enterohepatic circulation, MRP3) (86). Specificity of substrates is generally broad, making them susceptible for drug-drug interactions (84). Broad classifications can however still be made, e.g. substrates of MRP2 are generally hydrophobic uncharged molecules or water-soluble anionic compounds, while P-gp mostly recognizes large hydrophobic and positively charged compounds with a wide range of structures (87). Evidence for the involvement of transporters in the occurrence of DILI (mainly cholestatic injury) is accumulating, and is hypothesized to encompass a) transport of potentially hepatotoxic metabolites or drugs or b) inhibition of excretion of potentially hepatotoxic bile acids (88,89). For example, it was shown that reduced function of MRP2 was associated with a higher risk of diclofenac-induced hepatotoxicity in patients, presumably as a result of accumulation of CRMs (21). A well-known BSEP inhibitor is the antidiabetic drug troglitazone, for which its sulfate metabolite is an even more potent inhibitor (90,91).

(21)

Page | 20

1.2. Chemically reactive metabolites and liver toxicity

Although metabolism generally renders the parent drug more hydrophilic to facilitate elimination and reduce toxicity, it can also result in activation of the drug to CRMs. CRMs are usually electrophiles (e.g. epoxides, unstable conjugates, quinones) or radicals, which are short lived because they are unstable, readily detoxified (e.g. enzymatically or by GSH conjugation) or because they rapidly react with proteins, nucleic acids or lipids (94). The formation of CRMs is associated with the occurrence of oxidative stress and/or adduct formation with DNA or proteins, which may lead to genotoxicity, direct cell damage or immunoallergenic toxicity (Figure 7) (95). Importantly, the latter two are implicated in the occurrence of ADRs including (I)DILI (96,97). In a retrospective study performed by Stepan et al. it was shown that the formation of CRMs was linked to the initiation of toxicity for 62-69 % of drugs associated with IDILI. However, approximately half of the top 200 drugs in 2009 were also CRM positive, but not associated with IDILI (98). In terms of drug properties, the most obvious discriminating factor between ‘safe’ and ADR or (I)DILI drugs is the dose, as the therapeutic dose is generally higher for IDILI drugs. The combination of CRM formation and the daily dose alone is however not sufficient to single out (I)DILI drugs from safe drugs, and although still not predictive, inclusion of factors such as relative importance of metabolic pathways, covalent binding and/or inhibition of transporters can significantly improve discrimination (99–101). The general assumption is that CRM formation is especially troublesome when the systemic immune response dysfunctions and/or when detoxifying mechanism are overwhelmed by the levels and downstream effects of CRMs (102–104). The balance between bioactivation and detoxification ultimately determines the intracellular exposure which is more important than the formation of CRMs alone in the initiation of ADRs (Figure 8) (76). Moreover, a very important missing factor in assessment of CRM toxicity using nonclinical models is the human immune system, which is important in most hypotheses concerning ADRs and (I)DILI (section 1.1) (105).

1.2.1. Formation of reactive metabolites

(22)

carbamazepine, aflatoxin B1) (94,106–108) (Figure 9). These CYP generated CRMs are soft electrophiles, which can be trapped by the soft nucleophiles, e.g. GSH or thiols on proteins (109). Formation of hard electrophiles can also be catalyzed by CYP (e.g. the aldehydes acrolein from cyclophosphamide) which can react with hard nucleophiles such as DNA, or (enzymatically) with the soft nucleophile GSH (110,111). In some cases, the CRM is so reactive it does not leave the active site, but covalently binds an active site amino acid thereby inactivating the bioactivating CYP. This leads to downstream toxic effects such as drug-drug interactions and hapten formation (112). Bioactivation by CYP is common,

Figure 8. Downstream effects of CRMs formed in the liver. Intracellular concentrations of CRMs are determined by the balance between bioactivation and detoxification. CRMs are usually electrophiles or radicals, which can react with proteins or DNA resulting in direct damage, immunotoxicity or genotoxicity. Cell damage also occurs by direct increase of reactive oxygen species or by depletion of cellular GSH levels. GSH, glutathione, ROS, reactive oxygen species. Adapted from Gómez-Lechón et al., 2016 (106).

(23)

Page | 22

Figure 9. Examples of CRMs from DILI-associated drugs which are formed by CYPs.

Phase II-mediated bioactivation is less common. UGT-mediated bioactivation concerns the glucuronidation of carboxylate drugs, resulting in 1-O-β-acyl glucuronide conjugates (e.g. from diclofenac, Figure 10). 1-O-β-acyl glucuronide can directly react with nucleophiles as a result of its electrophilic acyl carbon, can be (enzymatically) hydrolyzed back to the parent molecule or, importantly, can undergo migration to 3 isomers: 2-O, 3-O or 4-O-β-acyl glucuronide. Although migration is reversible, it is energetically unfavorable. The 4-O-β-acyl glucuronide isomers cannot undergo hydrolysis and can ring-open to aldoses which are more reactive than 1-O-β-acyl glucuronide (116). Although all acyl glucuronides are

Figure 10. Glucuronidation of diclofenac to the CRM diclofenac 1-O-acyl glucuronide and subsequent hydrolysis, reactivity with nucleophiles (on proteins or GSH) and/or acyl migration followed by protein glycation. Adapted from Grillo et al., 2003 (120).

(24)

classification of safe and DILI drug, which was improved further when daily dose was taken into account (118). However, while in vitro proof of reactivity and resulting cell toxicity of acyl glucuronides is accumulating, the translation to the patient is difficult as acyl glucuronides are rapidly cleared from the hepatocytes and body. Therefore, even though it is well established that acyl-glucuronides form covalent protein adducts in the liver, the link to hepatotoxicity is less clear as the exact mechanism in vivo is not well defined (117,119).

Similar to UGTs, SULT-mediated bioactivation is uncommon. Reactivity of sulfate conjugates is dependent on the drug moiety, as sulfate has electron-withdrawing properties and is a good leaving group in particular chemical linkages (e.g. benzylic and allylic alcohols, aromatic hydroxylamines) (121). Sulfate groups in such conjugates can be displaced by a nucleophile (both in DNA and proteins) or undergo heterolytic cleavage resulting in a strong electrophilic cation (94,121). An example is the cancer chemotherapeutic drug tamoxifen. Tamoxifen is metabolized by CYP3A to α-hydroxytamoxifen, which can subsequently be conjugated by SULTs to a sulfate ester. This resulting reactive carbocation can directly form DNA adducts (122,123) (Figure 11).

