PRINCIPLES OF DRUG METABOLISM...

PRINCIPLES OF DRUG METABOLISM

BERNARD TESTA

Department of Pharmacy, University HospitalCentre, CH-1011 Lausanne, Switzerland

1. INTRODUCTION

Xenobiotic metabolism, which includes drugmetabolism, has become a major pharmacolo-gical science with particular relevance to biol-ogy, therapeutics and toxicology. Drug meta-bolism is also of great importance inmedicinalchemistry because it influences in qualitative,quantitative, and kinetic terms the deactiva-tion, activation, detoxification, and toxifica-tion of the vast majority of drugs. As a result,medicinal chemists engaged in drug discoveryand development should be able to integratemetabolic considerations into drug design. Todo so, however, requires a fair or even goodknowledge of xenobiotic metabolism.

This chapter, which is written by a medic-inal chemist for medicinal chemists, aims atoffering knowledge and understanding ratherthan encyclopedic information. Readerswant-ing to go further in the study of xenobioticmetabolism may consult various classical orrecent books [1–15], broad reviews, and bookchapters [16–27].

1.1. Definitions and Concepts

Drugs are but one category among the manyxenobiotics (Table 1) that enter the body buthavenonutritional or physiological value [14].The study of the disposition—or fate—of xe-nobiotics in living systems includes the con-sideration of their absorption into the organ-ism, how and where they are distributed andstored, the chemical and biochemical trans-formations they may undergo, and how andby which route(s) they are finally excretedand returned to the environment. As for“metabolism,” this word has acquired twomeanings, being synonymouswith disposition(i.e., the sumof the processes affecting the fateof a chemical substance in the body), and withbiotransformation as understood in thischapter.

In pharmacology, one speaks of pharmaco-dynamic effects to indicate what a drug does tothe body, and pharmacokinetic effects to indi-cate what the body does to a drug, two aspectsof the behavior of drugs that are strongly inter-dependent. Pharmacokinetic effects will ob-viously have a decisive influence on theintensity and duration of pharmacodynamiceffects, while metabolism will generate newchemical entities (metabolites) that mayhave distinct pharmacodynamic properties oftheir own. Conversely, by its own pharmaco-dynamic effects, a compound may affect thestate of the organism (e.g., hemodynamicchanges, enzyme activities, etc.) and hence theorganism’s capacity to handle xenobiotics.Onlyasystemicapproachcanhelponeappreci-ate the global nature of this interdependence.

1.2. Types of Metabolic Reactions AffectingXenobiotics

A first discrimination to be made among me-tabolic reactions is based on the nature of thecatalyst. Reactions of xenobiotic metabolism,like other biochemical reactions, are catalyzedby enzymes. However, while the vast majorityof reactions of xenobiotic metabolism are in-deed enzymatic ones, some nonenzymatic re-actions are also well documented. This is dueto the fact that a variety of xenobiotics havebeen found to be labile enough to react none-nzymatically under biological conditions ofpH and temperature [28]. But there is more.In a normal enzymatic reaction, metabolicintermediates exist en route to the product(s) anddonot leave the catalytic site.However,many exceptions to this rule are known, withthe metabolic intermediate leaving the activesite and reacting with water, with an endo-genous molecule or macromolecule, or with axenobiotic. Such reactions are also of a none-nzymatic nature but are better designated aspostenzymatic reactions.

The metabolism of drugs and other xeno-biotics is often a biphasic process in which thecompound may first undergo a functionaliza-tion reaction (phase I reaction) of oxidation,reduction, or hydrolysis. This introduces orunveils a functional group such as a hydroxy

Burger’s Medicinal Chemistry, Drug Discovery, and Development, Seventh Edition,edited by Donald J. Abraham and David P. RotellaCopyright � 2010 John Wiley & Sons, Inc.

405

or amino group suitable for coupling with anendogenous molecule or moiety in a secondmetabolic step known as a conjugation reac-tion (phase II reaction) [19,24]. In a number ofcases, phase I metabolites may be excretedprior to conjugation, while many xenobioticscan be directly conjugated. Furthermore, re-actions of functionalization may follow somereactions of conjugation, for example, someconjugates are hydrolyzed and/or oxidizedprior to their excretion.

Xenobiotic biotransformation thus pro-duces two types of metabolites, namely func-tionalization products and conjugates. Butwith the growth of knowledge, biochemistsand pharmacologists have progressively cometo recognize the existence of a third class ofmetabolites, namely xenobiotic–macromole-cule adducts, also called macromolecular con-jugates. Such peculiar metabolites are formedwhen a xenobiotic binds covalently to a biolo-gical macromolecule, usually following meta-bolic activation (i.e., postenzymatically). Bothfunctionalization products and conjugateshave been found to bind covalently to biologi-cal macromolecules, the reaction often beingtoxicologically relevant.

1.3. Specificities and Selectivitiesin Xenobiotic Metabolism

The words “selectivity” and “specificity” maynot have identical meanings in chemistry andbiochemistry. In this chapter, the specificity ofan enzyme is taken to mean an ensemble ofproperties, the description of which makes itpossible to specify the enzyme’s behavior. Incontrast, the term selectivity is applied tometabolic processes, indicating that a givenmetabolic reaction or pathway is able to selectsome substrates or products from a larger set.In other words, the selectivity of a metabolicreaction is the detectable expression of thespecificity of an enzyme. Such definitionsmaynot be universally accepted, but they have themerit of clarity.

What, then, are the various types of selec-tivities (or specificities) encountered in xeno-biotic metabolism? What characterizes an en-zyme from a catalytic viewpoint is first itschemospecificity, that is, its specificity interms of the type(s) of reaction it catalyzes.When two or more substrates aremetabolizedat different rates by a single enzyme underidentical conditions, substrate selectivity isobserved. In such a definition, the nature ofthe product(s) and their isomeric relationshipare not considered. Substrate selectivity isdistinct from product selectivity, which is ob-served when two or more metabolites areformed at different rates by a single enzymefrom a single substrate. Thus, substrate-se-lective reactions discriminate between differ-ent compounds, while product-selective reac-tions discriminate betweendifferent groups orpositions in a given compound.

The substrates beingmetabolized at differ-ent rates may share various types of relation-ships. They may be chemically dissimilar orsimilar (e.g., analogs), in which case the termof substrate selectivity is used in a narrowsense. Alternatively, the substrates may beisomers such as positional isomers (regioi-somers) or stereoisomers, resulting in sub-strate regioselectivity or substrate stereose-lectivity. Substrate enantioselectivity is a par-ticular case of the latter (see Section 4.2.1).

Products formed at different rates in pro-duct-selective reactions may also share var-ious types of relationships. Thus, they may be

Table 1. Major Categories of Xenobiotics(Modified from Ref. [14])

. Drugs

. Food constituents devoid of physiological roles

. Food additives (preservatives, coloring and fla-voring agents, antioxidants, etc.)

. Chemicals of leisure, pleasure, or abuse (ethanol,coffee and tobacco constituents, hallucinogens,doping agents, etc.)

.Agrochemicals (fertilizers, insecticides, herbicides,etc.)

. Industrial and technical chemicals (solvents, dyes,monomers, polymers, etc.)

. Pollutants of natural origin (radon, sulfur dioxide,hydrocarbons, etc.)

. Pollutants produced by microbial contamination(e.g., aflatoxins)

. Pollutants produced by physical or chemicaltransformation of natural compounds (polycyclicaromatic hydrocarbons by burning, Maillard re-action products by heating, etc.)

406 PRINCIPLES OF DRUG METABOLISM

analogs, regioisomers, or stereoisomers, re-sulting in product selectivity (narrow sense),product regioselectivity, or product stereose-lectivity (e.g., product enantioselectivity).Note that the product selectivity displayed bytwo distinct substrates in a given metabolicreaction may be different, in other words theproduct selectivitymaybe substrate-selective.The term substrate–product selectivity canbe used to describe such complex cases,which are conceivable for any type of selectiv-ity but have been reported mainly forstereoselectivity.

1.4. Pharmacodynamic Consequencesof Xenobiotic Metabolism

The major function of xenobiotic metabolismcanbeseenastheeliminationofphysiologicallyuseless compounds, some of which may beharmful, witness the tens of thousandsof toxins produced by plants. The function oftoxin inactivation justifies the designation ofdetoxification originally given to reactionsof xenobiotic metabolism. However, the possi-ble pharmacological consequences of biotrans-formationarenotrestrictedtodetoxification.Inthe simple case of a xenobiotic having a singlemetabolite, fourpossibilitiesexist[25],namely:

1. Both the xenobiotic and its metaboliteare devoid of biological effects (at leastin the concentration or dose range in-vestigated); such a situation has noplace in medicinal chemistry.

2. Only the xenobiotic elicits biologicaleffects, a situation that in medicinalchemistry is typical of drugs yieldingno bioactive metabolite, as seen forexample with soft drugs.

3. Both the xenobiotic and its metaboliteare biologically active, the two activ-ities being comparable or differenteither qualitatively or quantitatively.

4. The observed biological activity is dueexclusively to the metabolite, a situa-tion that in medicinal chemistry is ty-pical of prodrugs.

When a drug or another xenobiotic is trans-formed into a toxic metabolite, the reaction is

one of toxification. Such a metabolite may actor react in a number of ways to elicit a varietyof toxic responses at different biological levels.However, it is essential to stress that theoccurrence of a reaction of toxification (i.e.,toxicity at the molecular level) does not neces-sarily imply toxicity at the levels of organs andorganisms, as discussed later in this chapter.

1.5. Setting the Scene

In drug research and development, metabo-lism is of pivotal importance due to the inter-connectedness between pharmacokinetic andpharmacodynamic processes [14]. In vitrome-tabolic studies are now initiated early duringdiscovery and development to assess the over-all rate of oxidativemetabolism, to identify themetabolites, and to obtain primary informa-tion on the enzymes involved, and to postulatemetabolic intermediates. Based on these find-ings, themetabolitesmust be synthesized andtested for their own pharmacological and tox-icological effects. In preclinical and early clin-ical studies,manypharmacokinetic datamustbe obtained and relevant criteria must besatisfied before a drug candidate can enterlarge-scale clinical trials [29,30]. As a resultof these demands, the interest of medicinalchemists for drug metabolism has grown re-markably in recent years.

As will become apparent, the approachfollowed in this chapter is an analytical one,meaning that the focus is on metabolic reac-tions, the target groups they affect, and theenzymes by which they are catalyzed. Thisinformation provides the foundations of drugmetabolism, but itmust be complemented byasynthetic view to allow a broader understand-ing andmeaningful predictions. Two steps arerequired to approach these objectives, namely(a) the elaboration of metabolic schemeswhere the competitive and sequential reac-tions undergone by a given drug are orderedand (b) an assessment of the various biologicalfactors that influence such schemes bothquantitatively and qualitatively. As an exam-ple of a metabolic scheme, Fig. 1 presents thebiotransformation of propranolol (1) in hu-mans [31]. There are relatively few studies ascomprehensive and clinically relevant as thisone, which remains as current today as it was

INTRODUCTION 407

when published in 1985. Indeed, over 90% of adose were accounted for and consisted mainlyin products of oxidation and conjugation. Themissing 10% may represent other, minor andpresumably quite numerous metabolites, forexample, those resulting from ring hydroxyla-tion at other positions or from the progressivebreakdown of glutathione conjugates.

A large variety of enzymes and metabolicreactions are presented in Sections 2 and 3. Aswill become clear, some enzymes catalyze onlya single type of reaction (e.g., N-acetylation),whereas others use a basic catalytic mechan-ism to attack a variety ofmoieties and producedifferent types of metabolites (e.g., cyto-chromes P450). As an introduction to theseenzymes and reactions, we present in Table 2an estimate of their relative importance indrug metabolism. In this table, the correspon-dence between the number of substrates and

the overall contribution to drug metabolismdoes not need to be perfect, since some en-zymes show a limited capacity (e.g., sulfo-transferases) whereas others make a signifi-cant contribution to the biotransformation oftheir substrates (e.g., hydrolases).

2. FUNCTIONALIZATION REACTIONS

2.1. Introduction

Reactions of functionalization comprise oxida-tions (electron removal, dehydrogenation, andoxygenation), reductions (electron addition,hydrogenation, and removal of oxygen), andhydrations/dehydrations (hydrolysis and ad-dition or removal of water). The reactions ofoxidation and reduction are catalyzed by avery large variety of oxidoreductases, whilevarious hydrolases catalyze hydrations. A

O NH2

OH

O NH

OH

O NH

OH

OH

GLUC

SULF

GLUC GLUC

O NH2

OH

OH

1

O

OH

CHO O NH

OH

OCH3

HO

GLUC

SULF

O OH

OH

O

OH

COOH

O

OH

COOH

OH

O OH

OH

OH

GLUC

O COOH

SULFGLUC

GLUC

SULF

GLUC

NH2

Figure 1. The metabolism of propranolol (1) in humans, accounting for more than 90% of the dose;GLUC¼ glucuronide(s); SULF¼ sulfate(s) [31].


large majority of enzymes involved in xeno-biotic functionalization are briefly reviewed inSection 2.2 [14].Metabolic reactions and path-ways of functionalization constitute the mainbody of this section.

2.2. Enzymes Catalyzing FunctionalizationReactions

2.2.1. Cytochromes P450 Monooxygenationreactions are of major significance in drugmetabolism and are mediated by various en-zymes, which differ markedly in their struc-ture and properties. Among these, the mostimportant as far as xenobiotic metabolism isconcerned are the cytochromes P450 (EC1.14.14.1, also 1.14.13), a very large group ofenzymes belonging to heme-coupled monoox-ygenases [4–7,12,15–18,20,22,32–36]. The cy-tochrome P450 enzymes (CYPs) are encodedby the CYP gene superfamily and are classi-fied in families and subfamilies as summar-ized in Table 3. Cytochrome P450 is the majordrug-metabolizing enzyme system, playing akey role in detoxification and toxification, andis of additional significance in medicinalchemistry because several CYP enzymes are

drug targets, for example, thromboxanesynthase (CYP5) and aromatase (CYP19). Thethree CYP families mostly involved in xeno-biotic metabolism are CYP1 to CYP3, whoserelative importance is given in Table 4.

The present section focuses on the meta-bolic reactions, beginning with the catalyticcycle of cytochrome P450 (Fig. 2). This cycleinvolves a number of steps that can be simpli-fied as follows:

(a) The enzyme in its ferric (oxidized) formexists in equilibrium between two spinstates, a hexacoordinated low-spinform that cannot be reduced, and apentacoordinated high-spin form.Binding of the substrate to enzyme in-duces a shift to the reducible high-spinform (reaction a).

(b) A first electron enter the enzyme–sub-strate complex (reaction b).

(c) The enzyme in its ferrous form has ahigh affinity for diatomic gases such asCO (a strong inhibitor of cytochromeP450) and dioxygen (reaction c).

(d) ElectrontransferfromFe2þ toO2withinthe enzyme–substrate–oxygen ternarycomplexreducesthedioxygentoaboundmolecule of superoxide. Its possible lib-eration in the presence of compoundswithgoodaffinitybut lowreactivity (un-coupling) can be cytotoxic (reaction d).

(e) The normal cycle continues with a sec-ond electron entering via either FP1 orFP2 and reducing the ternary complex(reaction e).

(f) Electron transfer within the ternarycomplex generates bound peroxide an-ion (O2

2�).(g) The bound peroxide anion is split, lib-

erating H2O (reaction f).(h) The remaining oxygen atom is an oxene

species. This is the reactive form ofoxygen that will attack the substrate.

(i) The binary enzyme–product complexdissociates, thereby regenerating theinitial state of cytochrome P450 (reac-tion h).

Oxene is a rather electrophilic species,being neutral but having only six electrons in

Table 2. EstimateoftheRelativeContributionsof Major Drug-Metabolizing Enzymesa

Enzymes

OverallContribution

to DrugMetabolismb

Cytochromes P450 (Section 2.2.1) þ þ þ þDehydrogenases and reductases(Section 2.2.2)

þ þ þ

Flavin-containing monooxy-genases (Section 2.2.2)

þ

Hydrolases (Section 2.2.3) þ þMethyltransferases (Section 3.2) þSulfotransferases (Section 3.3) þGlucuronyltransferases(Section 3.4)

þ þ þ

N-Acetyltransferases (Section 3.5) þAcyl-coenzyme A synthetases(Section 3.6)

þ

Glutathione S-transferases(Section 3.7)

þ þ

Phosphotransferases (Section 3.8) (þ )

a (þ ) very low; þ low; þ þ intermediate; þ þ þ high;

þ þ þ þ very high.b Including drug metabolites.

FUNCTIONALIZATION REACTIONS 409

Table 3. The Human CYP Gene Superfamily: A Table of the Families and Subfamilies of GeneProducts [34–37]

Families Subfamilies (Representative Gene Products)

P450 1 Family (Aryl hydrocarbon hydroxylases; xenobiotic metabolism; inducible by polycyclic aromatichydrocarbons)

P450 1A Subfamily (CYP1A1, CYP1A2)P450 1B Subfamily (CYP1B1)

P450 2 Family (Xenobiotic metabolism; constitutive and xenobiotic-inducible)P450 2A Subfamily (CYP2A6, CYP2A7, CYP2A13)P450 2B Subfamily (CYP2B6)P450 2C Subfamily (CYP2C8, CYP2C9, CYP2C18, CYP2C19)P450 2D Subfamily (CYP2D6)P450 2E Subfamily (CYP2E1)P450 2F Subfamily (CYP2F1)P450 2J Subfamily (CYP2J2)P450 2R Subfamily (CYP2R1)P450 2S Subfamily (CYP2S1)P450 2U Subfamily (CYP2U1)P450 2W Subfamily (CYP2W1)

P450 3 Family (Xenobiotic and steroid metabolism; steroid-inducible)P450 3A Subfamily (CYP3A4, CYP3A5, CYP3A7, CYP3A43)

P450 4 Family (Peroxisome proliferator-inducible)P450 4A Subfamily (CYP4A11, CYP4A20, CYP4A22)P450 4B Subfamily (CYP4B1)P450 4F Subfamily (CYP4F2, CYP4F3, CYP4F8, CYP4F11, CYP4F12, CYP4F22)P450 4V Subfamily (CYP4V2)P450 4X Subfamily (CYP4X1)

P450 5 FamilyP450 5A Subfamily (CYP5A1)

P450 7 Family (Steroid 7-hydroxylases)P450 7A Subfamily (CYP7A1)P450 7B Subfamily (CYP7B1)

P450 8 FamilyP450 8A Subfamily (CYP8A1)P450 8B Subfamily (CYP8B1)

P450 11 Family (Mitochondrial steroid hydroxylases)P450 11A Subfamily (CYP11A1)P450 11B Subfamily (CYP11B1, CYP11B2)

P450 17 Family (Steroid 17b-hydroxylase)P450 17A Subfamily (CYP17A1)

P450 19 Family (CYP19)P450 20 Family (CYP20)P450 21 Family (CYP21)P450 24 Family (CYP24)P450 26 Family

P450 26A Subfamily (CYP26A1)P450 26B Subfamily (CYP26B1)P450 26C Subfamily (CYP26C1)

P450 27 Family (Mitochondrial steroid hydroxylases)P450 27A Subfamily (CYP27A1)P450 27B Subfamily (CYP27B1)P450 27C Subfamily (CYP27C1)

P450 39 Family (CYP39)P450 46 Family (CYP46)P450 51 Family (CYP51)

This list reports all 57 known human CYP gene products.


its outer shell. Its detailed reaction mechan-isms are beyond the scope of this chapter, butsome indications will be given when discuss-ing the various reactions of oxidation cata-lyzed by cytochromes P450.

