Chapter 5

Enzyme Inhibition and Inactivation

Chapter Outline
5.1. Why Inhibit an Enzyme? 208 5.2.4.2.4. D ual-Acting Drugs: Dual-Acting
5.2. Reversible Enzyme Inhibitors 210 Enzyme Inhibitors 230
5.2.1. Mechanism of Reversible Inhibition 210 5.2.4.3. Lovastatin (Mevinolin) and Simvastatin,
5.2.2. Selected Examples of Competitive Reversible Antihypercholesterolemic Drugs 232
Inhibitor Drugs 211 5.2.4.3.1. Cholesterol and Its Effects 232
5.2.2.1. Simple Competitive Inhibition 211 5.2.4.3.2. Lead Discovery 232
5.2.2.1.1. Epidermal Growth Factor 5.2.4.3.3. Mechanism of Action 233
Receptor Tyrosine Kinase 5.2.4.3.4. Lead Modification 234
as a Target for Cancer 211 5.2.4.4. Saxagliptin, a Dipeptidyl Peptidase-4
5.2.2.1.2. Discovery and Optimization Inhibitor and Antidiabetes Drug 234
of EGFR Inhibitors 212 5.2.5. Case History of Rational Drug Design of
5.2.2.2. Stabilization of an Inactive Conformation: an Enzyme Inhibitor: Ritonavir 235
Imatinib, an Antileukemia Drug 213 5.2.5.1. Lead Discovery 235
5.2.2.2.1. The Target: Bcr-Abl, 5.2.5.2. Lead Modification 236
a Constitutively Active Kinase 213 5.3. Irreversible Enzyme Inhibitors 238
5.2.2.2.2. Lead Discovery and 5.3.1. Potential of Irreversible Inhibition 238
Modification214 5.3.2. Affinity Labeling Agents 240
5.2.2.2.3. Binding Mode of Imatinib 5.3.2.1. Mechanism of Action 240
to Abl Kinase 215 5.3.2.2. Selected Affinity Labeling Agents 241
5.2.2.2.4. Inhibition of Other Kinases 5.3.2.2.1. Penicillins and Cephalosporins/
by Imatinib 216 Cephamycins241
5.2.2.3. Alternative Substrate Inhibition: 5.3.2.2.2. Aspirin 243
Sulfonamide Antibacterial Agents 5.3.3. Mechanism-Based Enzyme Inactivators 247
(Sulfa Drugs) 217 5.3.3.1. Theoretical Aspects 247
5.2.2.3.1. Lead Discovery 217 5.3.3.2. Potential Advantages in Drug Design
5.2.2.3.2. Lead Modification 217 Relative to Affinity Labeling Agents 247
5.2.2.3.3. Mechanism of Action 217 5.3.3.3. Selected Examples of Mechanism-
5.2.3. Transition State Analogs and Based Enzyme Inactivators 248
Multisubstrate Analogs 219 5.3.3.3.1. Vigabatrin, an
5.2.3.1. Theoretical Basis 219 Anticonvulsant Drug 248
5.2.3.2. Transition State Analogs 220 5.3.3.3.2. Eflornithine, an Antiprotozoal
5.2.3.2.1. Enalaprilat 220 Drug and Beyond 250
5.2.3.2.2. Pentostatin 221 5.3.3.3.3. Tranylcypromine, an
5.2.3.2.3. Forodesine and DADMe-ImmH 222 Antidepressant Drug 253
5.2.3.2.4. Multisubstrate Analogs 223 5.3.3.3.4. Selegiline (l-Deprenyl) and
5.2.4. Slow, T ight-Binding Inhibitors 225 Rasagiline: Antiparkinsonian Drugs 254
5.2.4.1. Theoretical Basis 225 5.3.3.3.5. 5-Fluoro-2′-deoxyuridylate,
5.2.4.2. Captopril, Enalapril, Lisinopril, and Other Floxuridine, and
Antihypertensive Drugs 225 5-Fluorouracil: Antitumor Drugs 256
5.2.4.2.1. Humoral Mechanism for Hypertension 225 5.4. General References 258
5.2.4.2.2. Lead Discovery 226 5.5. Problems 261
5.2.4.2.3. Lead Modification and References265
Mechanism of Action 226

The Organic Chemistry of Drug Design and Drug Action. http://dx.doi.org/10.1016/B978-0-12-382030-3.00005-2

Copyright © 2014 Elsevier Inc. All rights reserved. 207
208 The Organic Chemistry of Drug Design and Drug Action

effect. Prostaglandins (PGs) are important hormones that

5.1. WHY INHIBIT AN ENZYME?
are involved in the pathogenesis of inflammation and fever.
Many diseases, or at least the symptoms of diseases, arise Inhibition of PG synthase results in antiinflammatory, anti-
from a deficiency or excess of a specific metabolite in the pyretic, and analgesic effects. In the case of foreign organ-
body, from an infestation of a foreign organism, or from isms such as bacteria and parasites, or in the case of tumor
aberrant cell growth. If the metabolite deficiency or excess cells, inhibition of one of their essential enzymes can prevent
can be normalized, and if the foreign organisms and aber- important metabolic processes from taking place, resulting
rant cells can be destroyed, then these disease states will in inhibition of growth or replication of the organism or
be remedied. Many of these problems can be addressed by aberrant cell. Inhibition of the bacterial alanine racemase,
specific enzyme inhibition. the enzyme that makes d-alanine for incorporation into pep-
Any compound that slows down or blocks enzyme cataly- tidoglycan strands, for example, would prevent the biosyn-
sis is an enzyme inhibitor. If the interaction with the target thesis of the peptidoglycan and, therefore, the biosynthesis
enzyme (specific enzyme for inhibition) is irreversible (usu- of the bacterial cell wall. These compounds possess antibac-
ally covalent), then the compound is a special type of enzyme terial activity. The use of drugs to combat foreign organisms
inhibitor referred to as an enzyme inactivator. Of the 149 new and aberrant cells is called chemotherapy.
small molecule and biological drugs that went on the world Enzyme inhibition is a promising approach for the
drug market in the 6 years 2006–2011, one-third (51/149) rational discovery of new leads or drugs. Although there are
were enzyme inhibitors; if only small molecules (including numerous drugs that exert their therapeutic action by inhib-
peptides) are considered, then 41% (51/123) were enzyme iting specific enzymes, the mechanisms of action of some
inhibitors.[1] Of the 51 small molecule new drugs that are of these drugs were determined subsequent to the discovery
enzyme inhibitors, 12 inhibit kinase enzymes (10 of which of the therapeutic properties of the drugs. Target enzymes
are cancer treatments), 14 inhibit protease enzymes (six for selected for rational drug design are those whose inhibition
cardiovascular disease, five for diabetes, and three for viral in vivo would lead to the desired therapeutic effect.
infection), and 11 inhibit enzymes involved with deoxyribo- There are two general categories of target enzymes.
nucleic acid (DNA) or ribonucleic acid (RNA) synthesis or In most cases, a potential drug is designed for an enzyme
function (eight for viral or bacterial infection and three for whose inhibition is known to produce a specific pharma-
cancer or proliferative syndrome treatments). Representative cological effect, but existing inhibitors have certain unde-
­inhibitors of these enzyme classes are shown in Table 5.1. sirable properties such as lack of potency or specificity or
Consider what happens when an enzyme activity is exhibit side effects. A more daring approach is to design
blocked. The substrates for that enzyme cannot be metab- inhibitors of enzymes whose inhibition has not yet been
olized, and the metabolic products are not generated (that established to lead to a desired therapeutic effect. This cat-
is, unless there is another enzyme that can metabolize the egory of enzyme targets requires knowledge of the patho-
substrate, and unless there is another metabolic pathway physiology of disease processes and the ability to identify
that generates the same product). Why should these two important metabolites whose function or dysfunction results
outcomes be important to drug design? If a cell has a defi- in a disease state. While preliminary proof of concept may
ciency of the substrate for the target enzyme and as a result be obtained through such methods as decreasing enzyme
of that deficiency, a disease state results, then inhibition production through inhibitory RNA or inhibition of the
of that enzyme would prevent the degradation of the sub- enzyme with an antibody, not until a small molecule inhibi-
strate, thereby increasing its concentration. An example tor is obtained will it be possible to determine the real effect
of this is the onset of seizures that arises from diminished of inhibition of that enzyme on the metabolism of the organ-
γ-aminobutyric acid (GABA) levels in the brain. Inhibition ism under the conditions anticipated for the actual therapy.
of the enzyme that degrades GABA, namely, GABA amino- Once an enzyme target is identified, lead compounds must
transferase (GABA-AT), leads to an anticonvulsant effect. be prepared that can inhibit it completely and specifically.
If there is an excess of a particular metabolite that produces Of all the protein targets for potential therapeutic use,
a disease state, then inhibition of an enzyme on the path- including hormone and neurotransmitter receptors and
way for biosynthesis of that metabolite should diminish its transporter proteins, enzymes are the most promising for
concentration. Excess uric acid can lead to gout. Inhibition rational inhibitor design. Enzyme purification is generally a
of xanthine oxidase, the enzyme that catalyzes the conver- much simpler task than receptor purification; a ­homogeneous
sion of xanthine to uric acid, decreases the uric acid levels, enzyme preparation can be obtained for preliminary screen-
and results in an antihyperuricemic effect. If the product of ing purposes, and in some cases, may be used to elucidate
an enzyme reaction is required to carry out an important the active site structure, which is useful for computer-based
physiological function that the drug is supposed to block, drug design approaches (see Chapter 2, Section 2.2.6). Fur-
then inhibition of that enzyme decreases the concentration thermore, whereas effective receptor antagonists often bear
of that product and can interfere with the physiological little structural similarity to agonists, enzyme inhibitors are
TABLE 5.1 Representative Commercially Marketed Enzyme Inhibitors Launched between 2006 and 2010

Enzyme
Drug Structure Indication Inhibited Enzyme Class
Dasatinib Anticancer bcr-abl (includ- Kinase
ing mutants)

Pazopanib Anticancer Vascular endo- Kinase

thelial growth
factor receptor
kinase

Aliskiren Antihyperten- Renin Aspartyl pro-

sive tease

Saxagliptin H O Antidiabetic DPP-4 Serine protease

N
H
NH2 OH
CN
5.4

Rivaroxaban Anticoagulant Factor Xa Serine protease

Besifloxacin Antibacterial DNA gyrase/ Enzymes mod-

Cl topoisomerase ulating DNA/
N N RNA synthesis/
function
H2N OH
F
O O
5.6

Clevudine NH2 Antiviral DNA poly- Enzymes mod-

CH3 merase ulating DNA/
N RNA synthesis/
function
O N
OH
F O

5.7 OH

Vorinostat Anticancer Histone deacety- Enzymes mod-

lase ulating DNA/
RNA synthesis/
function
210 The Organic Chemistry of Drug Design and Drug Action

frequently able to be designed starting from substrates or be reversible covalent interactions. An irreversible enzyme
products of the target enzyme. In addition, knowledge of inhibitor, also called an enzyme inactivator, is one that pre-
enzyme mechanisms can be used in the design of transition vents the return of enzyme activity for an extended period,
state analogs and multisubstrate inhibitors (Section 5.2.3), suggesting the involvement of a covalent bond. This also is
slow tight-binding inhibitors (Section 5.2.4), and mecha- not strictly the case; it is possible for noncovalent interac-
nism-based enzyme inactivators (Section 5.3.3). tions to be so effective that the enzyme–inhibitor complex
To minimize side effects, there are certain properties that is, for all intents and purposes, irreversibly formed.
ideal enzyme inhibitors and/or enzyme targets should pos-
sess. An ideal enzyme inhibitor should be totally specific for
the one target enzyme. Because this is rare, if attained at all, 5.2. REVERSIBLE ENZYME INHIBITORS
highly selective inhibition is a more realistic objective. By
adjustment of the dose administered, essentially specific inhi-
5.2.1. Mechanism of Reversible Inhibition
bition may be possible. In some cases, such as infectious dis- The most common enzyme inhibitor drugs are the reversible type,
eases, enzyme targets can be identified because of biochemical particularly ones that compete with the substrate for active site
differences in essential metabolic pathways between foreign binding. These are known as competitive reversible inhibitors,
organisms and their hosts.[2] In other instances, there are sub- typically compounds that have structures similar to those of the
strate specificity differences between enzymes from the two substrates or products of the target enzymes and which bind at
sources that can be utilized in the design of selective enzyme the substrate binding sites, thereby blocking substrate binding.
inhibitor drugs. Unfortunately, when dealing with some Typically, these inhibitors establish their binding equilibria with
organisms, and especially with tumor cells, the enzymes that the enzyme rapidly, so that inhibition is observed as soon as the
are essential for their growth also are vital to human health. enzyme is assayed for activity, although there are some cases
Inhibition of these enzymes can destroy human cells as well. (see Sections 5.2.4, 5.3.2, and 5.3.3) in which ­inhibition can be
Nonetheless, this approach is taken in various types of che- relatively slow.
motherapy. The reason this approach can be effective is that As in the case of the interaction of a substrate with an
foreign organisms and tumor cells replicate at a much faster enzyme, an inhibitor (I) also can form a complex with an
rate than do most normal human cells (those in the gut, the enzyme (E) (Scheme 5.1). The equilibrium constant Ki
bone marrow, and the mucosa are exceptions). Consequently, (koff/kon) is a dissociation constant for breakdown of the E·I
rapidly proliferating cells have an elevated requirement for complex; therefore, as discussed for the Kd of drug–receptor
essential metabolites. Antimetabolites, compounds whose complexes (see Chapter 3, Section 3.2.1), the smaller the
structures are similar to those of essential metabolites and Ki value for an inhibitor I, the more potent the inhibitor.
which disrupt metabolic processes (typically by blocking or Another common measurement for inhibition, which is
acting as an alternative substrate for an enzyme in a metabolic quicker to determine, but is more of an estimate, is the IC50,
pathway), are taken up by the rapidly replicating cells and, the inhibitor concentration that produces 50% enzyme inhi-
therefore, these cells are selectively inhibited. The selective bition. An IC50 value is not a constant, but rather depends
toxicity in this case derives from a kinetic difference rather on the substrate concentration used in the assay. Therefore,
than a qualitative difference in the metabolism. the IC50 is useful for comparing the potency of inhibitors
An ideal enzyme target in a foreign organism or aber- so long as all the inhibitors were tested in the presence of
rant cell would be one that is essential for its growth, but the same concentration of substrate. The IC50 value can be
which is either nonessential for human health, or, even bet- converted into an ­approximate Ki value by Eqn (5.1).[3]
ter (in the case of foreign organisms), not even present in
( )
humans. This type of selective toxicity would destroy only [S]
the foreign organism or aberrant cell, and would not require IC50 = 1+ Ki
Km (5.1)
the careful administration of a drug that is necessary when
the inhibited enzyme is important to human metabolism as When the inhibitor binds at the active site (the substrate
well. The penicillins, for example, which inhibit the bacte- binding site), then it is a competitive inhibitor. Formation
rial peptidoglycan transpeptidase essential for the biosyn- of the E·I complex prevents the binding of substrate and,
thesis of the bacterial cell wall, are nontoxic to humans,
since human cells penicillins do not have cell walls (nor do kon
viruses, so penicillins are nontoxic to them as well). E + I E•I
Enzyme inhibitors can be grouped into two general koff
categories: reversible and irreversible inhibitors. As the -S +S
name implies, inhibition of enzyme activity by a ­reversible
­inhibitor is reversible, suggesting that noncovalent interac- E•S E•P E + P
tions are involved. This is not strictly the case; there also can SCHEME 5.1 Kinetic scheme for competitive enzyme inhibition
Chapter | 5 Enzyme Inhibition and Inactivation 211

therefore, blocks the catalytic conversion of the substrate to 5.2.2.1. Simple Competitive Inhibition
product. In some cases, the inhibitor may also act as a sub-
The most common type of inhibition is simple competi-
strate, whereby it may be converted to a metabolically useless
tive inhibition, in which the inhibitor rapidly and revers-
product. In the context of drug design, this is generally not a
ibly binds to the active site of the enzyme. Frequently the
favorable process because the product formed may be toxic
structure of the inhibitor resembles that of the substrate or
or may lead to other toxic metabolites. Nonetheless, there are
product of the target enzyme.
drugs that function by this mechanism (see Section 5.2.2.3).
Interaction of the inhibitor with the enzyme can occur at
a site other than the substrate-binding site (i.e., at an allo- 5.2.2.1.1. Epidermal Growth Factor Receptor
steric binding site) and still result in inhibition of substrate Tyrosine Kinase as a Target for Cancer
turnover. When this occurs, often as a result of an inhibi- For many years, various combinations of surgery, radiation,
tor-induced conformational change in the enzyme to give and chemotherapy have constituted mainstream treatments
a form of the enzyme that does not bind the substrate prop- for cancer. Chemotherapy is generally designed to inhibit
erly, then the inhibitor is a noncompetitive reversible inhibi- processes related to cell division, based on the premise that
tor. The discussion in this chapter will focus on the design rapidly dividing cells, i.e., those in tumors, will be selec-
and mechanism of action of competitive enzyme inhibitors. tively affected by inhibitors of cell division (see discussions
The equilibrium shown in Scheme 5.1, and therefore the in Section 5.3.3.3.5 and in Chapter 6, Section 6.1.1). While
EI concentration, will depend on the concentrations of the this strategy can be effective, normal cells, and particu-
inhibitor and the substrate, as well as the Ki for the inhibitor lar normal cells that divide relatively rapidly, such as hair
and the Km for the substrate. As the concentration of the inhib- cells and cells in the gastrointestinal tract, rarely escape
itor is increased, it drives the equilibrium toward the E·I com- the effects of the treatment, and side effects such as hair
plex. Because the substrate and the competitive inhibitor bind loss and gastrointestinal distress (for example, nausea) fre-
to the enzyme at the same site, they both cannot interact with quently occur.
the enzyme simultaneously. When the inhibitor concentration Since the early 2000s, much emphasis has been placed on
diminishes, the E·I complex concentration diminishes, and the targeted therapy. In the context of cancer, this term implies the
effect of the inhibitor can be overcome by the substrate. ability to identify and inhibit a biological target that exhibits
If the enzyme inhibitor is a drug, the maximal pharma- aberrant behavior that specifically promotes the disease. Sev-
cological effect will occur when the drug concentration is eral prominent examples of targeted therapies have emerged
maintained at a saturating level at the target enzyme active in the past decade, resulting in lifesaving therapy for patients
site. As the drug is metabolized (see Chapter 8), and the with certain types of cancers. A number of these examples
concentration of I diminishes, additional administration of are protein kinase inhibitors, and in particular, inhibitors
the drug is required to maintain the prevalence of the E·I of protein kinases that are abnormal, and as a result of this
complex. Drugs that are rapidly cleared from the circulation abnormality cause uncontrolled proliferation of certain types
by metabolism or other means may need to be taken several of cells. Recall from Chapter 2, Scheme 2.1 and associated
times a day. To increase the potency of reversible inhibi- discussion, that protein kinases catalyze the transfer of the
tors, and thereby reduce the dosage of the drug, the binding terminal phosphate group of adenosine triphosphate (ATP) to
interactions with the target enzyme must be optimized (that the hydroxyl group on the side chain of serine, threonine, or
is, the inhibitor should have a low Ki value). tyrosine residues of proteins. The reason that protein kinases
When a drug is designed to be an enzyme inhibitor, it often appear in the context of uncontrolled cell proliferation
frequently will be a competitive inhibitor because the lead is that in normal cells, these kinases are usually present in
compound often will be the substrate for the target enzyme. the biological pathways that communicate to cells when to
Because an enzyme is just a specific type of receptor, an grow and divide. In normal cells, these pathways are tightly
analogy can be made between agonists, partial agonists, controlled so that signals to proliferate are only turned on
and antagonists with good substrates, poor substrates, and at specified times. However, it has been found that aberra-
­competitive inhibitors, respectively. tions in specific kinases in these pathways can circumvent
the regulatory control and the signal to proliferate is therefore
5.2.2. Selected Examples of Competitive enhanced, leading to uncontrolled cell proliferation.
Epidermal growth factor (EGF) is a polypeptide that binds
Reversible Inhibitor Drugs
to receptors on the surface of targeted cells to stimulate the
In this section, we will take a look at five different approaches growth, proliferation, and survival of these cells. This mecha-
to the design of competitive reversible inhibitors: simple nism is used, for example, to regulate the growth of new cells in
competitive inhibition, stabilization of an inactive enzyme the lining of the intestine. The epidermal growth factor recep-
conformation, alternative substrate inhibition, transition tor (EGFR) is a transmembrane receptor wherein the intra-
state analog inhibition, and slow, tight binding inhibition. cellular domain functions as a kinase enzyme (Figure 5.1).
212 The Organic Chemistry of Drug Design and Drug Action

intracellular kinase activity. The natural ligand at the recep-

tor and the natural substrate(s) of the kinase are polypep-
EGF (ligand) tides and, therefore, difficult to use as leads, particularly
toward orally administered compounds (antibodies that
block the extracellular receptor domain have found clini-
Receptor domain cal use but are administered by injection). However, kinases
Dimerized receptor
bound to EGF (extracellular) have a second substrate, namely, ATP, which is the source
of the phosphoryl group that is transferred to the protein
Cell membrane substrate. Since there are many human kinases (and at least
500 human protein kinases) that all use ATP as a substrate,
Kinase domain the design of a kinase inhibitor that is specific for a single
(intracellular) kinase based on the structure of ATP as a lead has been a
significant challenge. However, there are sufficient differ-
Trans-phosphorylation
ences in binding regions in and around the ATP-binding site
of tyrosine residues of different kinases that specificity for individual kinases,
or at least for a comparatively small number of kinases, has
indeed proven possible.
Signal propagation
to cell nucleus
5.2.2.1.2. Discovery and Optimization of EGFR
Inhibitors
Gene transcription Scientists at Zeneca (now AstraZeneca) and other pharma-
Cell growth ceutical companies set out to discover inhibitors of EGFR
Cell proliferation kinase.[5] Mechanistic studies indicated that catalysis was
Cell survival effected through a ternary complex of enzyme, ATP, and
the peptide substrate, and that phosphate transfer occurred
FIGURE 5.1 Schematic depiction of epidermal growth factor (EGF) and
its receptor, activation of which leads to cell growth and proliferation. In its
through direct nucleophilic attack of the tyrosine hydroxyl
activated state, the EGFR normally functions in a dimeric form, as shown. group on the phosphorus atom of the γ-phosphate group
Adapted from Pawson, T.; Jorgensen, C. Signal Transduction by Growth of ATP (see Chapter 4, Section 4.3.5). The lead discovery
Factor Receptors (Chapter 11) In The Molecular Basis of Cancer, 3rd ed. approach at Zeneca for this target was through virtual screen-
Mendelsohn, J.; Howley, P.M.; Israel, M. A.; Gray, J. W.; Thompson, C. B. ing (see Chapter 2, Section 2.1.2.3.4), which searched for
(Eds.), Elsevier, 2008, pp. 155–168.
compounds with the following characteristics (recall that one
of the main requirements for virtual screening was a hypoth-
Thus, interaction of EGF with the extracellular domain of the esis of features required for interaction with the target):
receptor results in activation of the kinase domain inside the
cell and represents one mechanism by which signals outside 1. A six-membered aromatic ring to mimic the phenolic
a cell can lead to specific responses inside the cell (this is one ring of tyrosine in the substrate.
example of how a receptor, in this case EGFR, propagates the 2. An oxygen or nitrogen atom at a suitable distance from
response to an agonist, in this case EGF). Note that EGFR is the above aromatic ring to mimic the phenolic hydroxyl
a member of the family of receptor tyrosine kinases, which group of tyrosine in the substrate.
function as receptors to mediate the transmission, via the 3. Any oxygen or nitrogen atom at a suitable distance
kinase domain, of signals of certain growth factors. Here we from the tyrosine hydroxyl group to mimic at least two
will be primarily concerned with the kinase activity of EGFR, of the nonbridging oxygens of the triphosphate group
that is, the enzymatic activity of the intracellular kinase of ATP.
domain of the receptor (see Figure 5.1).
Analysis of cells from a number of different tumor types, The screening compounds were a 1500-member subset
for example, from lung or colon cancer, revealed abnormal- of the Zeneca corporate collection, which totaled about
ities in EGF or EGFR.[4] In some cases, signals for growth 250,000 compounds at that time. The screening subset was
and proliferation were being transmitted into the nucleus of designed to be representative of the range of chemotypes
these cells even in the absence of EGF, whereas in normal in the broader collection. A single low-energy conforma-
cells EGF is required for initiation of these signals. Thus, tion was predicted for each compound in the screening set
inhibition of EGFR signaling was recognized as a possible using Concord software, and Aladdin software was used
approach to slow the rate of growth of tumors harboring to computationally compare the test compounds with the
aberrant signaling via EGFR. Possible approaches include requirements of the hypothesis (the search query). The
antagonists at the receptor-binding site or inhibition of the search query matched 152 compounds, three of which in
Chapter | 5 Enzyme Inhibition and Inactivation 213

fact inhibited EGFR kinase with reasonable potency. Addi- An X-ray crystal structure of erlotinib in complex with
tional compounds from the corporate collection were then EGFR kinase shows a likely hydrogen bonding interac-
tested on the basis of their similarity to these inhibitors. In tion between the quinazoline N1 and the backbone NH of
this way, compound 5.9 was discovered, which exhibited Met-793 of the enzyme (Figure 5.2).[8] An X-ray crystal
an IC50 of 40 nM. Kinetic studies showed that this inhibitor structure has also been solved for the complex of EGFR
is not competitive with peptide substrate (in contrast to the with ATP analog adenosine-5′[β,γ-imido]triphosphate
hypothesis used in the virtual screen) but is competitive with (AMP-PNP, 5.14, Figure 5.3). AMP-PNP is identical to
ATP. At about the same time, Parke-Davis Pharmaceuticals ATP, except that the distal phosphoric anhydride oxygen
(subsequently acquired by Pfizer, then closed) reported has been replaced by NH (ionized to N− at physiological
that research starting with a mass screening campaign had pH) to stabilize the terminal phosphate linkage. As shown
yielded compounds 5.10 and 5.11 as EGFR kinase inhibi- in Figure 5.3, AMP-PNP also forms a hydrogen bond with
tors, which exhibited IC50 values of 0.31 and 0.025 nM, Met-793 of EGFR; a comparison of Figures 5.2 and 5.3
respectively.[6] While differences in assay conditions used suggests that the erlotinib-binding region, at least partially,
by the two groups do not allow a direct comparison of the overlaps with the ATP-binding site, which strongly supports
potencies of 5.9 and 5.10, it is apparent that the 6,7-dime- the conclusion deduced from enzyme kinetic studies.
thoxyl functionality of 5.10 contributes substantially to
increased potency. This effect may, at least in part, be the 5.2.2.2. Stabilization of an Inactive
result of the electron-donating properties of the methoxyl Conformation: Imatinib, an Antileukemia Drug
groups leading to increased electron density on the quinazo-
line nitrogen atoms. Subsequent lead optimization studies 5.2.2.2.1. The Target: Bcr-Abl, a Constitutively Active
focused on varying the substituents on the benzo portion of Kinase
the quinazoline ring as well as on the phenyl ring of the ani- Chronic myelogenous leukemia (CML) is a disease char-
lino appendage. These efforts led to two marketed drugs to acterized by excessive proliferation in the bone marrow of
treat lung and other cancers: gefitinib (5.12, Iressa) from the specific types of white blood cells and their precursors. It
Zeneca efforts and erlotinib (5.13, Tarceva) from the Parke- was found that most cases of CML were associated with
Davis lead compound. Note that in both cases, lead opti- the Philadelphia chromosome, a chromosomal abnormal-
mization led to incorporation of polar functionality in the ity that led to the expression of an abnormal protein called
substituents on the benzo portion of the quinazoline ring, Bcr-Abl. Abl, short for Abelson (Abl) kinase, is a tyrosine
no doubt intended to increase solubility of the compounds kinase normally expressed in many cells, and which like
for improved pharmacokinetic and pharmaceutical prop- other tyrosine kinases catalyzes the phosphorylation of phe-
erties. The fact that inhibition of EGFR kinase by 5.13 is nolic hydroxyl groups on the side chain of tyrosine residues
competitive with ATP (IC50 = 2 nM), in analogy to its earlier of other proteins. Bcr (short for breakpoint cluster region) is
analogs, was confirmed by kinetic studies.[7] a polypeptide fragment that is fused to Abl as coded by the

HN X HN Cl
HN Cl CH3O O
N N N
N
O
CH3O N CH3O N
N

5.9 5.10, X = Cl Gefitinib

5.11, X = Br 5.12

HN C
C
CH3 O H
O N
O
CH3 O N

Erlotinib
5.13
214 The Organic Chemistry of Drug Design and Drug Action

(A) O
(B)
O
Met-793
O
N
H O
N O
H

N NH

Erlotinib
5.13

FIGURE 5.2 (A) Schematic drawing of the interactions of erlotinib (5.13) with the active site of EGFR kinase. Modified image from PoseView
(University of Hamburg, Germany) created from Protein Data Bank ID 1M17. EGFR residue numbering corresponds to that in PDB 2ITX. (B) Three-
dimensional structure of erlotinib (green) bound to EGFR kinase (selected residues shown in blue). Data from Protein Data Bank ID 1M17 modified using
Accelrys DS Viewer software. (PDB ID:1M17). Stamos, J., Sliwkowski, M.X., Eigenbrot, C. Structure of the epidermal growth factor receptor kinase
domain alone and in complex with a 4-anilinoquinazoline inhibitor. J. Biol. Chem. 2002, 277, 46265–46272.

FIGURE 5.3 Schematic drawing of the interactions of stable ATP analog AMP-PNP (5.14) with the active site of EGFR kinase. Modified image from
PoseView (University of Hamburg, Germany) created from Protein Data Bank ID 2ITX. PDB ID: 2ITX (Yun, C.-H., Boggon, T.J., Li, Y., Woo, S., Greulich,
H., Meyerson, M., Eck, M.J. Structures of lung cancer-derived EGFR mutants and inhibitor complexes: mechanism of activation and insights into dif-
ferential inhibitor sensitivity. Cancer Cell 2007, 11, 217–227.)

Philadelphia chromosome. The effect of the fused Bcr seg- kinase. However, in contrast to EGFR, Abl is a soluble
ment is to significantly enhance the kinase activity of Abl. In kinase found in the cytosol and, therefore, does not have
1990, three different groups reported that mice transfected a transmembrane receptor domain. On the basis of the
with the bcr-abl gene developed symptoms closely resem- interactions of compounds like AMP-PNP to kinases
bling CML in humans.[9] (cf. Figure 5.3), many aminopyrimidine-like molecules
designed to mimic one or more of these key interactions
were synthesized and tested against various kinases as
5.2.2.2.2. Lead Discovery and Modification potential inhibitor leads.
With validation of Bcr-Abl as a target for CML at hand, The Novartis team had originally identified anilinopyrimi-
scientists at Ciba–Geigy (now Novartis since the Ciba- dine 5.15 as a lead structure for inhibition of a different kinase,
Geigy merger with Sandoz in 1996) set out to discover platelet-derived growth factor receptor (PDGFR) kinase.[11]
an inhibitor.[10] Like EGFR kinase, Abl is a tyrosine Key objectives of that effort were to enhance potency for
Chapter | 5 Enzyme Inhibition and Inactivation 215

H H H H H
N N N N N N N N

N N O N O
H 3C

N N N

5.15 5.16 5.17

IC50 (µM) IC50 (µM)

PDGFR 5 PDGFR 0.1
PKC 1.2 PKC 72

PDGFR kinase as well as increase selectivity of PDGFR kinase

over protein kinase C (PKC). Addition of a benzamido moi- 5.2.2.2.3. Binding Mode of Imatinib to Abl Kinase
ety gave 5.16 having low micromolar potency against PDGFR Shortly after imatinib was approved for treatment of CML
kinase, but poor selectivity against PKC. Remarkably, among in 2001, a crystal structure of the drug bound to Abl kinase
a number of analogs synthesized, compound 5.17, bearing one was published.[12] In the bioactive conformation of ima-
additional methyl group on the central phenyl ring, resulted tinib, the central phenyl ring is positioned approximately
in a 50-fold increase in potency against PDGFR kinase and a 90° out of the plane of the neighboring pyrimidine ring; the
60-fold decrease in potency against PKC! A possible role for flag methyl group promotes this conformation by destabi-
this methyl group (called the “flag methyl” by the Ciba-Geigy lizing the planar conformation shown in 5.18 because of
team) is discussed in Section 5.2.2.2.3. steric hindrance.
After identification of Abl kinase as an important therapeu- In the same paper, the complex of Abl kinase with
tic target, 5.17 was found to be a potent inhibitor of the enzyme, another inhibitor, PD173955 (5.19), was also reported.
but it was not yet an ideal drug since it exhibited poor bioavail- Interestingly, the conformation of the enzyme was shown
ability. Since the aqueous solubility of 5.17 was poor, it was to be markedly different in the two structures (Figure 5.4).
hypothesized that increasing the solubility would also lead to In Figure 5.4, the position of the activation loop (red in (A);
improved bioavailability, on the basis of the following rationale. blue in (B)) prominently defines the two different confor-
When a compound is poorly soluble, poor absorption from the mations. Only the conformation in (A) is able to bind ATP
gastrointestinal tract may result, since the compound must be (when PD173955 is absent), and is therefore called the
in solution prior to passing through gut membranes. In addi- “active conformation”. On the other hand, imatinib cannot
tion, once a compound reaches the bloodstream from the gut, it bind to the active conformation, since the extended portion
must next pass through the liver, where metabolic enzymes tar- of the molecule (see arrow in Figure 5.4) would clash with
get hydrophobic compounds for metabolic conversion to more a portion of the activation loop. In contrast, the inactive con-
soluble derivatives to promote elimination in the urine. There- formation in (B) contains an extended region that not only
fore, the next objective was to increase the solubility of 5.17. accommodates but also forms specific interactions with the
Appending hydrophilic groups such as N-methylpiperazine, extended portion of imatinib. Therefore, imatinib binds to
in this case through a methylene linker, proved successful in the inactive conformation and stabilizes it, thereby prevent-
improving bioavailability leading to imatinib (5.18, Gleevec). ing ATP from binding and causing inhibition of the enzyme.

Cl
N
H H
N N N N
CH3 N
N O H3 C Cl
H3C S N N N O
H
CH3

N imatinib PD173955
5.18 5.19
IC50 (µM)
Abl 0.1 - 0.3 It is assumed that the different conformational states
PDGFR 0.1 of the apoenzyme (e.g., Ablactive and Ablinactive in Scheme
c-KIT 0.1 5.2) are present and in equilibrium in a cell at rest (see
216 The Organic Chemistry of Drug Design and Drug Action

FIGURE 5.4 Ribbon representations of the structure of the Abl kinase domain (green) in complex with (A) PD173955 and (B) imatinib (originally
known as STI-571). The protein strand colored red in (A) and blue in (B) is the “activation loop” of the kinase; the arrow points to the N-methylpiperazine
portion of imatinib. Adapted by permission from the American Association for Cancer Research: Nagar, B. et al. Crystal structures of the kinase domain
of c-Abl in complex with the small molecule inhibitors PD173955 and imatinib (STI-571). Cancer Res. 2002, 62, 4236–4243.