Figure 11. Sequential CYP and SULT mediated bioactivation of tamoxifen in rat liver. Following α-hydroxylation, sulfation results in an unstable sulfate conjugate which collapse into a reactive carbocation. This metabolite can react with nucleophilic sites in DNA, thereby causing DNA adducts. Adapted from Brown, 2009 (125)

(25)

Page | 24

The prodrug azathioprine was specifically designed as a prodrug that is activated by conjugation with thiols followed by the release of 6-mercaptopurine, which interferes with DNA synthesis. It was originally believed that bioactivation of azathioprine mainly resulted from chemical GSH conjugation. However, Eklund and colleagues demonstrated in 2006 that in human tissues bioactivation was mainly catalyzed by GST and that chemical conjugation only contributed to 1% of total bioactivation (124).

1.3. Inter-individual variation in drug metabolism

(26)

Table 1. Reported associations with DILI. Most associations are found in a small subset of patients and/or need conformation in a repetition study.

Gene Drug Reference

Phase I CYP2B6 Ticlopidine (126) CYP2C8 Diclofenac (21,132) CYP2C9 Bosentan (126) CYP2C19 Troglitazone (133) CYP2D6 Perhexiline (132,134) CYP2E1 Isoniazid (132)

CYP2E1 Anti-tuberculosis drugs (135)

Phase II GSTM1 Anti-tuberculosis drugs (132) GSTM1 Diverse drugs (132) GSTT1 Diverse drugs (132) GSTT1/GSTM1 Troglitazone (135,136) GSTT1/GSTM1 Tacrine (135,137) GSTT1/GSTM1 NSAIDs (131,138) GSTT1/GSTM1 Antibacterial drugs (131,138) UGT1A6 Tolcapone (139)

UGT1A9 COMT inhibitors (132)

UGT2B7 Diclofenac (21,132)

SULT Chlorpromazine (140)

NAT2 Anti-tuberculosis drugs (132)

NAT2 Isoniazid (131)

Phase III

ABCB1 (MDR1) Nevirapine (141)

ABCB11 (BSEP) Diverse drugs (135)

ABCC2 (MRP2) Diclofenac (21,135,142)

ABCC2 (MRP2) Diverse drugs/herbal plants (135,143)

1.3.1. Genetic variation in phase I metabolism

(27)

Page | 26

metabolizers (one functionally deficient allele), extensive metabolizers (normal phenotype) or ultra-rapid metabolizers (gain of function variants) (144,145). Defective CYP2C19 genes have been linked to troglitazone-induced liver injury (133) and CYP2D6 mutants have been associated with liver injury resulting from perhexiline use (134). In contrast, a recent study in a well-defined cohort of patients experiencing IDILI from various drugs (not including troglitazone) did not find a relation between CYP2C9 and CYP2C19 polymorphisms and IDILI (146). Loss of function or gain of function variants are rare for the high/relatively high polymorphic CYP1A1, CYP1A2, CYP2C8 and CYP2E1 (32,144). CYP3A4 is very conserved in the population, and only a limited number of (rare) polymorphisms have been identified (32). Most of these do not affect mRNA expression or are lacking functional characterization. So far, one variant was shown to have decreased activity and two loss-of-function mutations have been identified (147). The most important genetic variants of CYP enzymes are known and in some cases, adverse drug reactions can be related to the occurrence of CYP polymorphisms (32). However, only weak associations have been reported for CYP mutations and IDILI, e.g. CYP2B6 and ticlopidine, CYP2C9 and bosentan and CYP2E1 with Isoniazid. Replications in larger cohorts are however needed to verify these associations (126). Illustratively, a recent study using human liver microsomes from 105 genetically well-characterized donors showed only marginal effects of genotype on metabolic activity (148).

(28)

The NQO1 gene locus is not highly polymorphic but activity is very variable because of its highly inducible nature, especially during liver injury (149). In addition, variation can be ascribed to the existence two low-activity variants (NQO1*2 and NQO1*3). (150) The NQO2 gene is highly polymorphic, however most mutations are non-functional or located in the promoter region (151,152). No associations between DILI and NQO1 or NQO2 have been reported, but it is noteworthy that mutations in the phase I NQO2 gene locus have been associated to a higher risk for clozapine-induced agranulocytosis (151,153).

1.3.2. Genetic variation in phase II metabolism

Several polymorphisms have been described for phase II enzymes including UGT, SULT, GST and NAT (50). N-acetylation by NAT2 is strongly polymorph and mutations resulting in increased or abolished activity are relatively common (131,135,145). Reduced NAT2 activity has been associated with isoniazid-induced hepatotoxicity. It is postulated that deficient NAT2 expression results in reduced detoxification, pushing the balance to CYP2E1 mediated bioactivation (131). Indeed, isoniazid-induced hepatotoxicity has also been linked to CYP2E1 polymorphisms as well (154).

A few associations have been reported for UGT and DILI. Reactive drug metabolites from the IDILI-related drug diclofenac can be derived from phase I (formation of quinoneimines) and from phase II (acyl glucuronides). Interestingly, while CYP2C9 is the main enzyme responsible for phase I mediated bioactivation, an association with diclofenac-induced IDILI could not be confirmed (131). Instead, mutations in UGT2B7 (which is the main enzyme in diclofenac glucuronidation), CYP2C8 and the transporter MRP2 have been linked to diclofenac hepatotoxicity (21). The association with UGT2B7 was confirmed in a later study (142). Reduced detoxification of tolcapone as a result of polymorphisms in UGT1A6 has been linked to a higher risk for liver toxicity as well (139). Polymorphic variants of SULT enzymes have been less well described. In terms of liver toxicity, a strong correlation has been found for chlorpromazine-induced liver toxicity and reduced sulphoxidation capacity, although specific SULT isoforms are not identified (140). It is hypothesized that reduced SULT-mediated detoxification of chlorpromazine propels metabolism towards CYP-mediated bioactivation.

(29)

Page | 28

scavenge CRMs by enzymatic conjugation with GSH. Most GSTs are strongly genetically determined. GSTM1 and GSTT1 null genotypes are highly abundant and these are therefore well studied in terms of drug toxicity (74,135). An association study in a large cohort (154 DILI patients from various drugs and 250 healthy controls) revealed a strong association between a combined GSTT1- and GSTM1-null genotype and antibacterial drug- or NSAID-induced DILI (131,138). Furthermore, in smaller cohorts it was shown that the double null mutation of GSTT1 and GSTM1 increases the risk for troglitazone- or tacrine-induced liver toxicity (136,137). GSTM1 activity in detoxification of diclofenac- and troglitazone-derived CRMs was later shown in vitro (76,77).