2.2.2. Other Oxidoreductases Besides cyto-chromes P450, other monooxygenases of im-portance are the flavin-containing monooxy-genases (see Table 5), while dopamine b-monooxygenase (EC 1.14.17.1) plays only aminor role. Other oxidoreductases that can

b

OH2 f

e

e

e

NADPH

FP1

e

O2c

3+

3+hs

[ Fe ] XH

[ Fe ] XH( )O

2+[ Fe ] XH(O2 )

NADH FP2 cyt b5

g

] XH3+[ Fels

23+[ Fe ] (O )XH

2

XH

XOH

a

d[ Fe3+] O+XH

223+[ Fe ] XH(O )

2+[ Fe ] XH(O2)

CO

hs2+[ Fe ] XH 2+

hs[ Fe ] (CO)XH

ls

3+] 3+hs

[ Fe ][ Fe

Figure 2. Catalytic cycle of cytochrome P450 associated with monooxygenase reactions. [Fe3þ ]¼ ferricy-tochrome P450; hs¼high spin; ls¼ low spin; [Fe2þ ]¼ ferrocytochrome P450; FP1¼ flavoprotein 1¼NADPH-cytochrome P450 reductase; F P2¼NADH-cytochrome b5 reductase; cyt b5¼ cytochrome b5; XH¼ substrate(modified from Refs [4,14]).

Table 4. Levels and Variability of Human CYP Enzymes Involved in Drug Metabolism [16]

Percent of Drugs (or Other Xenobiotics)Interacting with Enzyme

CYPLevel of Enzyme inLiver (% of Total)

VariabilityRange As Substrates As Inhibitors As Inducers/Activators

1A1 3 (12) 3 (13) 6 (33)1A2 ca. 13 ca. 40-fold 10 (15) 12 (18) 3 (25)1B1 <1 1 (10) 1 (7) 1 (9)2A6 ca. 4 ca. 30- to 100-fold 3 (10) 2 (6) 2 (1)2B6 <1 ca. 50-fold 4 (9) 3 (5) 13 (no data)2C ca. 18 25- to 100-fold 25 (13) 27 (17) 21 (6)2D6 up to 2.5 >1000-fold 15 (2) 22 (8) 2 (2 activators)2E1 up to 7 ca. 20-fold 3 (16) 4 (10) 7 (7)3A4 up to 28 ca. 20-fold 36 (13) 26 (16) 45 (17)

1(þ ) very low; þ low; þ þ intermediate; þ þ þhigh; þ þ þ þ very high.


play a major or less important role in drugmetabolism are the two monoamine oxidasesthat are essentially mitochondrial enzymes,and the broad group of copper-containingamine oxidases. Cytosolic oxidoreductases arethe molybdenum hydroxylases, namely alde-hyde oxidase and xanthine oxidase.

Various peroxidases are progressivelybeing recognized as important enzymes indrug metabolism. Several cytochrome P450enzymes have been shown to have peroxidaseactivity. A variety of peroxidases may oxidizedrugs, for example, myeloperoxidase (MPO).Prostaglandin G/H synthase (prostaglandin-endoperoxide synthase) is able to use a num-ber of xenobiotics as cofactors in a reaction ofcooxidation.

The large and important ensemble of dehy-drogenases/reductases include alcohol dehy-drogenases (ADH) that are zinc enzymesfound in the cytosol of the mammalian liverand in various extrahepatic tissues. Mamma-lian liver alcohol dehydrogenases (LADHs)are dimeric enzymes. The human enzymesbelong to three different classes: Class I(ADH1), comprising the various isozymes thatare homodimers or heterodimers of the alpha(ADH1A, ADH2A, ADH3A), beta and gammasubunits.

Enzymes categorized as aldehyde dehydro-genases (ALDHs) include the NADþ - andNAD(P)þ -dependent enzymes. They exist inmultiple forms in the cytosol, mitochondria,and microsomes of various mammalian tis-sues. Among the aldo-keto reductases, we findaldehyde reductase, alditol dehydrogenase, anumber of hydroxysteroid dehydrogenases,and dihydrodiol dehydrogenase. These en-zymes are widely distributed in nature andoccur in a considerable number ofmammaliantissues. Their subcellular location is primarilycytosolic, and for somealsomitochondrial. Thecarbonyl reductases belong to the short-chaindehydrogenases/reductases and are of note-worthy significance in xenobiotic metabolism.There are many similarities, including somemarked overlap in substrate specificity, be-tween monomeric, NADPH-dependent alde-hyde reductase (AKR1), alditol dehydrogen-ase (AKR2), and carbonyl reductase (AKR3).

Other reductases that have a role to play indrug metabolism include the important qui-none reductase, alsoknownasDT-diaphorase.

2.2.3. Hydrolases Hydrolases constitute avery complex ensemble of enzymes many ofwhichareknownor suspected tobe involved inxenobiotic metabolism (Table 6). Relevant en-

Table 5. A Survey of Oxidoreductases Other than Cytochromes P450 Playing a Role in DrugMetabolism [14,22]

Enzymes EC NumbersGene Root (or Gene) and Major

Human Enzymes

Flavin-containingmonooxygenases

EC 1.14.13.8 FMO (FMO1 to FMO5)

Monoamine oxidases EC 1.4.3.4 MAO (MAO-A and MAO-B)Copper-containing amineoxidases

EC 1.4.3.6 AOC (DAO and SSAO)

Aldehyde oxidase EC 1.2.3.1 AOX1 (AO)Xanthine oxidoreductase EC 1.17.1.4 and 1.17.3.2 XOR (XDH and XO)Various peroxidases EC 1.11.1.7 and 1.11.1.8 For example, EPO (EPO),MPO (MPO) and

TPO (TPO)Prostaglandin G/H synthase EC 1.14.99.1 PTGS (COX-1 and COX-2)Alcohol dehydrogenases EC 1.1.1.1 ADH (ADH1A, 1B and 1C, ADH4, ADH5,

ADH6, and ADH7)Aldehyde dehydrogenases EC 1.2.1.3 and 1.2.1.5 ALDH (e.g., ALDH1A1, 1A2 and 1A3, 1B1,

2, 3A1, 3A2, 3B1, 3B2, 8A1, and 9A1)Aldo-keto reductases In EC 1.1.1 and 1.3.1 AKR (e.g., ALR1, ALR2, DD1, DD2, DD3,

DD4, AKR7A2, 7A3, and 7A4)Carbonyl reductases EC 1.1.1.184 CBR (CR1, CR3)Quinone reductases EC 1.6.5.2 and 1.10.99.2 NQO (NQO1 and NGO2)


zymes among the serine hydrolases includecarboxylesterases, arylesterases, cholinester-ase, and a number of serine endopeptidases(EC 3.4.21). The role of arylsulfatases, para-oxonase, b-glucuronidase, and epoxide hydro-lases is worth noting. Some metalloendopep-tidases (EC 3.4.24) and amidases (EC 3.5.1and 3.5.2) are also of interest [9,14,19,23].

2.3. Reactions of Carbon Oxidationand Reduction

When examining reactions of carbon oxida-tion (oxygenations and dehydrogenations)and carbon reduction (hydrogenations), it isconvenient from a mechanistic viewpoint todistinguish between sp3-, sp2-, and sp-carbonatoms.

2.3.1. sp3-Carbon Atoms Reactions of oxida-tion and reduction of sp3-carbon atoms (plussome subsequent reactions at carbonylgroups) are schematized in Fig. 3 and dis-cussed sequentially below. In the simplest

cases, a nonactivated carbon atom in an alkylgroup undergoes cytochrome P450-catalyzedhydroxylation. The penultimate position is apreferred site of attack (reaction 1b), but hy-droxylation can also occur in the terminalposition (reaction 1a) or in other positions incase of steric hindrance or with some specificcytochromes P450. Dehydrogenation by dehy-drogenases can then yield a carbonyl deriva-tive (reactions 1c and 1e) that is either analdehyde or a ketone. Note that reactions 1cand 1e act not only onmetabolites, but also onxenobiotic alcohols, and are reversible (i.e.,reactions 1d and 1f) since dehydrogenasescatalyze the reactions in both directions. Andwhile a ketone is very seldomoxidized further,aldehydes are good substrates for aldehydedehydrogenases or other enzymes, and leadirreversibly to carboxylic acid metabolites (re-action 1g). A classical example is that of etha-nol, which in the body exists in redox equili-brium with acetaldehyde, this metabolitebeing rapidly and irreversibly oxidized toacetic acid.

Table 6. A Survey of Hydrolases Playing a Role in Drug Metabolism [9,14,23]

Classes of HydrolasesExamples of Enzymes (With Some Gene Roots and

Human Enzymes)

EC 3.1.1: carboxylic ester hydrolases EC 3.1.1.1: Carboxylesterase (CES) CES1A1, CES2, CES3EC 3.1.1.2: Arylesterase (PON, see 3.1.8.1)EC 3.1.1.8: Cholinesterase (BCHE)

EC 3.1.2: thiolester hydrolases EC 3.1.2.20: Acyl-CoA hydrolaseEC 3.1.3: phosphoric monoester hydrolases EC 3.1.3.1: Alkaline phosphatase (ALP)

EC 3.1.3.2: Acid phosphatase (ACP)EC 3.1.6: sulfuric ester hydrolases EC 3.1.6.1: ArylsulfataseEC 3.1.8: phosphoric triester hydrolases EC 3.1.8.1: Paraoxonase (PON) PON1, PON2, PON3

EC 3.1.8.2: Diisopropyl-fluorophosphataseEC 3.2: glycosylases EC 3.2.1.31: b-Glucuronidase (GUSB)EC 3.3.2: ether hydrolases EC 3.3.2.9: Microsomal epoxide hydrolase (EPHX1) mEH

EC 3.3.2.10: Soluble epoxide hydrolase (EPHX2) sEHEC 3.4.11: aminopeptidases EC 3.4.11.1: Leucyl aminopeptidase (LAP)EC 3.4.13 and 3.4.14: peptidases acting on

di- and tripeptidesEC 3.4.14.5: Dipeptidyl-peptidase IV (DPP4)

EC 3.4.16 to 3.4.18: carboxypeptidases EC 3.4.16.2: Lysosomal Pro-Xaa carboxypeptidaseEC 3.4.17.1: Carboxypeptidase A (CPA)

EC 3.4.21 to 3.4.25: endopeptidases EC 3.4.21.1: Chymotrypsin (CTRB)EC 3.4.24.15: Thimet oligopeptidase (THOP)

EC 3.5.1: hydrolases acting on linearamides

EC 3.5.1.4: Amidase

EC 3.5.1.39: AlkylamidaseEC3.5.2: hydrolases acting on cyclic amides EC 3.5.2.1: Barbiturase

EC 3.5.2.2: Dihydropyrimidinase (DPYS)EC 3.5.2.6: b-Lactamase


(1) CHC 2

R'

R

R''

CH3

aCHC 2

R'

R''

R CH2OH CHC 2

R'

R''

R C

O

H

c

dCHC 2

R''

R'

R COOHg

CHOHC

R'

R''

R CH3 CC

OR'

R''

R CH3

e

f

b

(2)

(3)

Y = aryl or CC

R''R'

R

CCR'

(4) XR CH2 R'a

XR CHOH R'b

X = NH or N-alkyl or N-aryl, or O, or SR = H or alkyl or aryl; R' = H or alkyl or aryl

(5) XCH

R'

R

a R'

C

X

OH

R

bOC

R'

R

X = halogen

XHR + CR'

O

H

CHR 2 CH2 R' CHR CH R'

CHOHY CH3CH3CHY 2

R, R' and R'' = H or alkyl or aryl

or

(6) CR' C R''

R

H

R'''

X

CC

R'''

R''R'

RX = Cl, Br

(7) CR' X

R''

R

CR' H

R''

RX = Cl, Br

(8) CR' C R''

R

X

R'''

X

CC

R'''

R''R'

RX = Cl, Br

Figure 3. Major functionalization reactions involving an sp3-carbon in substrate molecules; some reactionsat carbonyl groups are also included since they often follows the former. The reactions shown here aremainlyoxidations (oxygenations and dehydrogenations) and reductions (hydrogenations), plus some postenzymaticreactions of hydrolytic cleavage.

For a number of substrates, the oxidation ofprimary and secondary alcohol and of alde-hyde groups can also be catalyzed by cyto-chrome P450. A typical example is theC(10)-demethylation of androgens and ana-logs catalyzed by aromatase (CYP19).

A special case of carbon oxidation, recog-nized only recently and probably of underes-timated significance, is desaturation of a di-methylene unit by cytochrome P450 to pro-duce an olefinic group (reaction 2). An inter-esting example is provided by testosterone,which among many cytochrome P450-cata-lyzed reactions undergoes allylic hydroxyla-

tion to 6b-hydroxytestosterone and desatura-tion to 6,7-dehydrotestosterone [38].

There is a known regioselectivity in cyto-chrome P450-catalyzed hydroxylations forcarbon atoms adjacent (alpha) to an unsatu-rated system (reaction 3) or a heteroatom suchas N, O, or S (reaction 4a). In the former cases,hydroxylation can easily be followed by dehy-drogenation (not shown). In the latter cases,however, the hydroxylated metabolite isusually unstable and undergoes a rapid, post-enzymatic elimination (reaction 4b). Depend-ing on the substrate, this pathway produces asecondary or primary amine, an alcohol or


phenol, or a thiol, while the alkyl group iscleaved as an aldehyde or a ketone. Reactions4 constitute avery commonand frequent path-way as far as drug metabolism is concerned,since it underlies some well-known metabolicreactions of N�C cleavage discussed later.Note that the actual mechanism of such reac-tions isusuallymore complex thanshownhereandmay involve intermediate oxidation of theheteroatom.

Aliphatic carbonatomsbearing oneormorehalogen atoms (mainly chlorine or bromine)can be similarlymetabolized by hydroxylationand loss of HX to dehalogenated products(reactions 5a and 5b) (see later). Dehalogena-tion reactions can also proceed reductively orwithout change in the state of oxidation. Thelatter reactions are dehydrohalogenations(usually dehydrochlorination or dehydrobro-mination) occurring nonenzymatically (reac-tion 6). Reductive dehalogenations involvereplacement of a halogen by a hydrogen (reac-tion 7), or vic-bisdehalogenation (reaction 8).Some radical species formed as intermediatesmay have toxicological significance.

Reactions 1a, 1b, 3, 4a, and5aare catalyzedby cytochromes P450. Here, the iron-boundoxene (Section 2.2.1) acts by a mechanismknownas “oxygen rebound”whereby aHatomis exchanged for a OH group. In simplifiedterms, the oxene atom attacks the substrateby cleaving a C�H bond and removing thehydrogen atom (hydrogen radical). This formsan iron-bound HO. species and leaves thesubstrate as a C-centered radical. In the laststep, the iron-boundHO. species is transferredto the substrate.

Halothane (2) offers a telling example of themetabolic fate of halogenated compounds ofmedicinal interest. Indeed, this agent under-goes two major pathways, oxidative dehalo-genation leading to trifluoroacetic acid (3) andreduction producing a reactive radical (4)(Fig. 4).

2.3.2. sp2- and sp-Carbon Atoms Reactions atsp2-carbons are characterized by their ownpathways, catalytic mechanisms, and pro-ducts (Fig. 5). Thus, the oxidation of aromaticrings generates a variety of (usually stable)metabolites. Their common precursor is oftena reactive epoxide (reaction 1a) that can either

be hydrolyzed by epoxide hydrolase (reaction1b) to adihydrodiol, or rearrangeunder protoncatalysis to a phenol (reaction 1c). The produc-tion of a phenol is a very common metabolicreaction for drugs containing one or morearomatic rings. The para-position is the pre-ferred position of hydroxylation for unsubsti-tuted phenyl rings, but the regioselectivity ofthe reaction becomes more complex with sub-stituted phenyl or with other aromatic rings.

Dihydrodiols are seldom observed, as arecatechol metabolites produced by their dehy-drogenation catalyzed by dihydrodiol dehy-drogenase (reaction 1d). It is interesting tonote that this reaction restores the aromati-city that had been lost upon epoxide forma-tion. The further oxidation of phenols andphenolic metabolites is also possible, the rateof reaction and the nature of products depend-ing on the ring and on the nature and positionof its substituents. Catechols are thus formedby reaction 1e, while hydroquinones are some-times also produced (reaction 1f).

In some cases, catechols andhydroquinoneshave been found to undergo further oxidationto quinones (reactions 1g and 1i). Such reac-tionsoccurby twosingle-electron stepsandcanbe either enzymatic or nonenzymatic (i.e., re-sulting from autoxidation and yielding as by-product the superoxide anion-radical O2

.�).The intermediate is this reaction is a semiqui-none. Both quinones and semiquinones arereactive, in particular toward biomolecules,and have been implicated in many toxificationreactions. For example, the high toxicity ofbenzene for bone marrow is believed to be dueto the oxidation of catechol and hydroquinonecatalyzed by myeloper-oxidase.

The oxidation of diphenols to quinones isreversible (reactions 1h and 1j), a variety of

CHClBrCF3

COOHCF3

2

3

4

CF3 CHCl.

Figure 4. The structure of halothane (2) and two ofits metabolites, namely trifluoroacetic acid (3) pro-duced by oxidation, and a reactive radical (4) pro-duced by reduction.


cellular reductants being able to mediate thereduction of quinones either by a two-electronmechanism or by two single-electron steps.The two-electron reduction can be catalyzedby carbonyl reductase and quinone reductase,while cytochrome P450 and some flavopro-teins act by single-electron transfers. Thenonenzymatic reduction of quinones can occurfor example in the presence of O2

.� or somethiols such as glutathione.

Olefinic bonds in xenobiotic molecules canalso be targets of cytochrome P450-catalyzedepoxidation (reaction 2a). In contrast to areneoxides, the resulting epoxides are fairly stableand canbe isolatedandcharacterized.But likearene oxides, they are substrates of epoxidehydrolase to yield dihydrodiols (reaction 2b).

This is exemplified by carbamazepine (5),whose 10,11-epoxide (6) is a major and phar-macologically active metabolite in humans,and is further metabolized to the inactivedihydrodiol [39] (Fig. 6).

The reduction of olefinic groups (reaction2c) is documented for a few drugs bearing ana,b-ketoalkene function. The reaction isthought to be catalyzed by various NAD(P)Hoxidoreductases.

The few drugs that contain an acetylenicmoiety are also targets for cytochrome P450-catalyzed oxidation. Oxygenation of the triplebond (reaction 3a) yields an intermediate thatdepending on the substrate can react in anumber of ways, for example binding cova-lently to the enzyme, or forming a highlyreactive ketene whose hydration produces asubstituted acetic acid (reactions 3b and 3c).

2.4. Reactions of Nitrogen Oxidationand Reduction

Themainmetabolic reactions of oxidation andreduction of nitrogen atoms in organic mole-cules are summarized in Fig. 7. The functionalgroups involved are amines and amides andtheir oxygenated metabolites, as well as 1,4-

R OROH

H

H

OH

R

OH

OH

R

O

O

R

g

h

dba

OH

R

OH

HO

R

O

O

R

j

f

ec

CC

R'''

R''R'

R

aO

R''R'

R R'''

bR' R'''

HO R''

R OH

c

CHR' CH R''

R R'''

CR C Ha b

CCH OR CHR 2COOHc

i

CC

O

HR CR C H

O+

_

Figure 5. Major functionalization reactions involving an sp2- or sp-carbon in substrate molecules. Thesereactions are oxidations (oxygenations and dehydrogenations), reductions (hydrogenations), and hydrations,plus some postenzymatic rearrangements.

N

O NH2

N

O NH2

O

5 6

Figure 6. The structure of carbamazepine (5) andits 10,11-epoxide metabolite (6).


dihydropyridines, hydrazines, and azo com-pounds. In many cases, the reactions can becatalyzed by cytochrome P450 and/or flavin-containing monooxygenases. The first oxyge-nation step in reactions 1–4 and 6 have fre-quently been observed.

Nitrogen oxygenation is an (apparently)straightforward metabolic reaction of tertiaryamines (reaction 1a), be they aliphatic or aro-matic. Numerous drugs undergo this reaction,the resulting N-oxide metabolite being morepolar and hydrophilic than the parent com-pound. Identical considerations apply to pyr-idines and analogous aromatic azahetero-cycles (reaction 2a). Note that these reactionsare reversible, a number of reductases being

able to deoxygenate N-oxides back to theamine (i.e., reactions 1b and 2b).

