Goodman and Gilman Manual of Pharmacology and Therapeutics

Section I
General Principles

chapter 6
Drug Metabolism


Substances foreign to the body, or xenobiotics, are metabolized by the same enzymatic pathways and transport systems that are used for normal metabolism of dietary constituents. Drugs are considered xenobiotics and most are extensively metabolized in humans. The capacity to metabolize xenobiotics has made development of drugs very time consuming and costly due in large part to:

• Interindividual variations in the capacity of humans to metabolize drugs

• Drug-drug interactions

• Metabolic activation of chemicals to toxic and carcinogenic derivatives

• Species differences in expression of enzymes that metabolize drugs, thereby limiting the use of animal models to predict effects in humans

Most xenobiotics are subjected to 1 or multiple enzymatic pathways that constitute phase 1 oxidation and phase 2 conjugation. Metabolism serves to convert these hydrophobic chemicals into more hydrophilic derivatives that can easily be eliminated from the body through the urine or the bile.

Many drugs are hydrophobic, a property that allows entry through the lipid bilayers into cells where the agents can interact with their target receptors or proteins. This property of hydrophobicity renders drugs difficult to eliminate, because in the absence of metabolism, they accumulate in fat and cellular phospholipid bilayers in cells. The xenobiotic-metabolizing enzymes convert drugs and other xenobiotics into derivatives that are more hydrophilic and thus easily eliminated through excretion into the aqueous compartments of the tissues.

Drug metabolism that leads to elimination also plays a major role in diminishing the biological activity of a drug. For example, (S)-phenytoin, an anticonvulsant used in the treatment of epilepsy, is virtually insoluble in water. Metabolism by the phase 1 cytochromes P450 (CYPs) followed by phase 2 uridine diphosphate-glucuronosyltransferases (UGTs) produces a metabolite that is highly water soluble and readily eliminated from the body (Figure 6–1). Metabolism also terminates the biological activity of the drug.


Figure 6–1 Metabolism of phenytoin by phase 1 cytochrome P450 (CYP) and phase 2 uridine diphosphate-glucuronosyltransferase (UGT). CYP facilitates 4-hydroxylation of phenytoin. The hydroxy group serves as a substrate for UGT that conjugates a molecule of glucuronic acid (in green) using UDP-glucuronic acid (UDP-GA) as a cofactor. This converts a very hydrophobic molecule to a larger hydrophilic derivative that is eliminated via the bile.

Paradoxically, these drug metabolizing enzymes can also convert certain chemicals to highly reactive, toxic, and carcinogenic metabolites or carcinogens. Depending on the structure of the chemical substrate, xenobiotic-metabolizing enzymes can produce electrophilic metabolites that react with nucleophilic cellular macromolecules such as DNA, RNA, and protein. Reaction of these electrophiles with DNA can sometimes result in cancer through the mutation of genes, such as oncogenes or tumor suppressor genes. This potential for carcinogenic activity makes testing the safety of drug candidates of vital importance, particularly for drugs that will be used chronically.


Xenobiotic metabolizing enzymes are grouped into those that carry out: phase 1 reactions, which include oxidation, reduction, or hydrolytic reactions; and the phase 2 reactions, in which enzymes catalyze the conjugation of the substrate (the phase 1 product) with a second molecule (Table 6–1). Oxidation by phase 1 enzymes adds or exposes a functional group, permitting the products of phase 1 metabolism to serve as substrates for the phase 2 conjugating or synthetic enzymes. While many phase 1 reactions result in the biological inactivation of the drug, phase 2 reactions produce a metabolite with improved water solubility, thereby facilitating drug elimination. The example of phase 1 and phase 2 metabolism of phenytoin is shown in Figure 6–1.

Table 6–1

Xenobiotic Metabolizing Enzymes


The phase 1 enzymes lead to the introduction of functional groups such as –OH, –COOH, –SH, –O–, or NH2The phase 1 oxidation reactions are carried out by CYPs, flavin-containing monooxygenases (FMOs), and epoxide hydrolases (EHs).

CYPs and FMOs comprise superfamilies of enzymes containing multiple genes. The addition of functional groups does little to increase the water solubility of the drug, but can dramatically alter the biological properties of the drug. Reactions carried out by phase 1 enzymes usually lead to the inactivation of a drug. However, in certain instances, metabolism, usually the hydrolysis of an ester or amide linkage, results in bioactivation of a drug. Inactive drugs that undergo metabolism to an active drug are called prodrugs. For example, the antitumor drug cyclophosphamide, is bioactivated to a cell-killing electrophilic derivative (see Chapter 61); clofibrate, used to reduce triglyceride levels, is converted in the cell from an ester to an active acidic metabolite.

Phase 2 enzymes facilitate the elimination of drugs and the inactivation of electrophilic and potentially toxic metabolites produced by oxidation. Phase 2 enzymes include several superfamilies of conjugating enzymes, such as the glutathione-S-transferases (GSTs), UDP-glucuronosyltransferases (UGTs), sulfotransferases (SULTs), N-acetyltransferases (NATs), and methyltransferases (MTs). These conjugation reactions usually require the substrate to have oxygen (hydroxyl or epoxide groups), nitrogen, or sulfur atoms that serve as acceptor sites for a hydrophilic moiety, such as glutathione, glucuronic acid, sulfate, or an acetyl group.

In the case of the UGTs, glucuronic acid is delivered to the functional group, forming a glucuronide metabolite that is more water soluble and is targeted for excretion either in the urine or bile. When the substrate is a drug, these reactions usually convert the original drug to a form that is not able to bind to its target receptor, thus attenuating the biological response to the drug.