1.3.3. Genetic variation in phase III transporters

(30)

1.3.4. Non-genetic variation in drug metabolism

Non-genetic host factors include age, gender and underlying liver disease, however their utility in assessing causality of IDILI is highly subjected to confounding factors and not very well supported with clinical data. Furthermore, available data suggest that possible influences of these factors are drug-dependent (159). For example, while older age has been linked to higher risk for isoniazid-induced DILI, valproate-induced DILI has been linked with younger age (154). Nevertheless, the idiosyncratic nature of these reactions make it likely that nongenetic factors contribute to a patient’s susceptibility. It is well established that liver enzyme and transporter expression/activity is affected by a wide range of nongenetic factors including the ones mentioned before but also more complex factors such as dietary or polypharmacy (160–164). Illustratively, a recent review by the FDA showed that of all new drug entities approved in 2013, 77% and 29% showed in vitro CYP inhibition or induction, respectively (165). A case study in 2006 suggested simvastatin-induced DILI as a result of CYP3A4 inhibition by amiodarone (166). Cholestatic liver toxicity arising from flucloxacillin, sulindac, terbinafine and bosentan treatment has been linked to BSEP inhibition (131). Non-drug related factors can also interfere with enzyme activity or expression, for example CYP1A induction by broiled beef, cabbage or caffeine; inhibition of CYP2E1 and induction of UGT by watercress; CYP2E1 induction by alcohol; and CYP3A4 induction by St. John’s wort (164,167). As such, obesity is associated with a higher risk for halothane-induced liver toxicity, alcohol abuse has been linked to APAP and isoniazid related DILI and chronic hepatitis B and C has been suggested to increase the risk for isoniazid-induced DILI (154). An exploratory topic in the field of nongenetic inter-individual differences is the modification of gene expression by microRNA or gene methylation (epigenetics) (145,168). Research has indicated significant contributions in gene expression. For example, GSTP1 expression was shown to be silenced by epigenetic regulation in many cancer subtypes (55). The role of these regulatory mechanisms in susceptibility for IDILI still remains to be investigated.

1.4. In vitro models to study effects of hepatic metabolism on drug

toxicity

(31)

Page | 30

hepatotoxic effects cannot be reproduced (169). Moreover, in the search for biomarkers for hepatotoxicity it has been argued that the difference between in vitro models and humans may be less than the difference between in vivo models and humans (170). In this context, it is important to realize that most test systems are designed to detect acute intrinsic hepatotoxicity. Moreover, screening for risk factors for idiosyncratic hepatotoxicity, such as formation of CRMs or BSEP inhibition, may lead to attrition of ‘good’ drug candidates for which these factors do not lead to hepatotoxicity in the clinic. An improved understanding of the underlying mechanisms of idiosyncratic hepatotoxicity is important to develop more targeted in vitro test systems. Current in vitro systems, in particular cellular models, may help to elucidate the mechanisms of drug induced hepatotoxicity (171).

1.4.1. Non-cellular in vitro systems

(32)

enzymes in metabolism of ADR-associated drugs. The few available examples include effects of polymorphisms in GSTP1 on detoxification CRMs derived from acetaminophen, diclofenac and clozapine, and effects of polymorphisms in UGT2B7 and UGT1A9 in glucuronidation of flurbiprofen (179,180).

Table 2. Examples of endpoints investigated using cellular models. Adjusted from Atienzar et al., 2016 and Vernetti et al., 2017 (181,182). Many approaches have been described to improve the predictive power more, e.g. by including cytokines or by means of co-culture with non-parenchymal cells like Kupffer cells.

Cell model Endpoints assessed

Primary human hepatocytes (plus cytokines)

 High content screening of cell viability  Biliary efflux inhibition

 Covalent binding of radiolabeled compounds to proteins  Metabolite ID

 Pharmacokinetics  Toxicity testing  Mechanisms of toxicity  Liver clearance

HepG2 cells  High content screening of cell viability  Mitochondrial injury

 Toxicity testing  Mechanisms of toxicity

 Stress response pathway identification (e.g. ToxTracker®)  Induction

HepaRG cells  High content screening of cell viability  Bile canaliculi dysfunction

 Intrahepatic cholestasis  Toxicity testing  Mechanisms of toxicity  Steatosis  Induction  Metabolites ID Hepatic cell lines expressing

human CYPs   Toxicity testing Metabolite ID

1.4.2. Cell-based in vitro systems

(33)

Page | 32

1.4.2.1. Primary human hepatocytes

(34)

assays in PHH suspension incubations only identified 5 out of 12 tested DILI-associated drugs while sandwich monolayer cultures correctly identified 11 out of 12 drugs (197).

Figure 13. Correlation of 4’-hydroxyaceclofenac formation in PHH and the amount of 4’-hydroxyaceclofenac excreted in urine, both derived from the same donors. Empty symbols are excluded from the linear regression of in vitro and in vivo experimental data. Confidence and prediction intervals are represented by dashed lines. The correlation is significant at the 0.01 level (Spearman’s correlation ρ=0.950). Adapted from Ponsoda et al., 2001 (188).

1.4.2.2. Hepatic cell lines

Hepatic cell lines are isolated from human hepatomas and have the main advantages of growing continuously, a relatively stable phenotype and almost unlimited lifespan. Consequently, culturing of cell lines is relatively simple and easily standardized, and availability is not limited. Moreover, hepatocyte-like features are partly retained. Cell lines are therefore usually the model of choice for screening of perturbation of hepatocellular functions or hepatotoxicity (171,198).

(35)

Page | 34

Figure 14. Activation of stress response reporters for oxidative stress (Srxn1), DNA damage (p21) and ER stress (Chop and BiP) by DILI test compounds and references. Each column represents a 24 hour time course and response magnitude is displayed as color intensity according to the legend (right). Bacterial artificial chromosome (BAC)-GFP HepG2 reporter cells are treated with selected DILI compounds at concentrations (μM) as listed in each row. Transporter activation strongly correlated with transcriptional responses as observed in cryopreserved PHH not shown in figure). Abbreviations: BSO, buthionine sulfoximine; APAP, acetaminophen; DEM, diethylmaleate; BHA, butylated hydroxyanisole. Adapted from Wink et at., 2017 (206).