Secondary and primary amines also under-go N-oxygenation, the first isolable metabo-lites being hydroxylamines (reactions 3a and4a, respectively). Again, reversibility is docu-mented (reactions 3b and 4b). These com-pounds can be aliphatic or aromatic amines,and the same metabolic pathway occurs insecondary and primary amides (i.e., R¼ acyl),while tertiary amides appear resistant toN-oxygenation. The oxidation of secondaryamines and amides usually stops at the hydro-xylamine/hydroxylamide level, but formationof short-lived nitroxides (not shown) has beenreported.

R NCH3

CH3

(1)(2)

(3) N HR'

RN OH

R'

R

R NH2 R NHOH R N O R NO2(4)

(5) CH NH2

R'

RC

R'

RN OH CH N

R'

RO

f

g

e h

CR'

RO + NH3 C

R'

RO + NH2OH

(6)

NH

H R

N

R

(7) R NH NH R' R N N R'

R NH2 + R' NH2

c

ec

d

d

e

R, R' = alkyl or aryl

a

b

a

b

a

b

a

b

a

b

a

b

c

d

N

RR N

CH3

CH3

O

CR'

RNH2

+

R N N R'

O

NR

O

Figure 7. Major functionalization reactions involving nitrogen atoms in substrate molecules. The reactionsshown here aremainly oxidations (oxygenations and dehydrogenations) and reductions (deoxygenations andhydrogenations).


As opposed to secondary amines andamides, their primary analogs can be oxidizedto nitroso metabolites (reaction 4c), butfurther oxidation of the latter compounds tonitro compounds does not seem to occur invivo. In contrast, aromatic nitro compoundscan be reduced to primary amines via reac-tions 4e, 4d, and finally 4b. This is the case fornumerous chemotherapeutic drugs such asmetronidazole.

Note that primaryaliphatic amineshavingahydrogen on the alpha-carbon can display addi-tional metabolic reactions shown as reactions 5in Fig. 7. Indeed, N-oxidation may also yieldimines (reaction 5a), whose degree of oxidationis equivalent to that of hydroxylamines [40].Imines can be further oxidized to oximes (reac-tion 5c), which are in equilibrium with theirnitroso tautomer (reactions 5f and 5g).

1,4-Dihydropyridines, and particularly cal-cium channel blockers such as nivaldipine (7in Fig. 8), are effectively oxidized by cyto-chrome P450. The reaction is one of aromati-zation (reaction 6 in Fig. 7), yielding the cor-responding pyridine.

Dinitrogen moieties are also targets of oxi-doreductases. Depending on their substitu-ents, hydrazines are oxidized to azo com-pounds (reaction 7a), some of which can beoxygenated to azoxy compounds (reaction 7d).Another important pathway of hydrazines istheir reductive cleavage to primary amines(reaction 7c). Reactions 7a and 7d are rever-sible, the corresponding reductions (reactions7band7e) beingmediatedby cytochromeP450and other reductases. A toxicologically signif-icant pathway thus exists for the reduction ofsome aromatic azo compounds to potentiallytoxic primary aromatic amines (reactions 7band 7c).

2.5. Reactions of Oxidation and Reductionof Sulfur and Other Atoms

A limited number of drugs contain a sulfuratom, usually as a thioether. The major redoxreactions occurring at sulfur atoms in organiccompounds are summarized in Fig. 9.

Thiol compounds canbe oxidized to sulfenicacids (reaction 1a), then to sulfinic acids(reaction 1e), and finally to sulfonic acids (re-

SHR SOHR SOR 2H SOR 3H(1)fea

SR S R'

bR'SH

(2) SR R' SOR R' SOR 2 R'

(3) CR R'

S

CR R'

SO

CR R'

SO2

COR R'

d

c

dR = alkyl or aryl; R' = alkyl

a

b

a

b

c

c

Figure 9. Major reactions of oxidation and reduction involving sulfur atoms in organic compounds.

NH

CH3

NO2

H

N

O

O

CH3O

O CH3

CH3

7

Figure 8. The structure of nivaldipine (7).


action 1f). Depending on the substrate, thepathway is mediated by cytochrome P450and/or flavin-containing monooxygenases.Another route of oxidation of thiols is to dis-ulfides either directly (reaction 1c via thiylradicals), or by dehydration between a thioland a sulfenic acid (reaction 1b). However, ourunderstanding of sulfur biochemistry is in-complete, and much remains to be learned.This is particularly true for reductive reac-tions. While reaction 1c is well-known to bereversible (i.e., reaction 1d), the reversibilityof reaction 1a is unclear, while reduction ofsulfinic and sulfonic acids appears unlikely.

The metabolism of sulfides (thioethers) israther straightforward. Besides S-dealkyla-tion reactions discussed earlier, these com-pounds can also be oxygenated by monooxy-genases to sulfoxides (reaction 2a) and then tosulfones (reaction 2c). Here, it is known withconfidencethatreaction2ais indeedreversible,as documented bymany examples of reductionofsulfoxides (reaction2b,whilethereductionofsulfones has never been found to occur.

Thiocarbonyl compounds are also sub-strates of monooxygenases, forming S-monox-ides (sulfines, reaction3a) and thenS-dioxides(sulfenes, reaction 3c). As a rule, these meta-bolites cannotbe identifiedas suchdue to theirreactivity. Thus, S-monoxides rearrange tothe corresponding carbonyl by expelling a sul-fur atom (reaction 3d). This reaction is knownas oxidative desulfuration and occurs in thioa-mides and thioureas (e.g., thiopental). As forthe S-dioxides, they react very rapidly withnucleophiles, andparticularlywithnucleophi-lic sites in biological macromolecules. Thiscovalent binding results in the formation ofadducts of toxicological significance. Such amechanism is believed to account for the car-cinogenicity of a number of thioamides.

Some other elements besides carbon, nitro-gen, and sulfur can undergo metabolic redoxreactions. The direct oxidation of oxygenatoms in phenols and alcohols is well docu-mented for some substrates. Thus, the oxida-tion of secondaryalcohols by someperoxidasescan yield a hydroperoxide and ultimately aketone. Some phenols are known to be oxi-dized by cytochrome P450 to a semiquinoneand ultimately to a quinone. A classical ex-ample is that of the antiinflammatory drug

paracetamol (8 in Fig. 10, acetaminophen), aminor fraction of which is oxidized byCYP2E1to the highly reactive and toxic quinoneimine9.

Additional elements of limited significancein medicinal chemistry able to enter redoxreactions are silicon, phosphorus, arsenic, andselenium, among others (Fig. 11). Note how-ever that the enzymology and mechanisms ofthese reactions are insufficiently understood.For example, a few silanes have been shown toyield silanols in vivo (reaction 1). The sameapplies to some phosphines that can be oxy-genated to phosphine oxides by monooxy-genases (reaction 2).

Arsenicals have received some attentiondue to their therapeutic significance. Bothinorganic and organic arsenic compounds dis-play anAs(III)–As(V) redox equilibrium in thebody. This is illustratedwith the arsine-arsineoxide and arsenoxide–arsonic acid equilibria(reactions 3a and 3b and reactions 4b and 4c,respectively). Another reaction of interest isthe oxidation of arseno compounds to arsen-oxides (reaction 4a), a reaction of importancein the bioactivation of a number of chemother-apeutic arsenicals.

The biochemistry of organoselenium com-pounds is of some interest. For example, a fewselenols have been seen to be oxidized to se-lenenic acids (reaction 5a) and then to seleni-nic acids (reaction 5b).

2.6. Reactions of Oxidative Cleavage

A number of oxidative reactions presented inthe previous sections yield metabolic inter-mediates that readily undergo postenzymaticcleavage of a C�X bond (X being an heteroa-tom). As briefly mentioned, reactions 4a and

NHCOCH3

OH

NCOCH3

O

98

Figure 10. The structure of paracetamol (8) and itstoxic quinoneimine metabolite (9).


4b in Fig. 3 represent important metabolicpathways that affectmany drugs.WhenX¼N(by far the most frequent case), the metabolicreactions are known as N-demethylations, N-dealkylations, or deaminations, depending onthe moiety being cleaved. This is aptly exem-plified by the metabolic fate of fenfluramine(10 in Fig. 12). This withdrawn drug under-goes N-deethylation to norfenfluramine (11),

an active metabolite, and deamination to (m-trifluoromethyl)phenylacetone (12), an inac-tive metabolite that is further oxidized to m-trifluoromethylbenzoic acid (13).

When X¼O or S in reactions 4 (Fig. 3), themetabolic reactions are known asO-dealkyla-tions or S-dealkylations, respectively. O-De-methylations are a typical case of the formerreaction. And when X¼halogen in reactions

10 11

12

13

CH3

F3C NH

CH3

CH3

F3C NH2

CH3

F3C O F3C COOH

Figure 12. Fenfluramine (10), norfenfluramine (11), (m-trifluoromethyl)phenylacetone (12), andm-trifluor-omethylbenzoic acid (13).

(1) SiR' H

R''

R

SiR' OH

R''

R

(2) PR'

R

R''

(3) AsR'

R

R'' R''

As

R

R' O

(4) AsR As R AsR Ob

cAsOR 3H2

(5) SeHR SeOHR SeOR 2H

a

b

a

a

b

PR'

R''

R

O

Figure 11. Some selected reactions of oxidation and reduction involving silicon, phosphorus, arsenic andselenium in xenobiotic compounds.


5a and 5b (Fig. 3), loss of halogen can alsooccur and is known as oxidativedehalogenation.

The reactions of oxidative C�X cleavagediscussed above result from carbon hydroxy-lation and are catalyzed by cytochrome P450.However, N-oxidation reactions followed byhydrolytic C�Ncleavage can also be catalyzedby cytochrome P450 (e.g., reactions 5e and 5hin Fig. 7). The sequence of reactions 5a and 5ein Fig. 7 is of particular interest since it is themechanismbywhichmonoamine oxidase dea-minates endogenous and exogenous amines.

2.7. Reactions of Hydration and Hydrolysis

Hydrolases catalyze the addition of amoleculeof water to a variety of functional moi-eties [9,14,23]. Thus, epoxide hydrolase hy-drates epoxides to yield trans-dihydrodiols(reaction 1b in Fig. 5). This reaction is docu-mented for many arene oxides, in particularmetabolites of aromatic compounds, and ep-oxides of olefins. Here, amolecule of water hasbeen added to the substrate without loss of amolecular fragment, hence the use of the term“hydration” sometimes found in theliterature.

Reactions of hydrolytic cleavage (hydroly-sis) are shown in Fig. 13. They are frequent fororganic esters (reaction 1), inorganic esterssuch as nitrates (reaction 2) and sulfates (re-action 3), and amides (reaction 4). These reac-tions are catalyzed by esterases, peptidases,or other enzymes, but nonenzymatic hydroly-sis is also known to occur for sufficiently labilecompounds under biological conditions of pHand temperature.Acetylsalicylic acid, glyceroltrinitrate, and lidocaine are three representa-tive examples of drugs undergoing extensive

cleavage of the organic ester, inorganic ester,or amide group, respectively. The reaction is ofparticular significance in the activation ofester prodrugs.

3. CONJUGATION REACTIONS

3.1. Introduction

As defined in Section 1, conjugation reactions(also infelicitously known as phase II reac-tions) result in the covalent binding of anendogenous molecule or moiety to the sub-strate. Such reactions are of critical signifi-cance in the metabolism of endogenous com-pounds, witness the impressive battery of en-zymes that have evolved to catalyze them.Conjugation is also of great importance in thebiotransformation of xenobiotics, involvingparents compounds or metabolites thereof[15,19,24].

Conjugation reactions are characterized bya number of criteria:

1. The substrate is coupled to an endo-genous molecule sometimes desig-nated as the endocon . . .

2. . . . which is usually polar . . .3. . . . of medium molecular weight (ca.

100–300Da) . . .4. . . . and carried by a cofactor.5. The reaction is catalyzed by an enzyme

known as a transferase (Table 7).

First and above all, an endogenous mole-cule (called the endogenous conjugating moi-ety, and sometimes abbreviated as the“endocon”) is coupled to the substrate. This

COOR R' R COOH + OHR'

ONOR 2 OHR HNO3+

OSOR 3H OHR + H2SO4

(4) CONHR'R R COOH + NHR' 2

(1)

(2)

(3)

Figure 13. Major hydrolysis reactions involving esters (organic and inorganic) and amides.

CONJUGATION REACTIONS 421

Table 7. A Survey of Transferases (EC 2) [15,19,24,32]

Methyltransferases (EC 2.1.1)S-Adenosyl-L-methionine (SAM)

EC 2.1.1.6: catechol O-methyltransferase (COMT)EC 2.1.1.1: nicotinamide N-methyltransferase (NNMT)EC 2.1.1.8: histamine N-methyltransferase (HNMT)EC 2.1.1.28: noradrenaline N-methyltransferase (PNMT)EC 2.1.1.49: arylamine N-methyltransferase, indolethylamineN-methyltransferase (INMT)EC 2.1.1.9: Thiol S-methyltransferase (TMT)EC 2.1.1.67: Thiopurine S-methyltransferase (TPMT)

Sulfotransferases (EC 2.8.2) (SULT)30-Phosphoadenosine50-phosphosulfate (PAPS)

EC 2.8.2.1: aryl sulfotransferase (SULT1A1, 1A2, and 1A3)Thyroid hormone sulfotransferase (SULT1B1)SULT1C1, and SULT1C2EC 2.8.2.4: estrogen sulfotransferase (SULT1E1)EC 2.8.2.14: alcohol/hydroxysteroid sulfotransferase (SULT2A1)EC 2.8.2.2: hydroxysteroid sulfotransferase (SULT2B1a and 2B1b)EC 2.8.2.15: steroid sulfotransferaseEC 2.8.2.18: cortisol sulfotransferaseEC 2.8.2.3: amine sulfotransferase (SULT3A1)

UDP-Glucuronosyltransferases (2.4.1.17) (UGT)Uridine-50-diphospho-a-D-glucuronic acid (UDPGA)

Subfamily UGT1: UGT1A1, 1A3, 1A4 to 1A10Subfamily UGT2A: UGT2A1 to 2A3Subfamily UGT2B: UGT2B4, 2B7, 2B10, 2B11, 2B15, 2B17, 2B28Subfamily UGT3A: UGT3A1, 3A2Subfamily UGT8A: UGT8

AcetyltransferasesAcetylcoenzyme A (AcCoA) EC 2.3.1.5: N-acetyltransferase (NAT) NAT1 and NAT2

EC 2.3.1.56: aromatic-hydroxylamine O-acetyltransferaseEC 2.3.1.118: N-hydroxyarylamine O-acetyltransferase

Acyl-CoA synthetasesCoenzyme A (CoA) EC 6.2.1.1: short-chain fatty acyl-CoA synthetase (ACSS)

EC 6.2.1.2: medium-chain acyl-CoA synthetaseEC 6.2.1.3: long-chain acyl-CoA synthetase (ACSL)EC 6.2.1.7: cholate-CoA ligaseEC 6.2.1.25: benzoyl-CoA synthetase

AcyltransferasesXenobiotic acyl-Coenzyme A

EC 2.3.1.13: glycine N-acyltransferase (GLYAT)EC 2.3.1.71: glycine N-benzoyltransferaseEC 2.3.1.14: glutamine N-phenylacetyltransferaseEC 2.3.1.68: glutamine N-acyltransferaseEC 2.3.1.65: cholyl-CoA glycine-taurine N-acyltransferase (BAAT)EC 2.3.1.20: diacylglycerol O-acyltransferaseEC 2.3.1.22: 2-acylglycerol O-acyltransferaseEC 2.3.1.26: sterol O-acyltransferase (ACAT)

Glutathione S-transferases (EC 2.5.1.18) (GST)(Glutathione) Microsomal GST superfamily (homotrimers): MGST: MGST1 to MGST3

Cytoplasmic GST superfamily (homodimers, and a few heterodimers):GSTA: Alpha class, GST A1-1, A1-2, A2-2, A3-3, A4-4, A5-5GSTK: Kappa class, GST K1-1GSTM: Mu class, GSTM1a-1a, M1a-1b, M1b-1b, M2-2, M3-3, M4-4, M5-5GSTO: Omega class, GST O1-1, O2GSTP: Pi class: GST P1-1GSTS: Sigma class: GST S1GSTT: Theta class, GST T1-1, T2GSTZ: Zeta class, GST Z1-1


is the absolute criterion of conjugation reac-tions. Second, this endogenous molecule ormoiety is generally polar (hydrophilic) or evenhighly polar, but there are exceptions. Third,the size of the endocon is generally in therange 100–300Da. Fourth, the endogenousconjugating moiety is usually carried by acofactor, with the chemical bond linking thecofactor and the endocon being a high energyone such that the Gibbs energy released uponits cleavage helps drive the transfer of theendocon to the substrate. Fifth, conjugationreactions are catalyzed by enzymes known astransferases (EC 2) that bind the substrateand the cofactor in such a manner that theirclose proximity allows the reaction to proceed.Themetaphor of transferases being a “nuptialbed” has not escaped some biochemists. It isimportant from a biochemical and practicalviewpoint to note that criteria 2 to 5 consid-ered separately are neither sufficient nor ne-cessary to define conjugations reactions. Theyare not sufficient, since in hydrogenation re-actions (i.e., typical reactions of oxidoreduc-tion) the hydride is also transferred from acofactor (NADPH orNADH). And they are notnecessary, since they all suffer from someimportant exceptions.

3.2. Methylation

3.2.1. Introduction Reactions ofmethylationimply the transfer of a methyl group from thecofactor S-adenosyl-L-methionine (14, SAM).As shown in Fig. 14, the methyl group in SAMis bound to a sulfonium center, giving it amarkedelectrophilic character andexplainingits reactivity. Furthermore, it becomes phar-macokinetically relevant to distinguish

methylated metabolites in which the positivecharge has been retained or lost as a proton.

A number of methyltransferases are ableto methylate small molecules (see Table 7)[15,19,24,41]. Thus, reactions of methylationfulfill only two of the three criteria definedabove, since the methyl group is small com-pared to the substrate. The main enzymeresponsible for O-methylation is catechol O-methyltransferase, which is mainly cytosolicbut also exists in membrane-bound form. Sev-eral enzymes catalyze reactions of xenobioticN-methylation with different substrate speci-ficities, for example, nicotinamide N-methyl-transferase, histamine methyltransferase,phenylethanolamine N-methyltransferase(noradrenalineN-methyltransferase), and thenonspecific arylamine N-methyltransferase.Reactions of xenobiotic S-methylation aremediated by the membrane-bound thiolmethyltransferase and the cytosolic thiopur-ine methyltransferase.

The above classification of enzymes makesexplicit the three types of functionalities un-dergoing biomethylation, namely hydroxyl(phenolic), amino and thiol groups.

3.2.2. Methylation Reactions Figure 15 sum-marizes the main methylation reactions seenin drug metabolism [15,19,24,41].O-Methyla-tion is a common reaction of compounds con-taining a catechol moiety (reaction 1), with ausual regioselectivity for the meta position.The substrates can be xenobiotics and parti-cularly drugs, L-DOPA being a classic exam-ple. More frequently, however,O-methylationoccurs as a late event in themetabolism of arylgroups, after they have been oxidized to cate-chols (reactions 1, Fig. 5). This sequence wasseen for example in the metabolism of theantiinflammatory drug diclofenac (15 inFig. 16), which in humans yielded 30-hydro-xy-40-methoxy-diclofenac as a major metabo-lite with a very long plasma half-life.

Three basic types of N-methylation reac-tions have been recognized (reactions 2–4,Fig. 15). A number of primary and secondaryamines (e.g., some phenylethanolamines andtetrahydroisoquinolines) have been shown tobe substrates of N-methyltransferase (reac-tion 2). However, such reactions are seldomof significance in vivo, presumably due to

14

+

N

N N

N

NH2

OH OH

S

O

CH3

NH2

COOH

Figure 14. S-Adenosyl-L-methionine (14).


effective oxidative N-demethylation. Acomparable situation involves theN�Hgroupin an imidazole ring (reaction 3), as exempli-fied by histamine. A therapeutically relevantexample is that of theophylline (16) whose N(9)-methylation is masked by N-demethyla-tion in adult but not newborn humans.