Xenobiotic metabolizing enzymes are found in most tissues in the body, with the highest levels located in the GI tract (liver, small and large intestines). The liver is the major “metabolic clearing house” for both endogenous chemicals (e.g., cholesterol, steroid hormones, fatty acids, and proteins) and xenobiotics. The small intestine plays a crucial role in drug metabolism because drugs that are orally administered are absorbed by the gut and taken to the liver through the portal vein. The xenobiotic-metabolizing enzymes located in the epithelial cells of the GI tract are responsible for the initial metabolic processing of most oral medications. The absorbed drug then enters the portal circulation for its first pass through the liver, where it can undergo significant metabolism. A portion of active drug escapes metabolism in the GI tract and liver and enters the systemic circulation; subsequent passes through the liver result in more metabolism of the parent drug until the agent is eliminated. Thus, drugs that are poorly metabolized remain in the body for longer periods of time (have longer elimination half-lives). Other organs that contain significant xenobiotic-metabolizing enzymes include tissues of the nasal mucosa and lung, which play important roles in the metabolism of drugs that are administered as aerosols.

At the cellular level, the phase 1 CYPs, FMOs, and EHs, and some phase 2 conjugating enzymes, notably the UGTs, are all located in the endoplasmic reticulum of the cell (Figure 6–2). Once subjected to oxidation, drugs can be directly conjugated by the UGTs (in the lumen of the endoplasmic reticulum) or by the cytosolic transferases such as GST and SULT. The metabolites can then be transported out of the cell into the bloodstream. Hepatocytes, which constitute >90% of the cells in the liver, carry out most drug metabolism and produce conjugated substrates that can also be transported through the bile canalicular membrane into the bile, from which they are eliminated into the gut (see Chapter 5).


Figure 6–2 Location of CYPs in the cell. The figure shows increasingly microscopic levels of detail, sequentially expanding the areas within the black boxes. CYPs are embedded in the phospholipid bilayer of the endoplasmic reticulum (ER). Most of the enzyme is located on the cytosolic surface of the ER. A second enzyme, NADPH-cytochrome P450 oxidoreductase, transfers electrons to the CYP where it can, in the presence of O2, oxidize xenobiotic substrates, many of which are hydrophobic and dissolved in the ER. A single NADPH-CYP oxidoreductase species transfers electrons to all CYP isoforms in the ER. Each CYP contains a molecule of iron-protoporphyrin IX that functions to bind and activate O2. Substituents on the porphyrin ring are methyl (M), propionyl (P), and vinyl (V) groups.



The CYPs are a superfamily of enzymes, all of which contain a molecule of heme that is noncovalently bound to the polypeptide chain (see Figure 6–2). The heme iron binds oxygen in the CYP active site, where oxidation of the substrates occurs. The H+ is supplied through the enzyme NADPH-cytochrome P450 oxidoreductase and its cofactor NADPH. Metabolism of a substrate by a CYP consumes 1 molecule of O2 and produces an oxidized substrate and a molecule of water. Depending on the nature of the substrate, the reaction is “uncoupled,” consuming more O2 than substrate metabolized and producing what is called activated oxygen or O2. The O2 is usually converted to water by the enzyme superoxide dismutase. Elevated O2, a reactive oxygen species (ROS), can give rise to oxidative stress that is detrimental to cells and associated with diseases.

Among the diverse reactions carried out by mammalian CYPs are N-dealkylation, O-dealkylation, aromatic hydroxylation, N-oxidation, S-oxidation, deamination, and dehalogenation (Table 6–2). CYPs are involved in the metabolism of dietary and xenobiotic chemicals, in the synthesis of endogenous compounds (e.g., steroids; fatty acid-derived signaling molecules, such as epoxyeicosatrienoic acids), and in the production of bile acids from cholesterol. In contrast to the drug-metabolizing CYPs, the CYPs that catalyze steroid and bile acid synthesis have very specific substrate preferences. For example, the CYP that produces estrogen from testosterone, CYP19 or aromatase, can metabolize only testosterone or androstenedione and does not metabolize xenobiotics. Specific inhibitors for aromatase, such asanastrozole, have been developed for use in the treatment of estrogen-dependent tumors (see Chapters 40 and 60-63). CYPs involved in bile acid production have strict substrate requirements and do not participate in xenobiotic or drug metabolism.

Table 6–2

Major Reactions Involved in Drug Metabolism



The CYPs that carry out xenobiotic metabolism can metabolize structurally diverse chemicals. This is due both to multiple forms of CYPs, to the capacity of a single CYP to metabolize many structurally distinct chemicals, to significant overlapping substrate specificity amongst CYPs, and to the capacity of CYPs to metabolize a single compound at different positions on the molecule. Indeed, CYPs are promiscuous in their capacity to bind and metabolize multiple substrates (see Table 6–2). This property sacrifices metabolic turnover rates; CYPs metabolize substrates at a fraction of the rate of more typical enzymes involved in intermediary metabolism and mitochondrial electron transfer. As a result, drugs have, in general, half-lives in the range of 2-30 h, while endogenous compounds have half-lives of the order of seconds or minutes (e.g., dopamine and insulin).

The extensive overlapping substrate specificities by the CYPs is one of the underlying reasons for the predominance of drug-drug interactions. When 2 coadministered drugs are both metabolized by a single CYP, they compete for binding to the enzyme’s active site. This can result in the inhibition of metabolism of 1 or both of the drugs, leading to elevated plasma levels. If there is a narrow therapeutic index for the drugs, the elevated serum levels may elicit unwanted toxicities. Drug-drug interactions are among the leading causes of adverse drug reactions (ADRs).

THE NAMING OF CYPs. Genome sequencing has revealed the existence of 57 putatively functional genes and 58 pseudogenes in humans. These genes are grouped, based on amino acid sequence similarity, into a superfamily composed of families and subfamilies with increasing sequence similarity. CYPs are named with the root CYP followed by a number designating the family, a letter denoting the subfamily, and another number designating the CYP form. Thus, CYP3A4 is family 3, subfamily A, and gene number 4.