(36)

activity can be included in HepG2-based assays by combined use of a bioactivating system (e.g. S9 fraction or liver microsomes) or by transfection/transduction of drug metabolizing enzymes (176,208–210). As such, it was shown that diclofenac toxicity in HepG2 cells is increased by inclusion of rat liver S9 mix or human liver microsomes (208,211). Similarly, toxicity of diclofenac in HepG2 was aggravated in HepG2 cells transfected with adenoviruses encoding CYP1A2, CYP2D6, CYP2C9, CYP2C19 and CYP3A4 (176).

In recent years, use of HepG2 cells is increasingly replaced by the use of other cell systems, in particular HepaRG cells, which retain the advantages of dividing in culture but overcome the limitations of HepG2 in terms of loss of hepatic functions (171,203). HepaRG cells differentiate in vitro in hepatocyte-like and biliary cells, in a ratio of 1:1 (212). Levels of most phase I and phase II enzymes in HepaRG are significantly higher compared to HepG2, and transporter levels are comparable to freshly isolated PHH (203,213). Important exceptions are for example CYP2D6 and CYP2E1 (106). Recent work describing global proteomic analysis of hepatic cell lines showed that protein expression profiles in HepaRG cells are closer to PHH compared to HepG2 (214). Besides inclusion of certain drug metabolizing enzymes, a major additional advantage of HepaRG cells is the possibility for long-term (up to 4 weeks) toxicity studies (215). HepaRG cells have been used for assessment of different mechanisms in DILI-related research, such as the study of cholestasis, the effect of inflammatory stress on cytotoxicity or gene expression, the effect of impairment of BSEP on drug-induced cytotoxicity and the effect of metabolism on stress responses or apoptosis (216–218). As a result of the more liver-like phenotype HepaRG cells are expected to be more sensitive than HepG2 cells in discriminating DILI-associated drugs from ‘safe’ drugs. Illustratively, Wu et al. recently compared cytotoxicity of DILI-associated drugs in hepatic cell lines. HepaRG cells were most sensitive (Table 3) which could be related to increased oxidative stress, mitochondrial damage and disorders of neutral lipid metabolism in response to drug exposure compared to HepG2 (219). However, recently Rowena et used a slightly different set of DILI- and non-DILI drugs and showed that the distinction between DILI- and non-DILI drugs was better using HepG2 cells than HepaRG cells (220).

(37)

Page | 36

interindividual variability in drug metabolism, which cannot be reflected using hepatic cell lines (106).

Table 3. IC50 (μM) and human serum Cmax (μM) values of DILI-associated drugs and negative control drugs in HepG2 and HepaRG cells. Incubations were performed in triplicate and cells were treated for 24 hours. Cell viability was assessed using the cell counting kit-8 (CCK-8). Abbreviations: FOC, fold of Cmax. Adapted from Wu et al., 2016 (219).

Drugs HepG2 HepaRG Human

IC50 FOC IC50 FOC Cmax

DILI-associated drugs Carbamazepine >1000 >19.67 788.67 15.51 50.85 Etoposide 242.45 59.86 26.33 6.50 4.05 Acetaminophen >1000 >7.58 >1000 >7.58 132 Isoniazid >1000 >12.99 >1000 >12.99 77 Diclofenac sodium >1000 >135.14 631.94 85.40 7.4 Amiodarone hydrochloride 0.43 0.20 148.29 67.40 2.2 Tetracycline hydrochloride >1000 >52.63 587.73 30.93 19 Sodium valproate >1000 >1.45 >1000 >1.45 690 Rifampicin 491.73 49.17 551.02 55.10 10 Danazol 95.10 1285.14 47.14 637.03 0.074 Labetalol hydrochlorde 450.78 186.27 859.22 355.05 2.42

Negative control drugs

Bambuterol hydrochloride >1000 >66667 >1000 >66667 0.015 Buspirone hydrochloride >1000 >219780 >1000 >219780 0.00455 Scopolamine butylbromide >1000 >1469508 >1000 >1469508 0.00068

1.4.2.3. Stem cell-derived hepatocytes

(38)

(225). Taking into account the costs associated with hiPSC-derived hepatocyte culture, it has been argued that these cells are currently not advantageous compared to standard cell lines like HepG2 cells (5). Nevertheless, when generation of mature hiPSC-derived hepatocytes is further improved, it would allow many new approaches in toxicological research. While the above mentioned in vitro models mainly allow investigation of drug-related factors of idiosyncratic toxicity, the use of hiPSC allows generation of hepatocytes from susceptible patients (5). In this context, it was recently shown that donor variation in drug metabolizing capacity is retained in hiPSC derived hepatocytes (Figure 15) (226,227). Another advantageous feature of hiPSC derived hepatocytes is the retention of CYP activity in cultures over time, which also allows long term toxicity studies (226,228). Moreover, hiPSC can be used to generate cells needed for co-culture with PHH or hiPSC-derived hepatocytes, thereby creating a more liver-like environment in vitro and allowing extended incubation times (171). The possibility to genetically modify stem cells, for instance by knock-in or knock-out of genes of interest or introduction of reporter constructs, will enhance their potential enormously.

Figure 15. Correlation of CYP1A1 (A), CYP2C9 (B) and CYP3A4 (C) activity in hiPSC-derived hepatocytes (x-axis) with the activity observed in their parental PHHs. Adapted from Takayama et al., 2014 (226).

1.4.2.4. Complex models

(39)

Page | 38

natural liver microenvironment. Phase I and II enzyme expression decreases slightly over time, however much slower compared to cultured primary human hepatocytes. However, the acquisition of (healthy) human liver tissue is scare and cryopreserved liver slices are not yet commercially available (229).

In order to allow high throughput screening of drugs, efforts are being made to add levels of complexity to conventional cellular models. For example, Cosgrove et al. included an inflammatory background in vitro, by co-incubating primary hepatocytes or HepG2 cells with hepatotoxicants with mixtures containing LPS and cytokines. Cytokine synergy was found for 19% of the known hepatotoxic drugs screened, while 3% synergy was found for safe drugs (230). Co-incubation with LPS or cytokines is therefore especially useful in the investigation of inflammation-associated idiosyncratic hepatotoxicants.