ATP Imatinib
Ablactive...ATP Ablactive Ablinactive Ablinactive...imatinib

Protein substrate

Phosphorylated
Protein substrate
SCHEME 5.2 Equilibrium between active and inactive conformations of Abl kinase, where the active conformation binds to ATP, leading to catalysis of
the normal enzymatic reaction; the inhibitor imatinib binds to and stabilizes the inactive form, pulling the equilibrium to the right.

Chapter 3, Section 3.1). Normal cells have regulatory A tangible example in support of this hypothesis is that
mechanisms that influence the position of this equilib- PD173955, which binds to the active conformation of Abl,
rium. For example, phosphorylation of one or more amino is also a potent inhibitor of the related kinase Src, whereas
acid hydroxyl groups of Ablinacive by another kinase serves imatinib is not, despite the high sequence and structural
to shift the equilibrium in Scheme 5.2 away from Ablinac- ­similarity of the two enzymes.
tive and toward Ablactive, resulting in increased enzymatic
activity. Phosphorylation by kinases (counterbalanced by 5.2.2.2.4. Inhibition of Other Kinases by Imatinib
dephosphorylation by phosphatases) has been established Despite the foregoing discussion, imatinib does inhibit
as a key mechanism for cellular regulation (in simple other enzymes in addition to Abl, including platelet-derived
terms, serving as on and off switches for many cellular growth factor receptor (PDGFR) and KIT. Both of these
processes).[13] The Bcr fusion in Bcr-Abl is an abnormal enzymes have been found to be overactive in certain types
mechanism for shifting the equilibrium in Scheme 5.2 of tumors. Overactivated KIT is a primary driver of gas-
toward Ablactive. trointestinal stromal tumors (GIST), and imatinib has also
The presence of multiple conformations, as character- been approved as a new treatment for this disease. Since
ized by the position of the activation loop, is a common the approval of imatinib for chronic myelogenous leuke-
theme among protein kinases. It has been hypothesized mia (CML) and GIST, targeting kinases for cancer has been
that the design of inhibitors that preferentially interact with a significant focus in the pharmaceutical industry. More
an inactive conformation, for example, in a manner simi- recently, research on kinase inhibitors to treat inflamma-
lar to imatinib, offers an approach to achieve selectivity of tion and autoimmune diseases,[15] which involve overpro-
the inhibitor among different kinases.[14] Support for the liferation of certain cells of the immune system, also has
hypothesis comes from the observation that the extended been initiated. The challenge of this area of research is that
region present in the inactive conformation is more dissimi- a higher margin of safety is generally required for such dis-
lar among kinases than is the site in the active conforma- eases, which may be debilitating, but usually are not life
tion, which was designed by nature to interact with ATP. threatening.
Chapter | 5 Enzyme Inhibition and Inactivation 217

5.2.2.3. Alternative Substrate Inhibition: a bacteriostatic drug inhibits further growth of the bacteria,
Sulfonamide Antibacterial Agents (Sulfa Drugs) thus allowing the host defenses to catch up in their fight
against the bacteria. Because microorganisms replicate
Competitive inhibition of an enzyme also can be attained rapidly (with Escherichia coli, for example, the number of
with a molecule that not only binds to the enzyme but also cells can double every 20–30 min), a bacteriostatic agent
acts as a substrate. In this case, however, the product pro- will interrupt this rapid growth and allow the immune sys-
duced is not a compound that is useful to the organism. tem to destroy the organism. Of course, with immunocom-
While the alternative substrate is being turned over, it is promised individuals, who are unable to contribute natural
preventing the actual substrate from being converted to the body defenses to fight their disease, a bacteriocidal agent is
product that the organism needs. The principal disadvantage necessary to prevent continuation of growth of the organism
to this approach is that the product generated could be toxic when the drug is withdrawn.
or cause an unwanted side effect. In the example below, this
is not the case.
H2 N SO2NH2
5.2.2.3.1. Lead Discovery
At the beginning of the twentieth century, Paul Ehrlich Sulfanilamide
5.21
showed that various azo dyes were effective agents against
trypanosomiasis in mice; however, none was effective in
man. In the early 1930s, Gerhard Domagk, head of bacte-
riological and pathological research at the Bayer Company 5.2.2.3.2. Lead Modification
in Germany, who was trying to find agents against strepto- The discovery of Prontosil marks the beginning of modern
coccal infections, tested a variety of azo dyes. One of the chemotherapy. During the next decade thousands of sulfon-
dyes, Prontosil (5.20), showed dramatically positive results, amides were synthesized and tested as antibacterial agents.
and successfully protected mice against streptococcal infec- These were the first structure–activity relationship stud-
tions.[16] However, Bayer was unwilling to move rapidly on ies (see Chapter 2, Section 2.2.3), which demonstrated the
getting Prontosil onto the drug market. As Albert[17] tells it, importance of molecular modification in drug design. Also,
when, in late 1935, Domagk’s daughter cut her hand and this was one of the first examples where new lead com-
was about to die of a streptococcal infection, her father gave pounds for other diseases were revealed from side effects
her Prontosil! Although she turned bright red from the dye, observed during pharmacological and clinical studies (see
her recovery was rapid,[18] and the effectiveness of the drug Chapter 1, Sections 1.2.4 and 1.2.5; Chapter 2, Section
became quite credible. In 1939, Domagk was awarded the 2.1.2.2). These early studies led to the development of new
Nobel Prize in Medicine for this achievement. antidiabetic and diuretic agents. Another important scientific
advance that was derived from work with sulfonamides was
NH2 a simple method for the assay of these compounds in body
fluids and tissues.[20] This method involved diazotization of
H2N N the aromatic amino group followed by diazo coupling with
N SO2NH2 dimethyl-α-naphthylamine to produce a dye that could be
measured colorimetrically. Furthermore, it was shown that
Prontosil
5.20 the antibacterial effect of sulfanilamide was proportional to
its concentration in the blood, and that at a given dose this
An unexpected property of Prontosil, however, was that varied from patient to patient. This was the beginning of the
it had no activity against bacteria in vitro. Tréfouël and monitoring of blood drug levels during chemotherapy treat-
coworkers[19] found that if a reducing agent was added to ment, which led to the initiation of the routine use of phar-
Prontosil, then it was effective in vitro. They suggested that macokinetics, the study of the absorption, distribution, and
the reason for the lack of in vitro activity, but high in vivo excretion of drugs, in drug development programs. Proper
activity, was that Prontosil was metabolized by reduction drug dosage requirements could now be calculated.
to the active antibacterial agent, namely, p-aminoben-
zenesulfonamide (also called sulfanilamide) (5.21; AVC); 5.2.2.3.3. Mechanism of Action
therefore, Prontosil is a prodrug (see Chapter 9), a com- On the basis of the work by Stamp,[21] who showed that
pound that requires metabolic activation to be effective. bacteria and other organisms contained a heat-stable sub-
Furthermore, they demonstrated that sulfanilamide was as stance that counteracted the antibacterial action of sulfon-
effective as Prontosil in protecting mice against strepto- amides, Woods[22] in 1940 reported a breakthrough in the
coccal infections, and that it exerted a bacteriostatic effect determination of the mechanism of action of this class of
in vitro. Unlike a bacteriocidal agent, which kills bacteria, drugs. He hypothesized that because enzymes are inhibited
218 The Organic Chemistry of Drug Design and Drug Action

by compounds whose structures resemble those of their sub- folic acid biosynthesis by sulfonamides was competitively
strates, the counteractive substance should be a substrate for reversed by p-aminobenzoic acid. Two enzymes from E. coli
an essential enzyme, and the substrate should have a structure were purified by Richey and Brown,[27] one that catalyzed the
similar to that of sulfanilamide. After various chemical tests, diphosphorylation of 2-amino-4-hydroxy-6-hydroxymethyl-
and a vague notion of the possible structure of this counter- 7,8-dihydropteridine (5.23) and the other (dihydropteroate
active substance, he deduced that it must be p-aminobenzoic synthase), which catalyzed the synthesis of dihydrofolate
acid (5.22) and proceeded to show that 5.22 potently coun- (5.25) from diphosphate 5.24 and p-aminobenzoic acid
teracted sulfanilamide-induced bacteriostasis. The results of (Scheme 5.3). The name of the enzyme stems from the fact
his experiments showed that sulfanilamide was competitive that folic acid is a derivative of the pterin ring system (5.26,
with p-aminobenzoic acid for microbial growth. To main- R = H). Because of the structural similarity of sulfanilamide
tain growth with increasing concentrations of sulfanilamide, to p-aminobenzoic acid, it is a potent competitive inhibitor
it is also necessary to increase the concentration of p-ami- of the second enzyme. The reversibility of the inhibition
nobenzoic acid. Selbie[23] found that coadministration of
p-­aminobenzoic acid and sulfanilamide into streptococcal- H2N N N
infected mice prevented the antibacterial action of the drug.
HN
N R
O
H2N COOH
5.26
5.22
was demonstrated by Weisman and Brown,[28] who sug-
The observation of competitive inhibition by sulfanil- gested that sulfonamides were incorporated into the dihy-
amide was the basis for Fildes[24] to propose his theory of drofolate. This was verified by Bock et al.,[29] who incubated
antimetabolites. An antimetabolite is a compound that dis- dihydropteroate synthase with diphosphate 5.24 and [35S]-
rupts metabolic processes, typically by blocking or acting as sulfamethoxazole (5.27) and identified the product as 5.28
an alternative substrate for an enzyme in a metabolic pathway. (Scheme 5.4). Therefore, this is an example of competitive
He proposed a rational approach to chemotherapy, namely, reversible inhibition in which the inhibitor also is a substrate.
enzyme inhibitor design, and suggested that the molecular However, the product (5.28) cannot produce dihydrofolate,
basis for enzyme inhibition was that either the inhibitor com- and, therefore, the organism cannot get the tetrahydrofolate
bines with the enzyme and displaces its substrate or coen- needed as a coenzyme to make purines (see Chapter 4, Sec-
zyme or it combines directly with the substrate or coenzyme. tion 4.3.2), which are needed for DNA biosynthesis. This is
In the mid-1940s, Miller and coworkers[25] demonstrated why the sulfonamides are bacteriostatic, not bacteriocidal.
that sulfanilamide inhibited folic acid biosynthesis, and in Inhibition of tetrahydrofolate biosynthesis only inhibits rep-
1948, Nimmo-Smith et al.[26] showed that the inhibition of lication; it does not kill the existing bacteria.

H ATP AMP H
H2N N N H2 N N N
O O
N OH N O P O P O–
N N
O– O–
HO HO
5.24
5.23

Dihydropteroate ..
synthase H 2N COO –

H
H2N N N

N
N
HO HN COO –

5.25
SCHEME 5.3 Biosynthesis of bacterial dihydrofolic acid
Chapter | 5 Enzyme Inhibition and Inactivation 219

H
H2N N N
O O N O
N
N
O P O P O– + H2N 35SO
2NH
CH3
O– O–
HO
5.24 5.27

O O Dihydropteroate
–O P O P O– synthase
O – O–

H
H 2N N N

N
N N O
HO HN 35SO
2NH
CH3
5.28
SCHEME 5.4 Dihydropteroate synthase use of sulfamethoxazole in place of para-aminobenzoic acid

Inhibitors of dihydropteroate synthase, however, have attain, namely, that resistance to inhibitors of the target not
no effect on humans, because we do not biosynthesize folic be easily acquired. Typically, it takes between 1 and 4 years
acid and, therefore, do not have that enzyme. Folic acid is a for resistance to an antibacterial drug to emerge; in the case
vitamin and must be eaten by humans. Furthermore, because of the sulfonamides, it was almost 7 years.[33] There are sev-
bacteria biosynthesize their folic acid, they do not have a eral mechanisms of resistance to sulfonamide antibacterial
transport system for it.[30] Consequently, we can eat all the drugs; these are discussed in Chapter 7.
folic acid we want, and the bacteria cannot utilize it. This
is another example of selective toxicity, inhibition of the
growth of a foreign organism without affecting the host, and 5.2.3. Transition State Analogs and
falls into the category of an ideal enzyme inhibitor (see Sec- Multisubstrate Analogs
tion 5.1). It is interesting to note that sulfonamides are not
5.2.3.1. Theoretical Basis
effective with pus-forming infections because pus contains
many compounds that are the end products of tetrahydrofo- As discussed in Chapter 4 (Section 4.1.2.2), an enzyme
late-dependent reactions, such as purines, methionine, and accelerates the rate of a reaction by stabilizing the tran-
thymidine. Therefore, inhibition of folate biosynthesis is sition state, which lowers the free energy of activation.
unimportant, and pus can c­ ontribute to bacterial sustenance. The enzyme achieves this rate enhancement by changing
A major deterrent to the use of sulfa drugs is their central its conformation so that the strongest interactions occur
nervous system (CNS) side effect profile, including head- between the substrate and enzyme active site at the transi-
aches, tremors, nausea, vomiting, and insomnia. The side tion state of the reaction. Some enzymes act by straining
effects are believed to derive from the off-target inhibition or distorting the substrate toward the transition state. The
of the enzyme sepiapterin reductase, an enzyme involved in catalysis-by-strain hypothesis led to early observations that
the biosynthesis of the coenzyme tetrahydrobiopterin (5.26 some enzyme inhibitors owe their effectiveness to a resem-
doubly reduced; R = (1R,2S)-dihydroxypropyl), which is blance to the strained species. Bernhard and Orgel[34] theo-
essential for producing several neurotransmitters, such as rized that inhibitor molecules resembling the transition state
serotonin, dopamine, norepinephrine, and epinephrine; species would be much more tightly bound to the enzyme
depletion of these neurotransmitters could cause the neuro- than would be the substrate; 11 years prior to that Pauling
logical side effects.[31] had mentioned that the best inhibitor of an enzyme would
Dihydropteroate synthase satisfies at least three of four be one that resembled the “activated complex”.[35] There-
important criteria for a good antimicrobial drug target: fore, why should we design inhibitors on the basis of the
(1) the target is essential to the survival of the microorgan- ground state substrate structure? A potent enzyme inhibi-
ism; (2) the target is unique to the microbe, so its inhibition tor should be a stable compound whose structure resembles
does not harm humans; and (3) the structure and function that of the substrate at a postulated transition state (or tran-
of the target is highly conserved across a variety of spe- sient intermediate) of the reaction rather than that at the
cies of that microbe so that inhibitors are broad-spectrum ground state. A compound of this type would bind much
agents.[32] The fourth criterion may be the most difficult to more tightly to the enzyme, and is called a transition state
220 The Organic Chemistry of Drug Design and Drug Action

analog inhibitor. Jencks[36] was the first to suggest the exis- of a reaction that breaks a bond to a 1H can be seven times
tence of transition state analog inhibitors, and cited several greater than that of the same reaction with a 2H, whereas the
possible literature examples; Wolfenden[37] and Lienhard[38] corresponding reaction with a 12C will only be 1.04 times
developed the concept further. Values for dissociation con- that with 13C.
stants (Ki) of 10−15 M for enzyme–transition state com- KIEs provide information regarding bonding and geo-
plexes may not be unreasonable given the normal range of metric differences between reactants and transition state
10−3–10−5 M for dissociation constants of enzyme–substrate structures. If KIEs are combined with computational meth-
complexes (Km). ods that correlate experimental values of van der Waals
For the design of a transition state inhibitor to be effec- geometry and electrostatic potential surfaces with pre-
tive, the mechanism of the enzyme reaction must be under- dicted chemical models, a three-dimensional depiction of
stood, so that a theoretical structure for the substrate at the the transition state of an enzyme-catalyzed reaction can be
transition state can be hypothesized. Because many enzyme- obtained, which permits the design of transition state ana-
catalyzed reactions have similar transition states (for exam- log inhibitors.[45] Because geometry and electrostatics are
ple, the different serine proteases), the basic structure of a the dominant properties for ligand binding, the more simi-
transition state analog for one enzyme can be modified to lar these properties are to those of the transition state,[46]
meet the specificity requirements of another enzyme in the the more tightly the molecule will bind. The ideal enzyme
same mechanistic class and, thereby, generate a transition target for transition state analog inhibitor design, therefore,
state analog for the other enzyme. This modification may be would catalyze reactions that have altered geometry and
as simple as changing an amino acid in a peptidyl transition charge on the conversion of reactants to the transition state.
state analog inhibitor for one protease so that it conforms An example of drug design that uses this approach is given
to the peptide specificity requirement of another protease. in Section 5.2.3.2.3.
Christianson and Lipscomb[39] have renamed reversible When more than one substrate is involved in the enzyme
inhibitors that undergo a bond-forming reaction with the reaction, a single stable compound can be designed that has
enzyme prior to the observation of the enzyme–inhibitor a structure similar to that of the two or more substrates at
complex as reaction coordinate analogs. the transition state of the reaction. This special case of a
Schramm[40] argues that transition states only have a transition state analog is termed a multisubstrate analog
lifetime of a few femtoseconds,[41] whereas enzymes have inhibitor. Because these compounds are a combination of
turnover numbers (kcat) on the millisecond timescale; there- two or more substrates, their structures are unique, and they
fore, the transition state occupies only about 10−12 of the are often highly specific. Their great binding affinity for the
enzymatic reaction cycle, and the transition state complex target enzyme arises because the free energy of binding is
will occupy the enzyme about one-trillionth of the time and roughly the product of the free energies of each independent
the remainder will be in the ground state. Therefore, it is not substrate that it mimics.
reasonable to expect a thermodynamic equilibrium explana-
tion, such as binding, to account for this short-lived chemical 5.2.3.2. Transition State Analogs
event. Rather, to determine transition states, kinetic isotope
effect (KIE) measurements of the chemical steps,[42] which 5.2.3.2.1. Enalaprilat
perturb bond vibrational states and influence the transition Enalaprilat (5.29, Figure 5.5) is a very potent slow, tight-
state, should be used[43]; binding isotope effects would only binding inhibitor of angiotensin-converting enzyme (ACE)
give insight into ground state interactions.[44] (see Section 5.2.4.2). The rationalization for its binding
For those of you who have not studied KIEs, a brief effectiveness was that it had multiple binding interactions
description follows. For any bond that is broken in the rate- with the substrate- and product-binding sites. The resem-
determining step, the rate of the reaction will be affected blance of enalaprilat to the substrate and products of the
by incorporation of an isotope for one of the atoms con- enzyme reaction (Figure 5.5) supports this notion.
nected to the bond that breaks (called a primary KIE). The Both of these rationalizations are ground state argu-
degree of this effect will be determined by the mass of ments, but transition state theory suggests that the most
the atom that is substituted because that is related to the effective interactions occur at the transition state of the
energy of the bond that breaks: the greater the mass, the reaction. With that in mind, let’s consider a potential mech-
lower the vibrational frequency, which means the lower the anism for ACE-catalyzed substrate hydrolysis (Scheme
zero-point energy and, therefore, the more energy needed to 5.5) and see if the transition state structure is relevant. It
break the bond. A C-2H bond, then, is a little stronger than a is not known if a general base mechanism, as shown in
C-1H bond and requires more energy to break. The substitu- Scheme 5.5, or a covalent catalytic mechanism is involved.
tion of a 2H for a 1H (mass difference of 100%) will have a Enalaprilat has been drawn beneath transition states 1
much greater effect than the substitution of a 13C atom for and 2 (‡1 and ‡2) to show how the structures are related.
12C (mass difference of only 1/12 or about 8%). The rate An enzyme conformational change at the transition state
Chapter | 5 Enzyme Inhibition and Inactivation 221

(shown as a blocked enzyme instead of a rounded enzyme of the enzyme adenosine deaminase (adenosine aminohy-
in Scheme 5.5) could increase the binding interactions. drolase).[47] The Ki is 2.5 × 10−12 M which is 107 times lower
than the Km for adenosine! As in the case of enalaprilat
S1 (Sections 5.2.3.2.1 and 5.2.4.2), pentostatin is a slow, tight-
binding inhibitor (see Section 5.2.4).[48] The kon with human
erythrocyte adenosine deaminase is 2.6 × 106 m−1s−1 , and
S1' B
S2' the koff is 6.6 × 10−6 s−1. The very small koff value is reflected
CH2
ZnII
CH3
H
+ in the very low Ki value (Ki = koff/kon).
O O
O
NH CH C NH CH C N CH C -O Substrate
HO H
N
HN
CH2 CH3 O N
N
O O HO O
NH CH C - NH2 CH N CH C - Product
O O

OH
Pentostatin

CH2
O - O CH3
5.30
C O O 5.29
CH2 CH NH CH C N CH C - Enalaprilat Pentostatin is an analog of the natural nucleoside
O
2′-deoxyinosine (5.31, Scheme 5.6), in which the purine is
FIGURE 5.5 Hypothetical interactions of enalaprilat with ACE modified to contain a seven-membered ring with two sp3
carbon atoms. It is believed that this compound mimics the
5.2.3.2.2. Pentostatin transition state structure of the substrates adenosine and
2′-Deoxycoformycin (pentostatin, 5.30; Nipent), an anti- 2′-deoxyadenosine (5.31, Scheme 5.6) during their hydro-
neoplastic agent isolated from fermentation broths of the lysis to inosine and 2′-deoxyinosine (5.34), respectively, by
bacterium Streptomyces antibioticus, is a potent inhibitor adenosine deaminase.

‡1
Zn++
Zn++ Zn++
R1 R1' R1 O δ− R1' R1 O- R1'
O
NHCH C NHCH NHCH C NH CH NH CH C NH CH
−
O H
Oδ H B
O H +
H H H δ+
:B B

Ph O - O
CH3 O
CH2 C
CH2 CH NH CH C Pro
Enalaprilat
‡2
Zn++
R1 − R1'
Zn++ Oδ
R1 R1'
O NH CH C NH CH
NHCH C - NH2CH O δ−
H δ+
O H B

HB Ph O - O
+ CH3 O
CH2 C
CH2 CH NH CH C Pro
Enalaprilat
SCHEME 5.5 Hypothetical mechanism for ACE-catalyzed peptide hydrolysis
222 The Organic Chemistry of Drug Design and Drug Action

238
238
His
His 238His

N NH
NH NH
N H 296Asp HN 296Asp
296Asp O
H O H
O 295Asp H H O 295Asp H O
H O
295Asp
H O H2N H O
O O NH2 H O O O NH2 H O
O
Zn2+ N N Zn2+ N N
Zn2+ N N H
O H O
O H 217
Glu N 217 N
217 N N Glu N
Glu N HO HO
HO O O O O
O O

OH OH
OH
5.33
5.31 5.32
HO H
OH
NH3 N
N N HN
N
N N
N HO O
HO
O

OH
OH Pentostatin
5.34 5.30
SCHEME 5.6 Hypothetical mechanism for adenosine deaminase-catalyzed hydrolysis of 2′-deoxyadenosine

A crystal structure to 2.4 Å resolution of the transition hydrolase.[52] Ribonucleotide reductase catalyzes the con-
state analog 6R-hydroxy-1,6-dihydropurine ribonucleo- version of ribonucleotides to the corresponding 2′-deoxyri-
side (5.35), complexed to adenosine deaminase, revealed bonucleotides and is essential for DNA biosynthesis. Thus,
a zinc ion cofactor and suggested a possible mechanism inhibition of this enzyme leads to inhibition of DNA bio-
for the enzyme.[49] In this case, the resemblance of pen- synthesis. S-Adenosylhomocysteine competitively inhibits
tostatin (5.30) to the intermediate 5.33 (Scheme 5.6) is most of the methyltransferases that utilize S-adenosylmethi-
clearer than its similarity to the transition state shown onine (SAM) as the methyl-donating agent. This, apparently,
(5.32); however, a late transition state would look more is a mechanism for the regulation of these methyltransfer-
like 5.33. ases. Inhibition of S-adenosylhomocysteine hydrolase, the
enzyme that degrades S-adenosylhomocysteine, results in
OH accumulation of S-adenosylhomocysteine, which inhibits
H the growth and replication of various tumors (and viruses),
N
N particularly those requiring a methylated 5′-cap structure
N N on their messenger ribonucleic acids. Furthermore, various
HO lymphocytic functions are suppressed by the accumulation
O of extracellular adenosine. Resistance to antipurines is dis-
cussed in Chapter 7.
HO OH
5.35 5.2.3.2.3. Forodesine and DADMe-ImmH
Examples of transition state analog inhibitor design com-
Pentostatin is most effective in lymphocytic disease, bining KIEs and computational methods are immucillin-
for example, hairy cell leukemia and chronic lymphocytic H (ImmH; forodesine; 5.36), in clinical trials for T-cell
leukemia.[50] It is known that adenosine deaminase lev- malignancies,[53] and 4′-deaza-1′-aza-2′-deoxy-1′-(9-
els are elevated in T-lymphocytes (white blood cells that methylene)immucillin-H (DADMe-ImmH; BCX4208;
mature in the thymus).[51] Since inhibition of an enzyme 5.37), in clinical trials for gout,[54] both potent inhibi-
can lead to accumulation of its substrate(s), one hypothesis tors of purine nucleoside phosphorylase (PNP), which
for the action of pentostatin is that it induces lymphotox- catalyzes the phosphate-dependent conversion of purine
icity via accumulation of 2′-deoxyadenosine, which inhib- nucleosides or deoxynucleosides to α-D-(deoxy)ribose-
its ribonucleotide reductase and S-Adenosylhomocysteine 1-phosphate and the purine base (Scheme 5.7).[55]
Chapter | 5 Enzyme Inhibition and Inactivation 223

HO HO
Purine base
O O
+ HPO42– + Purine base
OH OPO32–
OH
SCHEME 5.7 Reaction catalyzed by PNP

Inhibition of PNP causes elevated deoxyguanosine,[56] ‡

O
which is triphosphorylated by lymphocytes,[57] leading to the H
N O
accumulation of deoxyguanosine triphosphate, known to be HN H
OH N
toxic to T cells[58]; T-cell deficiency induces apoptosis (cell N NH
N
death).[59] Normally, there is no measurable deoxyguanosine 1.77 Å O δ+ HO H H
N
in human plasma or urine because of highly active PNP, so 3.0 Å N
HO
deoxyguanosine triphosphate does not accumulate. Diseases OH
O
that could be impacted by controlling T-cell function by O HO OH
P 5.36
inhibiting PNP include T-cell leukemia and lymphoma, tissue –
O OH
transplant rejection, and T-cell autoimmune diseases, such as
psoriasis, inflammatory bowel disease, rheumatoid arthritis,
Transition state structure
for bovine PNP
and insulin-dependent juvenile diabetes.[60] FIGURE 5.6 Hypothesized transition state structure for the reaction
Using KIE studies with bovine PNP (assumed to be simi- catalyzed by bovine PNP; structure of forodesine (5.36).
lar to human PNP because of 87% amino acid sequence iden-
tity) and inosine as the substrate, a transition state structure
shown in Figure 5.6 was deduced.[61] The computer-aided O
H
‡ O
comparison of the electrostatic image of the transition state N H
N
HN NH
and a series of analogs led to the design of forodesine (5.36,
N OH
Figure 5.6) as a transition state inhibitor; the Kd was found to N
HO N
be 23 pM![62] A crystal structure of PNP with 5.36 and phos- 3Å O
phate bound showed the presence of at least six new hydrogen δ+ N
3Å H
HO
bonds at the active site relative to the substrate, which could OH HO
O
account for its 106 times enhancement in binding.[63] How- O 5.37
P
ever, 5.36 has a Kd for human PNP of only 56 pM (which, of –
O OH
course, is still phenomenal), indicating that the transition state
for human PNP is a little different from that for bovine PNP,[64] Transition state structure
for human PNP
and the aim was to treat humans, not cows. KIE studies with
FIGURE 5.7 Hypothesized transition state structure for the reaction
human PNP indicated a more dissociated transition state, with catalyzed by human PNP; structure of DADMe-ImmunH; BCX4208; 5.37.
the positive charge on C1′ rather than at the sugar oxygen
and the purine leaving group about 3 Å away (Figure 5.7).
a general approach that should be applicable to the design of
The inhibitor that was designed for this transition state had
any transition state inhibitor.
the positive charge on the sugar ring moved over to the C1′
position with the purine leaving group moved farther from the
5.2.3.2.4. Multisubstrate Analogs
sugar ring by the addition of a methylene group; in addition,
as the target substrate is 2′-deoxyguanosine, a 2′-deoxysugar One of the first steps in the de novo biosynthesis of pyrimi-
ring mimic was incorporated. This led to DADMe-ImmunH dines is the condensation of carbamoyl phosphate (5.38) and
(BCX4208; 5.37); the Kd for 5.37 with human PNP was 9 l-aspartic acid, catalyzed by aspartate transcarbamylase,
pM![65] This insightful use of KIEs and computer modeling is which produces N-carbamoyl-l-aspartate (5.39, Scheme
5.8). Below the transition state structure is drawn N-phospho-
noacetyl-l-aspartate (5.40, PALA), which is a stable com-
O O
H H
pound (the isosteric exchange of a CH2 for the O prohibits
N N loss of the PO3−
NH NH 4 moiety) that resembles the transition state
HO H H for condensation of the two substrates.[66] Although PALA
N HO N
N is an effective inhibitor of aspartate transcarbamylase, it was
N ineffective in clinical trials as a result of tumor resistance. The
HO OH H
HO proposed mechanism of resistance is discussed in Chapter 7.
Design of multisubstrate analogs has not yet led to many
Forodesine
5.36 DADMe-ImmH
5.37 clinically successful drugs.[67] This is possibly because many
224 The Organic Chemistry of Drug Design and Drug Action

PO3= COO– PO3= ‡

COO–
O O COO–
..
NH2 H H
H NH
O NH2 NH2 O
COO– O COO–
NH2
5.38 COO– NH2

PO3= COO–
5.39
CH2
H
O N
H COO–

5.40
SCHEME 5.8 Hypothetical mechanism for the reaction catalyzed by aspartate transcarbamylase

O NHNH2 O N NH O N N
O NNH2
KatG KatG KatG
(peroxidase) (peroxidase) (peroxidase)
-1 e-, -H+ -1 e-, -H+ -1 e-, -H+
N N N
N Mechanism 2
Isoniazid
5.41
Mechanism 1 N2
HN=NH

N N
O

+1 e-
O H O H O
O O
N NH2 NH2
NH2
5.42
N = NAD+ N N
R R R
NH2 5.43
N O O
N
R=
N N CH2 O P O P O CH2

O- O- O
O
HO OH
HO OH
SCHEME 5.9 Proposed mechanism of action of isoniazid

of the multisubstrate inhibitors designed to date have had accepted mechanism of action of isoniazid (Scheme 5.9) pro-
high molecular weights or are charged molecules and there- poses the oxidation of 5.41 by the heme-containing peroxi-
fore do not penetrate cell membranes to reach their intended dase enzyme KatG to form acyl radical 5.42.[68] As shown in
targets. However, there are some intriguing examples of Scheme 5.9, two possible pathways have been proposed for
bisubstrate analogs that are formed in vivo from smaller peroxidase-catalyzed conversion of aryl hydrazides to acyl
precursors that do have suitable druglike properties. For radicals, one involving a one-electron oxidation followed
example, isoniazid (5.41, Scheme 5.9) is an antituberculo- by loss of diimide (mechanism 1), and the other involving
sis drug that was used clinically for many years before its three sequential one-electron oxidations followed by loss of
target was elucidated. Current evidence supports the hypoth- molecular nitrogen (mechanism 2).[69] Subsequent reaction
esis that the target is a reduced nicotinamide adenine dinu- of 5.42 with NAD+ followed by a one-electron reduction by
cleotide (NADH)-dependent enoyl reductase, called InhA, superoxide (possibly generated by KatG-catalyzed reduction
involved in fatty acid metabolism, which is essential for cell of O2)[70] produces the bisubstrate analog 5.43, which is a
wall synthesis in Mycobacterium tuberculosis. However, iso- potent inhibitor of InhA. A subsequently reported X-ray crys-
niazid itself is not active against this enzyme. The presently tal structure of 5.43 bound to a close relative of InhA showed
Chapter | 5 Enzyme Inhibition and Inactivation 225

that 5.43 occupied the binding site for NADH and the bind- Angiotensinogen, an α-globulin produced by the
ing region for the fatty acid substrates of the enzyme. liver,[80] is hydrolyzed by the proteolytic enzyme renin
to a decapeptide, angiotensin I (Scheme 5.10), which
has little, if any, biological activity. The C-terminal
5.2.4. Slow, T ight-Binding Inhibitors
histidylleucine dipeptide is cleaved from angiotensin I
5.2.4.1. Theoretical Basis by angiotensin-converting enzyme (ACE or dipeptidyl
­carboxypeptidase I) mainly in the lungs and blood ves-
As indicated above, the equilibrium between an enzyme and
sels to give the octapeptide angiotensin II. This peptide
a reversible inhibitor is typically established rapidly. With
is responsible for the increase in blood pressure by act-
slow-binding inhibitors, however, the equilibrium between
ing as a very potent vasoconstrictor[81] and by trigger-
enzyme and inhibitor is reached slowly, and inhibition is
ing release of a steroid hormone, aldosterone (5.44),
time-dependent, reminiscent of the kinetics for irreversible
which regulates the electrolyte balance of body fluids
inhibition (see Section 5.3.2.1). Tight-binding inhibitors
by promoting excretion of potassium ions and retention
are those inhibitors for which substantial inhibition occurs
of sodium ions and water. Both vasoconstriction and
when the concentrations of inhibitor and enzyme are compa-
sodium ion/water retention lead to an increase in blood
rable.[71] Slow, tight-binding inhibitors have both properties.
pressure. Angiotensin II is converted to another pep-
These inhibitors can bind noncovalently[72] or covalently[73];
tide hormone, angiotensin III, by aminopeptidase A[82];
when a covalent bond is formed, a slowly reversible adduct
angiotensin III is also made by aminopeptidase A hydro-
may be involved. Noncovalent slow, tight-binding inhibitors
lysis of angiotensin I to 5.45 (Scheme 5.10), then by
are the bridge between rapidly reversible and covalent irre-
ACE-catalyzed hydrolysis of 5.45 to angiotensin III.[83]
versible inhibitors. Depending on the tightness of binding,
This hormone also stimulates the secretion of aldoste-
these inhibitors can become functionally equivalent to cova-
rone and causes vasoconstriction.[84] Angiotensin II also
lent, irreversible inhibitors with half-lives (time for half of the
is h­ ydrolyzed by aminopeptidase N to a hexapeptide,
E·I complex to break down) of hours, days, or even months!
angiotensin IV, which may be involved in memory reten-
The reason for slow-binding inhibition is not definitively
tion and neuronal development, but it is unclear if it is
known. One possibility is that these inhibitors are such good
involved in vasopressin release; angiotensin III is also
analogs of the substrate that they induce a conformational
converted to angiotensin IV by aminopeptidase N.[85]
change in the enzyme that resembles the conformation asso-
To make matters even worse, in addition to cleaving
ciated with the transition state[74]; typically, these compounds
angiotensin I to angiotensin II, and 5.45 to angiotensin
are transition state analogs (see Section 5.2.3.1). If this is the
III, ACE catalyzes the hydrolysis of the two C-terminal
case, then inhibitor binding would be slow because it does
amino acids from the potent vasodilator nonapeptide
not have all the essential structural features of the substrate
bradykinin (Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe-Arg),
transition state geometry. The dissociation would be even
thereby destroying its vasodilation activity. Consequently,
slower because the dissociation rate is not enhanced by prod-
the action of ACE results in the generation of potent
uct formation. The conformational change may result from a
hypertensive agents (angiotensin II and a­ ngiotensin III),
change in the protonation state of the enzyme[75] or from the
which also stimulate the release of another hypertensive
displacement of an essential water molecule by the inhibi-
agent (aldosterone), and destroys a potent antihyper-
tor.[76] The tightness of binding is often measured by the
tensive agent (bradykinin). All these outcomes of ACE
residence time, the duration of the inhibition, or the dissocia-
action result in hypertension, an increase in blood pres-
tive half-life of the drug–protein complex.[77] The longer the
sure. There are compensatory pathways that decrease
residence time, the longer is the drug’s effect. Assuming ade-
the blood pressure as well (e.g., see Section 5.2.4.2.4),
quate pharmacokinetic properties, this should translate into
but when there is an imbalance in these systems where
greater in vivo effects.[78] To differentiate simple competitive
the net result is high blood pressure, an antihypertensive
inhibition from slow-binding inhibition, it is necessary to
drug is needed. ACE, therefore, is an important target for
carry out kinetic studies. Many simple competitive inhibi-
the design of antihypertensive agents; inhibition of ACE
tors may turn out to be slow- or slow, tight-binding inhibitors
would shut down its three hypertensive mechanisms.[86]
after the kinetic analysis.
OH
OH
5.2.4.2. Captopril, Enalapril, Lisinopril, and O O
Other Antihypertensive Drugs H
H 3C H
5.2.4.2.1. Humoral Mechanism for Hypertension
H
The elucidation of the molecular details of the renin- O
angiotensin system, one of the humoral mechanisms Aldosterone
for blood pressure control, began over 60 years ago.[79] 5.44
226 The Organic Chemistry of Drug Design and Drug Action