N-Methylation of pyridine-type nitrogenatoms (reaction 4, Fig. 15) appears to be ofgreater in vivo pharmacological significancethan reactions 2 and 3, and this for two rea-sons. First, the resulting metabolites, beingquaternary amines, are more stable than ter-tiary or secondary amines towardN-demethy-

lation. And second, these metabolites are alsomore polar than the parent compounds, incontrast to the products of reactions 2 and 3.Good substrates are nicotinamide (17), pyri-dine and a number of monocyclic and bicyclicderivatives.

S-Methylation of thiol groups (reaction 5) isdocumented for such drugs as captopril (18)and 6-mercaptopurine. Other substrates aremetabolites (mainly thiophenols) resultingfrom the S�C cleavage of (aromatic) glu-tathione and cysteine conjugates (see below).Once formed, suchmethylthiometabolites canbe further processed to sulfoxides and sulfones

1615

N

CONH2

1817

HS

O

N

CH3

COOH

CH3N

N

N

N

O

O

CH3

H

NH

COOH

ClCl

3'

4'

Figure 16. Diclofenac (15), theophylline (16), nicotinamide (17), and captopril (18).

(1)

(2) N H

R'

RN CH3

R'

R

(3)

(4)

(5) R SH R S CH3

N H N CH3

N

RN CH3

R

+

OH

OH

R

OH

OCH3R

Figure 15. Major methylation reactions involving catechols, various amines, and thiols.


before excretion (i.e., reaction 2a and 2c inFig. 9).

From Fig. 15, it is apparent that methyla-tion reactions can be subdivided into twoclasses:

(a) Those where the substrate and the pro-duct have the same electrical state, aproton in the substrate having beenexchanged for a positively chargedmethyl group (reactions 1–3 and 5);

(b) Thosewhere theproduct has acquired apositive charge, namely has become apyridine-type quaternary ammonium(reaction 4).

3.3. Sulfonation

3.3.1. Introduction Sulfonation reactionsconsist in an SO3 moiety being transferredfrom the cofactor 30-phosphoadenosine 50-phosphosulfate (19, PAPS) (Fig. 17) to thesubstrate under catalysis by a sulfotransfer-ase [15,19,24,42,43]. The three criteria of con-jugation are met in these reactions. Sulfo-transferases, which catalyze a variety of phy-siological reactions, are soluble enzymes (seeTable 7). The phenol sulfotransferases includearyl sulfotransferases, thyroid hormone sulfo-transferase, SULT1C1 and SULT1C2, andestrogen sulfotransferase. The alcohol sulfo-

transferases include alcohol/hydroxysteroidsulfotransferase, hydroxysteroid sulfotrans-ferase, steroid sulfotransferase, and cortisolsulfotransferase. There is also an aminesulfotransferase.

The sulfate moiety in PAPS is linked to aphosphate group by an anhydride bridgewhose cleavage is exothermic and suppliesenthalpy to the reaction. The electrophilic�OH or�NH� site in the substrate will reactwith the leaving SO3moiety, forming an estersulfate or a sulfamate (Fig. 18). Some of theseconjugates are unstable under biological con-ditions and will form electrophilic intermedi-ates of considerable toxicological significance.

3.3.2. Sulfonation Reactions The sulfoconju-gation of alcohols (reaction 1 in Fig. 18) leads

(1)

(2)

(3)R'

NR

OH

R'CH

ROH

R'N

RH(4)

R'CH

R

+R'

CHR

OSO3

_

OSO3

R

_OH

R

R'N

ROSO3

_ R'N

R

+

R'N

RSO3

_

Figure 18. Major sulfonation reactions involving primary and secondary alcohols, phenols, hydroxylaminesand hydroxylamides, and amines.

19P

OOH

O

N

NN

N

NH2

O OH

OPOS

O

OH

O

O

O

O

_

_

Figure 17. 30-Phosphoadenosine 50-phosphosul-fate (19, PAPS).


to metabolites of different stabilities. Endo-genous hydroxysteroids (i.e., cyclic secondaryalcohols) form relatively stable sulfates, whilesome secondary alcohol metabolites of allyl-benzenes (e.g., safrole and estragole) mayform genotoxic carbocations. Primary alco-hols, for example, methanol and ethanol, canalso form sulfateswhose alkylating capacity iswell known [54]. Similarly, polycyclic hydro-xymethylarenes yield reactive sulfates be-lieved to account for their carcinogenicity.

In contrast to alcohols, phenols form che-mically stable sulfate esters (reaction 2). Thereaction is usually of high affinity (i.e., rapid),but the limited availability of PAPS restrictsthe amounts of conjugate being produced. Ty-pical drugs undergoing limited sulfonationinclude paracetamol (8 in Fig. 10) and diflu-nisal (20 in Fig. 19).

Aromatic hydroxylamines and hydroxyla-mides are good substrates for some sulfotrans-ferases and yield unstable sulfate esters (re-action 3 in Fig. 18). Indeed, heterolytic N�Ocleavage produces a highly electrophilic nitre-nium ion. This is a mechanism believed toaccount for part or all of the cytotoxicity of

arylamines and arylamides (e.g., phenacetin).In contrast, significantlymore stable productsare obtained upon N-sulfoconjugation ofamines (reaction 4). A fewprimary, secondary,and alicyclic amines are known to yield sulfa-mates. The significance of these reactions inhumans is still poorly understood.

An intriguing and very seldom reaction ofconjugation occurs forminoxidil (21 in Fig. 19),an hypotensive agent also producing hairgrowth. This drug is anN-oxide, and the actualactive form responsible for the different ther-apeutic effects is the N,O-sulfate ester (22).

3.4. Glucuronidation and Glucosidation

3.4.1. Introduction Glucuronidation is ama-jor and very frequent reaction of conjugation.It involves the transfer to the substrate of amolecule of glucuronic acid from the cofactoruridine-50-diphospho-a-D-glucuronic acid (23,UDPGA) (Fig. 20). The enzymes catalyzingthis reaction are known asUDP-glucuronosyl-transferases and consist in a number of pro-teins coded by genes of the UGT superfamily(see Table 7). The human UDPGT known to

21

OH

F

F

COOH

20

22

N

N NH2NH2

N

OSO3

+

_

N

N NH2NH2

N

O

Figure 19. Diflunisal (20), minoxidil (21), and its N,O-sulfate ester (22).


metabolize xenobiotics are the products ofmain two gene families, UGT1 and UGT2.These enzymes include UGT1A1 (bilirubinUDPGTs) and several UGT1A, as well as nu-merous phenobarbital-inducible or constitu-tively expressed UGT2B [15,19,24,44,45].

In addition to glucuronidation, this sectionbriefly mentions glucosidation, a minor meta-bolic pathway seen for a few drugs. Thesereactions are also catalyzed by UGT-glucuronosyltransferases.

3.4.2. Glucuronidation Reactions Glucuronicacid exists in UDPGA in the 1a-configuration,but the products of conjugation are b-glucur-onides. This is due to the mechanism of thereaction being a nucleophilic substitutionwith inversion of configuration. Indeed, andas shown in Fig. 21, all functional groups ableto undergo glucuronidation are nucleophiles,a common characteristic they share despitetheir great chemical variety. As a consequenceof this diversity, the products of glucuronida-

(1)

(2) R' C OH

R''

R

R' C O

R''

R

GLU

(3) R COOH R CO

(4)

(5) N OHR

R'N O

R

R'GLU

(6) R CO N R'

H

R CO N R'

GLU

O GLU

N HR

R'N

R

R'COOH N

R

R'CO O GLU

OH

R

O GLU

R

Figure 21. Major glucuronidation reactions involving phenols, alcohols, carboxylic acids, carbamic acids,hydroxylamines and hydroxylamides, carboxamides, sulfonamides, various amines, thiols, dithiocarboxylicacids, and 1,3-dicarbonyl compounds.

23

O

OHOH

COOH

OH

ON

NH

OH OH

OPOP

O

O

O

OH

O

OHO

Figure 20. Uridine-50-diphospho-a-D-glucuronic acid (23, UDPGA).


tion are classified as O-, N-, S- and C-glucuronides.

O-Glucuronidation is shown in reactions1–5 (Fig. 21). A frequent metabolic reactionof phenolic xenobiotics or metabolites is theirglucuronidation to yield polar metabolites ex-creted in urine and/or bile.O-Glucuronidationis often in competition withO-sulfonation (seeabove),with the latter reactionpredominatingat low doses and the former at high doses. Inbiochemical terms, glucuronidation is a reac-tion of low affinity and high capacity, whilesulfonation displays high affinity and low ca-pacity. A typical drug undergoing extensive

glucuronidation is paracetamol (8 in Fig. 10).Another major group of substrates are alco-hols, be they primary, secondary, or tertiary(reaction 2, Fig. 21). Medicinal examples in-clude chloramphenicol and oxazepam. An-other important example is that of morphine,which is conjugated on its phenolic and sec-ondary alcohol groups to form the 3-O-glucur-onide (a weak opiate antagonist) and the 6-O-glucuronide (a strong opiate agonist),respectively.

An important pathway of O-glucuronida-tion is the formation of acyl glucuronides(reaction 3). Substrates are arylacetic acids

SOR 2 R'N

H

SOR 2 R'N

GLU

(7)

(8)

(9)

(10) HNR

R'N

R

R'GLU

(11) CHN 3

R

R'

(12) SHR SR GLU

(13) CSSHR CSR S GLU

(14) COR CH2 CO R'

GLU =

HO

X

HO

O

OH

COOH

N

R

NR'

H

R

NR'

GLU

R

COR CH

GLU

CO R'

R'

N

CH3R

GLU+

N GLU+

R

Figure 21. (Continued).


(e.g., diclofenac, 15 in Fig. 16) and aliphaticacids (e.g., valproic acid). Aromatic acids areseldom substrates, a noteworthy exceptionbeing diflunisal (20 in Fig. 19) that yields boththe acyl and the phenolic glucuronides. Thesignificance of acyl glucuronides has long beenunderestimated perhaps because of analyticaldifficulties. Indeed, these metabolites arequite reactive, rearranging to positional iso-mers and binding covalently to plasma andseemingly also tissue proteins [46]. Thus, acylglucuronide formation cannot be viewedsolely as a reaction of inactivation and deto-xification.

A special class of acyl glucuronides are thecarbamoyl glucuronides (reaction 4 in Fig. 21).A number of primary and secondary amineshave been found to yield this type of conjugate,while as expected the intermediate carbamicacids are not stable enough to be character-ized. Carvedilol (24 in Fig. 22) is one drugexemplifying the reaction, in addition to form-ing an O-glucuronide on its alcohol group anda carbazole-N-linked glucuronide (see below).Much remains to be understood concerningthe chemical and biochemical reactivity ofcarbamoyl glucuronides.

Hydroxylamines and hydroxylamides mayalso form O-glucuronides (reaction 5, Fig. 21).Thus, a few drugs and a number of aromaticamines are known to be N-hydroxylated andthen O-glucuronidated. The glucuronidationof N�OH groups competes with O-sulfation,

but the reactivity of N-O-glucuronides to un-dergoheterolytic cleavageand formnitreniumions does not appear to be well characterized.

Second in importance to O-glucuronidesare the N-glucuronides formed by reactions6–11 inFig. 21, that is, amides (reactions 6–7),amines ofmediumbasicity (reactions 8 and 9),and basic amines (reactions 10 and 11). TheN-glucuronidation of carboxamides (reaction 6)is exemplified by carbamazepine (5 in Fig. 6)and phenytoin (25 in Fig. 22). In the lattercase,N-glucuronidationwas found to occur atN(3). The reaction has special significance forsulfonamides (reaction 7) and particularlyantibacterial sulfanilamides such as sulfadi-methoxine (26 in Fig. 22) since it produceshighly water-soluble metabolites that showno risk of crystallizing in the kidneys.

N-Glucuronidation of aromatic amines(reaction 8, Fig. 21) has been observed in afew cases only (e.g., conjugation of the carba-zole nitrogen in carvedilol (24). Similarly,there are a number of observations that pyr-idine-type nitrogens and primary and second-ary basic amines can be N-glucuronidated(reactions 9 and 10, respectively). As far ashuman drug metabolism is concerned, an-other reaction of significance is the N-glucur-onidation of lipophilic, basic tertiary aminescontaining one or twomethyl groups (reaction11). More and more drugs of this type (e.g.,antihistamines andneuroleptics), are found toundergo this reaction to a marked extent in

NH

O NH

O

OHO

CH3

24

25 26

N

N

O

NHS

O CH3

CH3

NH2

O

O3

NH

NH

O O

Figure 22. Carvedilol (24), phenytoin (25), and sulfadimethoxine (26).


humans, for example, cyproheptadine (27 inFig. 23).

Third in importance are theS-glucuronidesformed from aliphatic and aromatic thiols(reaction 12 in Fig. 21), and from dithiocar-boxylic acids (reaction 13) such as diethyl-dithiocarbamic acid, a metabolite of disulfir-am. As for C-glucuronidation (reaction 14),this reaction has been seen in humans for1,3-dicarbonyl drugs such as phenylbutazoneand sulfinpyrazone (28 in Fig. 23).

3.4.3. Glucosidation Reactions A few drugshave beenobserved to be conjugated to glucosein mammals. This is usually a minor pathwayin some cases where glucuronidation is possi-ble. An interesting medicinal example is thatof some barbiturates such as phenobarbitalthat yield the N-glucoside.

3.5. Acetylation and Acylation

All reactions discussed in this section involvethe transfer of an acyl moiety to an acceptorgroup. In most cases, an acetyl is the acylmoiety being transferred, while the acceptorgroup may be an amino or hydroxylfunction [15,19,24,47].

3.5.1. Acetylation Reactions The major en-zyme system catalyzing acetylation reactionsisarylamineN-acetyltransferase (seeTable7).Two enzymes have been characterized, NAT1and NAT2, the latter as two closely relatedisoformsNAT2A andNAT2Bwhose levels areconsiderably reduced in the liver of slow acet-ylators. The cofactor of N-acetyltransferase isacetylcoenzyme A (CoA-S-Ac, 29 with R¼acetyl) (Fig. 24) where the acetyl moiety isbound by a thioester linkage.

Two other activities, aromatic hydroxyla-mine O-acetyltransferase and N-hydroxyary-lamine O-acetyltransferase, are also involvedin the acetylation of aromatic amines andhydroxylamines (see below). Other acetyl-transferases exist, for example, diamineN-acetyltransferase (putrescine acetyltrans-ferase; EC 2.3.1.57) and aralkylamineN-acet-yltransferase (serotonin acetyltransferase;EC 2.3.1.87), but their involvement in xeno-biotic metabolism does not appear to bedocumented.

The substrates of acetylation, as schema-tized in Fig. 25, are mainly amines of mediumbasicity. Very few basic amines (primary orsecondary) of medicinal interest have beenreported to formN-acetylatedmetabolites (re-action 1), and when so the yields were low. Incontrast, a large variety of primary aromaticamines are N-acetylated (reaction 2). Thus,several drugs such as sulfonamides and para-aminosalicylic acid (30, PAS) (Fig. 26) are

29

OPO

O

OH

O OH

O

N

N N

N

NH2

P

OOH

O

P

O

OH

O

CH3OH

NH O

CH3

NH

OS

R

_

Figure 24. Acetylcoenzyme A (29, whenR¼ acetyl).

27

28

N N

OO

H

S

N

CH3

Figure 23. Cyproheptadine (27) and sulfinpyra-zone (28).


acetylated to large extents, not to mentionvarious carcinogenic amines such asbenzidine.

Arylhydroxylamines can also be acety-lated, but the reaction is one of O-acetylation(reaction 3a in Fig. 25). This is the reactionformally catalyzed by EC 2.3.1.118 with acet-yl-CoA acting as the acetyl donor, the N-hy-droxy metabolites of a number of arylaminesbeing known substrates. The same conjugates

can be formed by intramolecular N,O-acetyltransfer, when an arylhydroxamic acid (anN-aryl-N-hydroxy-acetamide) is substrate of,for example, EC 2.3.1.56 (reaction 3b). In ad-dition, such an arylhydroxamic acid can trans-fer its acetyl moiety to an acetyltransferase,which can then acetylate an arylamine or anarylhydroxylamine (intermolecularN,N- orN,O-acetyl transfer).

Besides amines, other nitrogen-containingfunctionalities undergo N-acetylation, hydra-zines, and hydrazides being particularly goodsubstrates (reaction 4, Fig. 25). Medicinal ex-amples include isoniazid (31 in Fig. 26) andhydralazine (32).

3.5.2. Other Acylation Reactions A limitednumber of studies have shown N-formylationto be a genuine route of conjugation for somearylalkylamines and arylamines, and particu-larly polycyclic aromatic amines. There is evi-dence to indicate that the reaction is catalyzedby arylformamidase (EC 3.5.1.9) in the pre-sence of N-formyl-L-kynurenine.

A very different type of reaction is repre-sented by the conjugation of xenobiotic

N

N

NHNH2

32

N

CONHNH2

31

OH

NH2

COOH

30

Figure 26. para-Aminosalicylic acid (30), isoniazid(31), and hydralazine (32).

(1)

(2)

N H

R'

R

(3)a

(4)

b

R NH NH2 R NHNH COCH3

N

R'

R

COCH3

NH COCH3

R

NH2

R

N H

O COCH3

R

NHOH

R

N

OH

COCH3

R

Figure 25. Major acetylation reactions involving aliphatic amines, aromatic amines, arylhydroxylamines,hydrazines and hydrazides.


alcohols with fatty acids, yielding highly lipo-philic metabolites accumulating in tissues.Thus, ethanol and haloethanols form esterswith, for example, palmitic acid, oleic acid,linoleic acid, and linolenic acid; enzymes cat-alyzing such reactions are cholesteryl estersynthase (EC 3.1.1.13) and fatty-acyl-ethyl-ester synthase (EC 3.1.1.67) [48,49]. Largerxenobiotics such as tetrahydrocannabinolsand codeine are also acylated with fatty acids,possibly by sterol O-acyltransferase (EC2.3.1.26).

3.6. Conjugation with Coenzyme Aand Subsequent Reactions

3.6.1. Conjugation with Coenzyme A The re-actions described in this section all have incommon the fact that they involve xenobioticcarboxylic acids forming an acyl-CoA meta-bolic intermediate (29 in Fig. 24, R¼ xenobio-tic acyl moiety). The reaction requires ATPand is catalyzed by various acyl-CoA synthe-tases of overlapping substrate specificity, forexample, short-chain fatty acyl-CoA synthe-tase, medium-chain acyl-CoA synthetase,long-chain acyl-CoA synthetase, and benzo-ate-CoA ligase (EC 6.2.1.25) (see Table 7).

The acyl-CoA conjugates thus formed areseldom excreted, but they can be isolated andcharacterized relatively easily in in vitrostudies. They may also be hydrolyzed back tothe parent acid by thiolester hydrolases (EC3.1.2). In a number of cases, such conjugateshave pharmacodynamic effects and may evenrepresent the active forms of some drugs, forexample, hypolipidemic agents. In the presentcontext, the interest of acyl-CoA conjugates istheir further transformation by a considerablevariety of pathways (Table 8) [15,19,24]. Themost significant routes are discussed below.

3.6.2. Formation of Amino Acid ConjugatesAmino acid conjugation is a major route fora number of small aromatic acids and involvesthe formation of an amide bond between thexenobiotic acyl-CoA and the amino acid. Gly-cine is the amino acidmost frequently used forconjugation (reaction 1 in Fig. 27), while a fewglutamine conjugates (reaction 2) and sometaurine conjugates (reaction 3) have beencharacterized in humans. The enzymes cata-

lyzing these transfer reactions are variousN-acyltransferases (see Table 7), for exampleglycine N-acyltransferase, glycine N-benzoyl-transferase, glutamine N-phenylacetyltrans-ferase and glutamine N-acyltransferase, andcholyl-CoAglycine-taurineN-acyltransferase.In addition, other amino acids can be used forconjugation in various animal species, for ex-ample, alanine, as well as some dipeptides [3].

The xenobiotic acids undergoing aminoacid conjugation are mainly substituted ben-zoic acids. In humans for example, hippuricacid and salicyluric acid (33 in Fig. 28) are themajormetabolites of benzoic acid and salicylicacid, respectively. Similarly, m-trifluoro-methylbenzoic acid (13 in Fig. 12), a majormetabolite of fenfluramine (10), is excreted asthe glycine conjugate. Phenylacetic acid deri-vatives can yield glycine and glutamine con-jugates. Some drugs containing a carboxylicgroupdo formthe taurine conjugateasaminormetabolite.