Much recent research has focused on culturing of hepatic cells in more complex configurations. Examples include co-cultures of PHH with non-parenchymal cells (e.g. fibroblasts) which help maintaining hepatic functions over longer periods of time and can therefore be used for repeated dose studies mimicking chronic exposure in vivo (106,231,232). Alternatively, the use of 3D (co-)cultures of PHH or hepatic cell lines is being explored, which also stabilizes function over time and results in improvement of morphology and functionality towards more in vivo like features (197,212,233–236). Very recently, the concept of Organs on a Chip is being developed which allows study of organoid interaction in 3D (237–239). These in vitro models have a high level of technical and biological complexity, and although still in its infancy, it is reasonable to assume that in future refined models will result in better sensitivity and specificity compared to the traditional models. However, it may take many years for the next generation of 3D models to be routinely used (5,171,181).

1.5. From in vitro to the susceptible patient

(40)

clinic. In addition, in contrast to pharmacokinetic modeling, the (diverse) mechanisms of (I)DILI toxicity are not well understood. A contributing limitation of interpretation of in vitro data is the lack of harmonization and description of cellular functionalities including drug metabolism (181). When experimental conditions are more standardized, it would aid in understanding the performance of current models. Importantly, a better understanding of current in vitro models will contribute to the discussion of the positioning of in vitro models within the drug discovery phase, and as mentioned in the introduction of this chapter, “testing the right things at the right time to make a diligent decision instead of testing everything early” may be a better approach (6,181). The emergence of complex in vitro models offer many opportunities which cannot be addressed by conventional models. In particular the possibility of long-term dosing, mimicking chronic in vivo exposure to parent or (low levels of) CRMs, may greatly contribute to better risk assessment. In risk assessment of (I)DILI individualized in vitro models may be required (242). Implementation of host-specific factors, such as polymorphisms of drug metabolizing enzymes or transporters or sensitization by inflammation, can accelerate the mechanistic understanding of patient susceptibility. Drug specific factors are well investigated using current models, however for better risk assessment it is essential to also focus on host-specific factors (171).

It should be noted that current in vitro models are not only limited in predictive value, information derived from the readout technologies is often limited as well. Since DILI-related drugs are expected to induce subtle cellular changes over a period of time which eventually will lead to DILI, simple readouts such as cytotoxicity assays overlook these distinct cellular effects. Newer methods such as toxicogenomics can improve these readouts (171).

(41)

Page | 40

regimen in interpretation of data from a battery of in vitro assays has been applied in DILI research by a number of groups, leading to a higher sensitivity and selectivity in distinguishing DILI from non-DILI drugs (100,243–246).

1.5.1. The MIP-DILI consortium

Pharma, SMEs and academia have joined forces in the search for biomarkers and better predictive systems of DILI in a number of consortia. One of these is the IMI European Consortium MIP-DILI, partnering 26 EFPIAs, universities/non-profit groups and SMEs. The primary aim of this consortium was to improve the understanding of DILI by refining the panel of test systems in prediction of DILI risk, specifically by 1) evaluation and harmonization of existing and novel models, 2) evaluation of biomarkers of cell injury and 3) mechanistic studies complemented by mathematical modeling approaches (Figure 16).

Figure 16. Schematic representation of the strategy within MIP-DILI to evaluate current models in risk assessment of DILI (right) and potential and selected training compounds within the MIP-DILI consortium (left). Compounds in the right panel include potential training compounds (blue), selected training compounds (green), negative controls (white), and still to be decided compounds (light green). Mechanisms of DILI in man are indicated by numbers: 1, mitochondrial; 2, reactive metabolites; 3, lysosomal impairment; 4, BSEP inhibition; 5, immune-mediated. Adapted from Dragovic et al., 2016 (247).

(42)

work described in this thesis is part of the MIP-DILI consortium, supported by the Innovative Medicines Initiative Grant Agreement number 115336.

1.6. Scope and objectives

Prospective identification of susceptible patients for (I)DILI remains a big challenge. The emphasis is put on cellular in vitro models as these are, in terms of hepatotoxicity biomarkers, closer to the human situation compared to in vivo animal models (170). Current in vitro models are however not well defined in terms of (I)DILI prediction, partly by the lack of harmonization among protocols and missing descriptions of essential factors that contribute to the outcome of readouts, such as variability in phase I and phase II enzyme activity.

The main objective of this thesis was to research the importance of inclusion of both phase I and phase II activity in interpretation and/or outcome of in vitro assays. During this project this aim subdivided into two research lines.

Cellular models often include multiple drug metabolizing enzymes, of which each has the potential to significantly alter the outcome of the experimental readout. Moreover, enzyme activity can vary greatly between cellular models. Various factors attribute to this variability, including differences in experimental conditions, cell line sources or donor characteristics. Characterization of cellular models in terms of phase I and phase II activity is therefore important for proper interpretation of the results. Therefore the first aim of this thesis was:

1) To assess the impact of primary human hepatocyte culture conditions on phase I and/or phase II (iso)enzyme activity and/or on donor variability (chapters 2-3).

Activities of drug metabolizing enzymes are also highly variable in the human population. Because of the key role of enzymes in bioactivation or inactivation of drugs, altered enzyme activity may contribute to the risk of developing ADRs such as (I)DILI. The second aim of this thesis was therefore:

2) To investigate variability of drug metabolizing enzyme activity as a risk factor for increased exposure to CRMs and resulting drug (cyto)toxicity (chapters 4-7).

(43)

Page | 42

Variation of phase I, II and III drug metabolism and transport as a risk factor for (I)DILI is discussed next. Furthermore, in vitro models which are currently used in the search for underlying mechanisms of (I)DILI are discussed, with the focus on drug metabolism and possibilities of incorporation of drug- and host dependent factors in (I)DILI. Finally, the translation of current in vitro models to the susceptible patient is reviewed.

The first research line is examined in chapter 2, two primary human hepatocyte culture conditions, monolayer and suspension cultures, are compared in terms of metabolic activity with three DILI drugs. Effects of culture conditions were different for phase I and phase II activity, and showed high inter-donor variability. To explore this further, UGT and CYP activities in primary human hepatocytes and respective human liver microsomes are characterized in chapter 3. To this end, a method was developed to characterize glucuronidation activity of the six major drug UGT isoforms in intact cells. We confirmed the donor-dependent effects of culture condition on phase I and II activity, and pinpointed culturing effects on the specific CYP and UGT isoforms.

(44)

(45)

Page | 44

References

(1) Nakanishi, T.; Tamai, I. Interaction of Drug or Food with Drug Transporters in Intestine and Liver. Curr. Drug Metab., 2015, 16, 753–764.

(2) Poulin, P.; Haddad, S. Toward a New Paradigm for the Efficient in Vitro-in Vivo Extrapolation of Metabolic Clearance in Humans from Hepatocyte Data. J. Pharm. Sci., 2013, 102, 3239–3251.