A DOZEN CYPs SUFFICE FOR METABOLISM OF MOST DRUGS. In humans, 12 CYPs (CYP1A1, 1A2, 1B1, 2A6, 2B6, 2C8, 2C9, 2C19, 2D6, 2E1, 3A4, and 3A5) are known to be important for metabolism of xenobiotics. The liver contains the greatest abundance of xenobiotic-metabolizing CYPs, thus ensuring efficient first-pass metabolism of drugs. CYPs are also expressed throughout the GI tract, and in lower amounts in lung, kidney, and even in the CNS. The most active CYPs for drug metabolism are those in the CYP2C, 2D, and 3A subfamilies. CYP3A4, the most abundantly expressed in liver, is involved in the metabolism of over 50% of clinically used drugs (Figure 6–3A). The CYP1A, 1B, 2A, 2B, and 2E subfamilies are not significantly involved in the metabolism of therapeutic drugs, but they do catalyze the metabolic activation of many protoxins and procarcinogens.


Figure 6–3 The fraction of clinically used drugs metabolized by the major phase 1 and phase 2 enzymes. The relative size of each pie section represents the estimated percentage of drugs metabolized by the major phase 1 (panel A) and phase 2 (panel B) enzymes, based on studies in the literature. In some cases, more than a single enzyme is responsible for metabolism of a single drug. CYP, cytochrome P450; DPYD, dihydropyrimidine dehydrogenase; GST, glutathione-S-transferase; NAT, N-acetyltransferase; SULT, sulfotransferase, TPMT, thiopurine methyltransferase; UGT, UDP-glucuronosyltransferase.

CYPs AND DRUG-DRUG INTERACTIONS. Differences in the rate of metabolism of a drug can be caused by drug interactions. Most commonly, an interaction occurs when 2 drugs (e.g., a statin and a macrolide antibiotic or antifungal agent) are metabolized by the same enzyme and affect each other’s metabolism. Thus, it is important to determine the identity of the CYP that metabolizes a particular drug and to avoid coadministering drugs that are metabolized by the same enzyme. Some drugs can also inhibit CYPs independently of being substrates for a CYP.

For example, the common antifungal agent, ketoconazole (NIZORAL), is a potent inhibitor of CYP3A4 and other CYPs, and coadministration of ketoconazole with an anti-HIV viral protease inhibitor reduces the clearance of the protease inhibitor and increases its plasma concentration and the risk of toxicity. For most drugs, information found on the package insert lists the CYP that metabolizes the drug and determines the potential for drug interactions.

Some drugs are CYP inducers that can increase not only their own rates of metabolism, but also induce metabolism of other coadministered drugs (see below and Figure 6–8).


Figure 6–8 Induction of drug metabolism by nuclear receptor–mediated signal transduction. When a drug such as atorvastatin (the Ligand) enters the cell, it can bind to a nuclear receptor such as the pregnane X receptor (PXR). PXR then forms a complex with the retinoid X receptor (RXR), binds to DNA upstream of target genes, recruits coactivator (which binds to the TATA box binding protein, TBP), and activates transcription by RNA polymerase II (RNAP II). Among PXR target genes are CYP3A4, which can metabolize the atorvastatin and decrease its cellular concentration. Thus, atorvastatin induces its own metabolism. Atorvastatin undergoes both ortho and para hydroxylation.

For example, steroid hormones and herbal products such as St. John’s wort can increase hepatic levels of CYP3A4, thereby increasing the metabolism of many orally administered drugs. Indeed, St. John’s wort can induce hepatic metabolism of the steroid components of birth control pills, rendering the standard dose ineffective in preventing pregnancy.

Drug metabolism can also be influenced by diet.

Components found in grapefruit juice (e.g., naringin, furanocoumarins) are potent inhibitors of CYP3A4, and thus grapefruit juice can increase the bioavailability of certain drugs that are substrates forCYP3A4. Terfenadine, a once popular antihistamine, was removed from the market because its metabolism was inhibited by CYP3A4 substrates such as erythromycin and grapefruit juice. Terfenadine was actually a prodrug that required oxidation by CYP3A4 to its active metabolite, and at high doses the parent compound caused the potentially fatal arrhythmia, torsades de pointes. Thus, as a result ofCYP3A4 inhibition by a coadministered agent, plasma levels of the parent drug could become dangerously elevated, causing ventricular tachycardia in some individuals; this led to terfenadine’s withdrawal from the market. Subsequently, the metabolite was developed as a drug, fexofenadine, which retains the therapeutic properties of the parent compound but avoids the step involving CYP3A4.

In addition, interindividual differences in drug metabolism are significantly influenced by heritable polymorphisms in CYPs.

Several human CYP genes exhibit polymorphisms, including CYP2A6, CYP2C9, CYP2C19, and CYP2D6. The CYP2D6 polymorphism has led to the withdrawal of several clinically used drugs (e.g., debrisoquine and perhexiline) and the cautious use of others that are known CYP2D6 substrates (e.g., encainide and flecainide [anti-arrhythmics], desipramine and nortriptyline [antidepressants], and codeine).


Epoxides are highly reactive electrophiles that can bind to cellular nucleophiles found in protein, RNA, and DNA, resulting in cell toxicity and transformation. Two forms of epoxide hydrolase carry out hydrolysis of epoxides, most of which are produced by CYPs. The soluble epoxide hydrolase (EH) is expressed in the cytosol while the microsomal epoxide hydrolase (mEH) is localized to the membrane of the endoplasmic reticulum. Thus, epoxide hydrolases participate in the deactivation of potentially toxic metabolites generated by CYPs.

The antiepileptic drug carbamazepine is a prodrug that is converted to its pharmacologically active derivative, carbamazepine-10, 11-epoxide by a CYP. This metabolite is efficiently hydrolyzed to a dihydrodiol by mEH, resulting in inactivation of the drug (Figure 6–4). The tranquilizer valnoctamide and anticonvulsant valproic acid inhibit mEH, resulting in clinically significant drug interactions with carbamazepine by causing elevation of the active derivative.