Asp•Arg•Val•Tyr•IIe•His•Pro•Phe•His•Leu•Val•IIe•His•Asn••••
Angiotensinogen (human)

Renin

Aminopeptidase A
Asp•Arg•Val•Tyr•IIe•His•Pro•Phe•His•Leu
Angiotensin I -Asp

Angiotensin-converting
- His•Leu
enzyme (ACE)

Asp•Arg•Val•Tyr•IIe•His•Pro•Phe
Angiotensin II
Aminopeptidase N Aminopeptidase A
-Asp•Arg -Asp

Aminopeptidase N ACE
Val•Tyr•IIe•His•Pro•Phe Arg•Val•Tyr•IIe•His•Pro•Phe Arg•Val•Tyr•IIe•His•Pro•Phe•His•Leu
Angiotensin IV -Arg Angiotensin III - His•Leu 5.45
SCHEME 5.10 Renin-angiotensin system

5.2.4.2.2. Lead Discovery in the penultimate position. An aromatic amino acid is pre-
In 1965, Ferreira[87] reported that a mixture of peptides ferred in the antepenultimate position.
in the venom of the South American pit viper Bothrops When the search for a potent inhibitor of ACE was ini-
jararaca potentiated the action of bradykinin by inhibition tiated at Squibb (now Bristol-Myers Squibb (BMS)) and
of some bradykininase activity. Bakhle and coworkers[88] Merck pharmaceutical companies, the enzyme had not yet
subsequently showed that these peptides also inhibited the been purified. Because the enzyme was inhibited by eth-
conversion of angiotensin I to angiotensin II. Nine active ylenediaminetetraacetic acid and other chelating agents,
peptides were isolated from this venom; the structure of a particularly bidentate ligands, it was believed to be a metal-
pentapeptide (Pyr-Lys-Trp-Ala-Pro, where Pyr is l-pyro- loenzyme. In fact, ACE purified to homogeneity from rab-
glutamate) was identified.[89] This peptide was shown to bit lung[97] was shown to contain 1 gram-atom of zinc ion
inhibit the conversion of angiotensin I to II and bradykinin per mole of protein. The zinc ion is believed to be a cofac-
degradation in vitro[90] and in vivo.[91] The structures of tor that assists in the catalytic hydrolysis of the peptide
six more of the peptides were determined by Ondetti and bond by both coordination to the carbonyl oxygen, mak-
coworkers.[92] The peptide with the greatest in vitro activ- ing the carbonyl more electrophilic, and by coordination
ity was the pentapeptide,[93] but a nonapeptide (Pyr-Trp- to a water molecule, making the water more nucleophilic.
Pro-Arg-Pro-Gln-Ile-Pro-Pro) called teprotide had the Coordination of both molecules to the zinc ion lowers the
greatest in vivo potency[94] and was effective in lowering activation energy for attack of the water on the scissile pep-
blood pressure.[95] Five other active peptides were isolated tide bond (Figure 5.8). Because the structure of the enzyme
from the venom of the Japanese pit viper Agkistrodon was not known, it was not obvious what peptidelike struc-
halys blomhoffii.[96] Because these compounds were pep- tures would be the best inhibitors. It was hypothesized that
tides, they were not effective when administered orally, the mechanism and active site of ACE may resemble those
but they laid the foundation for the design of orally active of carboxypeptidase A, another zinc-containing peptidase
ACE inhibitors. whose X-ray structure was known.[98] Three important
binding interactions between carboxypeptidase A and pep-
5.2.4.2.3. Lead Modification and Mechanism of tides are a carboxylate-binding group, a group that binds
Action the C-terminal amino acid side chain, and the zinc ion that
The fact that N-acylated tripeptides are substrates of ACE
indicated that it may be possible to prepare a small orally Zn++
-OH
active ACE inhibitor. After testing numerous peptides as R2 O R1 O R1' O R2 '
competitive inhibitors of ACE, it was concluded that pro- H
NH CH C NH CH C NH CH C N CH COO-
line was best in the C-terminal position and alanine was best FIGURE 5.8 Function of the Zn(II) cofactor in ACE catalysis
Chapter | 5 Enzyme Inhibition and Inactivation 227

coordinates to the carbonyl of the penultimate (the scissile) Although the results were encouraging, all of these com-
peptide bond (Figure 5.9).[99] (R)-2-Benzylsuccinic acid, pounds were only weak inhibitors of ACE. To increase the
which can bind at all three of these sites, is a potent inhibi- potency of the compounds, a better Zn(II)-coordinating
tor of carboxypeptidase A.[100] The extreme potency of ligand, a thiol group, was substituted for the carboxylate
inhibition of carboxypeptidase A by (R)-2-benzylsuccinic (5.47).[66] These compounds were very potent inhibitors
acid was suggested to be derived from the resemblance of ACE. Figure 5.11 shows a hypothesized depiction of
of this inhibitor to the collected products (Figure 5.10) the inter­action of 5.46 and 5.47 with ACE. Note that car-
of hydrolysis of the substrate, and, therefore, it combines boxypeptidase A is a C-terminal exopeptidase (it cleaves
all of their individual binding characteristics into a single
molecule. With this as a model, and the known effective- O O
– OOC alkyl C N HS alkyl C N
ness of a C-terminal proline for ACE inhibition, a series of
peptidomimetic carboxyalkanoylproline derivatives (5.46)
CO2H CO2H
were tested as inhibitors of ACE. Note that to increase sta- 5.46 5.47
bility and decrease peptide-like character, the N-terminal
amino group was substituted by an isosteric CH2 group to the C-terminal amino acid), whereas ACE is a C-ter-
which the Zn(II)-coordinating carboxylate was attached. minal endopeptidase or, more precisely, a dipeptidyl

S1'

S1
+
Zn++

R CH2
O O-
NH CH C NH CH C Substrate
O

-O CH2
O-
(R)-2-benzylsuccinic
C CH2 CH C
acid
O O
FIGURE 5.9 Hypothetical active site of carboxypeptidase A. Adapted with permission from Cushman, D. W., Cheung, H. S., Sabo, E. F., Ondetti, M. A.
Biochemistry 1977 16, 5484. Copyright © 1977 American Chemical Society

S1'

S1
+
Zn++

R CH2
O- O-
NH CH C NH2 CH C Products of
hydrolysis
O O

-O CH2
O-
(R)-2-benzylsuccinic
C CH2 CH C
acid
O O
FIGURE 5.10 The collected products hypothesis of enzyme inhibition using inhibition of carboxypeptidase A as an example
228 The Organic Chemistry of Drug Design and Drug Action

B
S1 S1 ' S2 ' TABLE 5.2 Effect on Ki of Structural Modification of
Zn++ H Captopril
+
R3
O
R2
O - Analog Relative Ki
O
NH CH NH CH C N CH Substrate H 3C H 1.0
O HS N
- (Captopril)
O R2 O
O
-
O CO2H
CH2 CH C N CH 5.46
O O
H 3C H 12,500
R2 O
O
- HS N
-S CH2 CH C N CH 5.47
O O
FIGURE 5.11 Hypothetical binding of carboxyalkanoylproline and mer-
10
captoalkanoylproline derivatives to ACE. Adapted with permission from HS N
Cushman, D. W., Cheung, H. S., Sabo, E. F., Ondetti, M. A. Biochemistry
1977 16, 5484. Copyright © 1977 American Chemical Society O CO2H

12,000
carboxypeptidase (it cleaves a C-terminal dipeptide). There- N
HS
fore, the active site of ACE (Figure 5.11) has two additional
O CO2H
binding sites than carboxypeptidase A (Figure 5.9) has
between the Zn(II) and the group that interacts with the H 120
C-terminal carboxylate group. The compound that had the HS N CO2H
best binding properties was 5.48 (captopril), a competitive
inhibitor of ACE with a Ki of 1.7 × 10−9 M under standard O
assay conditions. Furthermore, captopril is highly specific H CH3 120
for ACE; the Ki values for captopril with carboxypeptidase HS N
A and carboxypeptidase B, two other Zn(II)-containing pep-
O
tidases, are 6.2 × 10−4 M and 2.5 × 10−4 M, respectively.[101] CO2H
Presumably, the reason for the specificity is that there are
H 3C H 1100
many functional groups in 5.48 that can regio- and stereo- HOOC N
specifically interact with groups at the active site of ACE,
but they cannot interact to the same degree or perhaps at all O CO2H
with groups in other peptidases (compare Figures 5.9 and
5.11). The carboxylate group of the inhibitor can be stabi-
lized by an electrostatic interaction with a cationic group on Captopril was the first ACE inhibitor in the drug market,
the enzyme, the amide carbonyl can be hydrogen bonded and it was shown to be effective for the treatment of both hyper-
to a hydrogen donor group, the sulfhydryl can be ligated to tension and congestive heart failure. Given alone, captopril can
the zinc ion, and the proline and (S)-methyl group can be normalize the blood pressure of about 50% of the hypertensive
involved in stereospecific hydrophobic and van der Waals population. When given in combination with a diuretic, such
interactions. as hydrochlorothiazide (5.49; Aldoril) (remember, angiotensin
II releases aldosterone, which causes water retention), this can
H3C H be extended to 90% of the hypertensive population. In more
HS N
severe cases, an antagonist for the β-adrenergic receptor (a β-
O CO2H blocker), which triggers vasodilation, may be used in a triple
Captopril therapy with captopril and a diuretic.
5.48

O O
All of these interactions must be important because dele-
NH2SO2 S
tion or alteration of any of these groups raises the Ki consider- NH
ably (Table 5.2). A myriad of analogs of this basic structure,
Cl N
including compounds with Zn(II)-coordinating ligands other H
than carboxylate and thiol groups, have been synthesized and Hydrochlorothiazide
tested as ACE inhibitors (see the general references). 5.49
Chapter | 5 Enzyme Inhibition and Inactivation 229

Two side effects were observed in some patients during The interactions of captopril and enalaprilat with puri-
the early usage of captopril, namely, rashes and loss of taste. fied rabbit lung ACE was studied in detail, and the kinet-
Both of these side effects were reversible on drug withdrawal ics indicated that these compound are slow, tight-binding
or reduction of the dose.[102] Considering the potential lethal- inhibitors.[104] For example, the rate constant for the for-
ity of hypertension, a minor rash or loss of taste would seem mation of the E·I complex (kon) with enalaprilat at pH 7.5
insignificant to the benefits of the therapy. However, hyper- was determined to be 2 × 106 m−1s−1, which is at least two
tension is a disease without a symptom (that is, until it is too orders of magnitude smaller than expected for a diffusion-
late); generally, a patient discovers he has this disease when controlled reaction. Steady-state kinetics gave the value of
his physician takes his blood pressure. Because of this lack the Ki as 1.8 × 10−10 M. Since Ki = koff/kon, the koff should be
of immediate discomfort, there may be difficulties getting 3.6 × 10−4 s−1, which is in satisfactory agreement with the
the patient to comply with the therapy, especially if unpleas- measured koff value of 1.6 × 10−4 s−1. The small koff value for
ant side effects arise when the drug is taken. this noncovalent E·I complex emphasizes the strong affinity
Consequently, a Merck group investigated the cause for of enalaprilat for ACE.
the side effects. Because similar side effects arise when peni- Enalaprilat, however, is poorly absorbed orally and,
cillamine is administered, it was hypothesized that the thiol therefore, must be given by intravenous (IV) injection.
group may be responsible.[103] Furthermore, deletion of this This problem was remedied simply by conversion of the
functional group should give inhibitors greater metabolic carboxyl group to an ethyl ester (5.50, R = (S)-PhCH2CH2,
stability because thiols undergo facile in vivo oxidation to R′ = CH3CH2), giving enalapril, which has excellent oral
disulfides. The approach taken was to attempt to increase the activity. Because the in vitro IC50 for enalapril is 103 times
previously found weak potency of the carboxyalkanoylpro- higher than that for enalaprilat, the ethyl ester group must
line analogs by adding groups that can interact with additional be hydrolyzed by esterases in the body to liberate the active
sites on the enzyme, that is, to increase the pharmacophore. form of the drug, namely, enalaprilat. Enalapril, then, is
If the carboxyalkanoylproline derivatives are collected an example of a prodrug, a compound that requires meta-
product inhibitors (see Figure 5.10), then there are two fea- bolic activation for activity (see Chapter 9). The effect of
tures that can be built into these analogs to make them look esterification is to lower the pKa of the NH group (pKa 5.5
more productlike. One is to make them structurally more in enalapril, but 7.6 in enalaprilat)[105] and to remove the
similar to product dipeptides by substituting an NH for a charge of the carboxylate, both of which would increase
CH2 such as 5.50 (R = R′ = H). Disappointingly, however, membrane permeability. A prodrug related to enalapril is
this compound had less than twice the potency of the iso- trandolapril (5.51, Mavik), which is highly lipophilic with
stere with a CH2 in place of NH. The reason for this could prolonged ACE inhibitory activity, even at a dose of 2 mg
be compensatory factors. An NH (or its protonated form) is once a day.
much more hydrophilic than a CH2 group and might make
the molecule too hydrophilic. If that occurs, then an addi-
tional hydrophobic group could be added to counterbalance
EtO2C H H
this hydrophilic effect. When a methyl group was appended H3C H
(5.50, R = CH3, R′ = H), the potency increased about 55-fold. N
Because the other feature that could make these compounds H N
H O CO2H
structurally more similar to the collected products would be
Trandolapril
to append a group that might interact with the substrate S1 5.51
subsite, the R group of 5.50 was modified further, and 5.50
(R = (S)-PhCH2CH2, R′ = H), called enalaprilat (see Section
5.2.3.2.1), emerged as the viable drug candidate (note, in the Another compound, prepared by the Merck chem-
early studies with peptides, the antepenultimate site favored ists as an alternative to enalaprilat, was the lysylproline
an aromatic group). The IC50 for enalaprilat under standard analog called lisinopril (5.52; Prinivil). Note the stereo-
assay conditions is 19 times lower than that for captopril, chemistry shown (S,S,S) is that found in the most potent
suggesting that there are increased interactions of enalaprilat isomer of lisinopril, and is also the stereochemistry of
with ACE. These may be hydrophobic interactions of the enalaprilat. Lisinopril is more slowly and less completely
­phenylethyl group with the S1 subsite (see Figure 5.5). absorbed than enalapril, but its longer oral duration of
action, and the fact that it does not require metabolic
activation, makes it an attractive alternative. The crystal
R'OOC H3C H
N structure of lisinopril bound to ACE showed that it bears
R N little similarity to the structure of carboxypeptidase A,
H O CO2H which was the protein used as the basis for the design of
5.50 captopril.[106]
230 The Organic Chemistry of Drug Design and Drug Action

to the launch of captopril it was found that the loss of taste

and rash side effects initially observed were related to high
NH2 dosing, not to the sulfur atom hypothesized at Merck; at
CH2
a lower dose, captopril was just as effective, but without
CH2 CO2H (CH2)4
H those side effects. Homologation gave a compound that was
N potent in vitro for both enzymes, but not very potent in vivo
H N
H CO2H (5.54). The research group turned to conformationally
O
Lisinopril restricted analogs to increase potency. Compound 5.55 was
5.52 potent both in vitro and in vivo, and 5.56 had IC50 values of
5 and 17 nM for ACE and NEP, respectively, with oral activ-
ity greater than that of captopril in rats. A 7,6-fused bicy-
clic thiazepinone analog (5.57, omapatrilat, Vanlev) with
5.2.4.2.4. Dual-Acting Drugs: Dual-Acting Enzyme IC50 values of 5 and 8 nM for ACE and NEP, respectively,
Inhibitors was advanced to clinical trials.[109] The key advantages of
When there are two related enzymes whose inhibition would the dual-acting enzyme inhibitor omapatrilat over the ACE
give an enhanced effect compared to the sum of the effects inhibitors is its ability to lower both diastolic and systolic
of inhibiting either enzyme alone (see Drug Synergism, blood pressure better and its effectiveness in controlling
­Chapter 7), it may be beneficial to design a single inhibitor of subpopulations of patients, such as African-Americans and
both enzymes, a compound known as a dual-acting enzyme diabetics, where traditional drugs have been less effective.
inhibitor. There are several advantages to the design of one Unfortunately, omapatrilat was not approved by the Food
compound that inhibits two different enzymes rather than and Drug Administration (FDA) because of a larger inci-
two compounds, one for each enzyme: (1) with two drugs, dence of angioedema (swelling of the eyelids, lips, face,
two separate syntheses, two formulations, and two different tongue, and throat) than ACE inhibitors, which is caused by
metabolism studies (see Chapter 8) have to be developed; the increased bradykinin levels that occurs more with these
(2) two drugs will have different pharmacokinetic rates and inhibitors than ACE ­inhibitors.[110]
metabolic profiles, making it difficult for both to be optimal in
the same time frame; (3) the likelihood that both drugs would
progress to the clinic at the same rate is small; (4) the cost for O CH3
three sets of safety studies and three separate clinical trials HS N
(one for each drug plus one with the combination) would be N
H
enormous; and (5) the odds for a single drug just starting clini- Ph
O COOH
cal trials to be approved for the drug market is 1 in 10 and that 5.53
for two drugs entering the market would be 1 in 100! With a
single, dual-acting drug, none of these problems exists. Extra CH2
An example of dual-acting enzyme inhibition is related O CH3 O
to our discussion above of the “pril” antihypertensive HS
N HS N
N N
agents. Neutral endopeptidase (NEP) is another Zn++-­ H H O
O COOH
containing endopeptidase. It degrades and deactivates atrial Ph 5.54 Ph
5.55
COOH

natriuretic peptide (ANP), a 28-amino acid vasoactive pep-

S
tide hormone produced by the heart that causes vasodilation O
H
and inhibition of the ­formation of aldosterone, the steroid O HS N
N
hormone that regulates the electrolyte balance and leads to HS N H H O
N CH3
retention of sodium ions and water. Therefore, ANP acts H O Ph COOH
Omapatrilat
to lower the blood pressure; the hormone actions of ANP Ph 5.56
COOH
5.57
and angiotensin II, therefore, are functionally opposite. If
the formation of angiotensin II is blocked by the inhibition
of ACE, and the concentration of ANP is increased by the
inhibition of NEP, there should be a synergistic (or at least This dual-acting inhibitor approach for the design
additive) antihypertensive effect.[107] of new antihypertensive agents has been extended fur-
The lead compound devised at BMS (5.53), which had ther; there is a third zinc metalloprotease, endothelin-
a structure composed of part of captopril and enalapril, converting enzyme (ECE), which displays high amino
was found to have IC50 values for ACE and NEP of 30 and acid sequence identity with NEP, especially in the active
400 nM, respectively, which is excellent for a lead com- site,[111] and hydrolyzes a 38-amino acid inactive peptide
pound.[108] The thiol group was retained because subsequent into endothelin-1, a 21-amino acid peptide that is the most
Chapter | 5 Enzyme Inhibition and Inactivation 231

potent vasoconstrictor (constricts blood vessels, thereby 5.63 and 5.64 led to the design of 5.58, which ­significantly
raising blood pressure) known.[112] Several classes of reduces prostatic growth in rabbits and rats.
compounds that act as triple-acting enzyme inhibitors for
ACE, NEP, and ECE have been identified.[113] O
MeO N N N
A dual-acting drug does not have to be limited to com-
O
pounds that inhibit two different enzymes. It could also be a N
MeO .HCl
compound that inhibits one enzyme and acts as an antago-
NH2
nist for a receptor, a compound that is an antagonist for two
Terazosin HCl
different receptors, an agonist for two different receptors, or
5.59
any combination thereof.
An example of a dual-acting drug that acts as an inhibi-
tor of an enzyme and as an antagonist for a receptor is Z-350 HN
(5.58),[114] an inhibitor of steroid 5α-reductase and an antag- O
onist for the α1-adrenoceptor.[115] Benign prostatic hyper-
plasia (BPH) is a progressive enlargement of the prostate
gland, leading to bladder outlet obstruction. This obstruc- H
tion consists of a static component related to prostatic tissue O N
H H
mass and a dynamic component related to excessive contrac-
Finasteride
tion of the prostate and urethra.[116] Antagonists of the α1- 5.60
adrenoceptor, such as terazosin HCl (5.59, Hytrin), are used
to relax the smooth muscle of the prostate and urethra.[117]
O
Because dihydrotestosterone is known to be a dominant fac-
tor in prostatic growth, inhibitors of steroid 5α-reductase, N
H
the enzyme that converts testosterone into dihydrotestos- OCH3
O
terone, such as finasteride (5.60, Proscar), are also used
to treat BPH. A dual-acting agent was designed[118] on the
basis of the α1-adrenoceptor antagonist 5.61 (the structure is
drawn in a conformation to resemble 5.58) and the steroid N
5α-reductase inhibitor 5.62 by combining features of both N
molecules into one structure (5.63), a common approach
for the design of dual-acting agents. This compound was CH3O
a potent antagonist for the α1-adrenoceptor but needed 5.61
increased potency against steroid 5α-reductase. It was well
established that the lipophilic part and the butanoic acid O
moiety are essential for steroid 5α-reductase activity, but the N
benzanilide moiety could be replaced by an acyl indole, so H
O
the next structures included 5.64, which is as potent as 5.63 O CH3
as an antagonist for α1-adrenoceptor and more potent in ste- CH3
CH3
roid 5α-reductase inhibition.[119] Merging the structures of COOH

5.62

O
O

N
H
O N O
O

H3C N COOH COOH

H 3C N
N
N

CH3O
CH3O
Z-350
5.58 5.63
232 The Organic Chemistry of Drug Design and Drug Action

O one-half of all deaths in the United States can be attrib-

O uted to atherosclerosis,[123] which results from the
buildup of fatty deposits called plaque on the inner walls
O N of arteries. The major component of atherosclerotic
H3C plaque is cholesterol. In humans, more than one-half of
the total body cholesterol is derived from its de novo
N COOH biosynthesis in the liver.[124] Cholesterol biosynthesis
N
requires more than 20 enzymatic steps starting from ace-
tyl CoA. The rate-determining step is the conversion of
CH3O
3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA;
5.69) to mevalonic acid (5.70), catalyzed by HMG-CoA
5.64
reductase (Scheme 5.11). For the structure of coenzyme
A, see Chapter 4, 4.13b. Because hypercholesterolemia
An example of a dual-acting receptor agonist is a is a primary risk factor for coronary heart disease,[125]
compound that acts as an agonist for both the D2-recep- and the overall rate of cholesterol biosynthesis is a func-
tor and the β2-adrenoceptor for the treatment of airway tion of this enzyme, efforts were initiated to inhibit
diseases, such as chronic obstructive pulmonary disease HMG-CoA reductase as a means of lowering plasma
(COPD) and asthma. A D2-receptor agonist reduces ­cholesterol ­levels.[126]
reflex bronchoconstriction, dyspnea, cough, and mucus
production, but less likely diminishes bronchoconstric- 5.2.4.3.2. Lead Discovery
tor activity of locally released mediators of broncho-
Endo and coworkers[127] at the Sankyo Company in Tokyo
constriction.[120] β2-Adrenoceptor agonists are the most
tested 8000 strains of microorganisms for metabolites
commonly used antibronchoconstrictor agents,[121] but
that inhibited sterol biosynthesis in vitro and discovered
they have little effect on dyspnea, cough, and mucus pro-
three active compounds in the culture broths of the fungus
duction. A dual-acting agent for these receptors should
Penicillium citrinum. The most potent compound, called
combine all the desired features of this class of drugs.
mevastatin, was also isolated from broths of Penicillium
The structures of a weak β2-adrenoceptor agonist (5.65)
brevicompactum by Brown and coworkers[128] at Beecham
and a potent D2-receptor agonist (5.66) were hybrid-
Pharmaceuticals in England, who named it compactin (5.72,
ized to give 5.67. Modification of the side chain gave
R = H). A second, more potent, compound was ­ isolated
5.68,[122] which went into clinical trials for the treatment
by Endo[129] from the fungus Monascus ruber, which he
of the symptoms of COPD.
named monacolin K; the same compound was isolated by
5.2.4.3. Lovastatin (Mevinolin) and Simvastatin, a group at Merck from Aspergillus terreus,[130] which they
named mevinolin (5.72, R = CH3). Mevinolin is now known
Antihypercholesterolemic Drugs
as lovastatin (Mevacor). Several related metabolites were
An example of noncovalent slow, tight-binding inhibitors is also isolated from cultures of these fungi,[131] including
a family of drugs used to treat high cholesterol levels. dihydrocompactin from P. citrinum (5.73, R = H, R′ = (S)-
CH3CH2CH(CH3)CO2−), dihydromevinolin from A. ter-
5.2.4.3.1. Cholesterol and Its Effects reus (5.73, R = CH3, R′ = (S)-CH3CH2CH(CH3)CO2−), and
Coronary heart disease is the leading cause of death in dihydromonacolin L from a mutant strain of M. ruber (5.73,
the United States and other Western countries; about R = CH3, R′ = H).

O
H S
HO N
(CH2)6NH(CH2)2Ph HN N(CH2CH2CH3)2
HO
5.65 HO 5.66

O O
S S
H H
HN N HN N
(CH2)6NH(CH2)2Ph (CH2)2SO2(CH2)3O(CH2)2Ph

HO 5.67 HO 5.68
Chapter | 5 Enzyme Inhibition and Inactivation 233

2 NADPH 2 NADP+
H3C
H3C
HO COOH
HO COOH
O
HMG-CoA reductase OH
SCoA
5.69 5.70

NADP+

NADPH H 3C H3C
HO COOH -CoASH HO COOH
O H O NADPH
H
SCoA :B H
NADP+

5.71
SCHEME 5.11 HMG-CoA reductase, the rate-determining enzyme in de novo cholesterol biosynthesis

HO O in 5.72, but rather, the hydrolysis product, that is, the open
chain 3,5-dihydroxyvaleric acid form (5.74). This form mim-
O
O ics the structure of proposed intermediate 5.71 (Scheme 5.11)
in the reduction of HMG-CoA by HMG-CoA reductase, that
O
H3 C H H is, if the structure of CoA (see Chapter 4, 4.13b) is taken
CH3
into account. Enzyme studies with compactin and analogs
indicate that there are two important binding domains at the
R
active site, the HMG binding domain, to which the upper part
Mevastatin or compactin (R = H)
of 5.74 binds and a hydrophobic pocket located adjacent to
Monacolin K, mevinolin, or lovastatin (R = CH3)
the active site, to which CoA and the decalin (lower) part of
5.72
5.74 bind.[133] The high affinity of compactin and its analogs
to HMG-CoA reductase derives from the simultaneous inter-
HO O
actions of the two parts of these inhibitors, having an ethylene
O linker, with the two binding domains on the enzyme. This is
the same phenomenon as was responsible for the increased
R' binding of two inhibitors attached by a linker in SAR by NMR
H
CH3 (see Chapter 2, Section 2.1.2.3.6). As a result of the interac-
tions in two adjacent binding pockets, dissociation of the E·I
R complex is very slow. A kinetic analysis of the on and off rate
5.73 constants (kon and koff, respectively; see Scheme 5.1) with yeast
HMG-CoA reductase were 1.9 × 105 m−1s−1 and 0.11 s−1
for HMG-CoA and 2.7 × 107 m−1s−1 and 6.5 × 10−3 s−1 for com-
5.2.4.3.3. Mechanism of Action pactin, respectively. Therefore, compactin binds faster and dis-
Compactin[132] and lovastatin are potent competitive revers- sociates more slowly than does HMG-CoA, which accounts
ible inhibitors of HMG-CoA reductase. The Ki for com- for the difference in their Km and Ki (koff/kon) values. It is also
pactin is 1.4 × 10−9 M and for lovastatin is 6.4 × 10−10 M (rat possible to classify these inhibitors as transition state analogs
liver enzyme); for comparison, the Km for HMG-CoA is (see Section 5.2.3).
about 10−5 M. Therefore, the affinity of HMG-CoA reduc- H
tase for compactin and lovastatin is 7140 and 16,700 times, HO
3 COOH
respectively, greater than that for its substrate. Compactin
5 OH
and lovastatin do not affect any other enzyme in cholesterol O
H
synthesis except HMG-CoA reductase.
O
It may not be immediately obvious why lovastatin is a H 3C H H
CH3
competitive reversible inhibitor of HMG-CoA reductase,
given that it does not closely resemble the structure of the sub-
R
strate or product, which might be expected of a competitive
inhibitor. The reason is that the active form is not that shown 5.74
234 The Organic Chemistry of Drug Design and Drug Action

5.2.4.3.4. Lead Modification H

HO
Numerous structural modifications were made to com- COOH
pactin and lovastatin to determine the importance of the OH
lactone moiety and its stereochemistry, the ability of the H
lactone moiety to be opened to the dihydroxy acid, the F
optimal length and structure of the moiety bridging the lac-
tone and the lipophilic groups, and the size and shape of
O N
the lipophilic group. It was found that potency was greatly
reduced[134] unless a carboxylate anion could be formed
and the hydroxyl groups were left unsubstituted in an
erythro relationship. Insertion of a bridging unit other than
Cerivastatin
5.75
­ethylene or (E)-ethenyl between the 5-carbinol moiety and
the lipophilic moiety also diminishes the potency. Modifi-
Although side effects generally have a negative connota-
cations of the lower (lipophilic) part of compactin, in most
tion, another “side effect” of both the type 1 and type 2 statins
cases, led to compounds with considerably lower potencies,
is an enhancement in new bone formation in rodents.[141] This
except for certain substituted biphenyl analogs.[135] If the
new bone growth is associated with an increased expression of
substituted biphenyl rings were constrained as fluorenyli-
the bone morphogenetic protein-2 gene in bone cells, which
dene moieties, the corresponding potencies decreased.[136]
may have implications in the treatment of osteoporosis.
When the hydroxyl group in the lactone ring was replaced
The X-ray crystal structures of several of the statins
by an amino or thiol group, diminished potencies were
bound to HMG-CoA reductase confirm the biochemi-
observed.[137]
cal hypothesis and show that they occupy a portion of the
Modification of the 2(S)-methylbutyryl ester side chain
HMG-CoA binding site, thereby blocking access of the sub-
(CH3CH2CH (CH3) CO2−) of lovastatin indicated that intro-
strate to the active site.[142]
duction of an additional aliphatic group on the carbon α to
the carbonyl group increased the potency of lovastatin.[138]
To block ester hydrolysis and produce a compound with a
5.2.4.4. Saxagliptin, a Dipeptidyl Peptidase-4
much longer plasma half-life, a second methyl group was Inhibitor and Antidiabetes Drug
added, by synthesis,[139] to the side chain of the lower half Diabetes is a family of diseases characterized by high blood
of lovastatin (5.72 (R = CH3), but with ester side chain glucose levels resulting from defects in the body’s ability to
CH3 CH2 C(CH3 )2 CO2−) called simvastatin (Zocor), which produce and/or use insulin, a peptide hormone produced in the
has a potency about 2.5 times greater than that of lovas- pancreas that reduces blood sugar. The most common form,
tatin. The lactone epimer of lovastatin (the epimer of the type 2 diabetes, characterized by elevated plasma glucose
carbon adjacent to the lactone oxygen in 5.72) has less than and glycosylated hemoglobin, is the leading cause of blind-
10−4 times the potency of lovastatin.[140] Modifications in ness, kidney failure, and limb amputation worldwide.[143] As
the 3,5-dihydroxyvaleric acid moiety of analogs of 5.74 of 2011, over 8% of the US population had diabetes (with an
resulted in lower potencies, except when the 5-hydroxyl increasing number of children having the disease). This is a
group was replaced by a 5-keto group, in which case poten- growing chronic problem, which has been a major focus of
cies comparable to the parent compounds were observed. drug discovery research since the mid-1990s.
Presumably, the 5-keto group becomes reduced, but it is not Glucagon-like peptide-1 (GLP-1) and glucose-­dependent
known if this occurs by HMG-CoA reductase or by cyto- insulinotropic polypeptide are two incretins, gastrointestinal
chrome (see Chapter 8; Section 8.4.2.2.1) and, if it does, hormones that increase insulin by increasing insulin secre-
whether it occurs prior to inhibition or whether it is the tion from the pancreas (an insulin secretagogue), decreasing
cause for inhibition. glucagon secretion from the liver (a glucagon suppressor),
The generic names of this family of drugs end in the suf- increasing insulin gene expression, decreasing food intake,
fix “-statin”, so they are known collectively as the statins. and promoting insulin sensitivity. These polypeptides are the
The early analogs, such as compactin, lovastatin, and sim- natural safeguards against diabetes; however, the half-life of
vastatin, are classified as type 1 statins; the newer analogs, GLP-1 is only about 2 min because of dipeptidyl peptidase-4
known as type 2 statins, retain the HMG mimic of the mol- (DPP-4), the serine protease that degrades it. Therefore,
ecule, but the CoA mimic has been modified. Cerivastatin considerable effort by numerous laboratories to identify
(5.75, Baycol), a very potent anticholesterol drug (0.1– ­inhibitors of DPP-4 was initiated.[144]
0.15 mg taken once a day!), was launched by Bayer in 1999 BMS took as their starting point cyanopyrrolidine amide
but had to be recalled from the drug market in 2001 because 5.76, a compound reported by Ferring Research as a potent
of its association with deaths of some patients taking it who inhibitor of DPP-4,[145] which had moderate pharmacoki-
developed rhabdomyolysis, a side effect involving muscular netic duration and chemical instability as a result of an intra-
weakness. molecular cyclization that occurred (Scheme 5.12). To retard
Chapter | 5 Enzyme Inhibition and Inactivation 235