(3) van de Waterbeemd, H.; Gifford, E. ADMET in Silico Modelling: Towards Prediction Paradise? Nat. Rev. Drug Discov., 2003, 2, 192–204.

(4) Humphreys, W.G.; Will, Y.; Guengerich, F.P. Toxicology Strategies for Drug Discovery - Present and Future: Introduction. Chem. Res. Toxicol., 2016, 29, 437.

(5) Blomme, E.A.G.; Will, Y. Toxicology Strategies for Drug Discovery: Present and Future. Chem. Res. Toxicol., 2016, 29, 473–504.

(6) Hornberg, J.J.; Laursen, M.; Brenden, N.; Persson, M.; Thougaard, A. V; Toft, D.B.; Mow, T. Exploratory Toxicology as an Integrated Part of Drug Discovery. Part I: Why and How. Drug Discov. Today, 2014, 19, 1131– 1136.

(7) Zhang, D.; Luo, G.; Ding, X.; Lu, C. Preclinical Experimental Models of Drug Metabolism and Disposition in Drug Discovery and Development. Acta Pharm. Sin. B, 2012, 2, 549–561.

(8) Onakpoya, I.J.; Heneghan, C.J.; Aronson, J.K. Post-Marketing Withdrawal of 462 Medicinal Products because of Adverse Drug Reactions: A Systematic Review of the World Literature. BMC Med., 2016, 14, 10.

(9) Lasser, K.E.; Allen, P.D.; Woolhandler, S.J.; Himmelstein, D.U.; Wolfe, S.M.; Bor, D.H. Timing of New Black Box Warnings and Withdrawals for Prescription Medications. JAMA, 2002, 287, 2215–2220.

(10) Siramshetty, V.B.; Nickel, J.; Omieczynski, C.; Gohlke, B.-O.; Drwal, M.N.; Preissner, R. WITHDRAWN--a Resource for Withdrawn and Discontinued Drugs. Nucleic Acids Res., 2016, 44, D1080-6.

(11) Roth, R.A.; Ganey, P.E. Intrinsic versus Idiosyncratic Drug-Induced Hepatotoxicity--Two Villains or One? J. Pharmacol. Exp. Ther., 2010, 332, 692–697.

(12) Björnsson, E.S. Drug-Induced Liver Injury: An Overview over the Most Critical Compounds. Arch. Toxicol., 2015, 89, 327–334.

(13) Lammert, C.; Einarsson, S.; Saha, C.; Niklasson, A.; Bjornsson, E.; Chalasani, N. Relationship between Daily Dose of Oral Medications and Idiosyncratic Drug-Induced Liver Injury: Search for Signals. Hepatology, 2008, 47, 2003– 2009.

(14) Uetrecht, J.; Naisbitt, D.J. Idiosyncratic Adverse Drug Reactions: Current Concepts. Pharmacol. Rev., 2013, 65, 779– 808.

(15) Fontana, R.J. Pathogenesis of Idiosyncratic Drug-Induced Liver Injury and Clinical Perspectives. Gastroenterology,

2014, 146, 914–928.

(16) Fontana, R.J.; Seeff, L.B.; Andrade, R.J.; Björnsson, E.; Day, C.P.; Serrano, J.; Hoofnagle, J.H. Standardization of Nomenclature and Causality Assessment in Drug-Induced Liver Injury: Summary of a Clinical Research Workshop. Hepatology, 2010, 52, 730–742.

(17) Lee, W.M.; Squires, R.H.; Nyberg, S.L.; Doo, E.; Hoofnagle, J.H. Acute Liver Failure: Summary of a Workshop. Hepatology, 2008, 47, 1401–1415.

(18) Tailor, A.; Faulkner, L.; Naisbitt, D.J.; Park, B.K. The Chemical, Genetic and Immunological Basis of Idiosyncratic Drug-Induced Liver Injury. Hum. Exp. Toxicol., 2015, 34, 1310–1317.

(19) Watkins, P.B.; Seeff, L.B. Drug-Induced Liver Injury: Summary of a Single Topic Clinical Research Conference. Hepatology, 2006, 43, 618–631.

(20) Li, A.P. Accurate Prediction of Human Drug Toxicity: A Major Challenge in Drug Development. Chem. Biol. Interact., 2004, 150, 3–7.

(21) Daly, A.K.; Aithal, G.P.; Leathart, J.B.S.; Swainsbury, R.A.; Dang, T.S.; Day, C.P. Genetic Susceptibility to Diclofenac-Induced Hepatotoxicity: Contribution of UGT2B7, CYP2C8, and ABCC2 Genotypes. Gastroenterology,

2007, 132, 272–281.

(22) Licata, A. Adverse Drug Reactions and Organ Damage: The Liver. Eur. J. Intern. Med., 2016, 28, 9–16. (23) Uetrecht, J. Idiosyncratic Drug Reactions: Past, Present, and Future. Chem. Res. Toxicol., 2008, 21, 84–92.

(24) Amacher, D.E. The Primary Role of Hepatic Metabolism in Idiosyncratic Drug-Induced Liver Injury. Expert Opin. Drug Metab. Toxicol., 2012, 8, 335–347.

(25) Gunaratna, C. Drug Metabolism and Pharmacokinetics in Drug Discovery: A Primer for Bioanalytical Chemists, Part I. Curr Sep, 2000, 19, 17–23.

(26) Rendic, S.; Guengerich, F.P. Survey of Human Oxidoreductases and Cytochrome P450 Enzymes Involved in the Metabolism of Xenobiotic and Natural Chemicals. Chem. Res. Toxicol., 2015, 28, 38–42.

(46)

(28) Guengerich, F.P. Common and Uncommon Cytochrome P450 Reactions Related to Metabolism and Chemical Toxicity. Chem. Res. Toxicol., 2001, 14, 611–650.

(29) Neve, E.P. a; Ingelman-Sundberg, M. Intracellular Transport and Localization of Microsomal Cytochrome P450. Anal. Bioanal. Chem., 2008, 392, 1075–1084.

(30) Cribb, A.E.; Peyrou, M.; Muruganandan, S.; Schneider, L. The Endoplasmic Reticulum in Xenobiotic Toxicity. Drug Metab. Rev., 2005, 37, 405–442.