Figure 6–4 Metabolism of carbamazepine by CYP and microsomal epoxide hydrolase (mEH). Carbamazepine is oxidized to the pharmacologically active metabolite carbamazepine-10, 11-epoxide by CYP. The epoxide is converted to a transdihydrodiol by mEH. This metabolite is biologically inactive and can be conjugated by phase 2 enzymes.

The carboxylesterases superfamily catalyze the hydrolysis of ester- and amide-containing chemicals. These enzymes are found in both the endoplasmic reticulum and the cytosol of many cell types and are involved in detoxification or metabolic activation of various drugs, environmental toxicants, and carcinogens. Carboxylesterases also catalyze the activation of prodrugs to their respective free acids. For example, the prodrug and cancer chemotherapeutic agent irinotecan is a camptothecin analog that is bioactivated by plasma and intracellular carboxylesterases to SN-38 (Figure 6–5), a potent inhibitor of topoisomerase 1.


Figure 6–5 Metabolism of irinotecan (CPT-11). The prodrug CPT-11 is initially metabolized by a serum esterase (CES2) to the topoisomerase inhibitor SN-38, which is the active camptothecin analog that slows tumor growth. SN-38 is then subject to glucuronidation, which results in loss of biological activity and facilitates elimination of the SN-38 in the bile.


The FMOs are another superfamily of phase 1 enzymes involved in drug metabolism. Similar to CYPs, the FMOs are expressed at high levels in the liver and are bound to the endoplasmic reticulum. There are 6 families of FMOs, with FMO3 being the most abundant in liver. A genetic deficiency in this enzyme causes the fish-odor syndrome due to a lack of metabolism of trimethylamine N-oxide (TMAO) to trimethylamine (TMA). FMOs are considered minor contributors to drug metabolism and they almost always produce benign metabolites. FMOs are not induced by any of the xenobiotic receptors (see below) or easily inhibited; thus, in contrast to CYPs, FMOs are less involved in drug-drug interactions. In fact, this has been demonstrated by comparing the pathways of metabolism of 2 drugs used in the control of gastric motility, itopride, and cisapride. Itopride is metabolized by FMO3 while cisapride is metabolized by CYP3A4; thus, itopride is less likely to be involved in drug-drug interactions than is cisapride. CYP3A4 participates in drug-drug interactions through induction and inhibition of metabolism, whereas FMO3 is not induced or inhibited by any clinically used drugs. FMOs could be of importance in the development of new drugs. A candidate drug could be designed by introducing a site for FMO oxidation with the knowledge that favorable metabolism and pharmacokinetic properties could be accurately predicted.


The phase 2 conjugating enzymes, synthetic in nature, catalyze reactions that normally terminate the biological activity of drugs, although for drugs like morphine and minoxidil, glucuronide and sulfate conjugates, respectively, are more pharmacologically active than the parent. The contributions of different phase 2 reactions to drug metabolism are shown inFigure 6–3B. Two of the phase 2 reactions, glucuronidation and sulfation, result in the formation of metabolites with a significantly increased hydrophilicity. Characteristic of the phase 2 reactions is the dependency on the catalytic reactions for cofactors, such as UDP-glucuronic acid (UDP-GA) for UDP-glucuronosyltransferases (UGTs) and 3’-phosphoadenosine-5’-phosphosulfate (PAPS) for sulfotransferases (SULTs); these cofactors react with available functional groups on the substrates, reactive functional groups that are often generated by the phase 1 CYPs. All of the phase 2 reactions are carried out in the cytosol of the cell, with the exception of glucuronidation, which occurs on the luminal side of the endoplasmic reticulum.

The catalytic rates of phase 2 reactions are significantly faster than the rates of the CYPs. Thus, if a drug is targeted for phase 1 oxidation through the CYPs, followed by a phase 2 conjugation reaction, usually the rate of elimination will depend on the initial (phase 1) oxidation reaction.

GLUCURONIDATION. UGTs catalyze the transfer of glucuronic acid from the cofactor UDP-glucuronic acid to a substrate to form β-D-glucopyranosiduronic acids (glucuronides), metabolites that are sensitive to cleavage by β-glucuronidase. The generation of glucuronides can be formed through alcoholic and phenolic hydroxyl groups, carboxyl, sulfuryl, and carbonyl moieties, as well as through primary, secondary, and tertiary amine linkages. Examples of glucuronidation reactions are shown in Table 6–2 and Figure 6–5. The structural diversity in the many different types of drugs and xenobiotics that are processed through glucuronidation assures that most clinically efficacious therapeutic agents will be excreted as glucuronides.

The UGTs are expressed in a tissue-specific and often inducible fashion in most human tissues, with the highest concentration found in the GI tract and liver. Glucuronides are excreted into the urine or through active transport processes through the apical surface of the liver hepatocytes into the bile ducts where they are transported to the duodenum for excretion with components of the bile. Most of the bile acids that are conjugated are reabsorbed from the gut back to the liver via enterohepatic recirculation; many drugs that are glucuronidated and excreted in the bile can reenter the circulation by this process.

The expression of UGT1A1 assumes an important role in drug metabolism, because the glucuronidation of bilirubin by UGT1A1 is the rate-limiting step in assuring efficient bilirubin clearance, and this rate can be affected by both genetic variation and competing substrates (drugs). Bilirubin is the breakdown product of heme, 80% of which originates from circulating hemoglobin and 20% from other heme-containing proteins such as the CYPs. Bilirubin is hydrophobic, associates with serum albumin, and must be metabolized further by glucuronidation to assure its elimination. The failure to efficiently metabolize bilirubin by glucuronidation leads to elevated serum levels and a clinical symptom called hyperbilirubinemia or jaundice. There are >50 genetic lesions in the UGT1A1 gene that can lead to inheritable unconjugated hyperbilirubinemia. Crigler-Najjar syndrome type I is diagnosed as a complete lack of bilirubin glucuronidation; Crigler-Najjar syndrome type II is differentiated by the detection of low amounts of bilirubin glucuronides in duodenal secretions. These rare syndromes result from genetic polymorphisms in the UGT1A1 gene, resulting in abolished or highly diminished levels of functional protein.