3.6.3. Formation of Hybrid Lipids and SterolEsters Incorporation of xenobiotic acids intolipids forms highly lipophilic metabolites thatmay burden the body as long retained resi-dues. In the majority of cases, triacylglycerolanalogs (reaction 4 in Fig. 27) or cholesterolesters (reaction 5) are formed. The enzymescatalyzing such reactions are O-acyltrans-ferases (see Table 7), including diacylglycerol

Table 8. Metabolic Consequences of theConjugation of Xenobiotic Acids toCoenzyme A (CoA-SH) [15,24]

Depending on structure and other factors, theresulting R-CO-S-CoA intermediatemay enter thefollowing metabolic reactions:. Hydrolysis. Formation of amino acid conjugates. Formation of hybrid triglycerides. Formation of phospholipids. Formation of cholesteryl esters. Formation of bile acid esters. Formation of acyl-carnitines. Protein acylation. Unidirectional configurational inversion of aryl-propionic acids (profens)

. Dehydrogenation and b-oxidation

. Two-carbon chain elongation


O-acyltransferase, 2-acylglycerolO-acyltrans-ferase, and sterol O-acyltransferase (choles-terol acyltransferase). Some phospholipidanalogs, as well as some esters to the 3-hydro-xy group of biliary acids, have also beencharacterized [15,24,48–52].

The number of drugs and other xenobioticsthat are currently known to form glyceryl orcholesteryl esters is limited, but should in-creasedue to increasedawareness of research-ers. One telling example is that of ibuprofen(34 in Fig. 28), a much used antiinflammatorydrug whose (R)-enantiomer forms hybrid tri-glycerides detectable in rat liver and adiposetissue.

3.6.4. Configurational Inversion of Arylpropio-nic Acids Ibuprofen (34 in Fig. 28) and otherarylpropionic acids (i.e., profens) are chiraldrugs existing as the (þ )-(S) eutomer and the(�)-(R) distomer. These compounds undergoan intriguingmetabolic reaction such that the(R)-enantiomer is converted to the (S)-enan-tiomer, while the reverse reaction is negligi-ble. This unidirectional configurational inver-sion is thus a reaction of bioactivation, and itsmechanism isnowreasonablywell understood(Fig. 29) [53].

The initial step in the reaction is the sub-strate stereoselective formation of an acyl-CoA conjugate with the (R)-form but not with

33

CH

CH3

COOH

34

CO

OH

NHCH2COOH

Figure 28. Salicyluric acid (33) and ibuprofen (34).

R COOH R CO NHCH2COOH

R COOH R CO NH CH COOH

CH2CH2CONH2

R COOH

R COOH

OCOR

(6)

a

b

(5)

(4)

(2)

(1)

R COOH(CH2 CH2)n

R (CH2 CH2) COOHn+1

R (CH2 CH2)n-1 COOH

CH2

CH

CH2

O

O

O

Acyl

Acyl

COR

R COOH(3) R CO NH CH2CH2 SO3

_

Figure 27. Metabolic reactions involving acyl-CoA intermediates of xenobiotic acids, namely conjugationsand two-carbon chain lengthening or shortening. Other products of b-oxidation are shown in Fig. 30.


the (S)-form. This conjugate then undergoes areaction of epimerization possibly catalyzedby methylmalonyl-CoA epimerase (EC5.1.99.1), resulting in a mixture of the (R)-profenoyl- and (S)-profenoyl-CoA conjugates.The latter can then be hydrolyzed as shown inFig. 29, or undergo other reactions such ashybrid triglyceride formation (see below).

3.6.5. b-Oxidation and Two-Carbon ChainElongation In some cases, acyl-CoA conju-gates formed from xenobiotic acids can alsoenter the physiological pathways of fatty acidcatabolism or anabolism. A few examples areknown of xenobiotic alkanoic and arylalkanoicacids undergoing two-carbon chain elonga-tion, or two-, four- or even six-carbon chainshortening (reactions 6a and 6b in Fig. 27). Inaddition, intermediatemetabolites ofb-oxida-tion may also be seen, as illustrated by val-proic acid (35 in Fig. 30). Approximately 50metabolites of this drug have been character-ized; they are formed by b-oxidation, glucur-onidation, and/or cytochrome P450-catalyzeddehydrogenation or oxygenation. Fig. 30shows the b-oxidation of valproic acid seen inmitochondrial preparations [25,54]. The re-sulting metabolites have also been found inunconjugated form in the urine of humans oranimals dosed with the drug.

3.7. Conjugation and Redox Reactionsof Glutathione

3.7.1. Introduction Glutathione (36 inFig. 31, GSH) is a thiol-containing tripeptideof major significance in the detoxification and

toxification of drugs and other xenobio-tics [15,24,55–57]. In the body, it exists in aredox equilibrium between the reduced form(GSH) and an oxidized form (GS-SG). Themetabolism of glutathione (i.e., its synthesis,redox equilibrium and degradation) is quitecomplex and involves a number of enzymes.

Glutathione reacts in a variety of manners.First, the nucleophilic properties of the thiol(or rather thiolate) group make it an effectiveconjugating agent, as emphasized in this sec-tion. Second, and depending on its redox state,glutathione can act as a reducing or oxidizingagent (e.g., reducing quinones, organic ni-trates, peroxides and free radicals, or oxidiz-ing superoxide). Another dichotomy exists inthe reactions of glutathione, since these can beenzymatic (e.g., conjugations catalyzed by glu-tathione S-transferases, and peroxide reduc-tions catalyzed by glutathione peroxidase) ornonenzymatic (e.g., some conjugations andvarious redox reactions).

The glutathione S-transferases (seeTable 7) are multifunctional proteins codedby two gene superfamilies. They can act asenzymes as well as binding proteins. Theseenzymes are mainly localized in the cytosol ashomodimers and heterodimers, but microso-mal enzymes also exist [58–60]. The GST A1-2, A2-2 and P1-1 display selenium-indepen-dent glutathione peroxidase activity, a prop-erty also characterizing the selenium-contain-ing enzyme glutathione peroxidase (EC1.11.1.9). The GST A1-1 and A1-2 are alsoknown as ligandin when they act as bindingor carrier proteins, a property also displayed

CAryl

H CH3

COOH

CAryl

H CH3

SCO CoA

(R)

CAryl

CH3H

COOH

(S)

CAryl

CH3H

SCO CoA

epimerization

X

Figure 29. Mechanism of the unidirectional configurational inversion of some profens.


byM1a-1a andM1b-1b. In the latter function,these enzymes bind and transport a number ofactive endogenous compounds (e.g., bilirubin,cholic acid, steroid and thyroid hormones, andhematin), as well as some exogenous dyes andcarcinogens.

Thenucleophilic character of glutathione isdue to its thiol group (pKa 9.0) in its neutralform and even more to the thiolate form. Infact, an essential component of the catalyticmechanism of glutathione transferase is themarked increase in acidity (pKa 6–7) experi-enced by the thiol group upon binding of glu-tathione to the active site of the enzyme. As aresult, GSTs transfer glutathione to a verylarge variety of electrophilic groups; depend-ing on the nature of the substrate, the reac-tions can be categorized as nucleophilic addi-tions or nucleophilic additions–eliminations.And with compounds of sufficient reactivity,these reactions can also occur nonenzy-matically [61].

Once formed, glutathione conjugates (R-SG) are seldom excreted as such (they are bestcharacterized in vitro or in the bile of labora-tory animals), but usually undergo furtherbiotransformation prior to urinary or fecal

36

37O

OH

S R

NH

CH3

O

NH

HOOC

NH2

O

O

NH

COOH

SH

Figure 31. Glutathione (36) and N-acetylcysteineconjugates (37).

COOH

35

COSCoA

COSCoA

COSCoA

OH

COSCoA

O

COSCoA

COSCoA

Figure 30. Mitochondrial b-oxidation of valproic acid (35).


excretion. Cleavage of the glutamyl moiety byglutamyl transpeptidase (EC 2.3.2.2), and ofthe cysteinylmoietyby cysteinylglycinedipep-tidase (EC 3.4.13.6) or aminopeptidase M (EC3.4.11.2), leaves a cysteine conjugate (R-S-Cys) that is further N-acetylated by cy-steine-S-conjugate N-acetyltransferase (EC2.3.1.80) to yield an N-acetylcysteine conju-gate (37 inFig. 31,R-S-CysAc). The latter typeof conjugates are known asmercapturic acids,a name that clearly indicates that they werefirst characterized inurine. This however doesnot imply that the degradation of unexcretedglutathione conjugates must stop at this

stage, since cysteine conjugates can be sub-strates of cysteine-S-conjugate b-lyase (EC4.4.1.13) to yield thiols (R-SH). These in turncan rearrange as discussed below, or beS-methylated and then S-oxygenated to yieldthiomethyl conjugates (R-S-Me), sulfoxides(R-SO-Me), and sulfones (R-SO2-Me).

3.7.2. Reactions of Glutathione The majorreactions of glutathione, both conjugationsand redox reactions, are summarized inFig. 32. Reactions 1 and 2 are nucleophilicadditions and additions-elimination to sp3-carbons, respectively, while reactions 3–8 are

(1)

(2) CHR 2 X CHR 2 SG

X = Cl, Br, OSO3

-

S CysAcCHR 2

(3)

a

ba

CC

X

HR'

R

aCHC 2X

R

SG

R'b

CHC 2X

R

S

R'

CysAc

(4)

, ...

X = COR'', CN, ...

OH

XH

b

O

X

a

OH

X

SG

(5)

X = O, NR

CC

X

XR

X

aCC

SG

XR

X

cCC

SH

XR

X

CC S

R

X

CX C SG

R

H

X

X

dCX C SH

R

H

X

X

CX C

S

XR

H

X = halogen

b

e

g

f

bO

H

H

R

H

OH

SG

HR

S CysAc

R

Figure 32. Major reactions of conjugation of glutathione, sometimes accompanied by a redox reaction.


nucleophilic substitutions or additions at sp2-carbons, sometimes accompanied by a redoxreaction.Reactionsatnitrogen or sulfur atomsare shown in reactions 9–11.

The first reaction in Fig. 32 is nucleophilicaddition to epoxides (reaction 1a) to yield anonaromatic conjugate. This is followed byseveral metabolic steps (reaction 1b) leadingto an aromatic mercapturic acid. This is afrequent reaction of metabolically producedarene oxides (Fig. 5), as documented fornaphthalene and numerous drugs and xeno-biotics containing a phenyl moiety. Note thatthe same reaction can also occur readily forepoxides of olefins (not shown in Fig. 32).

An important pathway of addition-elimina-tion at sp3-carbons is represented in reaction2a, followed by the production of mercapturicacids (reaction 2b). Various electron-with-drawing leaving groups (X in reaction 2) maybe involved that are either of xenobiotic (e.g.,halogens) or metabolic origin (e.g., a sulfate

group). Such a reaction occurs for example atthe �CHCl2 group of chloramphenicol and atthe NCH2CH2Cl group of anticancer alkylat-ing agents.

The reactions at sp2-carbons are quite var-ied and complex. Addition to activated olefinicgroups (e.g., a,b-unsaturated carbonyls) isshown in reaction 3. A typical substrate isacrolein (CH2¼CH�CHO). Quinones (orthoand para) and quinoneimines react withglutathione by two distinct and competitiveroutes, namely nucleophilic addition to form aconjugate (reaction 4a), and reduction to thehydroquinone or aminophenol (reaction 4b). Atypical example is provided by the toxic qui-noneimine metabolite (9 in Fig. 10) of para-cetamol (8). Since in most cases quinones andquinoneimines are produced by the bio-oxida-tion of hydroquinones and aminophenols,respectively, their reduction by GSH can beseen as a futile cycle that consumes reducedglutathione. As for the conjugates produced by

(6)

X = Cl, Br, NO2, SOR', SO2R', ...

(7) CR

Cl

O

CR

SG

O

(8) NR C X

X = O, S

NHR C SG

X

(9)

(10) ONOR 2 GS NO2 + OHR

GSSG + HNO2

a

b

(11) SHR SR S G

bX

R

SG

R

S CysAc

R

N O

R

NH SO G

R

a

Figure 32. (Continued).


reaction 4a, they may undergo reoxidation toS-glutathionyl quinones or S-glutathionylquinoneimines of considerable reactivity.These quinone or quinoneimine thioethers areknown to undergo further GSH conjugationand reoxidation.

Haloalkenes are a special group of sub-strates of GS-transferases since they may re-act with GSH either by substitution (reaction5a) or by addition (reaction 5b). The formationof mercapturic acids occurs as for other glu-tathione conjugates, but in this case S–C clea-vage of theS-cysteinyl conjugates by the renalb-lyase (reactions 5c and 5d) yields thiols ofsignificant toxicity since they rearrange byhydrohalide expulsion to form highly reactivethioketenes (reaction 5e) and/or thioacyl ha-lides (reactions 5f and 5g) [62].

With good leaving groups, nucleophilic aro-matic substitution reactions also occur at aro-matic rings containing additional electron-withdrawing substituents and/orheteroatoms(reaction 6a). As for the detoxification of acylhalides with glutathione (reaction 7), a goodexample is providedbyphosgene (O¼CCl2), anextremely toxic metabolite of chloroform thatis inactivated to the diglutathionyl conjugateO¼C(SG)2.

The addition of glutathione to isocyanatesand isothiocyanates has received some atten-tion due in particular to its reversible char-acter (reaction 8) [63]. Substrates of the reac-tionare xenobiotics suchas the infamous toxinmethyl isocyanate, whose glutathione conju-gate behaves as a transport form able to car-bamoylate various macromolecules, enzymesand membrane components. The reaction isalso of interest from a medicinal viewpointsince anticancer agents such as methylforma-mide appear to work by undergoing activationto isocyanates and then to the glutathioneconjugate.

The reaction of N-oxygenated drugs andmetabolites with glutathione may also havetoxicological and medicinal implications.Thus, the addition of GSH to nitrosoarenes,probably followed by rearrangement, formssulfinamides (reaction 9) that have been pos-tulated to contribute to the idiosyncratic toxi-city of a few drugs such as sulfonamides [64].As for organic nitrate esters such as nitrogly-cerine and isosorbide dinitrate, the mechan-

ism of their vasodilating action is now knownto result from their reduction to nitric oxide(NO). Thiols, and particularly glutathione,play an important role in this activation. Inthe first step, a thionitrate is formed (reaction10a) whose N-reduction may occur by morethan one route. For example, a GSH-depen-dent reductionmay yield nitrite (reaction 10b)that undergoes further reduction to NO; S-nitrosoglutathione (GS-NO)hasalso beenpos-tulated as an intermediate.

The formation of mixed disulfides betweenGSH and a xenobiotic thiol (reaction 11) hasbeenobserved ina few cases, for example,withcaptopril.

Finally, glutathione and other endogenousthiols (including albumin) are able to inacti-vate free radicals (e.g., R., HO., HOO., ROO.)and have thus a critical role to play in cellularprotection [55–57]. The reactions involved arehighly complex and incompletely understood;the simplest are

GSHþX. !GS. þXH

GS. þGS. !GSSG

GS. þO2 !GS�OO.

GS�OO. þGSH!GS�OOHþGS.

3.8. Other Conjugation Reactions

The above Sections 3.2–3.7 present the mostcommon and important routes of xenobioticconjugation, but these are not the only ones. Anumber of other routes have been reportedwhose importance is at present restricted toa few exogenous substrates, or that have re-ceived only limited attention [15,24]. In thepresent section, two pathways of pharmaco-dynamic significance will be mentioned,namely phosphorylation and carbonyl conju-gation (Fig. 33). In both cases, the xenobioticsubstrates belong to narrowly defined chemi-cal classes.

Phosphorylation reactions are of great sig-nificance in the processing of endogenous com-pounds and macromolecules. It is thereforeastonishing that relatively few xenobiotics aresubstrates of phosphotransferases (e.g., EC


2.7.1) to form phosphate esters (reaction 1 inFig. 33). The phosphorylation of phenol is acuriosity observed by some workers. In con-trast, anumber of antiviral nucleosideanalogsare known to yield the mono-, di- and tripho-sphates in vitro and in vivo, for example,zidovudine (38 in Fig. 34, AZT) and numerousanalogs. These conjugates are active forms ofthe drugs, being in particular incorporated inthe DNA of virus-infected cells [65].

The second pathway of conjugation to bepresented here is the reaction of hydrazineswith endogenous carbonyls (reaction 2) [62].This reaction occurs nonenzymatically andinvolves a variety of carbonyl compounds,namely aldehydes (mainly acetaldehyde)and ketones (e.g., acetone, pyruvic acidCH3�CO�COOH, and a-ketoglutaric acidHOOC�CH2CH2�CO�COOH). The productsthus formed are hydrazones, which may beexcreted as such or undergo further transfor-mation. Isoniazid (31 in Fig. 26) and hydrala-zine (32 in Fig. 26) are two drugs that formhydrazones in the body. For example, thereaction of hydralazine with acetyldehyde oracetone is a reversible one, meaning that thehydrazones are hydrolyzed under biological

conditions of pHand temperature. Inaddition,the hydrazone of hydralazine with acetalde-hyde or pyruvic acid undergoes an irreversiblereaction of cyclization to another metabolite,methyltriazolophthalazine (39 in Fig. 34).

4. DRUG METABOLISM AND THEMEDICINAL CHEMIST

Having presented the various reactions ofdrug metabolism, we now briefly review a fewtopics of potential interest to medicinal che-mists, namely the influence of metabolism onactivity and toxicity, the prediction of drugmetabolism, and the biological factors thatmay influence biotransformation.

4.1. Metabolism and Bioactivity

The pharmacological and toxicological conse-quences of drug metabolism explain to a largeextent the immense significance this disci-pline has acquired. Indeed, such consequencesare of utmost relevance in fields such as drugdiscovery and development, clinical pharma-cology and toxicology, and therapeutics.

4.1.1. Pharmacological Activity An impor-tant information in any drug’s dossier is theactivity (or lack thereof) of its metabolites.What should be realized, however, is that“activity” is usually understood to imply thesame pharmacological target as the parentmolecule [67–69]. Here, we outline variouspossible combinations, from a drug having noactive metabolite to an intrinsically inactive“drug” (i.e., a prodrug) whose therapeuticeffects necessitate metabolism to an activemetabolite (i.e., bioactivation).

A number of drugs have no activemetabolites, for example the 3-hydroxylated

NH

N

O

O

CH3

OOH

N3 38

NN

CH3

N

N

39

Figure 34. Zidovudine (38) and methyltriazo-lophthalazine (39).

CHR 2OH CH2 PO OH

OH

O

R(1)

NHNHR 2 NHR N C

R'

R''

(2)

Figure 33. Additional conjugation reactions discussed in Section 3.8, namely formation of phosphate estersand of hydrazones.

DRUG METABOLISM AND THE MEDICINAL CHEMIST 439

benzodiazepines suchas lorazepam, oxazepam(40 inFig. 35), and temazepam,whichundergoO-glucuronidation and cleavage reactions. Acase of drugs designed to have neither activenor toxic metabolites is that of soft drugs, aconcept pioneered and extensively developedby Bodor and collaborators [70–72]. In practi-cal terms, soft drugs (a) are bioisosteres ofknown drugs, (b) contain a labile bridge (oftenan ester group), and (c) are cleaved to metabo-lites known to lack activity and toxicity. Inother words, soft drugs bear a close stereoelec-tronic analogy with the target drugs, but arapid breakdown to inactive metabolites isprogrammed into their chemical structure. Atypical example is succinylcholine (41 inFig. 35), although the discovery of this agentpredates by decades the concept and term ofsoft drugs. In most individuals, this curarimi-metic agent is very rapidly hydrolyzed to cho-line and succinic acid byplasma cholinesterasewith a half-life of about four minutes [73].

Drugs having both intrinsic activity andactive metabolites may be more numerousthan generally believed, but information isnot always available. Several benzodiaze-pines have one or more active metabolite(s).