(31) Zhou, L.; Erickson, R.R.; Hardwick, J.P.; Park, S.S.; Wrighton, S. a; Holtzman, J.L. Catalysis of the Cysteine Conjugation and Protein Binding of Acetaminophen by Microsomes from a Human Lymphoblast Line Transfected with the cDNAs of Various Forms of Human Cytochrome P450. J. Pharmacol. Exp. Ther., 1997, 281, 785–790.

(32) Ingelman-Sundberg, M. Human Drug Metabolising Cytochrome P450 Enzymes: Properties and Polymorphisms. Naunyn. Schmiedebergs. Arch. Pharmacol., 2004, 369, 89–104.

(33) Ding, X.; Kaminsky, L.S. Human Extrahepatic Cytochromes P450: Function in Xenobiotic Metabolism and Tissue-Selective Chemical Toxicity in the Respiratory and Gastrointestinal Tracts. Annu. Rev. Pharmacol. Toxicol.,

2003, 43, 149–173.

(34) Capoferri, L.; Verkade-Vreeker, M.C.A.; Buitenhuis, D.; Commandeur, J.N.M.; Pastor, M.; Vermeulen, N.P.E.; Geerke, D.P. Linear Interaction Energy Based Prediction of Cytochrome P450 1A2 Binding Affinities with Reliability Estimation. PLoS One, 2015, 10, e0142232.

(35) Zhao, Q.; Yang, X.L.; Holtzclaw, W.D.; Talalay, P. Unexpected Genetic and Structural Relationships of a Long-Forgotten Flavoenzyme to NAD(P)H:quinone Reductase (DT-Diaphorase). Proc. Natl. Acad. Sci. U. S. A., 1997, 94, 1669–1674.

(36) Jaiswal, A.K. Human NAD(P)H:quinone oxidoreductase2. Gene Structure, Activity, and Tissue-Specific Expression. J. Biol. Chem., 1994, 269, 14502–14508.

(37) Jaiswal, A.K.; Burnett, P.; Adesnik, M.; McBride, O.W. Nucleotide and Deduced Amino Acid Sequence of a Human cDNA (NQO2) Corresponding to a Second Member of the NAD(P)H:quinone Oxidoreductase Gene Family. Extensive Polymorphism at the NQO2 Gene Locus on Chromosome 6. Biochemistry, 1990, 29, 1899–1906. (38) Fu, Y.; Buryanovskyy, L.; Zhang, Z. Quinone Reductase 2 Is a Catechol Quinone Reductase. J. Biol. Chem., 2008,

283, 23829–23835.

(39) Vredenburg, G.; Elias, N.S.; Venkataraman, H.; Hendriks, D.F.G.; Vermeulen, N.P.E.; Commandeur, J.N.M.; Vos, J.C. Human NAD(P)H:quinone Oxidoreductase 1 (NQO1)-Mediated Inactivation of Reactive Quinoneimine Metabolites of Diclofenac and Mefenamic Acid. Chem. Res. Toxicol., 2014, 27, 576–586.

(40) Moffit, J.S.; Aleksunes, L.M.; Kardas, M.J.; Slitt, A.L.; Klaassen, C.D.; Manautou, J.E. Role of NAD(P)H:quinone Oxidoreductase 1 in Clofibrate-Mediated Hepatoprotection from Acetaminophen. Toxicology, 2007, 230, 197–206. (41) Powis, G.; See, K.L.; Santone, K.S.; Melder, D.C.; Hodnett, E.M. Quinoneimines as Substrates for Quinone

Reductase (NAD(P)H : (Quinone-Acceptor)oxidoreductase) and the Effect of Dicumarol on Their Cytotoxicity. Biochem. Pharmacol., 1987, 36, 2473–2479.

(42) Iskander, K.; Li, J.; Han, S.; Zheng, B.; Jaiswal, A.K. NQO1 and NQO2 Regulation of Humoral Immunity and Autoimmunity. J. Biol. Chem., 2006, 281, 30917–30924.

(43) Miettinen, T.P.; Björklund, M. NQO2 Is a Reactive Oxygen Species Generating off-Target for Acetaminophen. Mol. Pharm., 2014, 11, 4395–4404.

(44) Kumar, G.N.; Surapaneni, S. Role of Drug Metabolism in Drug Discovery and Development. Med. Res. Rev., 2001, 21, 397–411.

(45) Oda, S.; Fukami, T.; Yokoi, T.; Nakajima, M. A Comprehensive Review of UDP-Glucuronosyltransferase and Esterases for Drug Development. Drug Metab. Pharmacokinet., 2015, 30, 30–51.

(46) Achour, B.; Rostami-Hodjegan, A.; Barber, J. Protein Expression of Various Hepatic Uridine 5’-diphosphate Glucuronosyltransferase (UGT) Enzymes and Their Intercorrelations: A Meta-Analysis. Biopharm. Drug Dispos.,

2014, 35, 353–361.

(47) Achour, B.; Barber, J.; Rostami-Hodjegan, A. Expression of Hepatic Drug-Metabolizing Cytochrome p450 Enzymes and Their Intercorrelations: A Meta-Analysis. Drug Metab. Dispos., 2014, 42, 1349–1356.

(48) Riches, Z.; Stanley, E.L.; Bloomer, J.C.; Coughtrie, M.W.H. Quantitative Evaluation of the Expression and Activity of Five Major Sulfotransferases (SULTs) in Human Tissues: The SULT “pie”. Drug Metab. Dispos., 2009, 37, 2255–2261.

(49) Almazroo, O.A.; Miah, M.K.; Venkataramanan, R. Drug Metabolism in the Liver. Clin. Liver Dis., 2017, 21, 1–20. (50) Jancova, P.; Anzenbacher, P.; Anzenbacherova, E. Phase II Drug Metabolizing Enzymes. Biomed. Pap. Med. Fac.

Univ. Palacky. Olomouc. Czech. Repub., 2010, 154, 103–116.

(51) Di, L. The Role of Drug Metabolizing Enzymes in Clearance. Expert Opin. Drug Metab. Toxicol., 2014, 10, 379–393. (52) Court, M.H. Interindividual Variability in Hepatic Drug Glucuronidation: Studies into the Role of Age, Sex,

(47)

Page | 46

Rev., 2010, 42, 209–224.

(53) Evans, W.E.; Relling, M. V. Pharmacogenomics: Translating Functional Genomics into Rational Therapeutics. Science, 1999, 286, 487–491.

(54) James, M.O.; Ambadapadi, S. Interactions of Cytosolic Sulfotransferases with Xenobiotics. Drug Metab. Rev., 2013, 45, 401–414.