Gilbert syndrome is a generally benign condition present in up to 10% of the population; it is diagnosed clinically because circulating bilirubin levels are 60-70% higher than those seen in normal subjects. The most common genetic polymorphism associated with Gilbert syndrome is a mutation in the UGT1A1 gene promoter, which leads to reduced expression levels of UGT1A1. Subjects diagnosed with Gilbert syndrome may be predisposed to adverse drug reactions (ADRs) (Table 6–3) that result from a reduced capacity of UGT1A1 to metabolize drugs. If a drug undergoes selective metabolism by UGT1A1, competition for drug metabolism with bilirubin glucuronidation will exist, resulting in pronounced hyperbilirubinemia as well as reduced clearance of the metabolized drug. Gilbert syndrome also alters patient responses to irinotecan (CPT-11). Irinotecan, a prodrug used in chemotherapy of solid tumors (see Chapter 61), is metabolized to its active form, SN-38, by serum carboxylesterases (see Figure 6–5). SN-38, a potent topoisomerase inhibitor, is inactivated by UGT1A1 and excreted in the bile (Figure 6–6). Once in the lumen of the intestine, the SN-38 glucuronide undergoes cleavage by bacterial β-glucuronidase and reenters the circulation through intestinal absorption. Elevated levels of SN-38 in the blood lead to bone marrow toxicities characterized by leukopenia and neutropenia, as well as damage to the intestinal epithelial cells (see Figure 6–6), resulting in acute and life-threatening diarrhea. Patients with Gilbert syndrome who are receiving irinotecan therapy are predisposed to the hematological and GI toxicities resulting from elevated serum levels of SN-38.

Table 6–3

Drug Toxicity and Gilbert’s Syndrome



Figure 6–6 Cellular targets of SN-38 in the blood and intestinal tissues. Excessive accumulation of SN-38 can lead to bone marrow toxicities such as leukopenia and neutropenia, as well as damage to the intestinal epithelium. These toxicities are pronounced in individuals that have reduced capacity to form the SN-38 glucuronide, such as patients with Gilbert syndrome. Note the different body compartments and cell types involved. (Modified with permission from Tukey RH et al. Pharmacogenetics of human UDP-glucuronosyltransferases and irinotecan toxicity. Mol Pharmacol, 2002;62:446–450. Copyright © 2002 The American Society for Pharmacology and Experimental Therapeutics.)

SULFATION. The sulfotransferases (SULTs) are cytosolic and conjugate sulfate derived from 3’-phosphoadenosine-5’-phosphosulfate (PAPS) to the hydroxyl and, less frequently, amine groups of aromatic and aliphatic compounds. In humans, significant fractions of circulating catecholamines, estrogens, iodothyronines, and DHEA exist in the sulfated form. In humans, 13 SULT isoforms have been identified and classified into 4 families [SULT1 (8 members), SULT2 (3 members), SULT4 (1 member), and SULT6 (1 member) families]. SULTs play an important role in normal human homeostasis. For example, SULT2B1b is a predominant form expressed in skin, carrying out the catalysis of cholesterol. Cholesterol sulfate is an essential metabolite in regulating keratinocyte differentiation and skin development. SULT2A1 is very highly expressed in the fetal adrenal gland, where it produces the large quantities of dehydroepiandrosterone sulfate that are required for placental estrogen biosynthesis during the second half of pregnancy. SULT1A3 is highly selective for catecholamines; SULT1E1 sulfates estrogens.

Members SULT1 family are the major forms involved in xenobiotic metabolism, with SULT1A1 being the most important in the liver. It displays extensive diversity in its capacity to catalyze the sulfation of a wide variety of structurally heterogeneous xenobiotics with high affinity. SULT1B1 is similar to SULT1A1, but it is much more abundant in the intestine than the liver. Three SULT1C isoforms exist in humans, but little is known about their substrate specificity. SULT1C enzymes are expressed abundantly in human fetal tissues and decline in abundance in adults. SULT1E catalyzes the sulfation of steroids, and is localized in liver, as well as in hormone-responsive tissues such as the testis, breast, adrenal gland, and placenta. In the upper GI tract, SULT1A3 and SULT1B1 are particularly abundant. Drug metabolism through sulfation often leads to the generation of chemically reactive metabolites, where the sulfate is electron withdrawing and may be heterolytically cleaved, leading to the formation of an electrophilic cation. Examples of the generation by sulfation of a carcinogenic or toxic response in mutagenicity assays have been documented with chemicals derived from the environment or from food mutagens generated from well-cooked meat. Thus, it is important to understand whether genetic linkages can be made by associating known human SULT polymorphisms to cancers related to environmental sources.

GLUTATHIONE CONJUGATION. The glutathione-S-transferases (GSTs) catalyze the transfer of glutathione to reactive electrophiles, a function that protects cellular macromolecules from interacting with electrophiles that contain electrophilic heteroatoms (-O, -N, and -S). The cosubstrate in the reaction is the tripeptide glutathione, which is synthesized from l-glutamic acid, cysteine, and glycine (Figure 6–7). Glutathione exists in the cell as oxidized (GSSG) or reduced (GSH) forms, and the ratio of GSH:GSSG is critical in maintaining a cellular environment in the reduced state. In addition to affecting xenobiotic conjugation with GSH, a severe reduction in GSH content can predispose cells to oxidative damage, a state that has been linked to a number of human health issues.


Figure 6–7 Glutathione (GSH) as a cosubstrate in the conjugation of a drug or xenobiotic (X) by glutathione-S-transferase (GST).