H
R HO
Ph
N N N N
H2 N H2N N H2 N H2 N H2N
CN CN CN CN
O O O O
O CN
5.76 5.77 5.78 5.79 Saxagliptin
5.80

this cyclization, conformationally restricted analogs were competitive reversible inhibitor, one in which pharmacoki-
required. A potent DPP-4 inhibitor from another company, netics was of primary concern.
Phenomix, had just such an active scaffold (5.77), which
indicated that the cyclization could be prevented by confor- 5.2.5. Case History of Rational Drug Design
mational effects. However, to skirt the intellectual property
issue as well as to combat the cyclization problem, BMS
of an Enzyme Inhibitor: Ritonavir
incorporated a fused cyclopropane ring onto each position The genome of the human immunodeficiency virus-1 (HIV-
of the pyrrolidine ring in both stereochemistries to impact 1) encodes an aspartate protease (HIV-1 protease), which
the conformation of the pyrrolidine ring; the one that had the proteolytically processes the gag and gag-pol gene products
best pharmacodynamic and pharmacokinetic properties and into mature, functional proteins.[147] If these processing steps
solution stability was 5.78 (R = 1-methylpropyl). It was then are blocked, the progeny virions are immature and noninfec-
found that increasing the steric bulk (5.78, R = tert-butyl) tious. Consequently, HIV-1 protease should be an important
further enhanced solution stability and in vitro potency target for design of inhibitors to act as potential anti-acquired
(Ki = 7 nM) with excellent pharmacokinetic properties (includ- immune deficiency syndrome (AIDS) drugs. Several com-
ing >50% oral bioavailability in rats). Given the enhanced panies discovered very potent and effective inhibitors by
potency with increasing bulkiness of the R group, larger various lead discovery/modification approaches. What fol-
groups were added, and adamantyl was found to be most lows is the approach taken at Abbott Laboratories leading
potent (5.79, Ki = 0.9 nM). However, its in vivo stability was to ritonavir, a potent HIV-1 protease inhibitor with high oral
low with rapid metabolism and low bioavailability. The prob- bioavailability that was shown to be an effective drug for the
lem was found to be microsomal oxidation (see Chapter 8, treatment of AIDS. Similar approaches were taken by other
­Section 8.4.2.1.4) of the adamantyl ring to give 5.80 (saxa- pharmaceutical companies as well.
gliptin, Onglyza), which was found to be almost as potent as The general approach is first to find molecules that have
5.79 but with superior duration in vivo, slow metabolism with good potency (preferably in the nanomolar range), and then
microsomes, and with >60% bioavailability in rats. Kinetic use those molecules as starting points for solving phar-
analysis of 5.80 showed that it was a slow, tight-binding inhibi- macokinetic problems. There has to be constant monitor-
tor of DPP-4, which was dependent on two structural moieties: ing of potency while the pharmacokinetic issues are being
(1) the bulky adamantyl substituent to displace a water mol- addressed.
ecule at the active site, driving the slow entropic on rate and
preventing enzymatic hydrolysis and (2) the nitrile, which 5.2.5.1. Lead Discovery
formed a reversible, covalent bond with the active site serine,
driving the slow enthalpic off rate; both of these properties HIV protease is an unusual enzyme because the homodimer
resulted in the low Ki value. An X-ray crystal structure of 5.80 (two identical polypeptides come together to form the active
bound to DPP-4 revealed the covalent bond between Ser-630 enzyme) has C2 symmetry (a 180° rotation about an axis
and the nitrile carbon atom; however, upon dialysis, complete through the center gives the same structure). This symme-
enzyme activity was recovered, indicating the reversibility of try element was used as a key starting point in the design of
the bond.[146] novel inhibitor structures.[148] The initial plan was to design
Before we discuss covalent irreversible inhibitors, let’s C2 symmetric compounds that should show selectivity for
look in detail at another case history for the design of a HIV protease over other mammalian aspartate proteases
because of the lack of symmetry with other aspartate prote-
ases. The other design element was to make a transition state
H N N analog (or, more accurately, an intermediate analog). The
O O
N proposed tetrahedral intermediate structure for the hydro-
H2N H H
C N NH N O lysis of a good asymmetric substrate, such as -Phe-Pro-,
O H H
was bisected either through the scissile carbon (5.81,
N
5.76 Figure 5.12) or adjacent to it (5.82). However, to make a C2
SCHEME 5.12 Intramolecular cyclization of 5.76, which causes its short symmetric compound, the two amino acid residues must be
pharmacokinetic duration the same. Because the P region (the N-terminal side of the
236 The Organic Chemistry of Drug Design and Drug Action

scissile bond) had been shown to be more important than the major causes for poor peptidomimetic pharmacokinetic
P’ region (the C-terminal side of the scissile bond), the P′ profiles are high molecular weight, low aqueous solubility,
region was deleted and the proline was substituted by phe- susceptibility to proteolytic degradation, hepatic metabo-
nylalanine. A C2 symmetry operation was performed on lism, and biliary extraction.
the remainder of the substrate, generating two possible lead The aqueous solubilities of 5.83e and 5.84e were very
compounds, 5.83a and 5.84a (Figure 5.12), respectively. The poor. The crystal structure of HIV-1 protease with 5.83e
amino groups were acylated with a variety of N-­protecting bound[150] indicated that the Cbz groups were auxophoric,
groups to increase lipophilicity, including Ac (b), Boc (c), that is, they were not interacting with the protein as part
and Cbz (d). Compound 5.83b was very weakly inhibitory; of the pharmacophore and were also not interfering with
the stereoisomers of 5.84c, however, were good inhibitors. protein binding. As discussed in Chapter 2 (Section 2.2.1),
these are ideal groups to modify, because it is likely that
5.2.5.2. Lead Modification modification will not affect potency. The crystal structure
To increase the pharmacophore of these inhibitors, P2/ showed that there was room for structural modification, so
P2′ residues were added. The Cbz-Val analogs, 5.83e and the terminal Cbz phenyl groups were modified with polar,
5.84e (Figure 5.12), were low nanomolar and subnanomolar heterocyclic bioisosteres, such as pyridinyl (at the 2-, 3-,
inhibitors, respectively; those analogs with the 5.84 struc- and 4-positions of the pyridinyl ring) and thiazolyl (at the
ture were generally 10 times more potent than 5.83 analogs. 2- and 4-positions) groups.[151] The P1 phenyl (from Phe)
Varying the stereochemistry of the hydroxyl groups in 5.84e and P2 isopropyl (from Val) groups were embedded in
had little or no effect on inhibition. Compounds 5.84e (dif- lipophilic pockets, so modification of these groups would
ferent stereochemistries) were potent in vitro inhibitors of not be fruitful. Conversion of one of the Cbz phenyls to
HIV-1 protease in H9 cells (IC50 values 20–150 nM). The a pyridinyl group increased the aqueous solubility by
therapeutic indices were in the range 500–5000, which is 20-fold without a change in the inhibitory potency, but the
much better than is generally required for agents designed water solubility was still less than 1 µg/mL. More basic
to treat life-threatening diseases. nonaromatic heterocycles had much greater water solubil-
At this point in the lead modification process, relatively ities, but the potencies dropped. Substitution of both Cbz
potent inhibitors have been identified, so efforts can concen- phenyl groups by heteroaromatic groups had little effect
trate on pharmacokinetic problems. Because of generally on the inhibitory potencies, but gave dramatic increases
poor pharmacokinetic behavior of peptides, peptidomimet- in water solubilities. Some of the compounds showed oral
ics are generally sought (see Chapter 2, Section 2.2.4.5). bioavailabilities in the range of 10–20%, but these bio-
However, peptidomimetics also can suffer from low oral/ availabilities did not exceed concentrations required for
intestinal absorption and rapid hepatic elimination.[149] The effective anti-HIV activity in vitro. Compound 5.85 had

O H
HO OH N OH
H
N NHR a, R = H
RHN
N b, R = Ac
c, R = Boc
Ph d, R = Cbz
Ph Ph e, R = Cbz-Val
5.83
P1 C2 P 1'
5.81

Ph
O H HO OH OH
HO OH N
H
N RHN NHR RHN
N NHR

OH
Ph
Ph Ph Ph

P1 C2 P 1'
5.82 5.84
FIGURE 5.12 Lead design for C2 symmetric inhibitors of HIV protease based on the structure of the tetrahedral intermediate during hydrolysis of HIV
protease
Chapter | 5 Enzyme Inhibition and Inactivation 237

Ph
O OH Me
O
H H
N N N
N N N
N N H
H
O O
Me OH
Ph
5.85

the best combination of inhibitory potency and aqueous methyl substitution on the P3 pyridinyl group did not dimin-
solubility; unfortunately, it showed no oral bioavailability ish potency and often enhanced it. With regard to the phar-
in rats. In general, the compounds with the 5.83 frame- macokinetics, there was little correlation between aqueous
work, although less potent than the compounds with the solubility and oral bioavailability! Therefore, although aque-
5.84 framework, had consistently superior oral bioavail- ous solubility is a necessary (or at least helpful) parameter
abilities in animals. To take advantage of both of these for oral bioavailability, it certainly is not always a sufficient
properties, the next series of analogs investigated had gen- one. This makes sense. Aqueous solubility serves an impor-
eral structure 5.86, which combined the extended frame- tant role in helping to increase the exposure of the drug to the
work of 5.84 with the mono alcohol structure of 5.83.[152] gut membranes through which it must pass; however, solu-
bility can have less of an influence than other factors on how
Ph well the drug will pass through those membranes or how
OH well the drug will survive metabolism by the liver, which
AHN is one of the first hurdles that a drug faces on entering the
NHA
bloodstream by the oral route (see Chapter 8, Section 8.1).
Compounds with an N-methylurea linkage between the pyr-
Ph 5.86 idinyl group and the P2 aminoacyl residue generally showed
greater oral bioavailability and solubility than the ones with
Although these were still poorly bioavailable in rats, the 5.86 a carbamate linkage. Also, compounds in the 5.86 series
series was more potent than the 5.84 series (IC50 values were exhibited greater oral bioavailability than those in the 5.84
generally subnanomolar). With all of these modifications series. This is presumed to be because of improved absorp-
and still no oral bioavailability, it is time to consider Lipin- tion in the gut, but a decreased liability to liver metabolism
ski’s Rule of 5 (see Chapter 2, Section 2.1.2.3.2). One of or biliary excretion are other possible contributors.
his rules for oral bioavailability is that the molecular weight A good measure of the overall potential of HIV-1 protease
should not exceed about 500; the molecular weight of 5.85 inhibitors is the ratio of the maximum plasma levels achieved
is 794. To increase oral bioavailability, then, it was reason- in vivo (Cmax) to the concentration required for anti-HIV
able to try to decrease the size of the molecules. But to do activity in vitro (ED50). The compound that emerged from
that, and retain the C2 symmetry, both ends of 5.85 would this study, 5.87, has a Cmax/ED50 of 21.6 (4.11 µM/0.19 µM).
have to be deleted, but that brings the molecule back to the The aqueous solubility is 3.2 µg/mL, the oral bioavailabil-
5.83 and 5.84 series. That was when the C2 symmetry design ity is 32%, and the plasma half-life after a 5 mg/kg IV dose
had to be put aside, and unsymmetrically substituted deriva- was 2.3 h. This molecule is appreciably smaller than those
tives (different A groups at the N- and C-termini) of 5.84 in the 5.83 and 5.84 series (but the molecular weight is 653,
and 5.86 were investigated.[153] Several SAR observations still quite high), yet it maintained the submicromolar in vivo
were made that were incorporated into the later designs: (1) antiviral activity. However, the relatively short plasma half-
incorporation of a carbamate linkage (ROCONHR′) resulted life prohibits the maintenance of the plasma levels suffi-
in improved potency over the use of an N-alkylurea link- ciently in excess of the ED95 for viral replication observed
age (RNR′ CONHR"); (2) an N-ethylurea linkage was less in vitro. A pharmacokinetic study of the metabolism of 5.87
tolerated than N-methylurea; (3) there was no difference indicated that the probable cause for the short plasma half-
between 2-pyridinyl- and 3-pyridinyl groups at P3; and (4) life was the production of three metabolites: the N-oxide

Ph
OH O
O
H
N N
O N N O
H H
O
N
Ph
5.87
238 The Organic Chemistry of Drug Design and Drug Action

of the 2-­pyridinyl group, the N-oxide of the 3-pyridinyl group proximal to the P2 valine (5.89). By combining all of
group, and the bis(pyridine N-oxide). Using 14C-labeled these characteristics, the best analog was found to be 5.90
5.87, these metabolites corresponded to 92–95% of the total (ritonavir, Norvir), which has a solubility of 6.9 μg/mL at
bile radioactivity. Obviously, the next modification had to pH 4, a Cmax/ED50 of 105 (2.62 μM/0.025 μM), and an oral
be to minimize this metabolism. Attempts to hinder oxida-
tive metabolism sterically by the addition of substituents at Ph
O O
the 6-position of the pyridinyl group generally yielded com- H
N
pounds that showed lower Cmax values and oral bioavail- N
N N N O
S
H H
ability, although they did have greater potency than 5.87. O
S Me OH N
Attempts to modify the electronic nature of the P3 pyridinyl Ph
group by the addition of electron-donating groups, such as Ritonavir
methoxyl or amino, led to more potent and more soluble 5.90
analogs, but they had poorer pharmacokinetic profiles.
By diminishing the oxidation potential of the electron- bioavailability of 78%. The in vitro Ki for HIV-1 protease is
rich pyridinyl groups, drug metabolism should be dimin- 15 pM! The metabolic reactivity of pyridinyl groups versus
ished; consequently, the pyridinyl groups were replaced thiazolyl groups was elucidated by determining the metabo-
with other heteroaromatic bioisosteres.[154] The more elec- lism of 5.87 vs 5.88 vs 5.90. The relative rates of metabolism
tron-deficient (less basic) heterocycles, however, also had are 5.87 (1.0) > 5.88 (0.2) > 5.90 (0.05), as predicted. Oxa-
lower aqueous solubility and oral absorption. 5-Pyrimidinyl zolyl and isoxazolyl analogs had profiles similar to that of
substitution gave inhibitors of equal potencies to the pyridi- ritonavir, both in potency and pharmacokinetic behavior, but
nyl compounds, but with lower bioavailability; the furanyl plasma concentrations were maintained for shorter periods
analogs were more potent, but showed even lower bioavail- than ritonavir. It was later found that one additional reason
ability. However, 5-thiazolyl analogs, which generally for the improved pharmacokinetic properties of ritonavir is
increase aqueous solubility, showed excellent pharmacoki- that it is a potent inhibitor of the 3A4 isozyme of cytochrome
netic properties, although their solubilities were still low. P450, the enzyme responsible for the oxidative metabolism
Alkyl substitution on the P3 heterocycle led to increased of ritonavir.[155] Inhibition results from the binding of the
antiviral potency, but substitution on the P2′ heterocycle unhindered nitrogen atom on the P2′ 5-thiazolyl group to the
gave decreased potency. Later, the crystal structure of rito- heme cofactor in the active site of CYP3A4.[156]
navir bound to the enzyme showed that the alkyl group The general medicinal chemistry approach taken is to
participated in a hydrophobic interaction with Val-82; this start by modifying the lead to increase potency for binding
interaction was optimized with an isopropyl group. As to the appropriate receptor (pharmacodynamics). Once this
noted above, though, an N-methylurea linkage between is accomplished, modifications to improve the pharmaco-
the P3 and P2′ groups provided much higher solubility than kinetic behavior need to be carried out in conjunction with
the corresponding carbamate linkages. Consequently, P3 studies to determine if these modifications affect the phar-
N-methylurea analogs were made. Furthermore, it was found macodynamics. The discovery of ritonavir demonstrates
that in the N-methylurea series, but not the carbamate series, that these approaches can work very well[157] and also
the regioisomeric position of the hydroxyl group became shows that peptidomimetic analogs can be capable of high
significant; compounds in which the hydroxyl group was absorption, oral bioavailability, and slow hepatic clearance.
distal to the P2 valine residue (e.g., 5.88) were 10-fold more
potent than the corresponding analogs with the hydroxyl
5.3. IRREVERSIBLE ENZYME INHIBITORS
O
Ph
O
5.3.1. Potential of Irreversible Inhibition
H
N
N N N N O A reversible enzyme inhibitor is effective as long as a suit-
H H
Me
O OH N
able concentration of the inhibitor is present to drive the
S Ph equilibrium E + I ⇌ E·I to the right (see Section 5.2.1).
P3 thiazolyl N-Me Carbamate P ' 3-pyridinyl Therefore, a reversible inhibitor drug is effective only while
2
urea 5.88
the drug concentration is maintained at a high enough level
to sustain the enzyme–drug complex. Because of drug
Ph metabolism and excretion (see Chapter 8), repetitive admin-
O HO O
H
N
N istration of the drug is required.
N N N O
H O
H A competitive irreversible enzyme inhibitor, also known
Me N
S Ph as an active-site directed irreversible inhibitor or an enzyme
inactivator, is a compound whose structure is similar to
5.89
that of the substrate or product of the target enzyme and
Chapter | 5 Enzyme Inhibition and Inactivation 239

which generally forms a covalent bond to an active site [Nexium]; and lansoprazole, 5.93 [Prevacid]) are covalent
residue (a slow, tight-binding inhibitor, however, often is inhibitors of their biological targets. However, most of the
a noncovalent inhibitor that can be functionally irrevers- covalent drugs were not designed as such but their covalent
ibly bound). In the case of irreversible inhibition, it is not mechanisms were determined after their efficacy was discov-
necessary to sustain the inhibitor concentration to retain the ered, generally through a screening campaign.
enzyme–inhibitor interaction because this is an irreversible The anxiety about the use of covalent drugs stems from
reaction (usually covalent), and once the target enzyme has the fear of potential off-target reactions, resulting from
reacted with the irreversible inhibitor, the complex can- reactive functional groups, and immunogenic effects, which
not dissociate (again, there are exceptions). The enzyme, may have derived from studies in the 1970s that showed
therefore, remains inactive, even in the absence of addi- hepatotoxicity from compounds, such as acetaminophen,
tional inhibitor. This effect could translate into a require- that underwent metabolism to highly reactive intermediates,
ment for smaller and fewer doses of the drug. It could also which covalently bonded to liver proteins.[168] Modification
lead to lower toxicity because once the enzyme is inacti- of proteins, either by direct attack by the drug or by a reac-
vated, drug metabolism can rapidly clear the body of the tive metabolite, produces a hapten–protein complex, which
unbound drug, lowering potential toxicity from the drug, can elicit an immune response.[169] These early studies may
but retaining the pharmacological activity of the drug over not be relevant to many covalent drugs.
an extended period.[158] In fact, metabolic stability becomes The term irreversible is a loose one; either a very stable
less of an issue with an irreversible inhibitor because rapid covalent bond or a labile covalent bond may be formed
clearance may be advantageous, especially if covalent bond between the drug and the enzyme active site. A revers-
formation occurs rapidly.[159] Vigabatrin (Section 5.3.3.3.1) ible covalent inhibitor, which dissociates at a rate that is
is a good example of this phenomenon.[160] In rats[161] and faster than the physiological degradation of the protein, has
humans,[162] vigabatrin has a metabolic half-life of 1–3 and the advantage of longer lived inhibition than noncovalent
5 h, respectively, but a binding half-life to GABA-AT, the inhibitors and without the potential immunological effects
target of this drug, of several days. The return of GABA- of irreversible inhibitors. As pointed out earlier, some tight-
AT activity after vigabatrin is administered corresponds binding reversible inhibitors are also functionally irrevers-
to the rate of genetic synthesis of new GABA-AT.[163] In ible. As long as return of enzyme activity is slower than
some cases, however, particularly where genetic translation degradation of the enzyme, the enzyme is considered
of the target enzyme is extremely slow, it may be safer to irreversibly inhibited. The two principal types of enzyme
design reversible inhibitors whose effects can be controlled inactivators are reactive compounds called affinity labeling
more effectively by termination of their administration. For agents and unreactive compounds that are activated by the
diseases that require extended, high target occupancy, such target enzyme, known as mechanism-based enzyme inacti-
as cancer and microbial infections, irreversible inhibition vators.
might be the most effective therapy.[164]
Even though the target enzyme is destroyed by an irre-
versible inhibitor, it does not mean that only one dose of the O
drug would be sufficient to destroy the enzyme permanently. Cl H OCH3
Yes, it destroys that copy of the enzyme permanently, but our N
genes are constantly encoding more copies of proteins that
S
diminish. As the enzyme loses activity, additional copies of
Clopidogrel
the enzyme are synthesized, but this process can take hours 5.91
or even days (as is the case above with GABA-AT).[165] One
may wonder what the effect on metabolism would be if a
N O
particular enzyme activity were completely inhibited for an S N
extended period. Consider the case of aspirin, an irreversible MeO N
H
inhibitor of PG synthase (See Section 5.3.2.2.2). If the quan-
tity of aspirin consumed in the United States (about 80 bil- OMe
lion tablets a year in 1994)[166] were averaged over the entire Esomeprazole
5.92
population, then every man, woman, and child would be tak-
ing about 240 mg of aspirin every day, enough to shut down
human PG biosynthesis for the entire country permanently! N O
Suffice it to say that there are many irreversible enzyme inhib- S N
N
itor drugs in medical use. In fact, there is at least one example H
CF3
of an approved irreversible inhibitor drug for about one-third O
of all enzyme drug targets.[167] Three of the 10 top-selling Lansoprazole
drugs in 2009 (clopidogrel, 5.91 [Plavix]; esomeprazole, 5.92 5.93
240 The Organic Chemistry of Drug Design and Drug Action

5.3.2. Affinity Labeling Agents Furthermore, the inactivator may form an E·I complex with
other enzymes, but if there is no nucleophile near the reac-
5.3.2.1. Mechanism of Action tive functional group, no reaction will take place. Third, in
An affinity labeling agent is a reactive compound that has a the case of antitumor agents, mimics of DNA precursors
structure similar to that of the substrate for a target enzyme. are rapidly transported to the appropriate site and, therefore,
Subsequent to reversible E·I complex formation, it reacts with they are preferentially concentrated at the desired target.
active site nucleophiles (usually amino acid side chains), gen- The keys to the effective design of affinity labeling agents
erally by acylation or alkylation (SN2) mechanisms, thereby as drugs are specificity of binding and moderate reactivity
forming a stable covalent bond to the enzyme (Scheme 5.13). of the reactive moiety. If the molecule has a very low Ki for
Note that this reaction scheme is similar to that for the con- the target enzyme, then E·I complex formation with the tar-
version of a substrate to a product (see Chapter 4, Section get enzyme will be favored, and the selective reactivity will
4.1.2); instead of a kcat, the catalytic rate constant for product be enhanced. The effectiveness of modulating the activity of
formation, there is a kinact, the inactivation rate constant for the reactive functional group can be seen by comparing the
enzyme inactivation. On the assumption that the equilibrium relatively moderate reactivity of the functional groups in the
for reversible E·I complex formation (KI) is rapid, and the nontoxic affinity labeling agents described in Section 5.3.2.2
rate of dissociation of the E·I complex (koff) is fast relative to with the highly reactive functional groups in some of the can-
that of the covalent bond forming reaction, then kinact will be cer chemotherapeutic drugs in Chapter 6 (Section 6.3.2).
the rate-determining step. In this case, unlike simple revers- One approach that takes advantage of modulated reactiv-
ible inhibition, there will be a time-dependent loss of enzyme ity of an affinity labeling agent has been termed by Krantz[170]
activity (as is the case with slow, tight-binding reversible quiescent affinity labeling. In this case, the inactivator reac-
inhibitors because of the relatively small koff). tivity is so low that reactions with nucleophiles in solution at
The rate of inactivation is proportional to low concentra- physiological pH and temperature are exceedingly slow or
tions of inhibitor, but becomes independent at high concen- nonexistent. However, because of the exceptional nucleophi-
trations. As is the case with substrates, the inhibitor can also licity of groups in some enzymes that use covalent catalysis
reach enzyme saturation when the kinact is slow relative to as a catalytic mechanism (see Chapter 4, Section 4.2.2), the
the koff. Once all the enzyme molecules are tied up in an E·I poorly electrophilic sites in the quiescent affinity labeling
complex, the addition of more inhibitor will have no effect agents are reactive enough for nucleophilic reaction, but only
on the rate of inactivation. at the active site of the enzyme. High selectivity for inacti-
Because an affinity labeling agent contains a reactive func- vation of a particular enzyme can be built into a molecule
tional group, it can react not only with the active site of the tar- if the structure of the inactivator is designed so that it binds
get enzyme but also can react with thousands of nucleophiles selectively to the target enzyme. For example, peptidyl acyl-
associated with many other enzymes and biomolecules in the oxymethyl ketones (5.94, Scheme 5.14) have low chemi-
body. Consequently, these inactivators are potentially quite cal reactivity (the acyloxyl group is a weak leaving group),
toxic. In fact, many cancer chemotherapy drugs (see Chapter but are potent and highly selective inactivators of cathepsin
6) are affinity labeling agents, and they are quite toxic. B,[171] a cysteine protease that has been implicated in osteo-
There are several principal reasons why these reac- clastic bone resorption,[172] tumor metastasis,[173] and muscle
tive molecules, nonetheless, can be effective drugs. First, wasting in Duchenne muscular dystrophy.[174]
once the inactivator forms an E·I complex, a unimolecu- A third feature that would increase the potential effec-
lar reaction ensues (the E·I complex is now a single mol- tiveness of an affinity labeling agent can be built into the
ecule), which can be many orders of magnitude (108 times; molecule, if something is known about the location of the
see Chapter 4, Section 4.2.1) more rapid than nonspecific active site nucleophiles. In cases where a nucleophile is
bimolecular reactions with nucleophiles on other proteins. known to be at a particular position relative to the bound
substrate, the reactive functional group can be incorporated
KI kinact into the affinity labeling agent so that it is near that site
E+I E•I E -I when the inactivator is bound to the target enzyme. This
SCHEME 5.13 Basic kinetic scheme for an affinity labeling agent increases the probability for reaction with the target enzyme
by approximation (see Chapter 4, Section 4.2.1).
O Ar
R' O O R' O
H H
R N O Ar R N
N N
H H
O R" O O R" S
S
5.94

SCHEME 5.14 Peptidyl acyloxymethyl ketones as an example of a quiescent affinity labeling agent
Chapter | 5 Enzyme Inhibition and Inactivation 241

Because many enzymes involved in DNA biosynthesis is amoxicillin (Amoxil). The differences in these derivatives
utilize substrates with similar structures, high concentra- (other than structure) are related to absorption properties, sus-
tions of reversible inhibitors can block multiple enzymes. ceptibility to deactivation by penicillinases, and specificity for
A lower concentration of an irreversible inhibitor may be organisms for which they are most effective.[178]
more selective, if it only reacts with those enzymes having The structures of some cephalosporins and cephamycins
appropriately juxtaposed active site nucleophiles. are shown in 5.96; for example, 5.96a is cefazolin (injectable;
Even when all these design factors are taken into con- Ancef), 5.96b is cefoxitin (injectable; Mefoxin), 5.96c is
sideration, nonspecific reactions still can take place that cefaclor (oral; Ceclor), and 5.96d is ceftizoxime (injectable;
may result in side effects. Judicious use of low electro- Cefizox). The analogs where X = H are cephalosporins and
philic moieties can result in highly effective potential drug those where X = OCH3, such as cefoxitin, 5.96b, are called
candidates. Another approach, known as targeted covalent cephamycins. The penicillins, cephalosporins, and cephamy-
inhibition,[175] targets a noncatalytic nucleophilic residue cins all have in common the β-lactam ring and are known
that is poorly conserved across the target protein family, collectively as β-lactam antibiotics. Cephalosporins and
thereby giving selectivity for the desired target protein. Of cephamycins are classified by generations, which are based
course, in all cases, enhanced potency for the target protein on general features of antimicrobial activity.[179] Cefazolin
will lead to favorable selectivity and safety; the majority of is a first-generation cephalosporin, cefoxitin and cefaclor
drugs withdrawn from the market because of safety issues, are second-generation cephalosporins, and ceftizoxime is a
known as idiosyncratic drug toxicities,[176] which only third-generation cephalosporin. Modifications in the structure
occur in humans and are of unknown causes (although reac- of these antibiotics have been extensive,[180] so much so that
tive metabolites are suspected), were administered at doses essentially every atom excluding the lactam nitrogen has been
greater than 100 mg/day, whereas drugs taken at <10 mg/ replaced or modified in the search for improved antibiotics.
day seem to have generally good safety profiles.[177] The discovery of penicillin was described in Chapter 1,
Section 1.2.2.[181] Penicillins and cephalosporins/cephamy-
5.3.2.2. Selected Affinity Labeling Agents cins are bacteriocidal (they kill existing bacteria), unlike the
sulfonamides (see Section 5.2.2.3), which are bacteriostatic.
5.3.2.2.1. Penicillins and Cephalosporins/ They are ideal drugs in that they inactivate an enzyme that
Cephamycins is essential for bacterial growth, but which does not exist in
Penicillins have the general structure 5.95; for example, animals, namely, the peptidoglycan transpeptidase. As dis-
5.95a is penicillin G, 5.95b is penicillin V, 5.95c (R′ = H) is cussed in Chapter 4 (Section 4.2.7.), this enzyme catalyzes
oxacillin (Baetocill), 5.95c (R′ = Cl) is cloxacillin (Cloxapen), the cross-linking of the peptidoglycan to form the bacterial
5.95d (R′ = H) is ampicillin (Principen), and 5.95d (R′ = OH) cell wall. Scheme 4.13 shows that one (the acyl donor) of the

H H H a, R = PhCH2 – H2 N
R N S H H H
b, R = PhOCH2 – N S
O N
O R' O N
CO2H CH3 d, R = R' O
5.95 c, R = CO2H
5.95d
N O

H X H N N
R N S
a, R = N N CH2
X=H Y=
N N N CH2S S CH3
O Y
O
CO2H O
5.96 b, R = X = OCH3 Y= CH2OC
S CH2 NH2

c, R = CH X=H Y = Cl
NH2

C
HN N OCH3
d, R =
X=H Y=H
HN S
242 The Organic Chemistry of Drug Design and Drug Action

two substrates of the peptidoglycan transpeptidase contains in the dihydrothiazine ring also may activate the β-lactam car-
a peptide segment having d-alanyl-d-alanine at the C-termi- bonyl (Scheme 5.16).[187] It should be noted that β-lactam anti-
nus (Scheme 4.13, graphic A). According to Scheme 4.13, biotics, as well as all bactericidal antibiotics, have been shown
the enzyme initially acts analogously to a serine protease to to produce highly toxic hydroxyl radicals, which may be the
carry out the first half of the transpeptidase reaction. That ultimate mechanism of bacterial cellular death.[188]
is, nucleophilic attack of an enzyme serine residue on the On the basis of the structural similarity of penicillins
carbonyl group of the first d-Ala residue of the C-terminal to acyl d-alanyl-d-alanine (see Figure 5.13), Tipper and
d-Ala-d-Ala-OH segment results in formation of a serine Strominger predicted that 6-methylpenicillin (a methyl
ester linkage with the release of the C-terminal d-Ala-OH on the sp3 carbon adjacent to Na) would be a more potent
(Scheme 4.13, graphics A-C). The active site serine involved inhibitor than the parent molecule. However, both 6-meth-
in catalysis has been identified.[182] Scheme 4.13 (graphics ylpenicillin and 7-methylcephalosporin (the numbering is
C-F) further shows the reaction of the amino terminus of a different for cephalosporins because it has one more carbon
second peptidoglycan unit (the acyl acceptor) with the serine in the ring, but the methyl in 7-methylcephalosporin corre-
ester to form the cross-linked peptidoglycan. In the case of a sponds to that in 6-methylpenicillin) were synthesized and
serine protease, a water molecule would take the place of the were shown to be inactive.[189] Because the corresponding
second peptidoglycan segment to result in simple hydrolysis 7-methoxycephalosporins (that is, cephamycins) are better
of the serine ester. As discussed below, penicillins interfere inhibitors than the parent cephalosporins, it is not clear why
with these peptidoglycan cross-linking processes. the methyl analogs are poor inhibitors.
By comparison of a molecular model of penicillin with that The beauty of the penicillins (and cephalosporins) is that
of d-alanyl-d-alanine, Tipper and Strominger[183] suggested they are not exceedingly reactive; consequently, few non-
that penicillin could mimic the structure of the terminus of specific acylation reactions occur. Their modulated reactiv-
the peptidoglycan, and bind at the active site of the transpep- ity and nontoxicity make them ideal drugs. If it were not for
tidase (Figure 5.13). The Na to Nb distances (3.3 Å) and the allergic responses and problems associated with digestion
Nb to carboxylate carbon distances (2.5 Å) in both molecules and drug resistance (see Chapter 7), penicillins might be
are identical. The Na to carboxylate carbon distance is 5.4 Å in considered nutritious foods, composed of various carbox-
the penicillins and 5.7 Å in d-alanyl-d-alanine. The β-lactam ylic acid derivatives (the RCO side chains), cysteine, and
carbonyl may be further activated by torsional effects of the the essential amino acid valine!
thiazolidine ring in the penicillins. This carbonyl corresponds Because of the discovery of highly effective antibiotics,
to the carbonyl in the acyl d-alanyl-d-alanine that acylates the such as the penicillins and the sulfa drugs, and the realiza-
active site serine and, therefore, penicillins also could acyl- tion of even more effective synthetic analogs of these natural
ate the transpeptidase serine residue[184] (Scheme 5.15). This products, Nobel laureate immunologist F. Macfarlane Bur-
hypothesis was supported many years later in a crystal struc- net noted in 1962 that by the late twentieth century “the vir-
ture obtained to 1.2 Å resolution of a cephalosporin bound to a tual elimination of infectious disease as a significant factor
bifunctional serine type d-alanyl-d-alanine carboxypeptidase/ in social life” should be anticipated.[190] William Stewart, the
transpeptidase.[185] The bulk of the penicillin molecule, when it
is attached to the active site, precludes hydrolysis or transami- O R
O R
dation either for steric reasons or because it induces a confor-
mational change in the enzyme that prevents these processes NH NH
from occurring.[186] Covalent binding at the active site prevents O- O S
the peptidoglycan acyl donor substrate from binding. Similar O S
O N
N
arguments could be made for cephalosporins. The double bond B H H
+ COO –
COO –
O H
Me Nb H
H CO2H
O
Na Me SCHEME 5.15 Acylation of peptidoglycan transpeptidase by
H penicillins
Peptidoglycan