(55) Schnekenburger, M.; Karius, T.; Diederich, M. Regulation of Epigenetic Traits of the Glutathione S-Transferase P1 Gene: From Detoxification toward Cancer Prevention and Diagnosis. Front. Pharmacol., 2014, 5, 170.

(56) Mazerska, Z.; Mróz, A.; Pawłowska, M.; Augustin, E. The Role of Glucuronidation in Drug Resistance. Pharmacol. Ther., 2016, 159, 35–55.

(57) Tukey, R.H.; Strassburg, C.P. Human UDP-Glucuronosyltransferases: Metabolism, Expression, and Disease. Annu. Rev. Pharmacol. Toxicol., 2000, 40, 581–616.

(58) Burchell, B.; Lockley, D.J.; Staines, A.; Uesawa, Y.; Coughtrie, M.W.H. Substrate Specificity of Human Hepatic Udp-Glucuronosyltransferases. Methods Enzymol., 2005, 400, 46–57.

(59) Manevski, N.; Kurkela, M.; H??glund, C.; Mauriala, T.; Court, M.H.; Yli-Kauhaluoma, J.; Finel, M. Glucuronidation of Psilocin and 4-Hydroxyindole by the Human UDP-Glucuronosyltransferases. Drug Metab. Dispos., 2010, 38, 386–395.

(60) Patten, C.J. New Technologies for Assessing UDP-Glucuronosyltransferase (UGT) Metabolism in Drug Discovery and Development. Drug Discov. Today. Technol., 2006, 3, 73–78.

(61) Fisher, M.B.; Paine, M.F.; Strelevitz, T.J.; Wrighton, S.A. The Role of Hepatic and Extrahepatic UDP-Glucuronosyltransferases in Human Drug Metabolism. Drug Metab. Rev., 33, 273–297.

(62) Kiang, T.K.L.; Ensom, M.H.H.; Chang, T.K.H. UDP-Glucuronosyltransferases and Clinical Drug-Drug Interactions. Pharmacol. Ther., 2005, 106, 97–132.

(63) Williams, J.A.; Hyland, R.; Jones, B.C.; Smith, D.A.; Hurst, S.; Goosen, T.C.; Peterkin, V.; Koup, J.R.; Ball, S.E. Drug-Drug Interactions for UDP-Glucuronosyltransferase Substrates: A Pharmacokinetic Explanation for Typically Observed Low Exposure (AUC 1/AUC) Ratios. Drug Metab. Dispos., 2004, 32, 1201–1208.

(64) Serpe, L.; Canaparo, R.; Scordo, M.G.; Spina, E. Pharmacogenetics of Drug-Metabolizing Enzymes in Italian Populations. Drug Metab. Pers. Ther., 2015, 30, 107–120.

(65) James, M.O. Enzyme Kinetics of Conjugating Enzymes: PAPS Sulfotransferase. Methods Mol. Biol., 2014, 1113, 187–201.

(66) Chen, B.-H.; Wang, C.-C.; Hou, Y.-H.; Mao, Y.-C.; Yang, Y.-S. Mechanism of Sulfotransferase Pharmacogenetics in Altered Xenobiotic Metabolism. Expert Opin. Drug Metab. Toxicol., 2015, 11, 1053–1071.

(67) Glatt, H.; Engelke, C.E..; Pabel, U.; Teubner, W.; Jones, A.L.; Coughtrie, M.W..; Andrae, U.; Falany, C.N.; Meinl, W. Sulfotransferases: Genetics and Role in Toxicology. Toxicol. Lett., 2000, 112–113, 341–348.

(68) Blanchard, R.L.; Freimuth, R.R.; Buck, J.; Weinshilboum, R.M.; Coughtrie, M.W.H. A Proposed Nomenclature System for the Cytosolic Sulfotransferase (SULT) Superfamily. Pharmacogenetics, 2004, 14, 199–211.

(69) Freimuth, R.R.; Wiepert, M.; Chute, C.G.; Wieben, E.D.; Weinshilboum, R.M. Human Cytosolic Sulfotransferase Database Mining: Identification of Seven Novel Genes and Pseudogenes. Pharmacogenomics J., 2004, 4, 54–65. (70) Leyh, T.S.; Cook, I.; Wang, T. Structure, Dynamics and Selectivity in the Sulfotransferase Family. Drug Metab.

Rev., 2013, 45, 423–430.

(71) Ribas, V.; García-Ruiz, C.; Fernández-Checa, J.C. Glutathione and Mitochondria. Front. Pharmacol., 2014, 5, 151. (72) Raza, H. Dual Localization of Glutathione S-Transferase in the Cytosol and Mitochondria: Implications in

Oxidative Stress, Toxicity and Disease. FEBS J., 2011, 278, 4243–4251.

(73) Hayes, J.D.; Flanagan, J.U.; Jowsey, I.R. GLUTATHIONE TRANSFERASES. Annu. Rev. Pharmacol. Toxicol., 2005, 45, 51–88.

(74) Hayes, J.D.; Strange, R.C. Glutathione S-Transferase Polymorphisms and Their Biological Consequences. Pharmacology, 2000, 61, 154–166.

(75) Board, P.G.; Menon, D. Glutathione Transferases, Regulators of Cellular Metabolism and Physiology. Biochim. Biophys. Acta - Gen. Subj., 2013, 1830, 3267–3288.

(76) den Braver, M.W.; Zhang, Y.; Venkataraman, H.; Vermeulen, N.P.E.; Commandeur, J.N.M. Simulation of Interindividual Differences in Inactivation of Reactive Para-Benzoquinone Imine Metabolites of Diclofenac by Glutathione S-Transferases in Human Liver Cytosol. Toxicol. Lett., 2016, 255, 52–62.

(77) Okada, R.; Maeda, K.; Nishiyama, T.; Aoyama, S.; Tozuka, Z.; Hiratsuka, A.; Ikeda, T.; Kusuhara, H.; Sugiyama, Y. Involvement of Different Human Glutathione Transferase Isoforms in the Glutathione Conjugation of Reactive Metabolites of Troglitazone. Drug Metab. Dispos., 2011, 39, 2290–2297.

(78) Dekker, S.J.; Zhang, Y.; Vos, J.C.; Vermeulen, N.P.E.; Commandeur, J.N.M. Different Reactive Metabolites of Nevirapine Require Distinct Glutathione S -Transferase Isoforms for Bioinactivation. Chem. Res. Toxicol., 2016, acs.chemrestox.6b00250.

VU Research Portal