The formation of glutathione conjugates generates a thioether linkage with drug or xenobiotic to the cysteine moiety of the tripeptide. Because the concentration of glutathione in cells is usually very high, typically in the 10 mM range, many drugs and xenobiotics can react nonenzymatically with glutathione. However, the GSTs have been found to occupy up to 10% of the total cellular protein concentration, a property that assures efficient conjugation of glutathione to reactive electrophiles. More than 20 human GSTs have been identified and divided into 2 subfamilies: the cytosolic and the microsomal forms. The cytosolic forms have more importance in the metabolism of drugs and xenobiotics, whereas the microsomal GSTs are important in the endogenous metabolism of leukotrienes and prostaglandins. The high concentration of GSTs also provides the cells with a sink of cytosolic protein that sequesters compounds that are not substrates for glutathione conjugation. The cytosolic pool of GSTs binds steroids, bile acids, bilirubin, cellular hormones, and environmental toxicants, in addition to complexing with other cellular proteins.

The high concentrations of GSH in the cell and the plenitude of GSTs mean that few reactive molecules escape detoxification. However, there is always concern that some reactive intermediates will escape detoxification, and by nature of their electrophilicity, will cause toxicity. The potential for such an occurrence is heightened if GSH is depleted or if a specific form of GST is polymorphic. Reactive therapeutic agents that require large doses for clinical efficacy have the greatest potential to lower cellular GSH levels. Acetaminophen, which is normally metabolized by glucuronidation and sulfation, is also a substrate for oxidative metabolism by CYP2E1 and CYP3A4, which generate the toxic metabolite N-acetyl-p-benzoquinone imine (NAPQI) that, under normal dosing, is readily neutralized through conjugation with GSH. An overdose of acetaminophen can deplete cellular GSH levels, thereby increasing the potential for NAPQI to interact with other cellular components resulting in toxicity and cell death (see Figure 4–4).

The GSTs are polymorphic, and several of the polymorphic forms express a null phenotype; thus, individuals polymorphic at these loci are predisposed to toxicities by agents that are selective substrates for these GSTs. For example, the mutant GSTM1*0 allele is observed in 50% of the Caucasian population and has been linked genetically to human malignancies of the lung, colon, and bladder. Null activity in the GSTT1 gene has been associated with adverse side effects and toxicity in cancer chemotherapy with cytostatic drugs; the toxicities result from insufficient clearance of the drugs by GSH conjugation. Expression of the null genotype can be as high as 60% in Chinese and Korean populations. GST polymorphisms may influence efficacies and severity of adverse side effects of drugs. GSTs activities in cancerous tissues have been linked to the development of drug resistance toward chemotherapeutic agents. See Figures 6–10 and 6–11 in the 12th edition of the parent text for more on this subject.

N-ACETYLATION. The cytosolic N-acetyltransferases (NATs) are responsible for the metabolism of drugs and environmental agents that contain an aromatic amine or hydrazine group. The addition of the acetyl group from the cofactor acetyl-coenzyme A often leads to a metabolite that is less water soluble because the potential ionizable amine is neutralized by the covalent addition of the acetyl group. There are 2 functional NAT genes in humans, NAT1 and NAT2. NATs are among the most polymorphic of all the human xenobiotic drug-metabolizing enzymes. More than 25 allelic variants of NAT1 and NAT2have been characterized, and in individuals in whom acetylation of drugs is compromised, homozygous genotypes for at least 2 variant alleles are required to predispose a patient to lowered drug metabolism. The frequency of the slow acetylation patterns are attributed mostly to the polymorphism in the NAT2 gene. The characterization of an acetylator phenotype in humans was one of the first hereditary traits identified, and was responsible for the development of the field of pharmacogenetics (see Chapter 7). Following the discovery that isonicotinic acid hydrazide (isoniazid, INH) could be used in the cure of tuberculosis, a significant proportion of the patients (5-15%) experienced toxicities. Individuals suffering from the toxic effects of the drug excreted the largest amount of unchanged drug and the least amount of acetylated isoniazid. Pharmacogenetic studies led to the classification of “rapid” and “slow” acetylators, with the “slow” phenotype being associated with toxicity. Characterization of N-acetyltransferase revealed polymorphisms that correspond to the “slow” acetylator phenotype. See Figures 56–3 and 56–4 for details of isoniazid metabolism and NAT2 polymorphisms.

Drug substrates of NATs and their toxicities are listed in Table 6–4. If a drug is known to be metabolized through acetylation, determining an individual’s phenotype can be important in maximizing outcome in subsequent therapy. Several drugs, such as the sulfonamides, that are targets for acetylation have been implicated in idiosyncratic hypersensitivity reactions; in such instances, an appreciation of a patient’s acetylating phenotype is particularly important. Sulfonamides are transformed into hydroxylamines that interact with cellular proteins, generating haptens that can elicit autoimmune responses. Individuals who are slow acetylators are predisposed to drug-induced autoimmune disorders.

Table 6–4

Indications and Unwanted Side Effects of Drug Metabolized by N-Acetyltransferases


Tissue-specific expression patterns of NAT1 and NAT2 have a significant impact on drug metabolism and the potential for toxicity. NAT1 is expressed in most tissues, whereas NAT2 is found predominantly in the liver and GI tract. Both NAT1 and NAT2 can form N-hydroxy–acetylated metabolites from bicyclic aromatic hydrocarbons, a reaction that leads to the nonenzymatic release of the acetyl group and the generation of highly reactive nitrenium ions. Thus, N-hydroxy acetylation is thought to activate certain environmental toxicants. In contrast, direct N-acetylation of bicyclic aromatic amines is stable and leads to detoxification. Individuals who are NAT2 fast acetylators are able to efficiently metabolize and detoxify bicyclic aromatic amines through liver-dependent acetylation. Slow acetylators (NAT2 deficient) accumulate bicyclic aromatic amines, which are metabolized by CYPs to N-OH metabolites that are eliminated in the urine. In bladder epithelium, NAT1 is highly expressed and can catalyze theN-hydroxy acetylation of bicyclic aromatic amines and the formation of the mutagenic nitrenium ion, especially in NAT2-deficient subjects. Slow acetylators are predisposed to bladder cancer if exposed environmentally to bicyclic aromatic amines.