For example, triazolam (42 in Fig. 35) formsthe active 1-hydroxytriazolam, both agentshaving a short half-life compatible with theuse of the drug as a short-acting short-dura-tion hypnotic [74].

A drug whose activity owes much to meta-bolism is tamoxifen (43 in Fig. 35) as summar-ized here. This estrogen receptor antagonist isextensively used for endocrine treatment ofbreast cancer. Its metabolism is quite complexand leads to at least 13 oxidative metabolites,not to mention conjugation reactions. A phar-macologically significant route is aromatic oxi-dation to the active 4-hydroxytamoxifen. How-ever, themost importantmetabolic pathway inqualitative and quantitative terms is N-de-methylation to the secondary amine N-desmethyltamoxifen, a reaction catalyzedmainly by CYP3A4. This metabolite, whilereaching high plasma levels, is not as activeas 4-hydroxytamoxifen and the second genera-tion metabolite N-desmethyl-4-hydroxyta-moxifen (known as endoxifen). Indeed, a num-ber of investigations have demonstrated that4-hydroxytamoxifen and endoxifen are equipo-tent and many times more active as antiestro-gens than the parent drug [75]. When taking

40

NCl

NH O

OH3

41

OO

O

O

N(CH3)3

(CH3)3N+

+

CH3

O

N

CH3

CH3

43

442

NCl

NN

NCH3

Cl

1

Figure 35. Oxazepam (40) and succinylcholine (41) as examples of drugs having no active metabolite. Incontrast, part of the activity of triazolam (42) and tamoxifen (43) is due to one or more active metabolites.


plasma levels in consideration [76], it seemsthat themain contributor to therapeutic activ-ity in patients is endoxifen, and to a lesserextent tamoxifen and 4-hydroxy-tamoxifen.

An extreme case of activation by metabo-lism is that of prodrugs, namely intrinsicallyinactive (or poorly active) medicinal com-poundswhose biotransformationproduces theactive agent. This case is treated in a separatechapter and will not be considered here.

4.1.2. Toxicity The toxicological conse-quences of the metabolism of drugs and otherxenobiotics can be favorable (i.e., result indetoxification) or unwanted (i.e., result in tox-ification) [15,25]. The risks of toxification as-sociated with biotransformation [77–81] havenow become a major issue in drug discoveryand development, where minimizing meta-bolic toxification is given a high prior-ity [82–89], for example, screening for reactiveintermediates and assessing toxicity [90],with metabonomics [91,92] and toxicoge-nomics [93,94] being increasingly useful tools.

The main types of Averse Drug Reactions(ADRs) are summarized in Table 9[15,25,96–98]. On-target ADRs result from anexaggerated response caused by drug overdos-ing or too high levels of an active metabolite;they are predictable in principle and generallydose-dependent, and are labeled as Type A.Off-target ADRs result from the interaction ofthe drug or a metabolite with a nonintendedtarget such as a receptor or an enzyme. Theyalso are predictable in principle and generally

dose-dependent. These two types are pharma-cological in nature and fall outside the scope ofthe present work. ADRs caused by reactivemetabolites are the ones of greatest in ourcontext. They involve covalent binding tomacromolecules, and/or oxidative stress fol-lowing the formation of reactive oxygen spe-cies (ROSs). These ADRs are predictable (orrationalizable) in terms of the drug’s ormetabolite’s structure, and they are generallydose-dependent. They are often labeled asType C. Idiosyncratic drug reactions (IDRs)(also known as Type B ADRs) are rare to veryrare, unpredictable, and apparently dose-in-dependent. They are also poorly understood,yet appear to be usually due to reactivemetabolites.

Leaving aside on-target and off-target in-teractions, toxic responses are mostly due toreactive metabolites acting either directly onproteins and other macromolecules, or indir-ectly via reactive oxygen species, reactive ni-trogen species (RNSs), and reactive carbonylspecies (RCSs) [99–102]. Immunological reac-tions (not all of which should be classified asidiosyncratic) can be against self (autoimmu-nity) or against normal, harmless substances(allergy) [103–105]. In some cases, the immu-notoxic drug or chemical acts as a hapten (bydirect reaction with a protein). In most cases,however, the drug or chemical behaves as aprohapten since it needs metabolic activationto form adducts. In both cases, a hapten-car-rier conjugate is formed that may elicit animmune response.

Table 9. Types and Mechanisms of Adverse Drug Reactions (ADRs) [15,25]

On-target ADRs Predictable in principle and generally dose-dependent. Based on the pharmacologyof the drug and its metabolite(s), often an exaggerated response or a response in anontarget tissue

Off-target ADRs Predictable in principle and generally dose-dependent. Resulting from the inter-action of the drug or a metabolite with a nonintended target

ADRs involvingreactivemetabolites

Predictable in principle and generally dose-dependent. A major mechanism iscovalent binding to macromolecules (adduct formation) resulting in cytotoxic re-sponses, DNA damage, or hypersensitivity and immunological reactions. A distinct(andsynergetic)mechanism is the formationof reactiveoxygenspeciesandoxidativestress

Idiosyncratic drugreactions (IDRs)

Unpredictable, apparently dose-independent, and rare (<1 in 5000 cases). Theymight result from a combination of genetic and external factors, but their natureis poorly understood. IDRs include anaphylaxis, blood dyscrasias, hepatotoxicity,and skin reactions


Drug- or chemical-induced cytotoxicity canoccur throughapoptosis or necrosis,molecularmechanisms being immune injury, adduct for-mation, oxidative stress, and/orDNAdamage.Many organs are targets, but hepatotoxicity isof particular significance due to the high me-tabolizing capacity of the liver [106–111].Other organs include the kidneys (nephrotoxi-city), lungs and airways (pneumotoxicity), thecentral nervous system (neurotoxicity), repro-ductive organs, bone marrow (hematotoxi-city), and the skin (sensitization). As for ter-atogenesis, a number of embryotoxic chemi-cals are believed to be proteratogens activatedto reactive intermediates [112].

A look at toxicophoric groups (also calledtoxicophores, toxophores, or toxophoricgroups, see Table 10) is particularly illustra-tive of the unity that unlies their chemicaldiversity [95,113–115]. Indeed, the toxic po-tential of many toxicophores is explained bytheir metabolic toxification to electrophilicintermediates or to free radicals. In more de-tail, themajor functionalization reactions thatactivate toxophoric groups include oxidationto electrophilic intermediates, reduction tofree radicals, and autooxidation with oxygenreduction that leads to superoxide, otherreactive oxygen species, and reactive nitrogenspecies. The electrophilic intermediates andthe free radicals then react with bio(macro)molecules, mainly proteins and nucleic acids,producing chemical lesions. ROSs also reactwith unsaturated and mainly polyunsatu-rated fatty acids in membranes and else-

where, leading to lipid peroxidation. Of morerecent awareness is the fact that some con-jugation reactions may also lead to toxic me-tabolites, namely reactiveacyl glucuronides orconjugates with deleterious physicochemicalproperties [15,25].

However, it would be wrong to concludefrom the above that the presence of a toxopho-ric group necessarily implies toxicity. Realityis far less gloomy, as only potential toxicity isindicated. Given the presence of a toxophoricgroup in a compound, a number of factors willoperate to render the latter either toxic ornontoxic:

(a) The molecular properties of the sub-strate will increase or decrease its affi-nity and reactivity toward toxificationand detoxification pathways.

(b) Metabolic reactions of toxification arealways accompanied by competitiveand/or sequential reactions of detoxifi-cation that compete with the formationof the toxicmetabolite and/or inactivateit. A profusion of biological factors con-trol the relative effectiveness of thesecompetitive and sequential pathways.

(c) The reactivity and half-life of a reactivemetabolite control its sites of action anddetermine whether it will reach sensi-tive sites.

(d) Dose, rate, and route of entry into theorganism are all factors of knownsignificance.

Table 10. Major Toxophoric Groups and Their Metabolic Reactions of Toxification [15,25]

Functionalization reactions. Some aromatic systems that can be oxidized to epoxides, quinines, or quinonimines (Fig. 5, reaction 1). Ethynyl moieties activated by cytochrome P450 (Fig. 5, reaction 3). Some halogenated alkyl groups that can undergo reductive dehalogenation (Fig. 3, reaction 7). Nitroarenes that can be reduced to nitro anion-radicals, nitrosoarenes, nitroxides, and hydroxylamines(Fig. 7, reaction 4)

. Some aromatic amides that can be activated to nitrenium ions (reaction 3 in Fig. 7, followed by reaction 3 inFig. 18)

. Some thiocarbonyl derivatives, particularly thioamides, which can be oxidized to S,S-dioxide (sulfene)metabolites (Fig. 9, reaction 3)

. Thiols that can form mixed disulfides (Fig. 9, reaction 1)

Conjugation reactions. Some carboxylic acids that can form reactive acylglucuronides (Fig. 21, reaction 3). Some carboxylic acids that can form highly lipophilic conjugates (Fig. 27, reactions 4 and 5)


(e) Above all, there exist essentialmechan-isms of survival value that operate torepair molecular lesions, remove themimmunologically, and/or regenerate thelesioned sites.

In conclusion, the presence of a toxophoricgroup is not a sufficient condition for obser-vable toxicity, a sobering and often underem-phasized fact. Nor is it a necessary conditionsince other mechanisms of toxicity exist, forexample the acute toxicity characteristic ofmany solvents.

4.2. Structure–Metabolism Relationshipsand the Prediction of Metabolism

Two classes of factors will influence qualita-tively (how ? what ?) and quantitatively (howmuch ? how fast ?) the metabolism a givenxenobiotic in a given biological system. Thesecond class consists in the biological factorsbriefly presented below. The first class of fac-tors are the various molecular properties thatwill influence a metabolic reaction, most no-tably (a) global molecular properties such asconfiguration (e.g., chirality), electronicstructure, and lipophilicity and (b) local prop-erties of the target sites such as steric hin-drance, electron density, and reactivity.Whenspeaking of a metabolic reaction, however, weshould be clear about what is meant. Indeed,enzymekinetics allows ametabolic reaction tobe readily decomposed into a binding and acatalytic phase. In spite of its limitations,Michaelis–Menten analysis offers an informa-tive approach for assessing the binding andcatalytic components of a metabolic reac-tion [116]. Here, the Michaelis constant Km

represents the affinity, with Vmax being themaximal velocity, kcat the turnover number,and Km/Vmax the catalytic efficiency.

4.2.1. Chirality and Drug Metabolism The in-fluence of stereochemical factors in xenobioticmetabolism is a well-known and reviewedphenomenon [117]. Because metabolic reac-tions canproducemore than one response (i.e.,severalmetabolites), two basic types of stereo-selectivity are seen, namely substrate stereo-selectivity and product stereoselectivity (seeSection 1.3). Substrate stereoselectivity oc-

curs when stereoisomers are metabolized dif-ferently (in quantitative and/or qualitativeterms) and by the “same” biological systemunder identical conditions. Substrate stereo-selectivity is a well-known and abundantlydocumented phenomenon under in vivo andin vitro conditions. In fact, it is the rule formany chiral drugs, ranging from practicallycomplete to moderate. A complete absence ofsubstrate enantioselectivity has seldom beenseen.

Michaelis–Menten analysis suggests thatthe molecular mechanism of substrate enan-tioselectivity can occur in the binding step(different affinities Km), in the catalytic step(different reactivities, Vmax) or in both. Thereare cases in which stereoselective metabolismis of toxicological relevance. This situation canbe illustrated with the antidepressant drugmianserin that undergoes substrate stereose-lective oxidation. In human liver microsomes,mianserin (44 in Fig. 36) occurred by aromaticoxidation with amarked preference for its (S)-enantiomer, while N-demethylation was themajor route for the (R)-enantiomer. At lowdrug concentrations, cytotoxicity toward hu-man mononuclear leucocytes was due to (R)-mianserin more than to (S)-mianserin,and showed a significant correlation withN-demethylation [118]. Thus, the toxicity ofmianserin seemed associated with N-de-methylation rather than with aromatic oxida-tion. The chemical nature of the toxic inter-mediates was not established.

Product stereoselectivity occurs whenstereoisomeric metabolites are generated dif-ferently (in quantitative and/or qualitativeterms) and from a single substrate with asuitable prochiral center or face. Examples ofmetabolic pathways producing new centers ofchirality in substrate molecules include ke-tone reduction, reduction of carbon–carbondouble bonds, hydroxylation of prochiralmethylenes, oxygenation of tertiary aminesto N-oxides, and oxygenation of sulfides tosulfoxides. Product stereoselectivity may bedue to the action of distinct isoenzymes, or itmay result from different binding modes of aprochiral substrate to a single isoenzyme. Inthis case each productive binding mode willbring another of the two enantiotopic or dia-stereotopic target groups in the vicinity of the


catalytic site, resulting in diastereoisomericenzyme–substrate complexes. Thus, the ratioof products depends on the relative probabilityof the two binding modes. In addition, thecatalytic step, which involves diastereoiso-meric transition states, may also influence orcontrol product selectivity.

The potential pharmacological conse-quence of product-selective metabolism canbe illustrated with the antiemetic 5-HT3 an-tagonist dolansetron (45 in Fig. 36). This com-pound is prochiral by virtue of a center ofprochirality at the keto group featuring twoenantiotopic faces. This drug is reduced ra-pidly, extensively and stereoselectively in hu-mansto its (R)-alcohol (46 inFig. 36) [119,120].The enzymes involved are aldo-keto reduc-tases (AKR1C1, 1C2, and 1C4) and carbonylreductases. The metabolite proved to bemanyfold more active than dolasetron, allow-ing the latter to be viewed as a prodrug. Andsignificantly, the (R)-alcohol ismarkedlymoreactive than its (S)-enantiomer.

It thus appears that substrate and productstereoselectivities are the rule in the metabo-lism of stereoisomeric and stereotopic drugs.

But if the phenomenon is to be expected per se,it is not trivial to predict which enantiomerwill be the preferred substrate of a givenmetabolic reaction, orwhich enantiomericme-tabolite of a prochiral drug will predominate.This is due to the fact that the observed stereo-selectivity of a given reactionwill depend bothon molecular properties of the substrate andon enzymatic factors (e.g., the stereoelectronicarchitecture of the catalytic site in the variousisozymes involved). In fact, substrate and pro-duct stereoselectivities are determined by thebinding mode(s) of the substrate and by theresulting topography of the enzyme–substratecomplex.

4.2.2. Relations Between Metabolism and Lipo-philicity Comparing the overall metabolismof numerous drugs clearly reveals a globalrelation with lipophilicity [116]. Indeed, thereexist some highly polar xenobiotics known tobe essentially resistant to any metabolic reac-tion, for example, saccharin, disodium cromo-glycate, and zanamivir. Furthermore, manyin vivometabolic studies have demonstrated adependence of biotransformation on lipophili-

N

N

CH3

44

(R )-46

N

NH

O

O

O

N

OH

O

NH

O

H

45

Figure 36. Mianserine (44), dolasetron (45), and its active alcohol metabolite (46).


city, suggesting a predominant role for trans-port andpartitioningprocesses.Aparticularlyillustrative example is offered by b-blockers,where the more lipophilic drugs are exten-sively if not completely metabolized (e.g., pro-pranolol), whereas the more hydrophilic onesundergo biotransformation for only a fractionof the dose (e.g., atenolol).

This global trend is in line with the Darwi-nian rationale for xenobiotic metabolism,which is believed to have evolved in an ani-mal–plant “warfare,” with herbivores adap-tating to the emergence of protective chemi-cals (e.g., alkaloids) in plants [121]. The ex-ception to the global and direct relation be-tween extent of metabolism and lipophilicityis offered by the vast number of human-made,highly lipophilic polyhalogenated xenobiotics,which now pollute our entire biosphere. Suchcompounds include polyhalogenated insecti-cides (e.g., DDT), polyhalogenated biphenyls,and dioxins, which have a strong propensity toaccumulate in adipose tissues. In addition,these compounds are highly resistant to bio-transformation in animals due in part to theirvery high lipophilicity, and in part to the stericshielding against enzymatic attack providedby the halo substituents.

When the results of Michaelis–Mentenanalyses are examined for quantitative struc-ture–metabolism relationships (QSMRs), it isoften found that lipophilicity correlates withKm but not with Vmax or kcat. This is documen-ted for ester hydrolysis and oxidation of var-ious chemical series by monooxygenases andother oxidases [116]. Depending on the ex-plored property space, the relationships be-tween Km and lipophilicity are linear or para-bolic. Such results indicate that when rela-tions are found between rate of metabolismand lipophilicity, the energy barrier of thereaction is largely similar for all compoundsin the series, allowing lipophilicity to becomethe determining factor. A relevant example isprovided by the CYP2D6-catalyzed oxidationof a series of fluorinated propranolol deriva-tives. Their Km values spanned a 400-foldrange and showed a weak but real correlationwith the distribution coefficient. The samewas true for the catalytic efficiency kcat/Km,but only because the kcat values spanned anarrow sixfold range [122].

4.2.3. The Influence of Electronic FactorsStereoelectronic properties may influence thebinding of substrates to enzymatic active sitesin the same manner as they influence thebinding of ligands to receptors. For example,theKm values of the above-mentioned propra-nolol derivatives were highly correlated withtheir basicity (pKa), confirming that an ionicbond plays a critical role in the binding ofamine substrates to CYP2D6 [122].

Electronic properties are of particular in-terest in structure–metabolism relationshipssince they control the cleavage and formationof covalent bonds characteristic of a biotrans-formation reaction (i.e., the catalytic step).Correlations between electronic parametersand catalytic parameters obtained fromin vitro studies (e.g., Vmax or kcat) allow arationalization of substrate selectivity andsome insight into reaction mechanism.

An example of a quantitative SMR studycorrelating electronic properties and catalyticparameters is provided by the glutathioneconjugation of para-substituted 1-chloro-2-ni-trobenzene derivatives [123]. The values oflog k2 (second order rate constant of the none-nzymatic reaction) and log kcat (enzymatic re-action catalyzed by various glutathione trans-ferase preparations) were correlated with theHammett resonances-value of the substrates,ameasure of their electrophilicity. Regressionequationswith positive slopes and r2 values inthe range 0.88–0.98 were obtained. These re-sults quantitate the influence of substrateelectrophilicity on nucleophilic substitutionsmediated by glutathione, be they enzymatic ornonenzymatic.

Quantum mechanical calculations mayalso shed light on SMRs, revealing correla-tions between rates ofmetabolic oxidation andenergy barrier in cleavage of the target C�Hbond [124].

4.2.4. 3D-QSMRs and Molecular ModelingWhen lipophilicity and electronic parametersare used as independent variables and a me-tabolic parameter (often assessing affinity) asthe dependent variable, correlation equationsare obtained for rather limited and/or relatedseries of substrates, implying exploration of arelatively narrow structural diversity space.Also, the metabolic parameters are usually


obtained from relatively simple biological sys-tems, resulting in QSMR correlations that aretypically “local” methods (Table 11) of lowextrapolative capacity.

Quantum mechanical methods are alsoclassified as local in Table 11. Here, a wordof caution is necessary, since suchmethods areinprinciple applicable to any chemical system.However, they cannot handle more than onemetabolic reaction or catalyticmechanismatatime, and as such can only predict metabolismin simple biological systems, in contrast to theglobal methods presented below.

Three-dimensional (3D) methods are alsoof value in SMRs, namely 3D-QSARs and themolecular modeling of xenobiotic-metaboliz-ing enzymes (Table 11). Indeed, they repre-sent a marked progress in predicting the me-tabolic behavior (be it as substrates or inhibi-tors) of novel compounds. An obvious restric-tion, however, is that each enzyme requires aspecific model. 3D-QSMRs methods yield apartial view of the binding/catalytic site of agiven enzyme as derived from the 3D molecu-lar fields of a series of substrates or inhibitors(the training set). In other words, they yield a“photographic negative” of such sites, andwillallow a quantitative prediction for novel com-pounds structurally related to the training set.

Two popular methods in 3D-QSARs are CoM-FA (comparativemolecular field analysis) andcatalyst. Numerous applications can be foundin the literature.