O H H H
H Nb H S S
R N R N
O CO2H
N a S Me N
O Y O N Y
O O +
R H Me
CO2H CO2H
FIGURE 5.13 Comparison of the structure of penicillins with acyl d-ala-
nyl-d-alanine SCHEME 5.16 Activation of the β-lactam carbonyl of cephalosporins
Chapter | 5 Enzyme Inhibition and Inactivation 243

former Surgeon General of the United States, testified before prolonged use). Hoffmann synthesized many salicylate
Congress in 1967 that it was time to “close the book on infec- derivatives, and acetylsalicylic acid (aspirin, 5.97) was the
tious diseases”. Apparently, rampant resistance (Chapter 7) best. He gave it to his father, who responded well, and in
to these and all antibacterial drugs was not foreseen. 1899 Bayer introduced aspirin as an antipyretic (fevers),
antiinflammatory (arthritis), and analgesic (pain) agent.[194]
5.3.2.2.2. Aspirin Aspirin became the first drug ever to be tested in clinical
It is stated in the Papyrus Ebens (c. 1550 B.C.) that dried trials before registration. Also, because of its insolubility in
leaves of myrtle provide a remedy for rheumatic pain. Hip- water, it had to be sold in solid form, and, therefore, became
pocrates (460–377 B.C.) recommended the use of willow the first major medicine to be sold in the form of tablets.
bark for pain during childbirth.[191] A boiled vinegar extract The trade name aspirin was coined by adding an “a” for
of willow leaves was suggested by Aulus Cornelius Celsus acetyl to spirin for Spiraea, the plant species from which
(30 A.D.) for relief of pain.[192] In 1763, Reverend Edmund salicylic acid was once prepared. In the first 100 years since
Stone of England announced his findings that the bark of the its introduction on the market, it was estimated that one tril-
willow (Salix alba vulgaris) provided an excellent substi- lion (1012) aspirin tablets had been consumed.[195]
tute for Peruvian bark (Cinchona bark, a source of quinine)
COOH
in the treatment of fevers.[193] The connection between the
O CH3
two barks was discovered by his tasting them and making
the observation that they both had a similar bitter taste. The O
bitter active ingredient with antipyretic activity in the wil- Aspirin
low bark was called salicin, which was first isolated in 1829 5.97
by Leroux. Upon hydrolysis, salicin produced glucose and
salicylic alcohol, which was metabolized to salicylic acid. The mechanism of action was initially reported by
Sodium salicylate was first used for the treatment of rheu- Vane,[196] who shared the 1982 Nobel Prize in Medicine
matic fever and as an antipyretic agent in 1875. Toward the (with Sune Bergström and Bengt Samuelsson) for this dis-
end of the nineteenth century, the father of Felix H
­ offmann, covery, and by Smith and Willis,[197] to be the result of inhi-
a chemist employed by Bayer Company, suffered from bition of prostaglandin (PG) biosynthesis. PGs are derived
severe rheumatoid arthritis and pleaded with his son to from arachidonic acid (5.98) by the action of PG synthase
search for a less irritating drug than the sodium salicylate (also known as cyclooxygenase (COX)) (Scheme 5.17). At
he was using (salicylic acid not only tastes awful but also the time of the initial report there was evidence that vari-
causes ulcerations of the mouth and stomach linings with ous PGs were involved in the pathogenesis of inflammation

COOH Prostaglandin O COOH

O
Synthase
O2
5.98 OOH
PGG2

COOH O COOH
O
HO
OH OH
PGE2 PGH2

HO HO

COOH COOH

O HO
OH OH
PGD2 PGF2α
SCHEME 5.17 Biosynthesis of prostaglandins (PGs) from arachidonic acid
244 The Organic Chemistry of Drug Design and Drug Action

and fever. It is now known that all mammalian cells (except as less relaxation of blood vessels in the head, so fewer
erythrocytes) have microsomal enzymes that catalyze the headaches; fewer PGs in the hypothalamus means that the
biosynthesis of PGs, and that PGs are always released when body temperature set point is not changed, and no fever is
cells are damaged, and are in increased concentrations in induced.
inflammatory exudates. When PGs are injected into ani- Sheep vesicular gland PG synthase was shown to be
mals, the effects are reminiscent of those observed during irreversibly inactivated by aspirin.[198] When microsomes of
inflammatory responses, namely, redness of the skin (ery- sheep seminal vesicles were treated with aspirin tritiated in
thema) and increased local blood flow. A long-lasting vaso- the methyl of the acetyl group, acetylation of a single pro-
dilatory action is also prevalent. PGs can cause headache tein was observed. The same experiment carried out with
and vascular pain when infused in man. Elevation of body aspirin tritiated in the benzene ring resulted in no tritium
temperature during infection is also mediated by the release incorporation, suggesting that acetylation was occurring.
of PGs. Incubation of purified PG synthase with [3H-acetyl]aspirin
From the above discussion, it is apparent that inhibition led to irreversible inactivation with incorporation of one
of PG synthase, the enzyme responsible for the biosynthe- acetyl group per enzyme molecule.[199] Pepsin digestion of
sis of all the PGs and related compounds (PGH2 also is the tritiated enzyme gave a 22-amino-acid-labeled peptide;
converted by prostacyclin synthase to prostacyclin (5.99) tryptic digestion of this peptide gave a tritiated decapeptide
and by thromboxane synthase to thromboxane A2 (5.100)), in which the serine residue was acetylated.[200] Thermolysin
would be a desirable approach for the design of antiin- digestion of the tritiated enzyme gave the labeled dipeptide
flammatory, analgesic, and antipyretic drugs. Inhibition of Phe–Ser, where the hydroxyl group of Ser-530 had become
platelet COX is particularly effective in blocking PG bio- acetylated.[201] The most straightforward mechanism for
synthesis because, unlike most other cells, platelets can- acetylation would be a transesterification mechanism by
not regenerate the enzyme, because they have little or no aspirin acting as an affinity labeling agent; on the basis
capacity for protein biosynthesis. Therefore, a single dose of site-directed mutagenesis studies, it was proposed by
of 40 mg/day of aspirin is sufficient to destroy the COX Marnett and coworkers[202] that Tyr-385 and Tyr-348 in the
for the life of the platelet! When PG synthesis is blocked, active site hydrogen bond to the acetyl carbonyl of aspi-
there is less stimulation of the pain-sensitive nerve end- rin, which directs it specifically to Ser-530 for acetylation
ings, resulting in less aching of muscles and joints, as well (Scheme 5.18).
The principal side effect of aspirin and all nonsteroi-
HOOC
dal antiinflammatory drugs (NSAIDs) is that their chronic
use leads to ulceration of the stomach lining. In the late
O 1980s, it was found that COX activity was increased in cer-
tain inflammatory states and could be induced by inflam-
matory cytokines.[203] This suggested the existence of two
HO
different forms (isozymes) of COX, now known as COX-1
HO
and COX-2. COX-1, the constitutive form of the enzyme
5.99
(that is, present in the cells all the time), is responsible
for the physiological production of PGs that are important
COOH in maintaining tissues in the stomach lining and kidneys.
O
COX-2,[204] induced by cytokines in inflammatory cells,
O
is responsible for the elevated production of PGs dur-
HO
ing inflammation and is associated with inflammation,
5.100 pain, and fevers. The discovery of a second COX enzyme

Tyr-385

O H
Ser-530
H O
Tyr-348 CH3 CH3 OH
O O O
O +
H O COO
COO Ser-530
SCHEME 5.18 Hypothetical mechanism for acetylation of prostaglandin synthase by aspirin
Chapter | 5 Enzyme Inhibition and Inactivation 245

suggested that aspirin and other NSAIDs may inhibit both 350 compounds were tested for oral activity, of which seven
COX-1 and COX-2, leading to stomach irritation (and, candidates were selected. Compound 5.101 has an IC50 of
in some cases, ulcers); ideally, inhibition of only COX-2 60 nM and a selectivity of >1700 against COX-2 vs COX-
would produce the desired antiinflammatory effect without 1.[206] Series 5.102 has a range of IC50 values of 10–100 nM
stomach lining irritation.[205] and in vitro selectivities of 103–104 for COX-2 vs COX-
Both G. D. Searle (now defunct) and Merck initiated 1.[207] However, celecoxib (5.103, Celebrex), with an IC50 of
programs to discover a COX-2-selective (reversible) inhibi- 40 nM and in vitro selectivity of only 375 in favor of COX-2
tor. At Searle, more than 2500 compounds were synthesized inhibition, emerged as the commercial compound because
and almost 2000 screened of which 280 were potent. Almost of its overall favorable properties.[208] Its in vivo selectiv-
ity (inflammatory effect/stomach irritation; a therapeutic
index) was >1000. Merck also discovered a COX-2-selective
H 3C
inhibitor, rofecoxib (5.104, Vioxx),[209] which is about 5.5
N times more selective than celecoxib and was a once-a-day
treatment for osteoarthritis and rheumatoid arthritis. Searle
then discovered valdecoxib (5.105; Bextra), having an IC50
F SO2CH3 for COX-1 of 140 μM and for COX-2 of 5 nM or 28,000-
5.101 fold selectivity for COX-2 over COX-1.[210] In 2003, the
global sales of Vioxx reached $2.5 billion, but in 2004,
Merck voluntarily took the drug off the market after a study
CF3
found an increased incidence of heart attacks and strokes
N among patients taking the drug compared to those on pla-
N cebo.[211] This event engendered congressional hearings and
much finger-pointing as questions were asked about how
much and when the FDA and Merck knew about the poten-
X SO2R tially fatal side effects of the drug. Whether celecoxib and
5.102 valdecoxib might share this property was also questioned,
and sales of valdecoxib were discontinued in 2005, while as
of 2014, celecoxib was still on the market. It appears that
CF3
the lower selectivity of celecoxib relative to rofecoxib and
N valdecoxib may be key to its safer properties.[212] Inhibi-
N tion of COX-2 reduces production of prostacyclin (5.99),
and thereby attenuates prostacyclin’s athero-protective
properties. Inhibition of COX-1 decreases production of
H 3C SO2NH2 thromboxane A2 and thereby attenuates thromboxane A2’s
Celecoxib prothrombotic properties. A less selective inhibition may
5.103 lead to a more favorable balance between these opposing
effects.[213]
In addition to its oxygenation of arachidonic acid (5.98,
Scheme 5.17), COX-2 is also responsible for oxygenation
O O
of the endocannabinoids, which exert analgesic and anti-
inflammatory effects at cannabinoid receptors CB1 and
CB2; oxygenation of endocannabinoids turns off their pain-
relieving effects. Therefore, inhibition of the oxygenation
SO2CH3 of endocannabinoids should be a treatment for neuropathic
Rofecoxib pain.[214] The (S)-enantiomers of chiral NSAIDs ibuprofen
5.104
(5.106, Advil) and naproxen (5.107, Aleve) are the euto-
mers for inhibition of the oxygenation of arachidonic acid
O CH3 by COX-2; the (R)-enantiomers are inactive. However, it
N
was found that the (R)-enantiomers are potent substrate-
selective inhibitors of the endocannabinoid oxygenation
activity of COX-2,[215] suggesting their potential role in
SO2NH2 pain management.
Valdecoxib A third isozyme of COX called COX-3 (actually a vari-
5.105 ant of COX-1), found principally in the cerebral cortex, was
246 The Organic Chemistry of Drug Design and Drug Action

shown to be selectively inhibited by NSAIDs that have good selective inhibitors bind to COX-1, a substituent on the
analgesic and antipyretic activity, but low antiinflammatory inhibitor has a repulsive interaction with the Ile, but not
activity, such as acetaminophen (5.108, Tylenol).[216] Anal- with the smaller Val in COX-2, resulting in preferential
gesic/antipyretic drugs penetrate the blood–brain barrier binding to COX-2. Site-directed mutagenesis of these
well and may accumulate in the brain where they inhibit amino acids demonstrated that they were important to
COX-3. NSAIDs that contain a carboxylate group, such as the selective inhibitor binding.[223] This small difference
aspirin (5.97) and ibuprofen (5.106), cross the blood–brain in size between Ile and Val is not sufficient for structure-
barrier poorly, but are more potent inhibitors of COX-3, so based drug design approaches (see Chapter 2, Section
they also exhibit analgesic and antipyretic activities. COX-2 2.2.6), and in fact the COX-2-selective inhibitors were
selective inhibitor drugs inhibit inflammatory pain,[217] discovered without the aid of crystal structures. Note that
whereas COX-1-selective inhibitor drugs, such as aspirin, these COX-2-selective inhibitors are reversible inhibitors.
ibuprofen, and naproxen,[218] are superior to COX-2 inhibi- They are being discussed in this section of irreversible
tors against chemical pain stimulators.[219] The analgesic inhibitors only because they follow from our discussion
effect of COX-1-selective inhibitors occurs at lower doses of aspirin.
than those needed to inhibit inflammation.[220] Both penicillin and aspirin are examples of affinity label-
X-ray crystal structures of COX-1 and COX-2 are ing agents that involve acylation mechanisms. Other affinity
known, and there is very little difference in the active labeling agents function by alkylation or arylation mecha-
sites of these two isozymes. COX-1[221] has an active site nisms. An interesting example involves selective, covalent
isoleucine at position 523 (Ile-523) and COX-2[222] has modification of β-tubulin, leading to disruption of microtubule
a valine (Val-523) at that position. Because isoleucine is polymerization and cytotoxicity against multidrug-resistant
larger than valine (by one methylene group), when COX-2 tumors (multidrug resistance is discussed in Chapter 7).[224]
The pentafluorobenzene analog, 5.109, produces time-depen-
CH3 dent binding to β-tubulin; tritium-labeled 5.109 incorporates
tritium into a protein that comigrates on sodium dodecyl
COOH sulfate/polyacrylamide gel electrophoresis with β-tubulin.
Ibuprofen Covalent modification occurs at a conserved cysteine residue
5.106
(Cys-239) shared by β1, β2, and β4 tubulin isoforms. Replace-
ment of the para-fluoro group with other halogens causes a
H3C H substantial decrease in potency; replacement with a hydro-
COOH
gen abolishes the cytotoxicity of the compound.[225] These
results are consistent with a nucleophilic aromatic substitu-
MeO tion mechanism for covalent inactivation (Scheme 5.19).
Naproxen
5.107 H
O N OCH3
O S
O F F F
H3C N OH
H F F
Acetaminophen F
5.108 5.109

O H
O H O H N OCH3
N OCH3 N OCH3 S
O
O S S
O F
F F F F F F F F

F F F F F F
F F
S S
S

SCHEME 5.19 Hypothetical nucleophilic aromatic substitution mechanism for arylation of β-tubulin
Chapter | 5 Enzyme Inhibition and Inactivation 247

5.3.3. Mechanism-Based Enzyme 5.3.3.2. Potential Advantages in Drug Design

Inactivators Relative to Affinity Labeling Agents
5.3.3.1. Theoretical Aspects Because of the generally high reactivity of affinity labeling
agents, they can react with enzymes and biomolecules other
A mechanism-based enzyme inactivator is an unreactive than the target enzyme. When this occurs, toxicity and side
compound that bears a structural similarity to the sub- effects can arise. However, mechanism-based enzyme inacti-
strate for a specific enzyme. Once the inactivator binds vators are unreactive compounds, and this is the key feature
to the active site, the target enzyme, via its normal cata- that makes them so amenable to drug design. Consequently,
lytic mechanism, converts the compound into a product nonspecific alkylations and acylations of other proteins should
that inactivates the enzyme prior to escape from the active not be a problem. In the ideal case, only the target enzyme will
site. Therefore, these inactivators are acting initially as be capable of catalyzing the appropriate chemistry for conver-
substrates for the target enzyme. Although the product sion of the inactivator to the activated species, and inactivation
generally forms a covalent bond to the target enzyme, it is will result with every turnover, that is, the partition ratio will
not essential (a tight-binding inhibitor may form). How- be zero. This may be quite important for potential drug use. If
ever, the target enzyme must transform the inactivator the partition ratio is greater than zero, the released activated
into the actual inactivating species, and inactivation must species may react with other proteins, possibly resulting in a
occur prior to the release of this species from the active toxic effect. In this case, the inactivator is called a metaboli-
site. Because there is an additional step in the inactiva- cally activated inactivator.[227] Alternatively, the released spe-
tion process relative to that for an affinity labeling agent cies may be hydrolyzed by the aqueous medium or scavenged
(see Section 5.3.2), the kinetic scheme for mechanism- by a scavenger molecule such as glutathione prior to reaction
based inactivation differs from that for affinity labeling with other biomolecules, but the product formed may be toxic
(Scheme 5.20). Provided k4 is a fast step and the equi- or may be metabolized further to other toxic substances. Under
librium k1/k−1 is set up rapidly, k2 is the inactivation rate the ideal conditions mentioned above, the inactivator would
constant (kinact) that determines the rate of the inactivation be a strong drug candidate because it would be highly enzyme
process. The two key features of this type of inactivator specific and low in toxicity. In fact, α-difluoromethylornithine
that differentiate it from an affinity labeling agent are its (eflornithine), a specific mechanism-based inactivator of orni-
initial unreactivity and the requirement for the enzyme thine decarboxylase (see Section 5.3.3.3.2), has been adminis-
to catalyze a reaction on it, thereby converting it into a tered to patients in amounts of 30 g/day for several weeks with
product, that is, the actual inactivator species. Often this only minor side effects![228]
converted inactivator species is quite reactive, and there- There are quite a few drugs in current medical use that
fore acts as an affinity labeling agent that already is at are mechanism-based inactivators, but most were deter-
the active site of the target enzyme. Inactivation (k4 in mined ex post facto, rather than being designed, to be mech-
Scheme 5.20) does not necessarily occur every time the anism-based inactivators. A few known mechanism-based
inactivator is transformed into the inactivating species; inactivators not discussed in the section below include
sometimes this species escapes from the active site (k3 the antidepressant drug phenelzine sulfate (5.110, Nardil)
in Scheme 5.20). The ratio of the number of turnovers and the antihypertensive drug, hydralazine hydrochloride
that gives a released product per inactivation event, k3/k4, (5.111; Apresoline), both of which inactivate monoamine
is called the partition ratio. Ideally, the partition ratio oxidase (MAO); clavulanic acid (5.112; see Chapter 7,
should be zero to avoid release of potentially reactive spe- Section 7.2.2.1; in combination with amoxicillin called
cies, but this is rare.[226] Augmentin), a compound used to protect penicillins and
cephalosporins against bacterial degradation (inactivates
β-lactamases); the antiviral agent, trifluridine (5.113, Virop-
k1 tic), which inactivates thymidylate synthase; gemcitabine-
k2 k4
E +I E•I E•I' E - I '' HCl (5.114, Gemzar), an antitumor drug that inactivates
k-1 ribonucleotide reductase; the antihyperuricemic agent,
allopurinol (5.115, inactivates xanthine oxidase; Aloprim);
k3
and the antithyroid drug, methimazole (5.116, Tapazole),
which inactivates thyroid peroxidase. There are quite a few
E+P mechanism-based inactivators of cytochrome P450; how-
SCHEME 5.20 Kinetic scheme for simple mechanism-based enzyme ever, those inactivators do not derive their pharmacological
inactivation effects as a result of that inactivation.
248 The Organic Chemistry of Drug Design and Drug Action

Because the activation of mechanism-based inactiva- 5.3.3.3. Selected Examples of Mechanism-

tors depends on the catalytic mechanism of their target Based Enzyme Inactivators
enzyme, these inactivators can be designed by a rational
organic mechanistic approach. However, this approach 5.3.3.3.1. Vigabatrin, an Anticonvulsant Drug
to drug design was infrequently used for several reasons. Epilepsy is a disease that was described over 4000 years
First, knowledge of the catalytic mechanism of the target ago in early Babylonian and Hebrew writings. Full clini-
enzyme is required to use the mechanism in the design cal descriptions were written in Hippocrates’ monograph
of the inactivator. Also, this approach requires that a spe- On the Sacred Disease in about 400 BC.[229] Epilepsy is
cific molecule be synthesized, which may be time con- not a single disease, but a family of CNS diseases charac-
suming to do, just to test the mechanistic hypothesis. terized by recurring convulsive seizures. By that defini-
Even if the compound inactivates the enzyme, issues of tion then 1–2% of the world population has some form
potency, specificity, and pharmacokinetics have to be of epilepsy.[230] The disease is categorized as primary or
managed while maintaining the core structure for catalytic idiopathic when no cause for the seizure is known, and
turnover. Because of all of these complications, relatively secondary or symptomatic when the etiology has been
few mechanism-based inactivator drugs are commercial. identified. Symptomatic epilepsy can result from specific
Furthermore, by the last decade of the twentieth century, physiological phenomena such as brain tumors, syphilis,
high-throughput screening (see Chapter 2, Section 2.1.2.3) cerebral arteriosclerosis, multiple sclerosis, Buerger dis-
and combinatorial chemistry (see Chapter 2, Section ease, Pick disease, Alzheimer disease, sunstroke or heat
2.1.2.3.1.3.1) made random and focused screens popu- stroke, acute intoxication, lead poisoning, head trauma,
lar again, leaving the more rational approaches behind. vitamin B6 deficiency, hypoglycemia, and labor, inter
However, with the low success of the combinatorial alia. The biochemical mechanism leading to CNS elec-
chemistry approaches, the more rational approaches are trical discharges and epilepsy is unknown, but there may
finding renewed life. In the next section, several mech- be multiple mechanisms involved. However, it has been
anism-based inactivators are discussed, first in terms of shown that convulsions arise when there is an imbalance
the medicinal relevance of their target enzyme, and then in two principal neurotransmitters in the brain, l-glutamic
from a mechanistic point of view. The reason for the rela- acid, an excitatory neurotransmitter, and GABA, an inhib-
tively large number of examples of this type of enzyme itory neurotransmitter. The concentrations of l-glutamic
inhibitor discussed below, despite the lesser prominence acid and GABA are regulated by two PLP-dependent
of this approach in drug design, derives from our interest enzymes, l-glutamic acid decarboxylase (GAD), which
in organic mechanisms and the history of these drugs. converts glutamate to GABA, and GABA-AT, which
Chapter | 5 Enzyme Inhibition and Inactivation 249

CO2–
GAD + GABA-AT
+ H3 N CO2– OHC
H3N CO2– PLP CO2–
-CO2
PLP PMP
NADP+

SSADH
L-Glu α-KG
NADPH

–O
2C
CO2–
SCHEME 5.21 Metabolism of l-glutamic acid. SSADH, is semialdehyde dehydrogenase

degrades GABA to succinic semialdehyde with the regen- Scheme 5.22; Sabril). This is the first rationally designed
eration of glutamate (Scheme 5.21). Although succinic mechanism-based inactivator drug. The proposed inactiva-
semialdehyde is toxic to cells, there is no buildup of this tion mechanism is shown in Scheme 5.22. By compari-
metabolite, because it is efficiently oxidized to succinic son of Scheme 5.22 with Scheme 4.18 in Chapter 4 (in
acid by the enzyme succinic semialdehyde dehydroge- Scheme 4.18 an α-amino acid is used, but here a γ-amino
nase. GABA system dysfunction has been implicated in acid is the example), it is apparent that identical mecha-
the symptoms associated with epilepsy, Huntington’s dis- nisms are proposed up to compound 5.119 (Scheme 5.22)
ease, Parkinson’s disease, and tardive dyskinesia. When and compound 4.27 (Scheme 4.18). In the case of normal
the concentration of GABA diminishes below a threshold substrate turnover, hydrolysis of 4.27 gives pyridoxamine
level in the brain, convulsions begin. If a convulsion is 5′-phosphate (PMP) and the α-keto acid (Scheme 4.18).
induced in an animal, and GABA is injected directly into The same hydrolysis could occur with 5.119 to give the
the brain, the convulsions cease. It would seem, then, that corresponding products, PMP and 5.121. However, 5.119
an ideal anticonvulsant agent would be GABA; however, is a potent electrophile, a Michael acceptor, which can
peripheral administration of GABA produces no anticon- undergo conjugate addition by an active site nucleophile
vulsant effect. This was shown to be the result of the fail- (X−) and produce inactivated enzyme (5.120a or 5.120b).
ure of GABA, under normal circumstances, to cross the The active site nucleophile was identified as Lys-329, the
blood–brain barrier, a membrane that surrounds the cap- lysine residue that holds the PLP at the active site.[234] The
illaries of the circulatory system in the brain and protects mechanism in Scheme 5.22 (pathway a) appears to be rel-
it from passive diffusion of undesirable (generally hydro- evant for about 70% of the inactivation.[235] The other 30%
philic) chemicals from the bloodstream. Another approach of the inactivation is accounted for by an allylic isomeriza-
for increasing the brain GABA concentration, however, tion and enamine rearrangement leading to 5.122 (path-
would be to design a compound capable of permeat- way b). Note that vigabatrin is an unreactive compound
ing the blood–brain barrier that subsequently inactivates that is converted by the normal catalytic mechanism of the
GABA-AT, the enzyme that catalyzes the degradation of target enzyme into a reactive compound (5.119), which
GABA. Provided that GAD is also not inhibited, GABA attaches to the enzyme. This is the typical course of events
concentrations should rise. This, in fact, has been shown for a mechanism-based inactivator.
to be an effective approach to the design of anticonvulsant It may seem strange that GABA does not cross the
agents.[231] Compounds that both cross the blood–brain blood–brain barrier, but vigabatrin, which is also a small
barrier and inhibit GABA-AT in vitro have been reported charged molecule, can diffuse through that lipophilic
to increase whole brain GABA levels in vivo and possess membrane (albeit poorly). The attachment of a vinyl sub-
anticonvulsant activity. The anticonvulsant effect does not stituent to GABA, apparently, has two effects that permit
correlate with whole brain GABA levels, but it does cor- this compound to cross the blood–brain barrier. First, the
relate with an increase in the GABA concentration at the vinyl substituent increases the lipophilicity of the mole-
nerve terminals of the substantia nigra.[232] cule. Second, it is an electron-withdrawing substituent that
The mechanism for the PLP-dependent aminotransfer- would have the effect of lowering the pKa of the amino
ase reactions was discussed earlier (see Chapter 4, Section group. This would increase the concentration of the non-
4.3.1.3). On the basis of this mechanism, researchers at zwitterionic form (5.123b, Scheme 5.23), which is more
the former Merrell Dow Pharmaceuticals (now Sanofi)[233] lipophilic than the zwitterionic form (5.123a) because it
designed 4-amino-5-hexenoic acid (vigabatrin; 5.117, is uncharged.
250 The Organic Chemistry of Drug Design and Drug Action

COO- COO-
H
NH N

+
NH2 =O OH :NH2
OH 3PO
=O
3PO
N+
N+ H
H
5.122
COO-
COO-

.. + H +
NH N
.. NH2
NH2

OH OH
=
=O O3PO
3PO

N+ N+
H H

:B
Lys b COO-
HN + COO-
b
H2 N COO- H +B
H
OH a
=
O3PO 5.117 a + N+
NH
H
N+ a OH
H =O PO
OH =O PO
3
3
N
N+
H
H
5.118

X
X COO- X COO- COO-
COO-
+ NH NH N+ H2O
H
OH O
=O OH =O PO OH =
3PO 3 O3PO 5.121
N+ N+ N+
H
+ PMP
H H
5.120b 5.120a 5.119
SCHEME 5.22 Hypothetical mechanism for the inactivation of GABA aminotransferase by vigabatrin

5.3.3.3.2. Eflornithine, an Antiprotozoal Drug and

Beyond
+
H3N COO– H2N COOH The polyamines, spermidine (5.124) and spermine (5.125),
and their precursor, putrescine (5.126) are important regula-
5.123a 5.123b
tors of cell growth, division, and differentiation. The mech-
SCHEME 5.23 Zwitterionic and nonzwitterionic forms of vigabatrin
anisms by which they do this are unclear, but they appear to be
Chapter | 5 Enzyme Inhibition and Inactivation 251

H Spermidine (5.124) is produced by the spermidine synthase-

N NH2 catalyzed reaction of putrescine with S-adenosylhomocys-
H 2N
5.124 teamine (5.128), which is derived from the decarboxylation
of SAM (5.127) in a reaction catalyzed by SAM decarbox-
ylase. Another aminopropyltransferase, namely, spermine
H
synthase, catalyzes the reaction of spermidine with 5.128 to
N NH2
H2N N produce spermine (5.125).
H Because polyamines are important for rapid cell
5.125 growth, inhibition of polyamine biosynthesis should
be an effective approach for the design of antitumor and
NH2 antimicrobial agents. In fact, inactivation of ornithine
H 2N
decarboxylase by the potent mechanism-based inactivator
5.126 α-difluoromethylornithine eflornithine, 5.129, Ornidyl),
results in virtually complete reduction in the putrescine and
required for DNA synthesis. Rapidly growing cells have much spermidine content; however, the spermine concentration is
higher levels of polyamines (and ornithine decarboxylase; see only slightly affected. There are at least two reasons for this lat-
below) than do slowly growing or quiescent cells. When qui- ter observation. Inactivation of ornithine decarboxylase activity
escent cells are stimulated, the polyamine and ornithine decar- by eflornithine in vivo may not be complete because gene syn-
boxylase levels increase prior to an increase in the levels of thesis of ornithine decarboxylase has a half-life of only 30 min,
DNA, RNA, and protein. The polycationic nature of the poly- and, also, the endogenous ornithine that is present competi-
amines may be responsible for their interaction with cellular tively protects the enzyme from the inactivator. Furthermore,
structures that have negatively charged groups such as DNA. inactivation of ornithine decarboxylase induces an increase in
Ornithine decarboxylase, the enzyme that catalyzes the the levels of S-adenosylmethionine decarboxylase, which leads
conversion of ornithine to putrescine (5.126), is the rate- to higher levels of 5.128. This can be efficiently used to drive
limiting step in polyamine biosynthesis (Scheme 5.24). the existing putrescine and spermidine to spermine.[236] The

COOH Ornithine
decarboxylase NH2
NH2 H2 N ••
H 2N -CO2
Ornithine 5.126

COOH
+ +
H2N S SAM H 2N S
O Ad O Ad
Me decarboxylase Me
-CO2
HO OH HO OH
5.128
5.127

Spermidine synthase

H
N NH2
H2N
••
5.124

+
H 2N S
O Ad
Me Spermine synthase
H
N NH2
H2N N HO OH
H
5.125
SCHEME 5.24 Polyamine biosynthesis
252 The Organic Chemistry of Drug Design and Drug Action

inability of eflornithine to shut down all polyamine biosynthe- of unwanted facial hair in women (SkinMedica in the United
sis may be responsible for its discouragingly poor antitumor States and Vaniqa in Europe). The mechanism of action is
effects observed in clinical trials.[237] However, eflornithine not known, but it could involve inhibition of polyamine bio-
has been found to have great value in the treatment of certain synthesis important to hair growth.
protozoal infestations such as Trypanosoma brucei gambiense The mechanism for the PLP-dependent decarboxylases
(the more virulent strain, Trypanosoma brucei rhodesiense, is was discussed in Chapter 4 (see Section 4.3.1.2). Ornithine
resistant), which causes African sleeping sickness (300,000 decarboxylase catalyzes the decarboxylation of eflornithine
deaths per year)[238] and Pneumocystis carinii, the microorgan- (5.129), producing a reactive product that inactivates the
ism that produces pneumonia in AIDS patients.[239] Unfortu- enzyme; a possible inactivation mechanism is shown in
nately, large amounts of the compound (150 mg/kg) are needed Scheme 5.25.
every 4–6 h, which makes it very expensive, especially in Third According to the mechanism for PLP-dependent
World countries, where the disease is prevalent. decarboxylases (see Chapter 4, Scheme 4.16), the only
– OOC irreversible step is the one where CO2 is released. There-
CHF2
NH2
fore, if you incubate the decarboxylase with the prod-
H 2N uct amine, it catalyzes the reverse reaction up to the
Eflornithine loss of CO2, that is, the removal of the α-proton. This
5.129 is the microscopic reverse of the reaction that occurs
Another indication has been discovered for eflornithine once the CO2 is released. The principle of microscopic
hydrochloride, namely, as a topical cream for the reduction reversibility states that for a reversible reaction, the

H2N

-
O
Lys
O CHF2
HN+ -OOC NH+
CHF2
NH2
=O OH H2N =
OH
3PO O3PO
5.129
+N N+
H H - CO2

- CH CHF2 CHF2
H2N H2N H2N
F
NH NH NH
+ - + +
=O OH =O PO OH =O OH
3PO 3 3PO
..
+N +N N
H H H
5.130

-F-

X- X X
F
CHF C
H2N H2N H H2N
N+ NH + - F- N+
H H
=O PO
OH OH OH
3 =O =
3PO O3PO
N+ ..
N +N
H H H
SCHEME 5.25 Hypothetical mechanism for inactivation of ornithine decarboxylase by eflornithine
Chapter | 5 Enzyme Inhibition and Inactivation 253

H2N

+ B:
Lys H
CHF2
HN+ H CHF2 NH+
NH2
OH H2N =
OH
=
O3PO O3PO
5.131
+N N+
H H

H2N
- CH
H2N
CHF2
H2N
CHF2
F
NH NH NH
+ - + +
=O
OH =O PO
OH OH
=O PO
3PO 3 3
..
+N +N N
H H H
5.130

-F-

Inactivation as in Scheme 5.25

SCHEME 5.26 Hypothetical mechanism for inactivation of ornithine decarboxylase by α-difluoromethylputrescine

same mechanistic pathway will be followed in the for- et al.[242] showed that iproniazid was a potent inhibitor
ward and reverse reactions. Therefore, a mechanism- of MAO, and clinical studies were underway in the late
based inactivator that has a structure similar to that of 1950s.[243]
the product of the ornithine decarboxylase reaction (the O
amine) should undergo catalytic α-deprotonation, which N CNHNHCH(CH3)2
then would produce the same set of resonance struc-
tures produced by decarboxylation of the ornithine ana- Iproniazid
log. α-Difluoromethylputrescine (5.131, Scheme 5.26) 5.132
was shown to inactivate ornithine decarboxylase,[240] The brain concentrations of various biogenic (pressor)
presumably because deprotonation gives the same amines such as norepinephrine, serotonin, and dopamine
intermediate that is obtained by decarboxylation of eflo- were found to be depleted in chronically depressed indi-
rnithine (compare 5.130 in Scheme 5.25 with 5.130 in viduals. A correlation was observed between an increase in
Scheme 5.26). the concentrations of these brain biogenic amines and the
For any target enzyme that is reversible, mechanism- onset of an antidepressant effect.[244] This was believed to
based inactivators could be designed for both the forward be the result of MAO inhibition, because MAO is one of
(substratelike) and back (productlike) reactions. Depending the enzymes responsible for the catabolism of these bio-
on the metabolic pathway involved, enzyme selectivity may genic amines. By the early 1960s several MAO inactivators
be more favorable for one over the other. were being used clinically for the treatment of depression.
Unfortunately, it was found that in some cases there was a
5.3.3.3.3. Tranylcypromine, an Antidepressant Drug cardiovascular side effect that led to the deaths of several
The modern era of therapeutics for the treatment of patients. Consequently, these drugs were withdrawn from
depression began in the late 1950s with the introduction the drug market until the cause of death could be ascer-
of both the monoamine oxidase (MAO) inhibitors and the tained. Within a few months the problem was understood.
tricyclic antidepressants. The first MAO inhibitor was It was determined that all of those who died while taking
iproniazid (5.132), which initially was used as an antitu- an MAO inhibitor had two things in common: they had all
berculosis drug until it was observed that patients taking died from a hypertensive crisis and, prior to their deaths,
it exhibited excitement and euphoria.[241] In 1952, Zeller they had eaten foods containing high tyramine content (e.g.,
254 The Organic Chemistry of Drug Design and Drug Action