METHYLATION. In humans, drugs and xenobiotics can undergo O-, N-, and S-methylation. Methyltransferases (MTs) are identified by substrate and methyl conjugate. Humans express 3 N-methyltransferases, 1 catechol-O-methyltransferase (COMT), 1 phenol-O-methyltransferase (POMT), 1 thiopurine S-methyltransferase (TPMT), and 1 thiol methyltransferase (TMT). These MTs use S-adenosyl-methionine (SAM; AdoMet) as the methyl donor. With the exception of a signature sequence that is conserved among the MTs, there is limited conservation in sequence, indicating that each MT has evolved to display a unique catalytic function. Although all MTs generate methylated products, the substrate specificity is high.

Nicotinamide N-methyltransferase (NNMT) methylates serotonin and tryptophan and pyridine-containing compounds (e.g., nicotinamide and nicotine). Phenylethanolamine N-methyltransferase (PNMT) methylates norepinephrine to epinephrine; the histamine N-methyltransferase (HNMT) metabolizes drugs containing an imidazole ring (e.g., histamine). COMT methylates neurotransmitters containing a catechol moiety (e.g., dopamine and norepinephrine, methyldopa and drugs of abuse such as ecstasy).

Clinically, the most important MT may be thiopurine S-methyltransferase (TPMT), which catalyzes the S-methylation of aromatic and heterocyclic sulfhydryl compounds, including azathioprine (AZA), 6-mercaptopurine (6-MP), and thioguanine. AZA and 6-MP are used for inflammatory bowel disease (see Chapter 47) and autoimmune disorders such as systemic lupus erythematosus and rheumatoid arthritis. Thioguanine is used in the treatment of acute myeloid leukemia, and 6-MP is used for the treatment of childhood acute lymphoblastic leukemia (see Chapters 61-63). Because TPMT is responsible for the detoxification of 6-MP, a genetic deficiency in TPMT can result in severe toxicities in patients taking these drugs. The toxic side effects arise when a lack of 6-MP methylation by TPMT causes a buildup of 6-MP, resulting in the generation of toxic levels of 6-thioguanine nucleotides (see Figure 47–5). Tests for TPMT activity have made it possible to identify individuals who may be predisposed to the toxic side effects of 6-MP therapy.


Metabolism of drugs normally results in the inactivation of their therapeutic effectiveness and facilitates their elimination. The extent of metabolism can determine the efficacy and toxicity of a drug by controlling its biological t1/2. Among the most serious considerations in the clinical use of drugs are adverse drug responses (ADRs). If a drug is metabolized too quickly, it rapidly loses its therapeutic efficacy. A drug that is metabolized too slowly can accumulate in the bloodstream; the plasma clearance of the drug is decreased and the pharmacokinetic parameter AUC (area under the plasma concentration-time curve; see Figure 2–6) is elevated. An increase in AUC often results when specific xenobiotic-metabolizing enzymes are inhibited by diet or drug interactions.

For example, the consumption of grapefruit juice (which contains CYP3A4 inhibitors naringin and furanocoumarins) can inhibit intestinal CYP3A4, blocking the metabolism and altering the oral bioavailability of many classes of drugs, including, immunosuppressants, antidepressants, antihistamines, statins, and certain antihypertensives. Phenotypic changes in drug metabolism are also observed in individuals that are genetically predisposed to adverse drug reactions because of pharmacogenetic differences in the expression of xenobiotic-metabolizing enzymes (see Chapter 7). See, for example, the discussion of Gilbert syndrome, above (see Figures 6–5 and 6–6).

Nearly every class of therapeutic agent has been reported to initiate an ADR. In the U.S., the annual costs of ADRs have been estimated at >100,000 deaths and $100 billion. About 56% of drugs that are associated with adverse responses are subjected to metabolism by CYPs and UGTs. Because many of the CYPs and UGTs are subject to induction as well as inhibition by drugs, dietary factors, and other environmental agents, these enzymes play an important role in most ADRs. Thus, before a new drug application (NDA) is filed with the FDA, the route of metabolism and the enzymes involved in the metabolism must be known.

Xenobiotics can influence the extent of drug metabolism by activating transcription and expression of genes encoding drug-metabolizing enzymes. Thus, a drug may induce its own metabolism. One potential consequence of this is a decrease in plasma drug concentration over the course of treatment, resulting in loss of efficacy. Many ligands and receptors participate in this way to induce drug metabolism (Table 6–5). A particular receptor, when activated by a ligand, can induce the transcription of a battery of target genes, including CYPs and drug transporters, leading to drug interactions. The aryl hydrocarbon receptor (AHR) is a member of a superfamily of transcription factors. The AHR induces expression of genes encoding CYP1A1, CYP1A2, and CYP1B1, which metabolically activate chemical carcinogens, including environmental contaminants and carcinogens derived from food. Many of these substances are inert unless metabolized by CYPs. Induction of these CYPs by a drug could potentially result in an increase in the toxicity and carcinogenicity of procarcinogens.

Table 6–5

Nuclear Receptors and Ligands That Induce Drug Metabolism


For example, omeprazole, a proton pump inhibitor used to treat gastric and duodenal ulcers (see Chapter 45), is a ligand for the AHR and can induce CYP1A1 and CYP1A2, possibly activating toxins/carcinogens as well as drug-drug interactions in patients receiving agents that are substrates for either of these CYPs.