The same is true for the molecular model-ing of xenobiotic-metabolizing enzymes,which affords another approach to rationalizeand predict drug–enzyme interactions. Themethodology of molecular modeling is ex-plained in detail elsewhere in this work andwill not bepresentedhere. Suffice it to say thatits application to drug metabolism was madepossible by the crystallizationandX-ray struc-tural determination of the first bacterial cyto-chromesP450 in themid-1980s [125], followeda few years ago by human cytochromesP450 [126,127]; the crystal structure of nu-merous other drug-metabolizing enzymes isalso available. As more and more amino acidsequences of drug-metabolizing enzymes be-come available, their tertiary structure can bemodeled by homology superimposition withthe experimentally determined template of aclosely related protein (homology modeling).Known substrates and inhibitors can then bedocked in silico. Given the assumptions madein homology modeling and in scoring func-tions, such models cannot give quantitativeaffinity predictions. However, they can afford

Table 11. A Classification of In Silico Methods to Predict Biotransformation [14,21]

(A) “Local” methodsMethods applicable to series of compoundswith “narrow” chemical diversity, and/or to biological systems of“low” complexity:

. QSAR: linear, multilinear, multivariate (may predict affinities, relative rates, and so on, depending onthe physicochemical properties considered)

. Quantum mechanical: MO methods (may predict regioselectivity, mechanisms, relative rates, etc.)

. 3D-QSAR: CoMFA�, Catalyst�, GRID/GOLPE�, for example (may predict substrate behavior,relative rates, inhibitor behavior, etc.)

. Molecular modeling and docking (may predict substrate or inhibitor behavior, regioselectivity, etc.)

. Experts systems combining docking, 3D-QSAR and MO: MetaSite�, and so on

(B) “Global” methodsMethods applicable to series of compounds with “broad” chemical diversity (and, in the future, to biologicalsystems of “high” complexity):

. “Meta”-systems combining (A) docking, 3D-QSAR, MO, and B) a number of enzymes and otherfunctional proteins: MetaDrug�, and so on

. Databases: Metabolite�, Metabolism�, and so on (allow metabolites to be deduced by analogy)

. Rule-based expert systems: Meta�, MetabolExpert�, Meteor�, and so on (may predict metabolites,metabolic trees, reactive/adduct-forming metabolites, relative importance of these metabolites)


fairly reliable yes/no answers as to the affinityof test set compounds, and in favorable casesmay also predict the regioselectivity of meta-bolic attack (Table 11).

The last approach among “local” systemsare expert systems combining (a) 3D-modelsobtained by molecular modeling and (b) so-phisticated QSAR approaches based onmulti-variate analyses of parameters obtained frommolecular interaction fields (MIFs), as foundin theMetaSite algorithm [128,129].MetaSiteis a specific system in the sense that it iscurrently restricted to the major human cyto-chromes P450. At the end of the procedure theatoms of the substrate are ranked according totheir accessibility and reactivity. In otherwords,MetaSite takes the 3D stereoelectronicstructure of both the enzyme and the ligandinto account to prioritize the potential targetsites of CYP-catalyzed oxidation in themolecule.

4.2.5. “Global” Expert Systems to Predict Bio-transformation Whilemedicinal chemists arenot usually expected to possess a deep knowl-edge of the mechanistic and biological factorsthat influence drugmetabolism, they will findit quite useful to have a sufficient understand-ing of structure–metabolism relationships tobe able to predict reasonable metabolicschemes. A qualitative prediction of the bio-transformation of a novel xenobiotic shouldallow (a) the identification of all target groupsand sites of metabolic attack, (b) the listing ofall possible metabolic reactions able to affectthese groups and sites, and (c) the organiza-tion of the metabolites into a metabolic tree.The next desirable features would be a warn-ing for potentially reactive/adduct-formingmetabolites, and a bridge to a toxicity-predict-ing algorithm.

Given the available information, there ex-ists an ever increasing interest in expert sys-tems hopefully able to meet the abovegoals [130–132]. A few systems of this typeare now available, for example a systemknown as MetaDrug combining (a) docking,3D-QSAR, MO and (b) a number of enzymesand other functional proteins such as trans-porters [133,134]. Metabolic databases allowreasoning by analogy and can prove quiteuseful [130]. Rule-based systems are

perhaps the most versatile; they includeMeta[135,136], MetabolExpert [137,138], and Me-teor [139–141]. Such systems will make cor-rect qualitative or even semiquantitative pre-dictions for a number of metabolites, but therisk of false positives must be taken seriously.This is due to the great difficulty of devisingefficient filters to remove unlikely metabo-lites, based on the molecular properties ofsubstrates. Taking biological factors (tissues,animal species, etc.) into account, is still faraway.

4.2.6. Biological Factors Influencing Drug Me-tabolism A variety of physiological andpathological factors influence xenobiotic me-tabolism and hence the wanted and unwantedactivities associated with a drug. This issuecomplicates significantly the task ofmedicinalchemists, as optimizing the pharmacokineticproperty of a given lead for behavior in labora-tory animals often results in clinical failure.While it is not the objective of this chapter todiscuss biological factors in any depth, a briefoverview will nevertheless be presented forthe sake of clarity and to help medicinal che-mists prioritize the challenges they face. Astructured, comprehensive and amply illu-strated discussion of these factors can befound in recent works [15,26,27].

It can be helpful to distinguish betweeninterindividual and intraindividual factorsthat can influence the capacity of an indivi-dual tometabolize drugs (Table 12). The inter-individual factors are viewed as remainingconstant throughout the life span of an organ-ism and are the expression of its genome. Incontrast, the intraindividual factors may varydepending on time (age, even time of day),pathological states, or external factors (nutri-tion, pollutants, drug treatment).

The most significant interindividual factorto be considered in early drug development isclearly the species-related differences. Theuse of different species as a surrogate forinvestigating and predicting metabolism inhumans has been and remains an indispen-sable step in drug development. In parallel,there has been a steady switch during earlyphases of drug discovery from obtaining me-tabolism data with animal tissues and wholeanimals to screening studies with human


cDNA expressed enzymes, cell fractions (mi-crosomes), and cells. Nevertheless, a greaterunderstanding of the molecular, genetic, andphysiological differences between humansand animal models is still required for cor-rectly interpreting pharmacological safetystudies required by regulatory agencies [142].

Genetic factors may result in polymorphicbiotransformations that may cause seriousproblems at the stage of clinical trials andpostmarketing. As much as possible, lead op-timization may try to avoid affinity for cyto-chromes P450 known to exhibit phenotypicdifferences. After the specific CYP(s) involvedin the metabolism of a drug are known, howdoes this information relate to the overallmetabolism of a drug in a diverse population?The ability for an individual to metabolize adrug is dependent on the nature (genotype),location and amount of enzyme present.Studying such factors is the field of pharma-cogenetics and pharmacogenomics. At the le-vel of populations, ethnopharmacology inves-tigates the differences in proportions of“normal” and slow metabolizers observed indifferent populations, a study complicated bythe unavoidable influence of external factorssuch as nutrition and lifestyle cannot be ex-cluded. Comparable arguments are valid for

sex-related differences in metabolism, whichin humans are known for a limited number ofdrugs only [143].

Intraindividual factors are numerous andfor some difficult to investigate. Amajor one isage, since differences between the levels ofmetabolism enzymes for the fetal and neona-tal (first four weeks postpartum) liver versusthe adult liver have been observed in bothanimals and humans [144]. The problem isalso significant for the growing population ofelderly patients, where a marked reduction isusually observed. The origin of such differ-ences is quite complex and often related toother physiological changes such as reductionin liver blood flow and liver size. As the num-ber of older individuals increase in the popu-lation, especially those aged 80–90 years,adaptations in drug therapy and posology asrelated to metabolism become increasinglyimportant.

There is increasing evidence that drug me-tabolism may undergo some variations due todiurnal, monthly and yearly rhythms. Thus, itis well established that nearly all physiologi-cal functions and parameters can varywith the time of day, for example, heart rate,blood pressure, hepatic blood flow, urinarypH, and plasma concentrations of hormones,other signal molecules, glucose, and plasmaproteins [145].

A number of diseases may sometimes influ-ence drug metabolism, although no general-ization appears in sight. First and above all,diseases of the liver will have a direct effect onenzymatic activities in this organ, for example,cirrhosis, hepatitis, jaundice, and tumors. Dis-eases of extra-hepatic organs such as the lungsand the heart may also affect hepatic drugmetabolism by influencing blood flow, oxyge-nation, and other vital functions. Renal dis-eases have a particular significance due to adecrease of the intrinsic xenobiotic-metaboliz-ingactivityof thekidneys,andadecreasedrateof urinary excretion. Infectious and inflamma-tory diseases have been associated with differ-ential expression of various CYP enzymes.

Interference with the metabolism of adrug by inhibition of the primary enzymesinvolved in its metabolism can lead to seriousside-effects or therapeutic failure. Inhibitorsof the CYP enzymes are considered to be

Table 12. Biological Factors AffectingXenobiotic Metabolism [15,26,27]

InterindividualFactors

IntraindividualFactors

Constant for a givenorganism

Variable during thelifetime of a givenorganism

Animal species Physiological changes:AgeBiological rhythmsPregnancy

Genetic factors (geneticpolymorphism)

Pathological changes:DiseaseStress

Sex External influences:NutritionEnzyme induction byxenobioticsEnzyme inhibition byxenobiotics


the most problematic, being the primarycause of drug–drug and drug–food interac-tions [16,146].

A comparable situation is created by en-zyme induction, that is, when the amount andactivity of a metabolizing enzyme has in-creased following exposure to a drug or chemi-cal. In contrast to inhibition, induction is gen-erally a slow process caused by an increase inthesynthesis of theenzyme.Adrug that causesinduction may increase its own metabolism(autoinduction) or increase metabolism of an-other drug. Clinically, drug induction is con-sidered less of a problem than drug inhibition,but there are some marked exceptions. Pri-mary interest has been directed toward theCYP enzymes for which a number of classesof inducing agents have been identified.

5. CONCLUDING REMARKS

Drug discovery and development are becom-ing more complex by the day, with physico-chemical, pharmacokinetic and pharmacody-namic properties being screened and as-sessed as early and as simultaneously aspossible. But the real challenge lies with theresulting deluge of data, which must bestored, analyzed and interpreted. This callsfor a successful synergy between humans andalgorithmic machines, in other words be-tween human and artificial intelligence.However, making sense of the data, inter-preting their analyses, and rationally plan-ning subsequent steps is an entirely differentissue. The difference is a qualitative one, andit is the difference between information andknowledge.

This chapter is about both. Our first objec-tive in writing it was obviously to supply use-ful information, namely structured data asexemplified by the classification of metabolicreactions (Sections 2 and 3).

Information becomes knowledge when it isconnected to a context from which it receivesmeaning. This treatise is about medicinalchemistry, and indeed medicinal chemistry isthe context of our chapter and of all others. Bydiscussing the connection between drug me-tabolism and the medicinal chemistry context(Section 4), we have tried to be true to our

second objective thatwas to presentmedicinalchemists with meaningful information. Hadthis chapter been written for a treatise ofmolecular biology, much of the basic informa-tionwouldhavebeen the same, but the contextand the chapter’s meaning would have beendifferent. This also tells us that medicinalchemistry itself needs a context to acquiremeaning, the context of human welfare towhich responsible medicinal chemists areproud to contribute [147].

REFERENCES

1. TestaB, JennerP.DrugMetabolism.Chemicaland Biochemical Aspects. New York: Dekker;1976.

2. SilvermanRB. TheOrganic Chemistry of DrugDesign and Drug Action. San Diego: AcademicPress; 1992.

3. KauffmanFC, editor.Conjugation–Deconjuga-tion Reactions in Drug Metabolism and Toxi-city. Berlin: Springer Verlag; 1994.

4. Testa B. The Metabolism of Drugs and OtherXenobiotics—Biochemistry of Redox Reac-tions. London: Academic Press; 1995.

5. Ortiz de Montellano PR, editor. CytochromeP450. Structure, Mechanism, and Biochemis-try. 2nd ed. New York: Plenum Press; 1996.

6. Woolf TF, editor. Handbook of Drug Metabo-lism. New York: Dekker; 1999.

7. Ioannides C, editor. Enzyme Systems thatMetabolise Drugs and Other Xenobiotics. Chi-chester, UK: John Wiley & Sons; 2002.

8. Rodrigues AD, editor. Drug–Drug Interac-tions. New York: Dekker; 2002.

9. Testa B, Mayer JM. Hydrolysis in Drug andProdrug Metabolism—Chemistry, Biochemis-try and Enzymology. Zurich/Weinheim:VHCA/Wiley-VCH; 2003.

10. Boullata J, Armenti VT, editors. Handbook ofDrug–Nutrient Interactions. Totowa, NJ:Humana Press; 2004.

11. Lash LH, editor. Drug Metabolism andTransport. Totowa, NJ: Humana Press;2005.

12. Phillips IR, ShephardEA, editors. CytochromeP450 Protocols. Totowa NJ: Humana Press;2006.

13. Utrecht JP, Trager WF. Drug Metabolism—

Chemical and Enzymatic Aspects. New York:Informa; 2007.

REFERENCES 449

14. TestaB,Kr€amerSD.TheBiochemistry ofDrugMetabolism. Principles, Redox Reactions, Hy-drolyses Vol. 1. Zurich/Weinheim: VHCA/Wiley-VHC; 2008.

15. Testa B, Kr€amer SD. The Biochemistry of DrugMetabolism. Conjugation Reactions, Conse-quences ofMetabolism, InfluencingFactors,Vol.2. Zurich/Weinheim: VHCA/Wiley-VHC; 2009.

16. Rendic S. Summary of information on humanCYP enzymes: Human P450 metabolism data.Drug Metab Rev 2002;34:83–448.

17. Pelkonen O. Human CYPs: In vivo and clinicalaspects. Drug Metab Rev 2002;34:37–46.

18. Trager WF. Principles of drug metabolism 1:redox reactions. In: Testa B, van de Water-beemd H, editors. ADME-Tox Approaches. In:Taylor JB, TriggleDJ, editors. ComprehensiveMedicinal Chemistry, Vol. 5, 2nd ed. Oxford:Elsevier; 2007. p 87–132.

19. Testa B. Principles of drug metabolism 2: hy-drolysis and conjugation reactions. In: TestaB,van de Waterbeemd H, editors. ADME-ToxApproaches. In: Taylor JB, Triggle DJ, editors.Comprehensive Medicinal Chemistry, Vol. 5,2nd ed. Oxford: Elsevier; 2007. p 133–166.

20. Totah RA, Rettie AE. Principles of drug me-tabolism 3: enzymes and tissues. In: Testa B,van de Waterbeemd H, editors. ADME-ToxApproaches. In: Taylor JB, Triggle DJ, edi-tors. Comprehensive Medicinal Chemistry,Vol. 5, 2nd ed. Oxford: Elsevier; 2007. p167–191.

21. Testa B, Kr€amer SD. The biochemistry of drugmetabolism—an introduction. Part 1: Princi-ples and overview. Chem Biodiver 2006;3:1053–1101.

22. Testa B, Kr€amer SD. The biochemistry of drugmetabolism—an introduction. Part 2: RedoxReactions and their Enzymes. Chem Biodiver2007;4:257–405.

23. Testa B, Kr€amer SD. The biochemistry of drugmetabolism—an introduction. Part 3: Reac-tions of hydrolysis and their enzymes. ChemBiodiver 2007;4:2031–2122.

24. Testa B, Kr€amer SD. The biochemistry of drugmetabolism—an introduction. Part 4: Reac-tions of Conjugation and their Enzymes. ChemBiodiver 2008;5:2171–2336.

25. Testa B, Kr€amer SD. The biochemistry of drugmetabolism—an introduction. Part 5: Metabo-lism and Bioactivity. Chem Biodiver 2009;6:591–684.

26. Kr€amer SD, Testa B. The biochemistry of drugmetabolism—an introduction. Part 6: Inter-

individual factors affecting drug metabolism.Chem Biodiver 2008;5:2465–2578.

27. Kr€amer SD, Testa B. The biochemistry of drugmetabolism—an introduction. Part 7: Intra-individual factors affecting drug metabolism.Chem Biodiver 2009;6:1477–1660.

28. Testa B. Nonenzymatic contributions to xeno-biotic metabolism. Drug Metab. Rev. 1982;13:25–50.

29. Balant LP, Roseboom H, Guntert-Remy RM.Pharmacokinetic criteria for drug researchand development. In: Testa B, editor. Ad-vances in Drug Research. Vol. 19. London:Academic Press; 1990. p 1–138.

30. Gaviraghi G, Barnaby RJ, Pellegatti M. Phar-macokinetic challenges in lead optimization.In: Testa B, van de Waterbeemd H, Fokers G,GuyR, editors. Pharmacokinetic Optimizationin Drug Research—Biological, Physicochem-ical and Computational Strategies. Zurich/Weinheim; VHCA/Wiley-VCH; 2001. p 3–14.

31. Walle T, Walle UK, Olanoff LS. Quantitativeaccount of propranolol metabolism in urine ofnormal man. Drug Metab Dispos 1985;13:204–209.

32. Nomenclature Committee of the InternationalUnion of Biochemistry and Molecular Biology(IUBMB).EnzymeNomenclature.Availableathttp://www.chem.qmul.ac.uk/iubmb/enzyme/.Accessed 2008 Oct 27.

33. NelsonDR, ZeldinDC, Hoffman SMG,MaltaisLJ, Wain HM, Nebert DW. Comparison ofcytochrome P450 (CYP) genes from the mouseand human genomes, including nomenclaturerecommendations for genes, pseudogenes andalternative splice variants. Pharmacogenetics2004;14:1–18.

34. Nelson DR. Cytochrome P450 Homepage.Available at http://drnelson.utmem.edu/Cyto-chromeP450.html. Accessed 2008 Oct 27.

35. Directory of P450-containing Systems. Avail-able at http://www.icgeb.org/�p450srv/. Ac-cessed 2008 Oct 27.

36. Brenda:TheComprehensiveEnzyme Informa-tion System. Available at http://www.brenda-enzymes.info/. Accessed 2008 Oct 27.

37. ExPASyProteomics Server. Available atwww.expasy.org. Accessed 2008 Oct 27.

38. Korzekwa KR, Trager WF, Nagata K, Parkin-son A, Gillette JR. Isotope effect studies onthe mechanism of the cytochrome P-450IIA1-catalyzed formation of delta 6-testosteronefrom testosterone. Drug Metab Dispos1990;18:974–979.


39. Rambeck B, May T, Juergens U. Serum con-centrations of carbamazepine and its epoxideand diol metabolites in epileptic patients: theinfluence of dose and comedication. Ther DrugMonitor 1987;9:298–303.

40. Gorrod JW,RamanA. Imines as intermediatesin oxidative aralkylamine metabolism. DrugMetab Rev 1989;20:307–339.

41. WeinshilboumRM,OtternessDM, SzumlanskiCL. Methylation pharmacogenetics: catecholO-methyltransferase, thiopurine methyltrans-ferase, and histamine N-methyltransferase.Annu Rev Pharmacol Toxicol 1999; 39:19–52.

42. Nagata K, Yamazoe Y. Pharmacogenetics ofsulfotransferases. Annu Rev Pharmacol Tox-icol 2000;40:159–176.

43. Blanchard RL, Freimuth RR, Buck J, Wein-shilboum RM, Coughtrie MW. A proposed no-menclature system for the cytosolic sulfotrans-ferase (SULT) superfamily. Pharmacogenetics2004;14:199–211.

44. MackenziePI,Owens IS,BurchellB,BockKW,Bairoch A, B�elanger A, Fournel-Gigleux S,Green M, Hum DW, Iyanagi T, Lancet D,Louisot P, Magdalou J, Chowdhury JR, RitterJK, Schachter H, Tephly TR, Tipton KF, Ne-bert DW. The UDP glycosyltransferase genesuperfamily: recommended nomenclature up-date based on evolutionary divergence. Phar-macogenetics 1997;7:255–269.

45. Radominska-PandyaA, Czernik PJ, Little JM,Battaglia E, Mackenzie PI. Structural andfunctional studies of UDP-glucuronosyltrans-ferases. Drug Metab Rev 1999;31:817–899.

46. Spahn-Langguth H, Benet LZ. Acyl glucuro-nides revisited: is the glucuronidation processa toxification as well as a detoxification me-chanism?. Drug Metab Rev 1992;24:5–48.