Ph Fl Fl Fl FlH– S–
Ph .
NH Ph NH2 Ph NH2
.. 2 NH2 + +
H
5.133

S S
+H+
Ph NH2 Ph NH2
+
SCHEME 5.27 Hypothetical mechanism for the inactivation of monoamine oxidase by tranylcypromine

cheese, wine, beer, and chocolate). The connection between first MAO inactivators approved for clinical use; many other
these observations is that the ingested tyramine triggers the cyclopropylamine analogs show antidepressant activity.[248]
release of norepinephrine, a potent vasoconstrictor, which The one-electron mechanism for MAO, a flavoenzyme, was
raises the blood pressure. Under normal conditions, the discussed in Chapter 4 (Section 4.3.3.3; Scheme 4.32). If
excess norepinephrine is degraded by MAO and catechol- tranylcypromine acts as a substrate, and one-electron trans-
amine O-methyltransferase. If the MAO is inactivated, then fer from the amino group to the flavin occurs, the result-
the norepinephrine is not degraded fast enough, the blood ing cyclopropylaminyl radical will undergo rapid cleavage
pressure keeps rising, and this can lead to a hypertensive (Scheme 5.27),[249] and the benzylic radical produced could
crisis. This series of events has been termed the cheese combine with an active site radical[250]. However, it was
effect because of the high tyramine content found in many suggested[251] that an adduct with an active site cysteine
cheeses. Because the MAO inhibitors were not toxic, except may have formed. In Scheme 5.27, a mechanism for forma-
when taken with certain foods, these drugs were allowed tion of a cysteine adduct is shown.
to return to the drug market, but they were prescribed with
strict dietary regulations. Because of this inconvenience,
and the discovery of the tricyclic antidepressants (which H
block the reuptake of biogenic amines at the nerve termi-
nals), MAO inhibitors are not the drugs of choice, except H NH2
in those types of depression that do not respond to tricyclic Tranylcypromine
antidepressants or when treating phobic anxiety disorders, 5.133
which respond well to MAO inhibitors.
The resurgence of interest of the pharmaceutical indus-
try in MAO inhibitors in the late 1980 and 1990s[245] came 5.3.3.3.4. Selegiline (l-Deprenyl) and Rasagiline:
because MAO exists in at least two isozymic forms,[246] Antiparkinsonian Drugs
termed MAO A and MAO B. The main difference in these Parkinson’s disease, the second most common neurodegenera-
two isozymes is their selectivity for the oxidation of the tive disease, afflicting more than one-half million people in the
various biogenic amines. Because the antidepressant United States, is characterized by chronic, progressive motor
effect is related to increased concentrations of brain sero- dysfunction resulting in severe tremors, rigidity, and akinesia.
tonin and norepinephrine, both of which are MAO A sub- The symptoms of Parkinson’s disease arise from the degen-
strates, compounds that selectively inhibit MAO A possess eration of dopaminergic neurons in the substantia nigra and a
antidepressant activity; selective inhibitors of MAO B marked reduction in the concentration of the pyridoxal 5′-phos-
show potent antiparkinsonian properties.[247] To have an phate-dependent aromatic l-amino acid decarboxylase, the
antidepressant drug without the cheese effect, however, it enzyme that catalyzes the conversion of l-dopa to the inhibi-
is necessary to inhibit brain MAO A selectively without tory neurotransmitter dopamine. Because dopamine is metabo-
inhibition of peripheral MAO A, particularly MAO A in lized primarily by MAO B in man (see Section 5.3.3.3.3), and
the gastrointestinal tract and sympathetic nerve terminals. Parkinson’s disease is characterized by a reduction in the brain
Inhibition of brain MAO A increases the brain serotonin dopamine concentration (see Chapter 9, Section 9.2.2.10),
and norepinephrine concentrations, which leads to the selective inactivation of MAO B has been shown to be an
antidepressant effect; the peripheral MAO A must remain effective approach to increase the dopamine concentration
active to degrade the peripheral tyramine and norepineph- and, thereby, treat this disease. Actually, an MAO B-selective
rine that cause the undesirable cardiovascular effects. inactivator is used in combination with the antiparkinsonian
Tranylcypromine (5.133; Parnate), a nonselective MAO drug l-dopa (see Chapter 9, Section 9.2.2.10). Selective inhibi-
A/B inactivator that exhibits a cheese effect, was one of the tion of MAO B does not interfere with the MAO A-catalyzed
Chapter | 5 Enzyme Inhibition and Inactivation 255

degradation of tyramine and norepinephrine; therefore, no car- rats, so it was thought that MPTP was not responsible for
diovascular side effects (the cheese effect; see previous section) the neurological effects. However, when the young California
are observed with the use of an MAO B-selective inactivator. drug addicts were observed to have similar symptoms, tests
The earliest MAO B-selective inactivator, selegiline (5.134; in primates[256] and mice[257] showed that MPTP produced
Eldepryl),[252] was approved in 1989 by the FDA for the treat- the same neurological symptoms and histological changes in
ment of Parkinson’s disease in the United States; in 2006, the substantia nigra as those observed with idiopathic par-
another MAO B-selective inactivator, rasagiline (5.135; kinsonism. Rats, however, it is now known, are remarkably
Azilect),[253] was approved. resistant to the neurotoxic effects of MPTP, which explains
why the earlier tests in rats were negative. The observation
CH3
that an industrial chemist developed Parkinson’s disease after
N
synthesizing large amounts of MPTP as a starting material
H CH3 led to the suggestion that cutaneous absorption or vapor inha-
HN
lation may be significant pathways for introduction of the
Selegiline
neurotoxin. Because of this, it can be hypothesized that Par-
Rasagiline
5.134 5.135
kinson’s disease is, at least partially, an environmental disease
Little was known about the cause for this disease until arising from long-term slow degeneration of dopaminergic
1976, when a previously healthy 23-year-old man, who was neurons by ingested or inhaled neurotoxins similar to MPTP.
a chronic street drug user, was referred to the National Insti- Because the symptoms of Parkinson’s disease do not appear
tute of Mental Health for investigation of symptoms of what until 60–80% of the dopaminergic neurons are destroyed, it
appeared to be Parkinson’s disease.[254] Although the patient is reasonable that this disease is associated with the elderly.
responded favorably to the usual treatment for Parkinson’s Opponents of this hypothesis note that the interregional and
disease (l-dopa/carbidopa; see Chapter 9), the speed with subregional patterns of striatal dopamine loss by MPTP differ
which, and the age at which, the symptoms developed from those of idiopathic Parkinson’s disease.[258] However,
were inconsistent with the paradigm of Parkinson’s disease results of surveys taken throughout the world suggest that
as a regressive geriatric disease. When this man later died environmental toxins may have an important role in the etiol-
of a drug overdose, an autopsy showed the same extensive ogy of Parkinson’s disease.[259] Furthermore, it is interesting
destruction of the substantia nigra that is found in idiopathic that many frequently used medicines have structures related
parkinsonism. In 1982, seven young Californians who had to MPTP; some, particularly neuroleptics, produce parkinso-
tried some “synthetic heroin” also developed symptoms of nian side effects.[260] Also, genetic factors do not appear to be
an advanced case of Parkinson’s disease, including near important in most cases of Parkinson’s disease.
total immobility.[255] It was found that the drug they had
been taking was a designer drug, a synthetic narcotic that Me
has a structure designed to be a variation of an existing con- N
Me
trolled drug; at that time, because the new structure was not N
listed as a controlled substance, it was not illegal to possess Ph
O
O
it (this law has since been changed to include variants of O
controlled substances). Meperidine
Ph O
In this case, the “designers” of the street drug were 5.136 5.137
using the controlled analgesic meperidine (5.136; Demerol)
as the basis for their structure modification. The compound Me
synthesized, 1-methyl-4-phenyl-4-propionoxypiperidine N
(5.137), was referred to as a “reverse ester” of meperidine
(note the ester oxygen in 5.137 is on the opposite side of
the carbonyl from that in 5.136). However, by “reversing” Ph
the ester, the drug designers converted a stable ethoxycar- 5.138
bonyl group of meperidine to a propionoxy group, a good
leaving group, in 5.137. In fact, 5.137 decomposes on heat- Once a connection was made between MPTP and Par-
ing or in the presence of acids with elimination of propio- kinson’s disease, a vast amount of research with MPTP was
nate to give 1-methyl-4-phenyl-1,2,5,6-tetrahydropyridine initiated, and it soon was realized that the neurotoxic agent
(MPTP; 5.138). Analysis of several samples of the designer was actually a metabolite of MPTP. Pretreatment of ani-
drug (called “new heroin”) revealed the MPTP contamina- mals with the MAO B-selective inactivator selegiline was
tion. This same contamination was identified in the samples shown to protect the animals from the neurotoxic effects
of drugs taken by the 23-year-old in 1976, but the drug sam- (both the disease symptoms and the damage to the substan-
ples at that time were found to exhibit no neurotoxicity in tia nigra) of MPTP, whereas the MAO-A selective agent
256 The Organic Chemistry of Drug Design and Drug Action

clorgyline (5.139) did not.[261] This indicated that the neu- It is apparent, then, that, by inactivation of MAO B, sele-
rotoxic metabolite was generated by MAO B oxidation of giline is important both for the prevention of the oxidation
MPTP; the lower levels of MAO B in rat brain could explain of neurotoxin precursors such as MPTP and for preventing
the negative effect on rats when the street drug was initially the degradation of dopamine. The mechanism of inactiva-
tested.[262] The two metabolites produced by MAO B are tion of MAO by a simpler analog of selegiline and rasagi-
1-methyl-4-phenyl-2,3-dihydropyridinium ion (5.140) and line, 3-dimethylamino-1-propyne (5.142), has been studied;
1-methyl-4-phenylpyridinium ion (MPP+; 5.141).[263] The this compound becomes attached to the N5 position of the fla-
latter compound (5.141) accumulates in selected areas in the vin.[267] Because of the evidence for a one-electron mechanism
brain, and therefore, is believed to be the actual neurotoxic for MAO-catalyzed oxidations (see Chapter 4, Section 4.3.3.3),
agent.[264] Because MPTP is a neutral molecule, it can cross the inactivation mechanisms for selegiline and rasagiline,[268]
the blood–brain barrier and enter the brain; once it is oxidized shown in Scheme 5.28 for rasagiline, seem most reasonable.
by MAO B, the pyridinium ion (5.141) cannot diffuse out of The covalent N5-adduct depicted in Scheme 5.28 was con-
the brain. Selective toxicity of MPTP appears to be the result firmed by X-ray crystallography.[269] Rasagaline, which has a
of the transport of MPP+, but not MPTP, into dopamine neu- very similar pharmacological action to that of selegiline,[270] is
rons via an amine uptake system.[265] These studies suggest 10 times more potent than selegiline and is not metabolized to
that Parkinson’s disease may be more than just a geriatric l-amphetamine or l-methampethamine as is selegiline, which
disease; it may be caused by molecules in the environment, could account for selegiline’s sympathomimetic activity.[271]
possibly pesticides, that are consumed and metabolized to
CH3
molecules that are neurotoxic.[266]
N
CH3 H3 C
Cl
O N 5.142

Cl 5.3.3.3.5. 5-Fluoro-2′-deoxyuridylate, Floxuridine,

5.139 and 5-Fluorouracil: Antitumor Drugs
Me Me Cancer is a family of diseases characterized by abnormal and
+N +N uncontrolled cell division. Neither the etiology nor the way
in which it causes death is fully understood in most cases.
One important approach to antineoplastic (antitumor) agents
Ph Ph is the design of an antimetabolite (see Section 5.2.2.3.3),
5.140 5.141 whose structure is related to those of pyrimidines and purines

R R R
N N O N N O N N O
S S S
NH NH NH
N N N
•
O O– •
O–
H :B
N
N C N
H N
H H
H
Fl
Fl

FlH-
R R
N N O N N O FlH-
S
+H+ S
NH NH N
N N
O– H
O–
C
H
N
N H

SCHEME 5.28 Hypothetical mechanism for the inactivation of monoamine oxidase by selegiline
Chapter | 5 Enzyme Inhibition and Inactivation 257

involved in the biosynthesis of DNA. These compounds inter- biosynthesis of thymidylate, namely, the conversion of
fere with the formation or utilization of one of these essen- 2′-deoxyuridylate to 2′-deoxythymidylate (referred to as
tial normal cellular metabolites. This interference generally just thymidylate). The reaction catalyzed by thymidylate
results from the inhibition of an enzyme in the biosynthetic synthase is the only de novo source of thymidylate, which
pathway of the metabolite or from incorporation, as a false is an essential constituent of the DNA. Therefore, inhibition
building block, into vital macromolecules such as proteins of thymidylate synthase in tumor cells inhibits DNA bio-
and polynucleotides. Antimetabolites usually are obtained by synthesis and produces what is known as thymineless death
making a small structural change in the metabolite, such as of the cell.[273] Unfortunately, normal cells also require
a bioisosteric interchange (see Chapter 2, Section 2.2.4.3). thymidylate synthase for de novo synthesis of their thymi-
5-Fluorouracil (5.143, one name for systemic use is dylate. Nonetheless, inhibitors of thymidylate synthase are
Adrucil; for topical use is Efudex), its 2′-deoxyribonucle- effective antineoplastic agents. There are several reasons for
oside, floxuridine (5.144, FUDR), and its 2′-deoxyribo- this selective toxicity against tumor cells; all are related to
nucleotide, 5-fluoro-2′-deoxyuridylate (5.145), are potent the difference in the rates of cell division for normal and
antimetabolites of uracil and its congeners, and are also abnormal cells. Because aberrant cells replicate much more
potent antineoplastic agents. 5-Fluorouracil itself is not rapidly than do most normal cells, the rapidly proliferating
active, but it is converted in vivo to the 2′-deoxynucleotide tumor cells have a higher requirement for their DNA (and
(5.145), which is the active form. Fluorine is often used DNA precursors) than do the slower proliferating normal
as a replacement for hydrogen in medicinal chemistry[272] cells. This means that the activity of thymidylate synthase
(the van der Waals radius of fluorine is 1.35 Å vs 1.20 Å for is elevated in tumor cells relative to normal cells. Because
hydrogen), and 5-fluorouracil and its metabolites are recog- uracil is one of the precursors of thymidylate, it and 5-fluo-
nized by enzymes that act on uracil and its metabolites. There rouracil are taken up into tumor cells much more efficiently
are several pathways for this in vivo activation (Scheme than into normal cells. Finally, and possibly most impor-
5.29). A minor pathway to intermediate 5-fluoruouridylate tantly, enzymes that degrade uracil in normal cells also
(5.147) begins with the conversion of 5-fluorouracil (5.143) degrade 5-fluorouracil, and these degradation processes do
to 5-fluorouridine (5.146), catalyzed by ribose-1-phosphate not take place in cancer cells.[274] The adverse side effects
uridine phosphorylase, followed by further conversion accompanying the use of 5-fluorouracil in humans gener-
of 5.146 to 5.147, catalyzed by uridine kinase. The major ally arise from the inhibition of thymidylate synthase and
pathway to 5.147 is the direct conversion of 5-fluorouracil, destruction of the rapidly proliferating normal cells of the
which is catalyzed by orotate phosphoribosyltransferase. intestines, the bone marrow, and the mucosa. The effects of
5-Fluoro-2′-deoxyuridylate (5.145) is produced from 5.147 anticancer drugs are discussed in more detail in Chapter 6.
by the circuitous route shown in Scheme 5.29 or by direct Unlike other tetrahydrofolate-dependent enzymes (see
conversion of 5.143 to its 2′-deoxyribonucleoside, floxuri- Chapter 4, Section 4.3.2), thymidylate synthase utilizes
dine (5.144), catalyzed by uridine phosphorylase, followed methylenetetrahydrofolate as both a one-carbon donor and
by 5′-phosphorylation, which is catalyzed by thymidine as a reducing agent (Scheme 5.30).[275] An active site cyste-
kinase (Scheme 5.29). However, when 5.144 is administered ine residue undergoes Michael addition to the 6-position of
rapidly, it is converted back to 5.143 faster than it is phos- 2′-deoxyuridylate (5.148, dRP is deoxyribose phosphate) to
phorylated to 5.145. Under these circumstances, attempts give an enolate (5.149), which attacks 5,10-methylenetetra-
to use floxuridine to bypass the long metabolic route for hydrofolate (more likely, in one of the more reactive open,
conversion of 5.143 to 5.145 are unsuccessful. Continuous iminium forms; see Chapter 4, Scheme 4.23), and forms a
intra-arterial infusion of floxuridine, however, enhances the ternary complex (5.150) of the enzyme, the substrate, and
direct conversion of 5.144 to 5.145. the coenzyme. Enzyme-catalyzed removal of the C-5 proton
leads to β-elimination of tetrahydrofolate (5.151). Oxidation
O O of the tetrahydrofolate (a hydride mechanism is shown, but a
F F one-electron mechanism, that is, first transfer of an N-5 nitro-
HN HN
gen nonbonded electron to the alkene followed by hydrogen
O N O N
O =
atom transfer, is possible) gives dihydrofolate (5.152) and the
HO O3PO enzyme-bound thymidylate enolate (5.153). Reversal of the
F O O
HN
first step releases the active site cysteine residue and produces
O N thymidylate (5.154). This reaction changes the oxidation state
H HO HO of the coenzyme. Because of this, another enzyme, dihydrofo-
5.143 5.144 5.145 late reductase, is required to reduce the dihydrofolate back to
tetrahydrofolate (see Chapter 4, Scheme 4.21).
The principal site of action of 5.145 is thymidylate syn- 5-Fluoro-2′-deoxyuridylate (5.145) inactivates thymi-
thase, the enzyme that catalyzes the last step in de novo dylate synthase because once the ternary complex (5.155)
258 The Organic Chemistry of Drug Design and Drug Action

O
O O O
F ATP ADP ATP ADP
F HN F F
HN Ribose-1-P HN HN
Uridine O N Uridine kinase UMP kinase
O N phosphorylase O N O O N
H
HO O = =O POPO
O3PO O 3 O
5.143
O–
HO OH HO OH HO OH
Uridine phosphorylase

dR-1-P 5.146 5.147

PRPP
Pi Orotate phosphoribosyltransferase Ribonucleotide
reductase

O O O
F F
HN ATP ADP F
HN HN
O N
Thymidine kinase O N O N
HO O
O = =O POPO
O3PO O 3 O
O–
HO
HO HO
5.144 5.145
di

se
ph

na
ATP
dU hoh
o

ki
sp
TP yd

P
PPi

D
dU
ro

O ADP
la

H2 O
se

F
HN

O O O N
=O
3POP-O-PO O
O – O–

HO
SCHEME 5.29 Metabolism of 5-fluorouracil

forms, there is no C-5 proton that the enzyme can remove 5.4. GENERAL REFERENCES
to eliminate the tetrahydrofolate (Scheme 5.31).[276] Conse-
quently, the enzyme remains as the ternary complex. Copeland, R. A. Evaluation of Enzyme Inhibitors in
Note that in this mechanism the inactivator is not con- Drug Discovery: A Guide for Medicinal Chemists and
verted into a reactive compound that attaches to the enzyme. Pharmacologists, 2nd Edition, John Wiley & Sons:
Instead, it first attaches to the enzyme, then requires con- Hoboken, NJ, 2013.
densation with 5,10-methylenetetrahydrofolate to generate Lu, C.; Li, A. P. Enzyme Inhibition in Drug Discovery and
a stable complex. A mechanism-based inactivator, there- Development. The Good and the Bad, John Wiley & Sons:
fore, does not necessarily require formation of a reactive Hoboken, NJ, 2010.
species. It only requires the enzyme to catalyze a reaction
on it that leads to inactivation prior to release of the product. Sulfonamides
If the enzyme were inactivated without the requirement of Weidner-Wells, M. A.; Macielag, M. J. Sulfonamides, Kirk-
5,10-methylenetetrahydrofolate, that is, by simple Michael Othmer Encyclopedia of Chemical Technology, 5th ed.,
addition of the active site cysteine to the C-6 position of the 2007, 23, 493–513.
inactivator, then 5.145 would be an affinity labeling agent. Dibbern, D. A., Jr.; Montanaro, A. Allergies to sulfonamide
Next, in Chapter 6, we will consider drug interactions antibiotics and sulfur-containing drugs. Ann. Allergy, Asthma,
with another type of receptor, DNA. Immunol. 2008, 100, 91–100.
Chapter | 5 Enzyme Inhibition and Inactivation 259

H H H
H2N N N H 2N N N H2 N N N
N N N
N N N
HO NR HO NHR HO NHR
+
O H B O O
- :B
HN HN HN H

O N – O N S O N S
S dRP
dRP dRP
5.150
5.148 5.149

H
H H2N N N
O H 2N N N
- CH3
HN N ••
+ N N
N
HO H H
O N S HO NHR
NHR O
dRP 5.152
5.153 HN

O N S
dRP
O 5.151
CH3
HN

O N –
S
dRP
5.154
SCHEME 5.30 Hypothetical mechanism for thymidylate synthase (dRP is deoxyribose phosphate)

H H
H2N N N H 2N N N H
H2 N N N
N N
N N N
N
HO NR HO NHR
+ HO NHR
O H B O O
F - F
HN :B
HN HN F

O N - O N S O N S
S
dRP dRP dRP
5.145 5.155
SCHEME 5.31 Hypothetical mechanism for the inactivation of thymidylate synthase by 5-fluoro-2′-deoxyuridylate

Supuran, C. T. Diuretics: from classical carbonic anhydrase diabetes and its complications. Curr. Med. Chem. 2012, 19,
inhibitors to novel applications of the sulfonamides. Curr. 3578–3604.
Pharmaceut. Des. 2008, 14, 641–648.
Kalgutkar, A. S.; Jones, R.; Sawant, A. Sulfonamide as Slow, Tight-Binding Inhibitors
an essential functional group in drug design. RSC Drug Silverman, R. B. Enzyme inhibition. Wiley Encyclopedia
Discovery Series 2010, 1, 210–274. Chem. Biol. 2009, 1, 663–681.
Chen, X.; Hussain, S.; Parveen, S.; Zhang, S.; Yang, Szedlacsek, S. E.; Duggleby, R. G. Kinetics of slow and
Y.; Zhu, C. Sulfonyl group-containing compounds tight-binding inhibitors. Meth. Enzymol. 1995, 249,
in the design of potential drugs for the treatment of (Enzyme Kinetics and Mechanism, Part D), 144–180.
260 The Organic Chemistry of Drug Design and Drug Action

Morrison, J. F.; Walsh, C. T. The behavior and significance Mandell, G. L.; Sande, M. A. In Goodman and Gilman’s The
of slow-binding enzyme inhibitors. Adv. Enzymol. 1988, 61, Pharmacological Basis of Therapeutics, 7th ed., Gilman, A.
201–301. G.; Goodman, L. S.; Rall, T. W.; Murad, F., Eds.; Macmillan,
Schloss, J. V. Significance of slow-binding enzyme inhibition New York, 1985; p. 1115.
and its relationship to reaction-intermediate analogs. Acc. Morin, R. B.; Gorman, M. (Eds.), Chemistry and Biology
Chem. Res. 1988, 21, 348–353. of ß-Lactam Antibiotics; Academic: New York, 1982.
Sculley, M. J.; Morrison, J. F. The determination of kinetic
Aspirin
constants governing the slow, tight-binding inhibition of
De Caterina, R.; Renda, G. Clinical use of aspirin in ischemic
enzyme-catalysed reactions. Biochim. Biophys. Acta 1986,
heart disease: past, present, and future. Curr. Pharmaceut.
874, 44–53.
Des. 2012, 18, 5215–5223.
Transition State Analogues Ugurlucan, M.; Caglar, I. M.; Caglar, F. N. T.; Ziyade, S.;
Schramm, V. L.; Tyler, P. C. Transition state analogue inhibi- Karatepe, O.; Yildiz, Y.; Zencirci, E.; Ugurlucan, F. G.;
tors of N-ribosyltransferases. In Iminosugars, Compain, P.; Arslan, A. H.; Korkmaz, S.; et al. Aspirin: from a historical
Martin, O. R. (Eds.), Wiley, Southern Gate, Chichester, 2007, perspective. Recent Patents Cardiovasc. Drug Disc. 2012, 7,
pp. 177–208. 71–76.
Smyth, T. P. Substrate variants versus transition state ana- Thun, M. J.; Jacobs, E. J.; Patrono, C. The role of aspirin
logues as noncovalent reversible enzyme inhibitors. Bioorg. in cancer prevention. Nature Rev. Clin. Oncol. 2012, 9,
Med. Chem. 2004, 12(15), 4081–4088. 259–267.
Schramm, V. L. Enzymic transition states and transi- Lordkipanidze, M. Advances in monitoring aspirin therapy.
tion state analog design. Annu. Rev. Biochem. 1998, 67, Platelets 2012, 23, 526–536.
693–720. Choubey, A. K. Aspirin: a wonder drug. J. Pharm. Res. 2011,
Andrews, P. R.; Winkler, D. A. In Drug Design: Fact or 4, 3803–3805.
Fantasy? Jolles, G.; Wooldridge, K. R. H. (Eds.), Academic: Flower, R. J.; Moncada, S.; Vane, J. R. In Goodman and
London, 1984, p. 145. Gilman’s The Pharmacological Basis of Therapeutics,
Wolfenden, R. Transition state analog inhibitors and enzyme 7th ed., Gilman, A. G.; Goodman, L. S.; Rall, T. W.;
catalysis. Annu. Rev. Biophys. Bioeng. 1976, 5, 271–306. Murad, F. (Eds.), Macmillan, New York, 1985; p. 674.
Wolfenden, R. Transition state analogs as potential affinity Mechanism-Based Enzyme Inactivators
labeling reagents. Meth. Enzymol. 1977, 46, 15–28. Silverman, R. B. Mechanism-Based Enzyme Inactivation:
Multisubstrate Analogues Chemistry and Enzymology; CRC: Boca Raton, FL, 1988,
Le Calvez, P. B.; Scott, C. J.; Migaud, M. E. Multisubstrate Vol. 1–2.
adduct inhibitors: drug design and biological tools. J. Enz. Silverman, R.B. The Potential Use of Mechanism-Based
Inhib. Med. Chem. 2009, 24(6), 1291–1318. Enzyme Inactivators in Medicine. J. Enz. Inhib. 1988, 2,
Radzicka, A.; Wolfenden, R. Transition state and mul- 73–90.
tisubstrate analog inhibitors. Meth. Enzymol. 1995, 249 Silverman, R. B. Mechanism-Based Enzyme Inactivators.
(Enzyme Kinetics and Mechanism, Part D), 284–312. Methods Enzymol. 1995, 249, 240–283.
Page, M. I. Enzyme inhibition In Comprehensive Medicinal Polyamines
Chemistry, Hansch, C.; Sammes, P. G.; Taylor, J. B. (Eds.), Bachmann, A. S.; Levin, V. A. Clinical applications of poly-
Pergamon Press, Oxford, 1990, Vol. 2, pp. 61–87. amine-based therapeutics. RSC Drug Disc. Series 2012, 17
Broom, A. D. Rational design of enzyme inhibitors: mul- (Polyamine Drug Discovery), 257–276.
tisubstrate analogue inhibitors. J. Med. Chem. 1989, 32, Huang, Y.; Marton, L. J.; Woster, P. M. The design and
2–7. development of polyamine-based analogs with epigenetic
Covalent Inhibitors targets. RSC Drug Disc. Series 2012, 17 (Polyamine Drug
Mehdi, S. Covalent enzyme inhibition in drug discovery and Discovery), 238–256.
development, In Enzyme Technologies: Pluripotent Players Goodwin, A. C.; Murray-Stewart, T. R.; Casero, R. A., Jr.
in Discovering Therapeutic Agents. Yang, H.-C.; Yeh, W.-K.; Targeting the polyamine catabolic enzymes spermine oxi-
McCarthy, J. R. (Eds.), John Wiley & Sons, Hoboken, NJ, dase, N1-acetylpolyamine oxidase and spermidine/sperm-
2014. ine N1-acetyltransferase. RSC Drug Disc. Series 2012, 17
(Polyamine Drug Discovery), 135–161.
Penicillins and Cephalosporins Pegg, A. E. Inhibitors of polyamine biosynthetic enzymes.
Bush, K.; Mobashery, S. How beta-lactamases have driven RSC Drug Disc. Series 2012, 17 (Polyamine Drug Discovery),
pharmaceutical drug discovery. From mechanistic knowl- 78–103.
edge to clinical circumvention. Adv. Exp. Med. Biol. 1998, Clark, K.; Niemand, J.; Reeksting, S.; Smit, S.;
456, 71–98. Brummelen, A. C.; Williams, M.; Louw, A. I.; Birkholtz, L.
Chapter | 5 Enzyme Inhibition and Inactivation 261

Functional consequences of perturbing polyamine Bush, K.; Macielag, M. J. New β-lactam antibiotics and
metabolism in the malaria parasite, Plasmodium falci- β-lactamase inhibitors. Exp. Opin. Therapeut. Patents 2010,
parum. Amino Acids 2010, 38, 633–644. 20, 1277–1293.
Wallace, H. M. Targeting polyamine metabolism: a viable Cartwright, S. J.; Waley, S. G. Beta-lactamase inhibitors.
therapeutic/preventative solution for cancer? Exp. Opin. Med. Res. Rev. 1983, 3, 341–382.
Pharmacother. 2007, 8, 2109–2116. Knowles, J. R. Penicillin resistance: the chemistry
Heby, O.; Persson, L.; Rentala, M. Targeting the polyamine of β-lactamase inhibition. Acc. Chem. Res. 1985, 18,
biosynthetic enzymes: a promising approach to therapy of 97–104.
African sleeping sickness, Chagas’ disease, and leishmani-
asis. Amino Acids 2007, 33, 359–366.
Tabor, C. W.; Tabor, H. Polyamines. Annu. Rev. Biochem.
1984, 53, 749–790. 5.5. PROBLEMS (ANSWERS CAN BE FOUND
IN THE APPENDIX AT THE END OF THE
Monoamine Oxidase Inhibitors
Poewe, W.; Mahlknecht, P.; Jankovic, J. Emerging therapies
BOOK)
for Parkinson’s disease. Curr. Opin. Neurol. 2012, 25, 448–
1. Why would you want to design a drug that is an enzyme
459.
inhibitor?
Jankovic, J.; Poewe, W. Therapies in Parkinson’s disease.
2. If you wanted to inhibit an enzyme in a microorganism
Curr. Opin. Neurol. 2012, 25, 433–447.
that is also in humans, what approaches would you take
Carradori, S.; Secci, D.; Bolasco, A.; Chimenti, P.;
in your research?
D’Ascenzio, M. Patent-related survey on new monoamine
3. S-Adenosylmethionine (SAM) is biosynthesized from
oxidase inhibitors and therapeutic potential. Exp. Opin.
methionine and ATP, catalyzed by methionine adenosyl
Therapeut. Patents 2012, 22(7), 759–801.
transferase. The mechanism is shown below.

O O O
CH3 CH3
O P O P O P O
– –
OOC OOC
S O Ade S
O O
O Ade
O
+ +
H3 N H3 N
OH OH
OH OH

a.  Design a competitive reversible inhibitor for this

Helguera, A. M.; Perez-Machado, G.; Cordeiro, M. N.
enzyme.
D. S.; Borges, F. Discovery of MAO-B inhibitors-pres-
b.  Design a multi-substrate analogue inhibitor, and
ent status and future directions part I: oxygen heterocy-
show the basis for your design.
cles and analogs. Mini-Rev. Med. Chem. 2012, 12(10),
4. What advantage does a slow, tight-binding inhibitor
907–919.
have over a simple reversible inhibitor?
Flockhart, D. A. Dietary restrictions and drug interactions
5. Thromboxane A2 (TXA2) is biosynthesized in humans
with monoamine oxidase inhibitors: an update. J. Clin.
from prostaglandin H2 (PGH2) by the enzyme throm-
Psych. 2012, 73 (Suppl. 1), 17–24.
boxane A2 synthase. Binding of TXA2 to a receptor
Hoy, S. M.; Keating, G. M. Rasagiline: a review of its use in
causes vasoconstriction (raises the blood pressure) and
the treatment of idiopathic Parkinson’s disease. Drugs 2012,
platelet aggregation; therefore, TXA2 has been impli-
72, 643–669.
cated as a causative factor in a number of cardiovascu-
Schapira, A. H. V. Monoamine oxidase B inhibitors for the
lar and renal diseases.
treatment of Parkinson’s disease: a review of symptomatic
a.  Without looking at part b of this question,
and potential disease-modifying effects. CNS Drugs 2011,
briefly describe three approaches you could take
25, 1061–1071.
regarding TXA2 to design a new antihypertensive
ß-Lactamase Inhibitors agent.
Biondi, S.; Long, S.; Panunzio, M.; Qin, W. L. Current trends b.  Thromboxane A2 synthase inhibitors, such as 2,
in β-lactam based β-lactamase inhibitors. Curr. Med. Chem. have not been successful, possibly because of the
2011, 18, 4223–4236. buildup of PGH2, the substrate for thromboxane
262 The Organic Chemistry of Drug Design and Drug Action

A2 synthase, which is also a potent agonist 8. 5-Lipoxygenase requires a ferric ion as a cofactor.
at the TXA2 receptor. PGH2 is converted to PGI2, Show how 4 acts as an inhibitor of 5-lipoxygenase.
a potent vasodilator and platelet inhibitory agent,
by PGI2 synthase. TXA2 receptor antagonists,
OH
such as 3, block the binding of both PGH2 and H
N N
TXA2 to the TXA2 receptor, but because TXA2
synthase is not inhibited, there is no buildup of O
PGH2 and therefore no conversion to PGI2. What 4
approach would you take to get around these
problems?
9. AG 7088 was designed to inhibit a picornavirus
­protease and is a potent antiviral agent.
NHSO 2
O
NH
N D
N O
O B

CH3 N N
N H H
O N O OEt
E
O
COOH O
COOH A
F
2 3
AG 7088 C

 Why do you think moieties A-E were incorporated into

c. 
Design a molecule that uses the principle you pro-
the structure?
pose in part b.
10.  a. Two isoforms of an enzyme were discovered; iso-
6. Show the transition state for the reaction below, and
form-1 produces a hormone that causes muscle
draw a reasonable transition state analog inhibitor.
spasms and isoform-2 makes another hormone
from the same substrate that lowers cholesterol
levels. What would you do to prevent muscle
:B spasms without raising cholesterol levels?
H b. If the active sites of isoform-1 and -2 are the same
except isoform-1 has a cysteine residue and iso-
form-2 has a phenylalanine residue at that same
position, what two approaches would you take for
=
[TS]‡ a muscle spasm drug without a cholesterol level
OPO3 increase side effect?
11. Fosfomycin (5) is a potent antibacterial agent (see
Chapter 7, Figure 7.3) that interferes with cell wall bio-
synthesis at the enzyme called MurA.