Another important induction mechanism is caused by type 2 nuclear receptors that are in the same superfamily as the steroid hormone receptors. Figure 6–8 shows the scheme by which a drug may interact with nuclear receptors to induce its own metabolism. Many of these receptors were originally termed “orphan receptors,” because they had no known endogenous ligands. The type 2 nuclear receptors of most importance to drug metabolism and drug therapy include the pregnane X receptor (PXR), constitutive androstane receptor (CAR), and the peroxisome proliferator activated receptors (PPARs).

PXR is activated by a number of drugs including, antibiotics (rifampicin and troleandomycin), Ca2+ channel blockers (nifedipine), statins (mevastatin), antidiabetic drugs (troglitazone), HIV protease inhibitors (ritonavir), and anticancer drugs (paclitaxel). Hyperforin, a component of St. John’s wort, an over-the-counter herbal remedy used for depression, also activates PXR. This activation is thought to be the basis for the increase in failure of oral contraceptives in individuals taking St. John’s wort: Activated PXR induces CYP3A4, which can metabolize steroids found in oral contraceptives. PXR also induces the expression of genes encoding certain drug transporters and phase 2 enzymes including SULTs and UGTs. Thus, PXR facilitates the metabolism and elimination of xenobiotics, including drugs, with notable consequences.

The nuclear receptor CAR was discovered based on its ability to activate genes in the absence of ligand. Steroids such as androstenol, the antifungal agent clotrimazole, and the antiemetic meclizine are inverse agonists that inhibit gene activation by CAR, while the pesticide 1,4-bis[2-(3,5-dichloropyridyloxy)]benzene, the steroid 5-β-pregnane-3,20-dione, are agonists that activate gene expression when bound to CAR. Genes induced by CAR include those encoding several CYPs (CYP2B6, CYP2C9, and CYP3A4), phase 2 enzymes (including GSTs, UGTs, and SULTs), and drug and endobiotic transporters. CYP3A4 is induced by both PXR and CAR and thus its level is highly influenced by a number of drugs and other xenobiotics. There are species differences in the ligand specificities of these receptors. For example, rifampin activates human PXR, but not mouse or rat PXR, while meclizine activates mouse CAR, but inhibits gene induction by human CAR. These findings further underscore that data from rodent model systems do not always reflect the response of humans to drugs.

Members of a receptor family do not always show similar activities toward xenobiotics. One peroxisome proliferator activated receptor, PPARα, is the target for the fibrate class of hyperlipidemic drugs (e.g., gemfibrozil and fenofibrate). Activation of PPARα) results in induction of genes encoding fatty acid metabolizing enzymes (result: ↓serum triglycerides) and of CYP4 enzymes that oxidize fatty acids and compounds with fatty acid side chains, such as leukotriene and arachidonic acid analogs. Another family member, PPARγ, is the target for the thiazolidinedione class of agents for type 2 diabetes (e.g., rosiglitazone and pioglitazone); PPARγ does not induce xenobiotic metabolism.

The UGT genes, especially UGT1A1, are inducible via a host of transcriptional activation pathways, including AHR, Nrf2 (a transcriptional regulator of cytoprotective genes that is induced by an antioxidant response), PXR, CAR, and PPARα. Because the UGTs are abundant in the GI track and liver, regulation of the UGTs by drug-induced activation of these receptors could affect the pharmacokinetic parameters of many orally administered therapeutics.

ROLE OF DRUG METABOLISM IN THE DRUG DEVELOPMENT PROCESS. There are 2 key elements associated with successful drug development: efficacy and safety. Both depend on drug metabolism. It is necessary to determine which enzymes metabolize a potential new drug candidate in order to predict whether the compound may cause drug-drug interactions or be susceptible to marked interindividual variation in metabolism because of genetic polymorphisms. Computational chemical systems biology and metabolomic approaches can enhance these studies.

Historically, drug candidates have been administered to rodents at doses well above the human target dose to predict acute toxicity. For drug candidates to be used chronically in humans, long-term carcinogenicity studies are carried out in rodent models. For determination of metabolism, the compound is subjected to analysis by human liver cells or extracts from these cells that contain the drug-metabolizing enzymes. If a CYP is involved, a panel of recombinant CYPs can be used to determine which CYP predominates in the metabolism of the drug. If a single CYP, such as CYP3A4, is found to be the sole CYP that metabolizes a drug candidate, then a decision can be made about the likelihood of drug interactions. Interactions arise when multiple drugs are simultaneously administered, for example, in elderly patients, who on a daily basis may take prescribed anti-inflammatory drugs, 1 or 2 cholesterol-lowering drugs, several classes of blood pressure medications, a gastric acid suppressant, an anticoagulant, and a number of over-the-counter medications. Ideally, the best drug candidate would be metabolized by several CYPs so that variability in expression levels of one CYP or drug-drug interactions would not alter its metabolism and pharmacokinetics.

Similar studies can be carried out with phase 2 enzymes and drug transporters. In addition to the use of recombinant human xenobiotic-metabolizing enzymes in predicting drug metabolism, human receptor-based (PXR and CAR) systems or cell lines expressing these receptors are used to determine whether a particular drug candidate could be a ligand or activator of PXR, CAR, or PPARα. For example, a drug that activates PXR may result in rapid clearance of other drugs that are CYP3A4 substrates, thus decreasing their bioavailability and efficacy.

Traditional toxicity studies in animals can be a bottleneck in the drug development process of lead compound optimization. A new technology of high-throughput screening for biomarkers of toxicity is being adopted for drug development using metabolomics, the systematic identification and quantification of all metabolites in a given organism or biological sample. Analytical platforms such as 1H-NMR and liquid or gas chromatography coupled to mass spectrometry, in conjunction with chemometric and multivariate data analysis, allow the simultaneous determination and comparison of thousands of chemicals in biological fluids such as serum and urine, as well as the chemical constituents of cells and tissues. Metabolomics can be used to find biomarkers for drug efficacy and toxicity that can be of value in clinical trails to identify responders and nonresponders.