47. Grant DM, Blum M, Meyer UA. Polymorph-isms ofN-acetyltransferase genes. Xenobiotica1992;22:1073–1081.

48. Dodds PF. Xenobiotic lipids: the inclusion ofxenobiotic compounds in pathways of lipidsynthesis. Prog Lipid Res 1995;34:219–247.

49. Kaphalia BS, Fritz RR, Ansari GAS. Purifica-tion and characterization of rat liver microso-mal fatty acid ethyl and 2-chloroethyl estersynthase and their relationship with carbox-ylesterase (pI 6.1). Chem Res Toxicol 1997;10:211–218.

50. Fears R. Lipophilic xenobiotic conjugates: thepharmacological and toxicological conse-quences of the participation of drugs and otherforeign compounds as substrates in lipid bio-synthesis. Prog Lipid Res. 1986;24:177–195.

51. Vickery S, Dodds PF. Incorporation of xenobio-tic carboxylic acids into lipids by cultured 3T3-L1adipocytes.Xenobiotica2004;34:1025–1042.

52. Williams K, Day R, Knihinicki R, Duffield A.The stereoselective uptake of ibuprofen enan-tiomers into adipose tissue. Biochem Pharma-col 1986;35:3403–3405.

53. Mayer JM, TestaB. Pharmacodynamics, phar-macokinetics and toxicity of ibuprofenenantiomers. Drugs of the Future 1997;22:1347–1366.

54. Bjorge SM, Baillie TA. Studies on the beta-oxidation of valproic acid in rat livermitochon-drial preparations. Drug Metab. Dispos.1991;19:823–829.

55. Sies H. Glutathione and its role in cellularfunctions. Free Radic. Biol. Med. 1999;27:916–921.

56. Dickinson DA, Forman HJ. Cellular glu-tathione and thiols metabolism. BiochemPharmacol 2002;64:1019–1026.

57. PompellaA, VisvikisA, Paolicchi A, DeTataV,CasiniAF. The changing faces of glutathione, acellular protagonist. Biochem Pharmacol2003;66:1499–1503.

58. Hayes JD, Flanagan JU, Jowsey IR. Glu-tathione transferases. Annu Rev PharmacolToxicol 2005;45:51–88.

59. Armstrong RN. Structure, catalytic mechan-ism, and evolution of the glutathione trans-ferases. Chem Res Toxicol 1997;10:2–18.

60. van der Aar EM, Tan KT, Commandeur JNM,Vermeulen NPE. Strategies to characterizethe mechanisms of action of glutathioneS-transferases: a review. Drug Metab. Rev.1998;30:569–643.

61. KettererB. The role of nonenzymatic reactionsof glutathione in xenobiotic metabolism. DrugMetab. Rev. 1982;13:161–187.

62. Anders MW, Dekant W. Glutathione-depen-dent bioactivation of haloalkenes. Annu RevPharmacol Toxicol 1998;38:501–537.

63. Baillie TA, Slatter JG. Glutathione: a vehiclefor the transport of chemically reactive meta-bolites in vivo. AccChemRes1991;24: 264–270.

64. Cribb AE, Miller M, Leeder JS, Hill J, Spiel-berg SP. Reactions of the nitroso and hydro-xylamine metabolites of sulfamethoxazolewith reduced glutathione—implications foridiosyncratic toxicity. Drug Metab Dispos1991;19:900–906.

65. De Clercq E. The discovery of antiviral agents:ten different compounds, ten different stories.Med Res Rev 2008;28:929–953.

REFERENCES 451

66. O’Donnell JP. The reaction of amines withcarbonyls: its significance in the nonenzymaticmetabolism of xenobiotics. Drug Metab. Rev.1982;13:123–159.

67. Fura A, Shu YZ, Zhu M, Hanson RL, RoongtaV, Humphreys WG. Discovering drugs thoughbiological transformation: role of pharmacolo-gically active metabolites in drug discovery. JMed Chem 2004;47:4339–4351.

68. Caccia S, Garattini S. Formation of activemetabolites of psychotropic drugs. An updatedreview of their significance. Clin Pharmacokin1990;18:434–459.

69. Garattini S. Active drug metabolites. An over-view of their relevance in clinical pharmacoki-netics. Clin Pharmacokin 1985;10:216–227.

70. Bodor N, Buchwald P. Soft drug design: gen-eral principles and recent applications. MedRes Rev 2000;20:58–101.

71. Bodor N. Novel approaches to the design ofsafer drugs: soft drugs and site-specific chemi-cal delivery systems. In: Testa B, editor. Ad-vances in Drug Research, Vol. 13. London:Academic Press; 1984. p. 255–331.

72. Bodor N. Soft drugs: principles and methodsfor the design of safe drugs. Med Res Rev1984;4:449–469.

73. Durant NN, Katz RL. Suxamethonium. Br JAnaesth 1982;54:195–207.

74. Greenblatt DJ, Divoli M, Abernethy DR, OchsHR, Shader RI. Clinical pharmacokinetics ofthe newer benzodiazepines. Clin Pharmacokin1983;8:233–252.

75. Lim YC, Desta Z, Flockhart DA, Skaar TC.Endoxifen has anti-estrogenic effects in breastcancer cells with potency similar to 4-hydroxy-tamoxifen. Canc Chemother Pharmacol2005;55:471–478.

76. FarlanutM,Franceschi L, PasqualE, Bacchet-ti S, Poz D, Giorda G, Cagol P. Tamoxifen andits main metabolites serum and tissue concen-trations in breast cancer women. Ther DrugMonit. 2007;29:349–352.

77. JosephyPD,MannervikB.MolecularToxicology.2nd ed. Oxford: Oxford University Press; 2006.

78. Anders MW, editor. Bioactivation of ForeignCompounds. London: Academic Press; 1985.

79. Oesch-BartlomowiczB,OeschF.Mechanismsof toxification and detoxification that chal-lenge drug candidates and drugs. In: TestaB, van deWaterbeemdH, editors. ADME-ToxApproaches. In: Taylor JB, Triggle DJ, edi-tors. Comprehensive Medicinal Chemistry,

Vol. 5, 2nd ed. Oxford: Elsevier; 2007. p.193–214.

80. Macherey AC, Dansette PM. Biotransforma-tions leading to toxic metabolites: chemicalaspects. In: Wermuth CG, editor. The Practiceof Medicinal Chemistry. 3rd ed. London: Aca-demic Press; 2008. p 674–696.

81. Liebler DC. The poison within: applicationof toxicity mechanisms to fundamental diseaseprocesses, ChemToxicol Res 2006;19: 610–613.

82. Baillie TA. Future of toxicology—metaboliteactivation and drug design: challenges andopportunities in chemical toxicology. ChemRes Toxicol 2006;19:889–893.

83. Doss GA, Baillie TA. Addressing metabolicactivation as an integral component of drugdesign. Drug Metab. Rev. 2006;38:641–649.

84. Kumar S, Kassahun K, Tschirret-Guth RA,Mitra K, Baillie TA. Minimizing metabolicactivation during pharmaceutical lead optimi-zation: progress, knowledge gaps and futuredirections. Curr Opin Drug Discov Devel2008;11:43–52.

85. Shu YZ, Johnson BM, Yang TJ. Role of bio-transformation studies in minimizing metabo-lism-related liabilities in drug discovery.AAPS J. 2008;10:178–192.

86. Guengerich FP, MacDonald JS. Applying me-chanisms of chemical toxicity to predictdrug safety. Chem Res Toxicol 2007;20:344–369.

87. HumphreysWG,UngerSE.Safety assessmentof drugmetabolites: characterization of chemi-cally stable metabolites. Chem Res Toxicol2006;19:1564–1569.

88. Hop CECA, Kalgutkar AS, Soglia JR. Impor-tance of early assessment of bioactivation indrug discovery. Annu RepMed Chem 2006;41:369–381.

89. Stevens JL. Future of toxicology—mechan-isms of toxicity and drug safety: where do wego from here? Chem Res Toxicol 2006;19:1393–1401.

90. Caldwell GW, Yan Z. Screening for reactiveintermediates and toxicity assessment in drugdiscovery. Curr Opin Drug Discov Devel 2006;9:47–60.

91. Wilson ID, Nicholson JK. Metabonomics. In:Testa B, van de Waterbeemd H, editors.ADME-Tox Approaches. In: Taylor JB, TriggleDJ, editors. Comprehensive Medicinal Chem-istry, Vol. 5. 2nd ed. Oxford: Elsevier; 2007. p.989–1007.


92. Mortishire-Smith RJ, Skiles GL, LawrenceJW, Spence S, Nicholls AW, Johnson BA, Ni-cholson JK. Use of metabonomics to identifyimpaired fatty acidmetabolismas themechan-ism of drug-induced toxicity. ChemResToxicol2004;17:165–173.

93. Khor TO, Ibrahim S, Kong ANT. Toxicoge-nomics in drug discovery and drug develop-ment: potential applications and future chal-lenges. Pharm Res 2006;23:1659–1664.

94. Ulrich R, Friend SH. Toxicogenomics and drugdiscovery: will new technologies help us pro-duce better drugs? Nature Rev Drug Discov2002;1:84–88.

95. Williams DP, Naisbitt DJ. Toxicophores:groups and metabolic routes associated withincreased safety risks. Curr Opin DrugDiscov.Devel. 2002;5:104–115.

96. Schulz M, Schmoldt A. Therapeutic and toxicblood concentrations of more than 500 drugs.Pharmazie 1997;52:895–911.

97. PirmohamedM, Park BK. Genetic susceptibil-ity to adverse drug reactions. Trends Pharma-col. Sci. 2001;22:298–305.

98. Park BK. Prediction of adverse drug reactions.Drug Metab Rev 2006;38(S1) 14–15.

99. Darley-Usmar V, Halliwell B. Blood radicals:reactive nitrogen species, reactive oxygen spe-cies, transition metal ions, and the vascularsystem’. Pharm Res 1996;13:649–662.

100. Lancaster JR, Jr. Nitroxidation, nitrosative,and nitrative stress: kinetic predictions of re-active nitrogen species chemistry under biolo-gical conditions. Chem Res Toxicol2006;19:1160–1174.

101. Buettner GR. The pecking order of free radi-cals and antioxidants: lipid peroxidation,a-toxopherol, and ascorbate. Arch BiochemBiophys 1993;300:535–543.

102. Aldini G, Dalle-Donne I, Maffei Facino R, Mil-zani A, Carini M. Intervention strategies toinhibit protein carbonylation by lipoperoxida-tion-derived reactive carbonyls. Med Res Rev2007;27:817–868.

103. Uetrecht JP. New concepts in immunologyrelevant to idiosyncratic drug reactions: the“danger hypothesis” and innate immune sys-tem. Chem Res Toxicol 1999;12:387–395.

104. Uetrecht JP. Evaluation of which reactive me-tabolite, if any, is responsible for a specificidiosyncratic reaction. Drug Metab Rev 2006;38:745–753.

105. Esser C. Immunotoxicology In. TestaB, van deWaterbeemd H, editors. ADME-Tox Ap-

proaches. In: Taylor JB, Triggle DJ, editors.Comprehensive Medicinal Chemistry, Vol. 5.2nd ed. Oxford: Elsevier; 2007. p. 215–229.

106. Gram TE, Okine LK, Gram RA. The metabo-lism of xenobiotics by certain extrahepaticorgans and its relation to toxicity. Annu RevPharmacol Toxicol 1986;26:259–291.

107. ParkBK, KitteringhamNR,Maggs JL, Pirmo-hamed M, Williams DP. The role of metabolicactivation in drug-induced hepatotoxicity.Annu Rev Pharmacol Toxicol 2005;45:177–202.

108. Halegoua-De Marzio D, Navarro VJ. Drug-in-duced hepatotoxicity in humans’. Curr OpinDrug Discov Develop 2008;11:53–59.

109. Welch KD,Wen B, Goodlett DR, Yi EC, Lee H,Reilly TP, Nelson SD, Pohl LR. Proteomicidentification of potential susceptibility fac-tors in drug-induced liver disease. Chem ResToxicol 2005;18:924–933.

110. Hultin-Rosenberg L, Jagannathan J, NilssonKC, Matis SA, Sj€ogren N, Huby RDJ, SalterAH, Tugwood JD. Predictive models of hepa-totoxicity using gene expression data fromprimary rat hepatocytes. Xenobiotica2006;36:1122–1139.

111. Dieterle F, Marrer E, Suzuki E, Grenet O,Cordier A, Vonderscher J. Monitoring kidneysafety in drug development: emerging technol-ogies and their implications. Curr Opin DrugDiscov Develop 2008;11:60–71.

112. Wells PG,Winn LM. Biochemical toxicology ofchemical teratogenesis. Crit Rev BiochemMolBiol 1996;31:1–40.

113. Masubuchi N, Makino C, Murayama N. Pre-diction of in vivo potential for metabolic acti-vation of drugs into chemically reactive inter-mediates: correlation of in vitro and in vivogeneration of reactive intermediates and invitro glutathione conjugate formation in ratsand humans. Chem Res Toxicol 2007;20:455–464.

114. Kazius J, McGuire R, Bursi R. Derivation andvalidation of toxicophores for mutagenicityprediction. J Med Chem 2005;48:312–320.

115. Kalgutkar AS, Gardner I, Obach RS, ShafferCL, Callegari E, Henne KR,Mutlib AE, DalvieDK, Lee JS, Nakai Y, O’Donnell JP, Boer J,Harriman SP. A comprehensive listing ofbioactivation pathways of organic functionalgroups. Curr Drug Metab 2005;6:161–225.

116. Testa B, Crivori P, Reist M, Carrupt PA. Theinfluence of lipophilicity on the pharmacoki-netic behavior of drugs: Concepts and

REFERENCES 453

examples. Perspect Drug Discov. Design2000;19:179–211.

117. Eichelbaum M, Testa B, Somogyi A, editors.Stereochemical Aspects of Drug Action andDisposition. Berlin: Springer Verlag; 2003.

118. Riley RJ, Lambert C, Kitteringham NR, ParkBK. A stereochemical investigation of the cy-totoxicity of mianserin metabolites in vitro. BrJ Clin Pharmacol 1989;27:823–830.

119. Martin HJ, Breyer-Pfaff U, Wsol V, Venz S,Block S, Maser E. Purification and character-ization of AKR1B10 from human liver: role incarbonyl reduction of xenobiotics. Drug MetabDispos 2006;34:464–470.

120. Dow J, Berg C. Stereoselectivity of the carbo-nyl reduction of dolasetron in rats, dogs andhumans. Chirality 1995;7:342–348.

121. Gonzalez FJ, Nebert DW. Evolution of theP450 gene superfamily: animal-plant“warfare”, molecular drive and human geneticdifferences in drug oxidation. Trends Genet.1990;6:182–186.

122. Upthagrove AL, Nelson WL. Importance ofamine pKa and distribution coefficient in themetabolism of fluorinated propranolol deriva-tives. Preparation, identification of metaboliteregioisomers, and metabolism by CYP2D6.Drug Metab. Dispos. 2001;29:1377–1388.

123. Morgenstern R, Lundqvist G, Hancock V, De-Pierre JW. Studies on the activity and activa-tion of rat liver microsomal glutathione trans-ferase, in particular with a substrate analogueseries. J Biol Chem 1988;263:6671–6675.

124. Hines RN, Luo Z, Cresteil T, Ding X, ProughRA, Fitzpatrick JL, Ripp SL, Falkner KC, GeNL, Levine A, Elferink CJ. Molecular regula-tion of genes encoding xenobiotic-metabolizingenzymes: mechanisms involving endogenousfactors.DrugMetab.Dispos. 2001;29:623–633.

125. Poulos TL, Finkel BC, Howard AJ. High-reso-lution crystal structure of cytochromeP450cam. J Mol Biol 1987;195:687–700.

126. Williams PA, Cosme J, Ward A, Angove HC,Vinkovic DM, Jhoti H. Crystal structure ofhuman cytochrome P450 2C9 with bound war-farin. Nature 2003;424:464–468.

127. Williams PA, Cosme J, Vinkovic DM, Ward A,Angove HC, Day PJ, Vonrhein C, Tickle IJ,Jhoti H. Crystal structure of human cyto-chrome P450 3A4 bound to metyrapone andprogesterone. Science 2004;305:683–686.

128. CrucianiG,CarosatiE,DeBoeckB,EthirajuluK, Mackie C, Howe T, Vianello R. MetaSite:Understanding metabolism in human cyto-

chromes from the perspective of the chemist.J Med Chem 2005;48:6970–6979.

129. Available at http://www.moldiscovery.com/.Accessed 2008 Dec 5.

130. Erhardt PW, editor. Drug Metabolism—Data-bases and High-Throughput Testing DuringDrugDesign andDevelopment. London: Inter-nationalUnion of Pure andAppliedChemistryand Blackwell Science; 1999.

131. Testa B, Cruciani G. Structure–metabolismrelations, and the challenge of predicting bio-transformation. In: Testa B, van de Water-beemd H, Folkers G, Guy R, editors. Pharma-cokinetic Optimization in Drug Research: Bio-logical, Physicochemical and ComputationalStrategies. Zurich/Weinheim: Verlag Helveti-ca Chimica Acta/Wiley-VCH; 2001. p. 65–84.

132. Hawkins DR. Comprehensive expert systemsto predict drugmetabolism. In: TestaB, vandeWaterbeemd H, editors. ADME-Tox Ap-proaches. In: Taylor JB, Triggle DJ, editors.Comprehensive Medicinal Chemistry, Vol. 5.2nd ed. Oxford: Elsevier; 2007. p. 795–807.

133. Ekins S, Andreyev S, Ryabov A, Kirillov E,Rakhmatulin EA, Bugrim A, Nikolskaya T.Computational prediction of human drug me-tabolism. Exp Opin Drug Metab Toxicol2005;1:303–324.

134. Available at http://www.genego.com/meta-drug.php/. Accessed 2008 Dec 5.

135. Klopman G, Tu M. META: a program for theprediction of the products of mammal metabo-lism of xenobiotics. In: Erhardt PW, editor.Drug Metabolism: Databases and High-Throughput Testing During Drug Design andDevelopment. London: International Union ofPure and Applied Chemistry and BlackwellScience; 1999. p. 271–276.

136. Available at http://www.multicase.com/. Ac-cessed 2008 Dec 5.

137. DarvasP,Marokh�aziS,KormosP,KulkarniG,Kal�asz H, Papp A. MetabolExpert: its use inmetabolism research and in combinatorialchemistry. In: Erhardt PW, editor. Drug Me-tabolism: Databases and High-ThroughputTesting During Drug Design and Develop-ment. London: International Union of PureandAppliedChemistry andBlackwell Science;1999. p. 237–270.

138. Available at http://www.compudrug.com/. Ac-cessed 2008 Dec 5.

139. Langowski JJ, Long A. Computer systems forthe prediction of xenobiotic metabolism. AdvDrug Deliv. Rev. 2002;54:407–415.


140. Testa B, Balmat AL, Long A, Judson P. Pre-dicting drugmetabolism—an evaluation of theexpert system METEOR. Chem Biodiv2005;2:872–885.

141. Available at http://www.lhasalimited.org/. Ac-cessed 2008 Dec 5.

142. Collins JM. Inter-species differences in drugproperties. Chem -Biol Interact 2001;134:237–242.

143. Tanaka E. Gender-related differencesin pharmacokinetics and their clinicalsignificance. J Clin Pharmacol Ther 1999;24:339–346.

144. Gow PJ, Ghadrial H, Smallwood RA, MorganDJ, ChingMS.Neonatal hepatic drug elimina-tion. Pharmacol Toxicol 2001;88:3–15.

145. LemmerB.Relevance for chronopharmacologyin practical medicine. Seminars Perinatology2000;24:280–290.

146. Lin JH, YuAYH. Interindividual variability ininhibition and induction of cytochrome P450enzymes. Annu Rev Pharmacol Toxicol2001;41:535–567.

147. Testa B. Missions and finality of drug re-search: a personal view. Pharm News 1996;3:10–12.

REFERENCES 455

PRINCIPLES OF DRUG METABOLISM...

Documents

Transcript of PRINCIPLES OF DRUG METABOLISM...