7. 5-Fluoro-2′-deoxyuridylate (5-FdUMP) inactivates thy-

midylate ­synthase and is an antitumor agent. What would
be a good choice for drug combination with 5-FdUMP?  Draw a potential chemical mechanism for how fosfo-
Why? mycin acts.
Chapter | 5 Enzyme Inhibition and Inactivation 263

12. Carfilzomib (5a, Kyprolis) is a proteasome inhibi- which causes apoptosis and inhibition of tumor cell
tor for the treatment of multiple myeloma and solid growth. The N-terminal threonine of the 20S pro-
tumors. It acts by irreversible inhibition of the 20S teasome reacts to form a morpholine adduct (5b).
proteasome, an enzyme that degrades unwanted Draw a reasonable inactivation mechanism that
proteins. Inhibition of this enzyme in the tumor rationalizes why acid- or base-catalysis would be
cell leads to a build-up of ubiquitinated proteins, favored in each step.

O O
H H
N N O
N N N
H H
O O O O
Ph
Ph 5a
HO
H
N
H2N
O

O O
H H
N N O
N N N
H H H
O O O HO N
N
Ph H
Ph 5b HO O

13. Note if any of the following drugs have the potential

to be irreversible inhibitors. If so, draw the most likely
reaction that they undergo.

N NH2
Ph O O
O OH N
H Cl
N N B N N
N NH2 O F
OH
H N N
N O
Cl
HN
Bortezomib N Vilazodone Crizotinib
H

EtO N
O
O C11H23 HN
O N
Me2N HN
Cl
H O
N
O O
Orlistat Neratinib
H N
264 The Organic Chemistry of Drug Design and Drug Action

14. GABA aminotransferase catalyzes a PLP-dependent convulsing, but he went into a coma. If the problem
conversion of GABA to succinic semialdehyde (see was that the GABA concentration became too high,
Section 5.3.3.3.1). mention two possible solutions to the problem.
a. Draw a mechanism for how 6 inactivates this 15. An excess of androgenic hormones such as testoster-
enzyme. one can cause benign prostatic hypertrophy (enlarged
prostate). Most of the androgenic activity appears to
be caused by a metabolite of testosterone (7), namely,
5α-dihydrotestosterone (8), produced from testoster-
one in a NADPH-dependent reaction catalyzed by ste-
roid 5α-reductase. The mechanism for this reductase is
shown below.
 Finasteride (9, Proscar) is a potent inhibitor of steroid
b. 
A hypothetical anticonvulsant drug that inhibits 5α-reductase. It has been proposed to be a mechanism-
GABA aminotransferase was given to a patient in based inhibitor. Draw a mechanism consistent with this
overdose quantities. Not only did the patient stop proposal.

OH OH

steroid
5α-reductase

O O
H
7 8

NH2 NH2
O O
H
:N R H NR
H OH OH

8
B H O
H O
B: H
B

O
N
H

O N 9
H H
Chapter | 5 Enzyme Inhibition and Inactivation 265

REFERENCES 13. Krebs, E. G. Historical perspectives on protein phosphorylation and a

1. Statistics taken from Annual Reports in Medicinal Chemistry, Aca- classification system for protein kinases. Philos. Trans. R. Soc. Lon-
demic Press, San Diego, CA, Vol. 42–47, 2007–2012. don B Biol. Sci. 1983, 302, 3–11.
2. Cohen, S. S. Comparative biochemistry and drug design for infec- 14. Liu, Y.; Gray, N. S. Rational design of inhibitors that bind to inactive
tious disease. Science, 1979, 205, 964–971. kinase conformations. Nature Chem. Biol. 2006, 2, 358–364.
3. (a) Segal, I. H. Enzyme Kinetics, John Wiley & Sons, New York, 15. (a) Bhagwat, S. S. Kinase inhibitors for the treatment of inflam-
1975, p. 106. (b) Burlingham, B. T.; Widlanski, T. S. An intuitive matory and autoimmune disorders. Purinergic Signal. 2009, 5,
look at the relationship of Ki and IC50: a more general use of the 107–115. (b) Rokosz, L. L.; Beasley, J. R.; Carroll, C. D.; Lin, T.;
Dixon plot. J. Chem. Educ. 2003, 80, 214–218. Zhao, J.; Appell, K. C.; Webb, M. L. Kinase inhibitors as drugs for
4. Normano, N.; De Luca, A.; Bianco, C.; Strozzo. L.; Mancino, M.; chronic inflammatory and immunological diseases: progress and
Maiello, M. R.; Carotenuto, A.; De Feo, G.; Caponigro, F.; Salomon, challenges. Exp. Opin. Ther. Targets 2008, 12, 883–903.
D. S. Epidermal growth factor receptor (EGFR) signaling in cancer. 16. Domagk, G. Chemotherapy of bacterial infections. Deut. Med.
Gene 2006, 366, 2–16. Wschr. 1935, 61, 250–253.
5. Ward, W. H. J.; Cook, P. N.; Slate, A. M.; Davies, D. H.; Holdgate, 17. Albert, A. Selective Toxicity, 7th ed., Chapman and Hall, London,
G. A.; Green, L. R. Epidermal growth factor receptor tyrosine kinase: 1985, p. 220.
investigation of catalytic mechanism, structure-based searching and 18. Domagk, G. Chemotherapy for streptococcus infection. Klin.
discovery of a potent inhibitor. Biochem. Pharmacol. 1994, 48, Wochenschr. 1936, 15, 1585–1590.
659–666. 19. Tréfouël, J.; Tréfouël, Mme. J.; Nitti, F.; Bovet, D. Action of p-ami-
6. (a) Fry, D. W.; Kraker, A. J.; McMichael, A.; Ambroso, L. A.; Nel- nophenylsulfamide in experimental streptococcus infections of mice
son, J. M.; Leopold, W. R.; Connors, R. W.; Bridges, A. J. A specific and rabbits. Compt. Rend. Soc. Biol. Paris, 1935, 120, 756–758.
inhibitor of the epidermal growth factor receptor tyrosine kinase. Sci- 20. (a) Marshall, E. K., Jr. Determination of sulfanilamide in blood and
ence 1994, 265, 1093–1095. (b) Rewcastle, G. W.; Denny, W. A.; urine. J. Biol. Chem. 1937, 122, 263–273. (b) Bratton, A. C.; Mar-
Bridges, A. J.; Zhou, H.; Cody, D. R.; McMichael, A.; Fry, D. W. shall, E. K., Jr. A new coupling component for sulfanilamide deter-
Tyrosine kinase inhibitors. 5. Synthesis and structure-activity rela- mination. J. Biol. Chem. 1939, 128, 537–550.
tionships for 4-[(phenylmethyl)amino]- and 4-phenylamino)quinazo- 21. Stamp, T. C. Bacteriostatic action of sulfanilamide in vitro. Influence
lines as potent adenosine 5’-triphosphate binding site inhibitors of of fractions isolated from hemolytic streptococci. Lancet 1939, ii,
the tyrosine kinase domain of the epidermal growth factor receptor. 10–17.
J. Med. Chem. 1995, 38, 3482–3487. 22. Woods, D. D. The relation of p-aminobenzoic acid to the mechanism
7. Moyer, J. D.; Barbacci, E. G.; Iwata, K. K; Arnold, L.; Boman, B.; of the action of sulfanilamide. Br. J. Exper. Pathol. 1940, 21, 74–90.
Cunningham, A.; DiOrio, C.; Doty, J.; Morin, M. J.; Moyer, M.; 23. Selbie, F. R. The inhibition of the action of sulfanilamide in mice by
Neveu, M.; Pollack, V. A.; Pustilnik, L. R.; Reynolds, M. M.; Sloan, p-aminobenzoic acid. Br. J. Exper. Pathol. 1940, 21, 90–93.
D.; Theleman, A.; Miller, P. Induction of apoptosis and cell cycle 24. Fildes, P. A rational approach to research in chemotherapy. Lancet,
arrest by CP-358,774, an inhibitor of epidermal growth factor recep- 1940, i, 955–957.
tor tyrosine kinase. Cancer Res. 1997, 57, 4838–4848. 25. Miller, A. K. Folic acid and biotin synthesis by sulfonamide-sensitive
8. Stamos, J.; Sliwkowski, M. X.; Eigenbrot, C. Structure of epider- and sulfonamide-resistant strains of E. coli. Proc. Soc. Exper. Pathol.
mal growth factor receptor kinase domain alone and in complex Med. 1944, 57, 151–153. Miller, A. K.; Bruno, P.; Berglund, R. M.
with a 4-anilinoquinazoline inhibitor. J. Biol. Chem. 2002, 277, The effect of sulfathiazole on the in vitro synthesis of certain vita-
46265–46272. mins by Escherichia coli. J. Bacteriol. 1947, 54, 9 (G20).
9. (a) Daley, G. O.; Van Etten, R. A.; Baltimore, D. Induction of chronic 26. Nimmo-Smith, R. H.; Lascelles, J.; Woods, D. D. The synthesis of
myelogenous leukemia in mice by the P210bcr/abl gene of the Phila- "folic acid" by Streptobacterium plantarum and its inhibition by sul-
delphia chromosome. Science 1990, 247, 824–830. (b) Kelliher, fonamides. Br. J. Exper. Pathol. 1948, 29, 264–281.
M. A.; McLaughlin, J.; Witte, O. N.; Rosenberg, N. Induction of a 27. Richey, D. P.; Brown, G. M. Biosynthesis of folic acid. IX.
chronic myelogenous leukemia-like syndrome in mice with v-abl and Purification and properties of the enzymes required for the
BCR/ABL. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 6649–6653. (c) formation of dihydropteroic acid. J. Biol. Chem. 1969, 244,
Heisterkamp, N.; Jenster, G.; ten Hoeve, J.; Zovich, D.; Pattengale, 1582–1592.
P. K.; Groffen, J. Acute leukaemia in bcr/abl transgenic mice. Nature 28. Weisman, R. A.; Brown, G. M. The biosynthesis of folic acid. V.
1990, 344, 251–253. Characteristics of the enzyme system that catalyzes the synthesis of
10. Capdeville, R.; Buchdunger, E.; Zimmerman, J.; Matter, A.C. Glivec dihydropteroic acid. J. Biol. Chem. 1964, 239, 326–331.
(STI571, imatinib), a rationally developed, targeted anticancer drug. 29. Bock, L.; Miller, G. H.; Schaper, K.-J.; Seydel, J. K. Sulfonamide
Nat. Rev. Drug Disc. 2002, 1, 493–502. structure-activity relations in a cell-free system. 2. Proof for the for-
11. Zimmerman, J.; Buchdunger, E.; Mett, H.; Meyer, T.; Lydon, N.; mation of a sulfonamide-containing folate analog. J. Med. Chem.
Traxler, P. Phenylamino-pyrimidine (PAP) – derivatives: a new class 1974, 17, 23–28.
of potent and highly selective PDGF-receptor autophosphorylation 30. Wood, R. C.; Ferone, R.; Hitchings, G. H. Relation of cellular per-
inhibitors. Bioorg. Med. Chem. Lett. 1996, 6, 1221–1226. meability to the degree of inhibition by amethopterin and pyrimeth-
12. Nagar, B.; Bornmann, W. G.; Pellicena, P.; Schindler, T.; Veach, D. amine in several species of bacteria. Biochem. Pharmacol. 1961, 6,
R.; Miller, W. T.; Clarkson, B.; Kuriyan, J. Crystal structure of the 113–124.
kinase domain of c-abl in complex with the small molecule inhibi- 31. Haruki, H.; Pedersen, M. G.; Gorska, K. E.; Pojer, F.; Johnsson, K.
tors PD173955 and imatinib (STI-571). Cancer Res. 2002, 62, Tetrahydrobiopterin biosynthesis as an off-target of sulfa drugs. Sci-
4236–4243. ence 2013, 340, 987–991.
266 The Organic Chemistry of Drug Design and Drug Action

32. Wong, K. K.; Pompliano, D. L. In Resolving the Antibiotic Paradox, 49. (a) Wilson, D. K.; Rudolph, F. B.; Quiocho, F. A. Atomic structure
Rosen, B. P.; Mobashery, S. (Eds.), Plenum Publ., New York, 1998, of adenosine deaminase complexed with a transition-state analog:
pp. 197–217. understanding catalysis and immunodeficiency mutations. Science
33. Davies, J. E. In Antibiotic Resistance: Origins, Evolution and Spread, 1991, 252, 1278–1284. (b) See PDB ID 2ADA (pdb.org) for a cor-
Chadwick, D. J.; Goode, J. (Eds.), Wiley, New York, 1997. rection to reference 49a.
34. Bernhard, S. A.; Orgel, L. E. Mechanism of enzyme inhibition by 50. Sauter, C.; Lamanna, N.; Weiss, M. A. Pentostatin in chronic lympho-
phosphate esters. Science 1959, 130, 625–626. cytic leukemia. Expert Opin. Drug Metab. Toxicol. 2008, 4, 1217–1222.
35. Pauling, L. Chemical achievement and hope for the future. Am. Sci. 51. Daddona, P. E.; Kelley, W. N. Control of adenosine deaminase levels
1948, 36, 51–58. in human lymphoblasts. Adv. Enzyme Regul. 1982, 20, 153–163.
36. Jencks, W. P. In Current Aspects of Biochemical Energetics; Ken- 52. Berne, R. M.; Rall, T. W.; Rubio, R. (Eds.), Regulatory Functions of
nedy, E. P., Ed., Academic, New York, 1966, p. 273. Adenosine; Martinus Nijhoff, Boston, 1983.
37. Wolfenden, R. Transition state analog inhibitors and enzyme cataly- 53. Balakrishnan, K.; Verma, D.; O’Brien, S.; Kilpatrick, M.; Chen, Y.;
sis. Annu. Rev. Biophys. Bioeng. 1976, 5, 271–306. Wolfenden, R. Tyler, B.; Bickel, S.; Santia, S.; Keating, M. Jabtarhuabm H. Phase
Transition state analogues for enzyme catalysis. Nature 1969, 223, 2 and pharmacodynamic study of oral forodesine in patients with
704–705. Wolfenden, R. Transition state analogs as potential affinity advanced, fludarabine-treated chronic lymphocytic leukemia. Blood
labeling regents. Meth. Enzymol. 1977, 46, 15–28. 2010, 116, 886–892.
38. Lienhard, G. E. Enzymic catalysis and transition-state theory. Sci- 54. Bantia, S. Parker, C.; Upshaw, R.; Cunningham, A.; Kotian, P.; Kil-
ence 1973, 180, 149–154. Lienhard, G. E. Transition state analogs as patrick, M.; Morris, P.; Chand, P.; Babu, Y. Potent orally bioavailable
enzyme inhibitors. Annu. Rep. Med. Chem. 1972, 7, 249–258. purine nucleoside phosphorylase inhibitor BCX4208 induces apop-
39. Christianson, D. W.; Lipscomb, W. N. Carboxypeptidase A. Acc. tosis in B- and T-lymphocytes–a novel approach for autoimmune
Chem. Res. 1989, 22, 62–69. diseases, organ transplantation and hematologic malignancies. Int.
40. Schramm, V. L. Enzymatic transition states: thermodynamics, dynam- Immunopharmacol. 2010, 10, 784–790.
ics and analogue design. Arch. Biochem. Biophys. 2005, 433, 13–26. 55. Kline, P. C.; Schramm, V. L. Pre-steady-state transition-state analysis
41. Saen-Oon, S.; Quaytman-Machleder, S.; Schramm, V. L.; Schwartz, of the hydrolytic reaction catalyzed by purine nucleoside phosphory-
S. D. Atomic detail of chemical transformation at the transition state lase. Biochemistry 1995, 34, 1153–1162.
of an enzymatic reaction. Proc. Natl. Acad. Sci. 2008, 105, 16543– 56. Cohen, A.; Gudas, L. J.; Ammann, A. J.; Staal, G. E. J.; Martin, Jr., D. W.
16548. Deoxyguanosine triphosphate as a possible toxic metabolite in the
42. (a) Cleland, W. W. Isotope effects: Determination of enzyme transi- immunodeficiency associated with purine nucleoside phosphorylase
tion state structure. Meth. Enzymol. 1995, 249, 341–373. (b) Sch- deficiency. J. Clin. Invest. 1978, 61, 1405–1409.
ramm, V. L. Enzymatic transition state poise and transition state 57. Tattersall, M. H.; Ganeshaguru, K.; Hoffbrand, A. V. The effect of
analogues. Acc. Chem. Res. 2003, 36, 588–596. (c) Northrup, D. B. external deoxyribonucleosides on deoxyribonucleoside triphosphate
The expression of isotope effects on enzyme-catalyzed reactions. concentrations in human lymphocytes. Biochem. Pharmacol. 1975,
Annu. Rev. Biochem. 1981, 50, 103–131. 24, 1495–1498.
43. Cleland, W. W. The use of isotope effects to determine the transition- 58. Mitchell, B. S.; Mejias, E.; Daddona, P. E.; Kelley, W. N. Purinogenic
state structure for enzymatic reactions. Meth. Enzymol. 1982, 87, immunodeficiency diseases: selective toxicity of deoxyribonucleo-
625–641. sides for T cells. Proc. Natl. Acad. Sci. U.S.A. 1978, 75, 5011–5014.
44. Lewis, B. E.; Schramm, V. L. Enzymatic binding isotope effects and 59. Bantia, S.; Ananth, S. L.; Parker, C. D.; Horn, L.L.; Upshaw, R.
the interaction of glucose with hexokinase. Isotope Effects in Chem- Mechanism of inhibition of T-acute lymphoblastic leukemia cells by
istry and Biology; Kohen, A.; Limbach, H.-H. (Eds.), CRC Press, PNP inhibitor-BCX-1777. Int. Immunopharmcol. 2003, 3, 879–887.
Boca Raton, FL, 2006; pp. 1019–1053. 60. Morris, P. E., Jr.; Montgomery, J. A. Inhibitors of the enzyme purine
45. Schramm, V. L. Enzymatic transition states, transition-state analogs, nucleoside phosphorylase. Expert Opin. Ther. Pat. 1998, 8, 283–299.
dynamics, thermodynamics, and lifetimes. Annu. Rev. Biochem. 61. Kline, P. C.; Schramm, V. L. Purine nucleoside phosphorylase. Cata-
2011, 80, 703–732. lytic mechanism and transition state analysis of the arsenolysis reac-
46. (a) Bagdassarian, C. K.; Schramm, V. L.: Schwartz, S. D. Molecular tion. Biochemistry 1993, 32, 13212–13219.
electrostatic potential analysis for enzymatic substrates, competitive 62. Miles, R. W.; Tyler, P. C.; Furneaux, R. H.; Bagdassarian, C. K.; Sch-
inhibitors and transition state inhibitors. J. Am. Chem. Soc. 1996, ramm, V. L. One-third-the-sites transition state inhibitors for purine
118, 8825–8836. (b) Bagdassarian, C. K.; Braunheim, B. B.; Sch- nucleoside phosphorylase. Biochemistry 1998, 37, 8615–8621.
ramm, V. L.; Schwartz, S. D. Quantitative measures of molecular 63. Fedorov, A; Shi, W.; Kicska, G.; Federov, E.; Tyler, P. C.; Furneaux, R.
similarity: measures to analyze transition-state analogs for enzymic H.; Hanson, J. C.; Gainsford, G. J.; Larese, J. Z.; Schramm, V. L. Transi-
reactions. Int. J. Quantum. Chem.: Quantum. Biol. Symp.1996, 60, tion state structure of purine nucleoside phosphorylase and principles of
73–80. atomic motion in enzymatic catalysis. Biochemistry 2001, 40, 853–860.
47. Loo, T. L.; Nelson, J. A. In Cancer Medicine, 2nd ed., Holland, J. F.; 64. Lewandowicz, A.; Schramm, V. L. Transition state analysis for
Frei, E. III. (Eds.), Lea & Febiger, Philadelphia, 1982, p. 790. McCor- human and Plasmodium falciparum purine nucleoside phosphory-
mack, J. J.; Johns, D. G. In Pharmacologic Principles of Cancer Treat- lases. Biochemistry 2004, 43, 1458–1468.
ment, Chabner, B. A. (Ed.), W. B. Saunders, Philadelphia, 1982, p. 213. 65. Lewandowicz, A.; Tyler, P. C. ; Tyler, P.; Evans, G., Furneaux, R.;
48. Agarwal, R. P.; Spector, T.; Parks, R. E., Jr. Tight-binding inhibitors. Schramm, V. Achieving the ultimate physiological goal in transition
IV. Inhibition of adenosine deaminases by various inhibitors. Bio- state analogue inhibitors for purine nucleoside phosphorylase. J.
chem. Pharmacol. 1977, 26, 359–367. Biol. Chem. 2003, 278, 31465–31468.
Chapter | 5 Enzyme Inhibition and Inactivation 267

66. (a) Stark, G. R.; Bartlett, P. A. Design and use of potent, specific 78. Lu, H.; England, K.; am End, C.; Truglio, J.; Luckner, S.; Reddy
enzyme inhibitors. Pharmacol. Ther. 1983, 23, 45–78. (b) Collins, B. G.; Marlenee N.; Knudson, S.; Knudson, D.; Bowen, R. Slow-
K. D.; Stark, G. R. Aspartate transcarbamylase. Interaction with onset inhibition of the FabI enoyl reductase from Francisella tular-
the transition state analog N-(phosphonacetyl)-L-aspartate. J. Biol. ensis: Residence time and in vivo activity. ACS Chem. Biol. 2009, 4,
Chem. 1971, 246, 6599–6605. 221–231.
67. Frantom, P. A.; Blanchard, J. S. Bisubstrate analog inhibitors. In 79. Espiner, E. A.; Nicholls, M. G. In The Renin-Angiotensin System,
Comprehensive Natural Products II Chemistry and Biology, Man- Robertson, J. I. S.; Nicholls, M. G. (Eds.), Gower Medical Publish-
der, L.; Liu, H.-W. (Eds.), Elsevier: Amsterdam, 2010, Vol. 8; ing, London, 1993; pp. 33.1–33.24.
pp. 689–717. 80. Tewksbury, D. A.; Dart, R. A.; Travis, J. The amino terminal amino
68. (a) Johnsson, K.; Schultz, P. G. Mechanistic studies of the acid sequence of human angiotensinogen. Biochem. Biophys. Res.
oxidation of isoniazid by the catalase peroxidase from Mycobac- Commun. 1981, 99, 1311–1315; Tewksbury, D. In Biochemical
terium tuberculosis. J. Am. Chem. Soc. 1994, 116, 7425–7426. Regulation of Blood Pressure; Soffer, R. L. (Ed.), Wiley, New York,
(b) Zhao, X.; Yu, H.; Yu, S.; Wang, F.; Sacchettini, J. C.; 1981; p. 95.
Magliozzo, R. S. Hydrogen peroxide-mediated isoniazid acti- 81. Moeller, I.; Allen, A. M.; Chai, S.-Y.; Zhuo, J.; Mendelsohn, F. A.
vation catalyzed by Mycobacterium tuberculosis catalase-per- O. Bioactive angiotensin peptides. J. Hum. Hypertens. 1998, 12,
oxidase (KatG) and its S315T mutant. Biochemistry 2006, 45, 289–293.
4131–4140. 82. Wilk, D.; Healy, D. P. Glutamyl aminopeptidase (aminopeptidase A),
69. Aitken, S. M.; Ouellet, M.; Percival, M. D.; English, A. M. Mecha- the BP-1/6C3 antigen. Adv. Neuroimmunol. 1993, 3, 195–207.
nism of horseradish peroxidase inactivation by benzhydrazide: a crit- 83. Larner, A.; Vaughan, E. D., Jr.; Tsai, B.-S.; Peach, M. J. Role of convert-
ical evaluation of arylhydrazides as peroxidase inhibitors. Biochem. ing enzyme in the cardiovascular and adrenal cortical responses to (des-
J. 2003, 375, 613–621. Asp1)-angiotensin I. Proc. Soc. Exp. Biol. Med. 1976, 152, 631–634.
70. Wiseman, B.; Carpena, X.; Feliz, M.; Donald, L. J.; Pons, M.; Fita, 84. (a) Zini, S.; Fournie-Zaluski, M.-C.; Chauvel, E.; Roques, B. P.;
I.; Loewen, P. C. Isonicotinic acid hydrazide conversion to isonic- Corvol, P.; Llorens-Cortes, C. Identification of metabolic pathways
otinyl-NAD by catalase-peroxidases. J. Biol. Chem. 2010, 285, of brain angiotensin II and III using specific aminopeptidase inhibi-
26662–26673. tors: Predominant role of angiotensin III in the control of vasopressin
71. (a) Waley, S. G. The kinetics of slow-binding and slow, tight-binding release. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 11968–11973. (b)
inhibition: The effects of substrate depletion. Biochem. J. 1993, Blair-West, J. R.; Coghlan, J. P.; Denton, D. A.; Funder, J. W.; Scog-
294, 195–200. (b) Sculley, M. J.; Morrison, J. F. The determina- gins, B. A.; Wright, R. D. Effect of the heptapeptide (2-8) and hexa-
tion of kinetic constants governing the slow, tight-binding inhibition peptide (3-8) fragments of angiotensin II on aldosterone secretion.
of enzyme-catalysed reactions. Biochim. Biophys. Acta 1986, 874, J. Clin. Endocrinol. Metab. 1971, 32, 575–578. (c) Caldicott, W. J.
44–53. (c) Schloss, J. V. Significance of slow-binding enzyme inhibi- H.; Taub, K. J.; Hollenberg, N. K. Identical mesenteric, femoral and
tion and its relationship to reaction-intermediate analogs. Acc. Chem. renal vascular responses to angiotensins II and III in the dog. Life Sci.
Res. 1988, 21, 348–353. (d) Morrison, J. F.; Walsh, C. T. The behav- 1977, 20, 517.
ior and significance of slow-binding enzyme inhibitors. Adv. Enzy- 85. Palmieri, F. E.; Bausback, H. H.; Ward, P. E. Metabolism of vasoac-
mol. 1988, 61, 201–301. tive peptides by vascular endothelium and smooth muscle aminopep-
72. (a) Rich, D. H. Pepstatin-derived inhibitors of aspartic proteinases. tidase M. Biochem. Pharmacol. 1989, 38, 173–180.
A close look at an apparent transition-state analog inhibitor. J. Med. 86. (a) Powers B.; Greene L.; Balfe L. M. Updates on the treatment of
Chem. 1985, 28, 263–273. (b) Morrison, J. F.; Walsh, C. T. The essential hypertension: a summary of AHRQ’s comparative effective-
behavior and significance of slow-binding enzyme inhibitors. Adv. ness review of angiotensin-converting enzyme inhibitors, ­angiotensin
Enzymol. 1988, 61, 201–301. II receptor blockers, and direct renin inhibitors. J. Manag. Care
73. (a) Imperiali, B.; Abeles, R. H. Inhibition of serine proteases by pep- Pharm. 2011, 17(Suppl. 8), S1–S4. (b) White C. M.; Greene L. Sum-
tidyl fluoromethyl ketones. Biochemistry 1986, 25, 3760. (b) Stein, mary of AHRQ’s comparative effectiveness review of angiotensin-
R. L.; Strimpler, A. M.; Edwards, P. D.; Lewis, J. J.; Mauger, R. converting enzyme inhibitors, angiotensin II receptor blockers added
C.; Schwartz, J. A.; Stein, M. M.; Trainor, D. A.; Wildonger, R. A.; to standard medical therapy for treating stable ischemic heart disease.
Zottola, M. A. Mechanism of slow-binding inhibition of human leu- J. Manag. Care Pharm. 2011, 17(Suppl. 5), S1–S15. (c) Karthikeyan
kocyte elastase by trifluoromethyl ketones. Biochemistry 1987, 26, V. J.; Lip G. Y. H. Review: Angiotensin converting enzyme inhibitors
2682–2689. and angiotensin receptor blockers prevent atrial fibrillation.
74. Morrison, J. F. The slow-binding and slow, tight-binding inhibition of ­Evidence-based Med. 2006, 11(1), 15. (d) Sica, D. A. Pharmaco-
enzyme-catalyzed reactions. Trends Biochem. Sci. 1982, 7, 102–105. therapy review: angiotensin-converting enzyme inhibitors. J. Clin.
75. Bartlett, P. A.; Marlowe, C. K. Evaluation of intrinsic binding energy Hypertens. (Greenwich, Conn.) 2005, 7(8), 485–488.
from a hydrogen bonding group in an enzyme inhibitor. Science 87. Ferreira, S. H. A bradykinin-potentiating factor (BPF) present in the
(Washington, D. C.) 1987, 235, 569–571. venom of Bothrops jararaca. Br. J. Pharmacol. Chemother. 1965,
76. Rich, D. H.; Pepstatin-derived inhibitors of aspartic proteinases. 24, 163–169.
A close look at an apparent transition-state analog inhibitor. J. Med. 88. (a) Bakhle, Y. S. Conversion of angiotensin I to angiotensin II by
Chem. 1985, 28, 263–273. cell-free extracts of dog lung. Nature (London) 1968, 220, 919–921.
77. Copeland, R. A.; Pompliano, D. L.; Meek, T. D. Drug-target resi- (b) Bakhle, Y. S.; Reynard, A. M.; Vane, J. R. Metabolism of the
dence time and its implications for lead optimization. Nat. Rev. Drug angiotensins in isolated perfused tissues. Nature (London) 1969, 222,
Disc. 2006, 5, 730–739. 956–959.
268 The Organic Chemistry of Drug Design and Drug Action

89. Ferreira, S. H.; Bartelt, D. C.; Greene, L. J. Isolation of bradykinin- 105. Wyvratt, M. J.; Patchett, A. A. Recent developments in the design of
potentiating peptides from Bothrops jararaca venom. Biochemistry angiotensin-converting enzyme inhibitors. Med. Res. Rev. 1985, 4,
1970, 9, 2583–2593. 483–531.
90. Ferreira, S. H.; Greene, L. J.; Alabaster, V. A.; Bakhle, Y. S.; Vane, 106. Natesh, R.; Schwager, S. L. U.; Sturrock, E. D.; Acharya, K. R. Crys-
J. R. Activity of various fractions of bradykinin potentiating factor tal structure of the human angiotensin-converting enzyme-lisinopril
against angiotensin I converting enzyme. Nature (London) 1970, complex. Nature 2003, 421, 551–554.
225, 379–380. 107. (a) De Lombaert, S.; Chatelain, R. E.; Fink, C. A.; Trapani, A. J.
91. Stewart, J. M.; Ferreira, S. H.; Greene, L. J. Bradykinin potentiating Design and pharmacology of dual angiotensin-converting enzyme
peptide pyrrolidonecarbonyl-Lys-Trp-Ala-Pro. Inhibitor of the pul- and neutral endopeptidase inhibitors. Curr. Pharm. Des. 1996,
monary inactivation of bradykinin and conversion of angiotensin I to 2, 443–462. (b) Fink, C. A. Recent advances in the development
II. Biochem. Pharmacol. 1971, 20, 1557–1567. of dual angiotensin-converting enzyme and neutral endopepti-
92. Ondetti, M. A.; Williams, N. J.; Sabo, E. F.; Pluscec, J.; Weaver, E. dase inhibitors. Exp. Opin. Ther. Pat. 1996, 6, 1147–1164. (c)
R.; Kocy, O. Angiotensin-converting enzyme inhibitors from the Seymour, A. A.; Asaad, M. M.; Lanoce, V. M.; Langenbacher, K.
venom of Bothrops jararaca. Isolation, elucidation of structure, and M.; Fennell, S. A.; Rogers, W. L. Systemic hemodynamics, renal
synthesis. Biochemistry 1971, 10, 4033–4039. function and hormonal levels during inhibition of neutral endopep-
93. Cheung, H. S.; Cushman, D. W. Inhibition of homogeneous angio- tidase 3.4.24.11 and angiotensin-converting enzyme in conscious
tensin-converting enzyme of rabbit lung by synthetic venom peptides dogs with pacing-induced heart failure. J. Pharmacol. Exp. Ther.
of Bothrops jararaca. Biochim. Biophys. Acta 1973, 293, 451–463. 1993, 266, 872–883. (d) Pham, I.; Gonzalez, W.; El Amraani, A. I.;
94. Cushman, D. W.; Cheung, H. S. In Hypertension, Genest, J.; Koiw, E. Fournie-Zaluski, M. C.; Philippe, M.; Laboulandine, I.; Roques, B.
(Eds.), Springer: Berlin, 1972, p. 532. P.; Michel, J. B. Effects of converting enzyme inhibitor and neu-
95. Ondetti, M. A.; Cushman, D. W. In Biochemical Regulation of Blood tral endopeptidase inhibitor on blood pressure and renal function
Pressure, Soffer, R. L. (Ed.), Wiley, New York, 1981, p. 165. in experimental hypertension. J. Pharmacol. Exp. Ther. 1993, 265,
96. Kato, H.; Suzuki, T. Bradykinin-potentiating peptides from the 1339–1347.
venom of Agkistrodon halys. Isolation of five bradykinin potentiators 108. Robl, J. A.; Cimarusti, M. P.; Simpkins, L. M.; Brown, B.; Ryono, D.
and the amino acid sequences of two of them, potentiators B and C. E.; Bird, J. E.; Asaad, M. M.; Schaeffer, T. R.; Trippodo, N. C. Dual
Biochemistry 1971, 10, 972–980. metalloprotease inhibitors. 6. incorporation of bicyclic and substi-
97. Das, M.; Soffer, R. L. Pulmonary angiotensin-converting enzyme. II. tuted monocyclic azepinones as dipeptide surrogates in angiotensin-
Structural and catalytic properties. J. Biol. Chem. 1975, 250, 6762– converting enzyme (ACE)/neutral endopeptidase (NEP) inhibitors. J.
6768. Med. Chem. 1996, 39, 494–502.
98. Quiocho, F. A.; Lipscomb, W. N. Carboxypeptidase A: a protein and 109. Robl, J. A.; Sun, C. Q.; Stevenson, J.; Ryono, D. E.; Simpkins, L.
an enzyme. Adv. Protein Chem. 1971, 25, 1–78. M.; Cimarusti, M. P.; Dejneka, T.; Slusarchyk, W. A.; Chao, S.; Strat-
99. Cushman, D. W.; Cheung, H. S.; Sabo, E. F.; Ondetti, M. A. Design ton, L.; Misra, R. N.; Bednarz, M. S.; Asaad, M. M.; Cheung, H. S.;
of potent competitive inhibitors of angiotensin-converting enzyme. AbboaOffei, B. E.; Smith, P. L.; Mathers, P. D.; Fox, M.; Schaeffer,
Carboxyalkanoyl and mercaptoalkanoyl amino acids. Biochemistry, T. R.; Seymour, A. A.; Trippodo, N. C. Dual metalloprotease inhibi-
1977, 16, 5484–5491. tors: Mercaptoacetyl-based fused heterocyclic dipeptide mimetics as
100. Byers, L. D.; Wolfenden, R. Potent reversible inhibitor of carboxy- inhibitors of angiotensin-converting enzyme and neutral endopepti-
peptidase A. J. Biol. Chem. 1972, 247, 606–608. Byers, L. D.; dase. J. Med. Chem. 1997, 40, 1570–1577.
Wolfenden, R. Binding of the by-product analog benzylsuccinic acid 110. (a) Pickering, T. G. The rise and fall of omapatrilat. Medscape News
by carboxypeptidase A. Biochemistry 1973, 12, 2070–2078. Today 2002; http://www.medscape.com/viewarticle/443224. (b)
101. Ondetti, M. A.; Cushman, D. W.; Sabo, E. F.; Cheung, H. S. In Drug Venugopal, J. Pharmacological modulation of the natriuretic peptide
Action and Design: Mechanism-Based Enzyme Inhibitors. Kalman, system. Exp. Opin. Ther. Pat. 2003, 13, 1389–1409.
T. I. (Ed.), Elsevier/North Holland, New York, 1979, p. 271. 111. Turner, A. J.; Tanzawa, K. Mammalian membrane metallopeptidases:
102. Atkinson, A. B.; Robertson, J. I. S. Captopril in the treatment of clini- NEP, ECE, KELL, and PEX. FASEB J. 1997, 11, 355–364.
cal hypertension and cardiac failure. Lancet 1979, ii, 836–839. 112. (a) Rubanyi, G. M.; Polokoff, M. A. Endothelins: molecular biol-
103. Patchett, A. A.; Harris, E.; Tristram, E. W.; Wyvratt, M. J.; Wu, M. ogy, biochemistry, pharmacology, physiology, and pathophysiology.
T.; Taub, D.; Peterson, E. R.; Ikeler, T. J.; ten Broeke, J.; Payne, L. Pharmacol. Rev. 1994, 46, 325–415. (b)Patel, T. R. Therapeutic
G.; Ondeyka, D. L.; Thorsett, E. D.; Greenlee, W. J.; Lohr, N. S.; potential of endothelin receptor antagonists in cerebrovascular dis-
Hoffsommer, R. D.; Joshua, H.; Ruyle, W. V.; Rothrock, J. W.; Aster, ease. CNS Drugs 1996, 5, 293–310. (c) Benigni, A.; Remuzzi, G.
S. D.; Maycock, A. L.; Robinson, F. M.; Hirschmann, R.; Sweet, C. The renoprotective potential of endothelin receptor antagonists. Exp.
S.; Ulm, E. H.; Gross, D. M.; Vassil, T. C.; Stone, C. A. A new class Opin. Ther. Pat. 1997, 7, 139–149.
of angiotensin-converting enzyme inhibitors. Nature (London)1980, 113. (a) Loffler, B.-M. The renoprotective potential of endothelin receptor
288, 280–283. antagonists. J. Cardiovasc. Pharmacol. 2000, 35(Suppl. 2), S79–S82.
104. (a) Shapiro, R.; Riordan, J. F. Inhibition of angiotensin converting (b) Vemulapalli, S.; Chintala, M.; Stamford, A.; Watkins, R.; Chiu, P.;
enzyme: mechanism and substrate dependence. Biochemistry 1984, Sybertz, E.; Fawzi, A. B. Renal effects of SCH 54470: a triple inhibitor
23, 5225–5233. (b) Bull, H. G.; Thornberry, N. A.; Cordes, M. H. J.; of ECE, ACE, and NEP. Cardiovasc. Drug Rev. 1997, 15, 260–272.
Patchett, A. A.; Cordes, E. H. Inhibition of rabbit lung angiotensin- 114. Fukuda, Y.; Fukuta, Y.; Higashino, R.; Ogishima, M.; Yoshida, K.;
converting enzyme by Nα[(S)-1-carboxy-3-phenylpropyl]L-alanyl- Tamaki, H.; Takei, M. Z-350, a new chimera compound possessing α
L-proline and Nα -[(S)-1-carboxy-3-phenylpropyl]L-lysyl-L-pro- 1-adrenoceptor antagonistic and steroid 5α -reductase inhibitory actions.
line. J. Biol. Chem. 1985, 260, 2952–2962. Naunyn Schmiedebergs Arch. Pharmacol. 1999, 359, 433–438.
Chapter | 5 Enzyme Inhibition and Inactivation 269

115. (a) Furuta, S.; Fukuda, Y.; Sugimoto, T.; Miyahara, H.; Kamada, E.; and in vivo by ML-236A and ML-236B, competitive inhibitors of
Sano, H.; Fukuta, Y.; Takei, M.; Kurimoto, T. Pharmacodynamic 3-hydroxy-3-methylglutaryl-coenzyme A reductase. Eur. J. Bio-
analysis of steroid 5α -reductase inhibitory actions of Z-350 in rat chem. 1977, 77, 31.
prostate. Eur. J. Pharmacol. 2001, 426, 105–111. (b) Fukuta, Y.; 128. Brown, A. G.; Smale, T. C.; King, T. J.; Hasenkamp, R.; Thompson,
Fukuda, Y.; Higashino, R.; Yoshida, K.; Ogishima, M.; Tamaki, H.; R. H. Crystal and molecular structure of compactin, a new antifungal
Takei, M. Z-350, a novel compound with α 1-adrenoceptor antago- metabolite from Penicillium brevicompactum. J. Chem. Soc. Perkin
nistic and steroid 5α -reductase inhibitory actions: pharmacological Trans. I 1976, 1165–1170.
properties in vivo. J. Pharmacol. Exp. Ther. 1999, 290, 1013–1018. 129. (a) Endo, A. Monacolin K, a new hypocholesterolemic agent pro-
116. Kenny, B.; Ballard, S.; Blagg, J.; Fox, D. Pharmacological options in duced by a Monascus species. J. Antibiot. 1979, 32, 852–854. (b)
the treatment of benign prostatic hyperplasia. J. Med. Chem. 1997, Endo, A. Monacolin K, a new hypocholesterolemic agent that spe-
40, 1293–1315. cifically inhibits 3-hydroxy-3-methylglutaryl coenzyme A reductase.
117. Lepor, H.; Knap-Maloney, G.; Sunshine, H. A dose titration study J. Antibiot. 1980, 33, 334–336.
evaluating terazosin, a selective, once-a-day alpha 1-blocker for 130. Alberts, A. W.; Chen, J.; Kuron, G.; Hunt, V.; Huff, J.; Hoff-
the treatment of symptomatic benign prostatic hyperplasia. J. Urol. man, C.; Rothrock, J.; Lopez, M.; Joshua, H.; Harris, E.; Patch-
1990, 144, 1393–1397. ett, A.; Monaghan, R.; Currie, S.; Stapley, E.; Albers-Schonberg,
118. Yoshida, K.; Horikoshi, Y.; Eta, M.; Chikazawa, J.; Ogishima, M.; G.; Hensens, O.; Hirschfield, J.; Hoogsteen, K.; Liesch, J.;
Fukuda, Y.; Sato, H. Synthesis of benzanilide derivatives as dual Springer, J. Mevinolin: A highly potent competitive inhibitor
acting agents with α 1-adrenoceptor antagonistic action and steroid of hydroxymethylglutaryl-coenzyme A reductase and a cho-
5-α reductase inhibitory activity. Bioorg. Med. Chem. Lett. 1998, 8, lesterol-lowering agent. Proc. Natl. Acad. Sci. U.S.A. 1980, 77,
2967–2972. 3957–3961.
119. Sato, H.; Kitagawa, O.; Aida, Y.; Chikazawa, J.; Kurimoto, T.; Takei, 131. Endo, A. Compactin (ML-236B) and related compounds as poten-
M.; Fukuta, Y.; Yoshida, K. Dual-acting agents with α 1-adrenoceptor tial cholesterol-lowering agents that inhibit HMG-CoA reductase. J.
antagonistic and steroid 5α -reductase inhibitory activities. synthe- Med. Chem. 1985, 28, 401–405.
sis and evaluation of arylpiperazine derivatives. Bioorg. Med. Chem. 132. Tanzawa, K.; Endo, A. Kinetic analysis of the reaction catalyzed by
Lett. 1999, 9, 1553–1558. rat liver 3-hydroxy-3-methylglutaryl-coenzyme-A reductase using
120. (a) Missale, C.; Nash, S. R.; Robinson, S. W.; Jaber, M.; Caron, M. G. two specific inhibitors. Eur. J. Biochem. 1979, 98, 195–201.
Dopamine receptors: from structure to function. Physiol. Rev. 1998, 133. Nakamura, C. E.; Abeles, R. H. Mode of interaction of β -hydroxy-
78, 189–225. (b) Strange, P. G. In Advances in Drug Research, Testa, β -methylglutaryl coenzyme A reductase with strong binding inhibi-
B., Meyer, U. A. (Eds.),Academic Press, London, 1996, Vol. 28, pp. tors: compactin and related compounds. Biochemistry 1985, 24,
313–352. 1364–1376.
121. Andersen, G. P. In New Drugs for Asthma Therapy. Agents and 134. Stokker, G. E.; Hoffman, W. F.; Alberts, A. W.; Cragoe, E. J., Jr.;
Actions (Suppl. 34) Anderson, G. P., Chapman, I. D., Morley, J. Deana, A. A.; Gilfillan, J. L.; Huff, J. W.; Novello, F. C.; Prugh, J. D.;
(Eds.), Birkhauser Verlag, Basel, 1991, pp. 97–115. Smith, R. L.; Willard, A. K. 3-Hydroxy-3-methylglutaryl-coenzyme
122. Bonnert, R. V.; Brown, R. C.; Chapman, D.; Cheshire, D. R.; A reductase inhibitors. I. Structural modification of 5-substituted
Dixon, J.; Ince, F.; Kinchin, E. C.; Lyons, A. J.; Davis, A. M.; 3,5-dihydroxypentanoic acids and their lactone derivatives. J. Med.
Hallam, C.; Harper, S. T.; Unitt, J. F.; Dougall, I. G.; Jackson, D. M.; Chem. 1985, 28, 347–358.
McKechnie, K.; Young, A.; Simpson, W. T. Dual D2-receptor and 135. (a) Hoffman, W. F.; Alberts, A. W.; Cragoe, E. J., Jr.; Deana, A. A.;
β 2-adrenoceptor agonists for the treatment of airway diseases. 1. Evans, B. E.; Gilfillan, J. L.; Gould, N. P.; Huff, J. W.; Novello, F.
Discovery and biological evaluation of some 7-(2-aminoethyl)- C.; Prugh, J. D.; Rittle, K. E.; Smith, R. L.; Stokker, G. E.; Willard,
4-hydroxybenzothiazol-2(3H)-one analogs. J. Med. Chem. 1998, A. K. 3-Hydroxy-3-methylglutaryl-coenzyme A reductase inhibitors.
41, 4915–4917. 2. Structural modification of 7-(substituted aryl)-3,5-dihydroxy-
123. Witztum, J. L. In Goodman and Gilman’s The Pharmacological Basis 6-heptenoic acids and their lactone derivatives. J. Med. Chem. 1986,
of Therapeutics, 9th ed., Hardman, J. G.; Limbird, L. E.; Molinoff, P. 29, 159–169. (b)Stokker, G. E.; Alberts, A. W.; Anderson, P. S.; Cra-
B.; Ruddon, R. W.; Gilman, A. G. (Eds.), McGraw-Hill, New York, goe, E. J., Jr.; Deana, A. A.; Gilfillan, J. L.; Hirschfield, J.; Holtz, W.
1996, p. 875. J.; Hoffman, W. F.; Huff, J. W.; Lee, T. J.; Novello, F. C.; Prugh, J.
124. Grundy, S. M. Cholesterol metabolism in man. West. J. Med. 1978, D.; Rooney, C. S.; Smith, R. L.; Willard, A. K. J. Med. Chem. 1986,
128, 13–25. 29, 170.
125. (a) Stamler, J., Dietary and serum lipids in the multifactorial etiol- 136. Stokker, G. E.; Alberts, A. W.; Gilfillan, J. L.; Huff, J. W.; Smith, R.
ogy of atherosclerosis. Arch. Surg. 1978, 113, 21. (b) Havel, R. J.; L. 3-Hydroxy-3-methylglutaryl-coenzyme A reductase inhibitors. 5.
Goldstein, J. L.; Brown, M. S. In Metabolic Control and Disease, 6-(Fluoren-9-yl)- and 6-(fluoren-9-ylidene)-3,5-dihydroxyhexanoic
Bundy, P. K.; Rosenberg, L. E. (Eds.), W. B. Saunders, Philadel- acids and their lactone derivatives. J. Med. Chem. 1986, 29, 852–855.
phia, 1980, p. 393. 137. Bartmann, W.; Beck, G.; Granzer, E.; Jendralla, H.; Kerekjarto, B.
126. Vigna, G. B.; Fellin, R. Pharmacotherapy of dyslipidemias in V.; Wess, G. Convenient two-step stereospecific ­hydroxy-substitution
the adult population. Exp. Opin. Pharmacother. 2010, 11(18), with retention in β -hydroxy-δ-lactones. 4(R)-Heterosubstituted
3041–3052. mevinolin and analogs. Tetrahedron Lett. 1986, 27, 4709–4712.
127. (a) Endo, A.; Kuroda, M.; Tsujita, Y. ML-236A, ML-236B, and ML- 138. Hoffman, W. F.; Alberts, A. W.; Anderson, P. S.; Chen, J. S.; Smith,
236C, new inhibitors of cholesterogenesis produced by Penicillium R. L.; Willard, A. K. 3-Hydroxy-3-methylglutaryl-coenzyme A
citrinum. J. Antibiot. 1976, 29, 1346–1348. (b) Endo, A.; Tsujita, Y.; reductase inhibitors. 4. Side-chain ester derivatives of mevinolin. J.
Kuroda, M.; Tanzawa, K. Inhibition of cholesterol synthesis in vitro Med. Chem. 1986, 29, 849–852.
270 The Organic Chemistry of Drug Design and Drug Action

139. Thaper, R. K.; Kumar, Y.; Kumar, S. M. D.; Misra, S.; Khanna, J. M. A 151. Kempf, D. J.; Codacovi, L.; Wang, X. C.; Kohlbrenner, W. E.; Wide-
cost-efficient synthesis of simvastatin via high-conversion methylation burg, N. E.; Saldivar, A.; Vasavanonda, S.; Marsh, K. C.; Bryant, P.;
of an alkoxide ester enolate. Org. Process Res. Dev. 1999, 3, 476–479. Sham, H. L.; Green, B. E.; Betebenner, D. A.; Erickson, J.; Norbeck,
140. (a) Stokker, G. E.; Rooney, C. S.; Wiggins, J. M.; Hirschfield, J. Synthe- D. W. Symmetry-based inhibitors of HIV protease. Structure-activity
sis and x-ray characterization of 6(S)-epimevinolin, a lactone epimer. J. studies of acylated 2,4-diamino-1,5-diphenyl-3-hydroxypentane and
Org. Chem. 1986, 51, 4931–4934. (b) Heathcock, C. H.; Hadley, C. R.; 2,5-diamino-1,6-diphenylhexane-3,4-diol. J. Med. Chem. 1993, 36,
Rosen, T.; Theisen, P. D.; Hecker, S. J. J. Med. Chem. 1987, 30, 1858. 320–330.
141. Mundy, G.; Garrett, R.; Harris, S.; Chan, J.; Chen, D.; Rossini, G.; 152. Kempf, D. J.; Norbeck, D. W.; Codacovi, L.; Wang, X. C.; Kohl-
Boyce, B.; Zhao, M.; Gutierrez, G. Stimulation of bone formation brenner, W. F.; Wideburg, N. E.; Saldivar, A.; Craig-Kennard, A.;
in vitro and in rodents by statins. Science 1999, 286, 1946–1949. Vasavanonda, S.; Clement, J. J.; Erickson, J. Recent Advances in
142. Istvan, E. S.; Deisenhofer, J. Structural mechanism for statin inhibi- the Chemistry of Anti-Infective Agents.Bentley, P. H.; Ponsford,
tion of HMG-CoA reductase. Science 2001, 292, 1160–1164. R. (Eds.), Royal Society of Chemistry, Cambridge, 1993; pp.
143. American Diabetes Association. http://www.diabetes. Org/diabetes- 297–313.
basics/; World Health Organization. http://www.who.int/diabetes/ 153. Kempf, D. J.; Marsh, K. C.; Fino, L. C.; Bryant, P.; Craig-Kennard,
facts/en/, National Diabetes Education Program. http://ndep.nih.gov/ A.; Sham, H. L.; Zhao, C.; Vasavanonda, S.; Kohlbrenner, W. E.
diabetes-facts/index.aspx. Design of orally bioavailable, symmetry-based inhibitors of HIV
144. (a) van Genugten, R. E.; Raalte, D. H.; Diamant, M. Dipeptidyl protease. Bioorg. Med. Chem. 1994, 2, 847–858.
peptidase-4 inhibitors and preservation of pancreatic islet-cell func- 154. Kempf, D. J.; Marsh, K. C.; Denissen, J. F.; McDonald, E.; Vasava-
tion: a critical appraisal of the evidence and Rasagiline. Diabetes nonda, S.; Flentge, C. A.; Green, B. E.; Fino, L.; Park, C. H. ABT-
Obes. Metab. 2012, 14, 101–111. (b) Matteucci, E.; Giampietro, 538 is a potent inhibitor of human immunodeficiency virus protease
O. Dipetidyl peptidase-4 inhibition: linking chemical properties to and has high oral bioavailability in humans. Proc. Natl. Acad. Sci.
clinical safety. Curr. Med. Chem. 2011, 18, 4753–4760. (c) Ahren, U.S.A. 1995, 92, 2484–2488.
B. Inhibition of dipetidyl peptidase-4 (DPP-4): a target to treat type 2 155. Kumar, G. N.; Gravowski, B.; Lee, R.; Denissen, J. F. Hepatic drug-
diabetes. Curr. Enzym. Inhib. 2011, 7, 205–217. metabolizing activities in rats after 14 days of oral administration of
145. Ashworth, D. M.; Atrash, B.; Baker, G. R.; Baxter, A. J.; Jenkins, P, D.; the human immunodeficiency virus-type 1 protease inhibitor ritona-
Jones, D. M.; Szelke, M. 2-Cyanopyrrolidides as potent, stable inhib- vir (ABT-538). Drug Metab. Dispos. 1996, 24, 615–617.
itors of dipeptidyl peptidase IV. Bioorg. Med. Chem. Lett. 1996, 6, 156. Kempf, D. J.; Marsh, K. C.; Kumar, G.; Rodrigues, A. D.; Denissen,
1163–1166. J. F.; McDonald, E.; Kukulka, M. J.; Hsu, A.; Granneman, G. R.;
146. Robl, J. A.; Hamman, L. H. The discovery of the dipeptidyl dipep- Baroldi, P. A.; Sun, E.; Pizzuti, D.; Plattner, J. J.; Norbeck, D. W.;
tidase (DPP4) inhibitor Onglyza: from concept to market. In Leonard, J. M. Pharmacokinetic enhancement of inhibitors of the
Accounts in Drug Discovery: Case Studies in Medicinal Chemistry, human immunodeficiency virus protease by coadministration with
Barrish, J. C.; Carter, P. C.; Cheng, P. T. W.; Zahler, R. (Ed.), RSC ritonavir. Antimicrob. Agents Chemother. 1997, 41, 654–660.
Publishing, Cambridge, 2011, Chapter 1, pp. 1–24, http://dx.doi. 157. Kempf, D. J.; Sham, H. L.; Marsch, K. C.; Flentge, C. A.; Beteben-
org/10.1039/9781849731980-00001. ner, D.; Green, B. E.; McDonald, E.; Vasavanonda, S.; Saldivar, A.;
147. (a) Kramer, R. A.; Schaber, M. S.; Skalka, A. M.; Ganguly, K.; Wideburg, N. E.; Kati, W. M.; Ruiz, L.; Zhao, C.; Fino, L.; Patterson,
Wong-Staal, F.; Reedy, E. P. HTLV-III gag protein is processed in J.; Molla, A.; Plattner, J. J.; Norbeck, D. W. Discovery of ritonavir,
yeast cells by the virus pol-protease. Science 1986, 231, 1580–1584. a potent inhibitor of HIV protease with high oral bioavailability and
(b) Debouck, C.; Gorniak, J. G.; Strickler, J. E.; Meek, T. D.; Met- clinical efficacy. J. Med. Chem. 1998, 41, 602–617.
calf, B.W.; Rosenberg, M. Human immunodeficiency virus protease 158. Smith, A. J. T.; Zhang, X.; Leach, A. G.; Houk, K. N. Beyond pico-
expressed in Escherichia coli exhibits autoprocessing and specific molar affinities: quantitative aspects of noncovalent and covalent
maturation of the gag precursor. Proc. Natl. Acad. Sci. 1987, 84, binding of drugs to proteins. J. Med. Chem. 2009, 52, 225.
8903–8906. (c) Kohl, N. E.; Emini, E. A.; Schleif, W. A.; Davis, L. 159. Copeland, R.A. Irreversible enzyme inactivators In Evaluation of
J.; Heimbach, J. C.; Dixon, R. A. F.; Scolnick, E. M.; Sigal, I. S. Enzyme Inhibitors in Drug Discovery: A Guide for Medicinal Chem-
Active human immunodeficiency virus protease is required for viral ists and Pharmacologists, 2nd Edition, John Wiley & Sons: Hoboken,
infectivity. Proc. Natl. Acad. Sci. 1988, 85, 4686–4690. N.J., 2013, Chapter 9, pp. 345–382.
148. Kempf, D. J.; Norbeck, D. W.; Codacovi, L. M.; Wang, X. C.; Kohl- 160. Lewis, P. J.; Richen, E. Vigabatrin: a new antiepileptic drug. Br. J.
brenner, W. E.; Wideburg, N. E.; Paul, D. A.; Knigge, M. F.; Vasava- Clin. Pharmacol. 1989, 27(Suppl 1), 1S–12S.
nonda, S.; Craigkennard, A.; Saldivar, A.; Rosenbrook, W.; Clement, 161. Gram, L.; Larsson, O. M.; Johnsen, A.; Schoesboe, A. Experimental
J. J.; Plattner, J. J.; Erickson, J. Structure-based, C2 symmetric inhib- studies of the influence of vigabatrin on the GABA system. Br. J.
itors of HIV protease. J. Med. Chem. 1990, 33, 2687–2689. Clin. Pharmacol. 1989, 27(Suppl 1), 13S–17S.
149. Plattner, J. J.; Norbeck, D. W. In Drug Discovery Technologies; 162. Messenheimer, J. A. Lamotrigine. Clin. Neuropharmacol. 1994, 17,
Clark, R.; Moos, W. H., Eds.; Ellis Horwood Ltd: Chichester, 1990; 548–559.
pp. 92–126. 163. Browne, T. R; Mattson, R. H.; Penry, J. K.; Smith, D. B.; Wilder,
150. Erickson, J.; Neidhart, D. J.; Vandrie, J.; Kempf, D. J.; Wang, X. C.; B.J.; Treiman, D. M.; Ben-Menachem, E.; Miketta, R. M.; Sherry,
Norbeck, D. W.; Plattner, J. J.; Rittenhouse, J. W.; Turon, M.; Wideburg, K. M.; Szabo, G. K. A multicenter study of vigabatrin for drug
N.; Kohlbrenner, W. E.; Simmer, R.; Helfrich, R.; Paul, D. A.; resistant epilepsy. Br. J. Clin. Pharmacol. 1989, 27(Suppl. 1),
Knigge, M. Design, activity, and 2.8. ANG. crystal structure of a C2 95S–100S.
symmetric inhibitor complexed to HIV-1 protease. Science 1990, 164. Barf, T.; Kaptein, A. Irreversible protein kinase inhibitors: balancing
249, 527–533. the benefits and risks. J. Med. Chem. 2012, 55, 6243–6262.
Chapter | 5 Enzyme Inhibition and Inactivation 271

165. Lippert, B.; Jung, M. J.; Metcalf, B. W. Biochemical consequences 182. Yocum, R. R.; Rasmussen, J. R.; Strominger, J. L. The mechanism
of reactions catalyzed by GAD and GABA-T. Brain Res. Bull. 1980, of action of penicillin. Penicillin acylates the active site of Bacil-
5(Supp. 2), 375–379. lus stearothermophilus D-alanine carboxypeptidase. J. Biol. Chem.
166. Jeffreys, D. Aspirin: The remarkable story of a wonder drug, 1980, 255, 3977–3986.
Bloomsbury, New York, 2004. 183. Tipper, D. J.; Strominger, J. L. Mechanism of action of penicillins;
167. Robertson, J. G. Mechanistic basis of enzyme-targeted drugs. Bio- a proposal based on their structural similarity to acyl-D-alanyl-D-
chemistry 2005, 44, 5561–5571. alanine. Proc. Natl. Acad. Sci. U.S.A. 1965, 54, 1133–1141.
168. (a) Erve, J. C. Chemical toxicology: reactive intermediates and their 184. Izaki, K.; Matsuhashi, M.; Strominger, J. L. Biosynthesis of the pep-
role in pharmacology and toxicology. Exp. Opin. Drug Metab. 2006, tidoglycan of bacterial cell walls. xiii. peptidoglycan transpeptidase
2, 923–946. (b) Baillie, T. C. Future of toxicology-metabolic activa- and d-alanine carboxypeptidase: penicil. J. Biol. Chem. 1968, 243,
tion and drug design: challenges and opportunities in chemical toxi- 3180–3192.
cology. Chem. Res. Toxicol. 2006, 19, 889–893. 185. Lee, W.; McDonough, M. A.; Kotra, L. P.; Li, Z.-H.; Silvaggi, N. R.;
169. (a) Uetrecht, J. Immune-mediated adverse drug reactions. Chem. Takeda, Y.; Kelly, J. A.; Mobashery, S. A 1.2-.ANG. snapshot of the
Res. Toxicol. 2009, 22, 24–34. (b) Naisbitt, D. J.; Gordon, S. F.; Pir- final step of bacterial cell wall biosynthesis. Proc. Natl. Acad. Sci.
mohamed, M.; Park, B. K. Immunological principles of adverse drug U.S.A. 2001, 98, 1427–1431.
reactions. Drug Saf. 2000, 23, 483–507. 186. Kuzin, A.; Liu, H., Kelly, J. A.; Knox, J. R. Binding of cephalothin
170. Krantz, A. In Advances in Medicinal Chemistry, JAI Press, London, and cefotaxime to D-Ala-D-Ala-peptidase reveals a functional basis
1992, Vol. 1, pp. 235–261. of a natural mutation in a low-affinity penicillin-binding protein and in
171. Smith, R. A.; Copp, L. J.; Coles, P. J.; Pauls, H. W.; Robinson, V. extended-spectrum β -lactamases. Biochemistry 1995, 34, 9532–9540.
J.; Spencer, R. W.; Heard, S. B.; Krantz, A. New inhibitors of cys- 187. Sweet, R. M.; Dahl, K. F. Molecular architecture of the cephalospo-
teine proteinases. Peptidyl acyloxymethyl ketones and the quiescent rins. Insights into biological activity based on structural investiga-
nucleofuge strategy. J. Am. Chem. Soc. 1988, 110, 4429–4431. tions. J. Am. Chem. Soc. 1970, 92, 5489–5507.
172. Kominami, E.; Tsukahara, T.; Bando, Y.; Katunuma, N. Distribution 188. Kohanski, M. A.; Dwyer, D. J.; Hayete, B.; Lawrence, C. A.; Collins,
of cathepsins B and H in rat tissues and peripheral blood cells. J. J. J. A common mechanism of cellular death induced by bactericidal
Biochem. 1985, 98, 87–93. antibiotics. Cell 2007, 130, 797–810.
173. Sloane, B. F.; Lah, T. T.; Day, N. A.; Rozhin, J.; Bando, Y.; Honn, 189. Böhme, E. H. W.; Applegate, H. E.; Toeplitz, B.; Dolfini, J. E.; Goug-
K. V. In Cysteine Proteinases and Their Inhibitors, Turk, V. (Ed.), outas, J. Z. 6-methyl penicillins and 7-methyl cephalosporins. J. Am.
Walter de Gruyter: New York, 1986, pp. 729–749. Chem. Soc. 1971, 93, 4324–4326.
174. Prous, J. R. (Ed.), EST. Drugs Future 1986, 11, 927–930. 190. Kotra, L. P.; Golemi, D.; Vakulenko, S.; Mobashery, S. Bacteria fight
175. Singh, J.; Petter, R. C.; Baillie, T. A.; Whitty, A. The resurgence of back. Chem. Ind. 22 May 2000, 341–344.
covalent drugs. Nature Rev. Drug Discov. 2011, 10, 307–317. 191. Gross, M.; Greenberg, L. A. The Salicylates. A Critical Bibliographic
176. (a) Ulrich, R. Idiosyncratic toxicity: a convergence of risk factors. Annu. Review, Hillhouse, New Haven, 1948.
Rev. Med. 2007, 58, 17–34. (b) Uetrecht, J. Idiosyncratic drug reactions: 192. Margotta, R., In An Illustrated History of Medicine, Lewis, P. (Ed.),
past, present, and future. Chem. Res. Toxicol. 2008, 21, 84–92. Paul Hamlyn, London, 1968.
177. (a) Lammert, C.; Einarsson, S.; Saha, C.; Niklasson, A.; Bjornsson, 193. Stone, E. An account of the success of the bark of the willow in
E.; Chalasani, N. Relationship between daily dose of oral medica- the cure of ages. In a letter to the Right Honourable George Earl of
tions and idiosyncratic drug-induced liver injury: search for signals. Macclesfield, President of R. S. from the Rev. Mr. Edmund Stone,
Hepatology 2008, 47, 2003–2009. (b) Kalgutkar, A. S.; Gardner, I.; of Chipping-Norton in Oxfordshire. Philos. Trans. R. Soc. London
Obach, R. S.; Shaffer, C. L.; Callegari, E.; Henne, K. R.; Mutlib, 1963, 53, 195–200.
A. E.; Dalvie, D. K.; Lee, J. S.; Nakai, Y. A comprehensive list- 194. Martin, B. K., In Salicylates, An International Symposium, Dixon, A.
ing of bioactivation pathways of organic functional groups. Curr. St. J.; Martin, B. K.; Smith, M. V. H.; Wood, R. H. N. (Eds.), Little,
Drug Metab. 2005, 6, 161–225. (c) Nakayama, S. A zone classifica- Brown and Company, Boston, 1963, p. 6.
tion system for risk assessment of idiosyncratic drug toxicity using 195. Jourdier, S. A miracle drug. Chem. Br. 1999, 35, 33–35.
daily dose and covalent binding. Drug Metab. Dispos. 2009, 37, 196. Vane, J. R. Inhibition of prostaglandin synthesis as a mechanism
1970–1977. of action for aspirin-like drugs. Nature New Biol. 1971, 231,
178. Neuhaus, F. C.; Georgopapadakou, N. H. In Emerging Targets for 232–235.
Antibacterial and Antifungal Chemotherapy, Sutcliffe, J.; Georgo- 197. Smith, J. B.; Willis, A. L. Aspirin selectively inhibits prostaglan-
papadakou, N. H. (Eds.), Chapman and Hall, New York, 1992, pp. din production in human platelets. Nature New Biol. 1971, 231,
206–273. 235–237.
179. Mandell, G. L. In Principles and Practice of Infectious Diseases, 2nd 198. Roth, G. J.; Stanford, N.; Majerus, P. W. Acetylation of prostaglan-
ed., Mandell, G. L.; Douglas, R. G., Jr.; Bennett, J. E. (Eds.), Wiley: din synthase by aspirin. Proc. Natl. Acad. Sci. U.S.A. 1975, 72,
New York, 1985; p. 180. 3073–3076.
180. (a) Sammes, P. G. (Ed.), Topics in Antibiotic Chemistry, Ellis Hor- 199. (a) Hemler, M.; Lands, W. E. M.; Smith, W. L. Purification of
wood, Chichester, 1980; Vol. 4. (b) Brown, A. G.; Roberts, S. M. the cyclooxygenase that forms prostaglandins. Demonstration of
(Eds.), Recent Advances in the Chemistry of β-Lactam Antibiotics, two forms of iron in the holoenzyme. J. Biol. Chem. 1976, 251,
Royal Society of Chemistry, London, 1985. 5575–5579. (b) Van der Ouderaa, F. J.; Buytenhek, M.; Nugteren,
181. Bush, K.; Mobashery, S. How β -lactamases have driven pharma- D. H.; Van Dorp, D. A. Acetylation of prostaglandin endoperox-
ceutical drug discovery: From mechanistic knowledge to clinical cir- ide synthetase with acetylsalicylic acid. Eur. J. Biochem. 1980,
cumvention. Adv. Exp. Med. Biol. 1998, 456, 71–98. 109, 1–8.
272 The Organic Chemistry of Drug Design and Drug Action

200. Roth, G. J.; Machuga, E. T.; Ozols, J. Isolation and covalent structure I.; Tagari, P.; Therien, M.; Vickers, P.; Wong, E.; Xu, L. J.; Young,
of the aspirin-modified, active-site region of prostaglandin synthe- R. N.; Zamboni, R.; Boyce, S.; Rupniak, N.; Forrest, N.; Visco,
tase. Biochemistry 1983, 22, 4672–4675. D.; Patrick, D. The discovery of rofecoxib, (MK 966, Vioxx,
201. (a) Van der Ouderaa, F. J.; Buytenhek, M.; Nugteren, D. H.; Van 4-(4’-methylsulfonylphenyl)-3-phenyl-2(5H)-furanone), an orally
Dorp, D. A. Acetylation of prostaglandin endoperoxide synthe- active cyclooxygenase-2 inhibitor. Bioorg. Med. Chem. Lett. 1999, 9,
tase with acetylsalicylic acid. Eur. J. Biochem. 1980, 109, 1–8. (b) 1773–1778. (b) Chan, C.-C.; Boyce, S.; Brideau, C.; Charleson, S.;
DeWitt, D. L.; El-Harith, E. A.; Kraemer, S. A.; Andrews, M. J. Yao, Cromlish, W.; Ethier, D.; Evans, J.; Ford-Hutchinson, A. W.; Forrest,
E. F.; Armstrong, R. L.; Smith, W. L. The aspirin and heme-binding M. J.; Gauthier, J. Y.; Gordon, R.; Gresser, M.; Guay, J.; Kargman, S.;
sites of ovine and murine prostaglandin endoperoxide synthases. J. Kennedy, B.; Leblanc, Y.; Leger, S.; Mancini, J.; O’Neill, G. P.; Ouel-
Biol. Chem. 1990, 265, 5192–5198. let, M.; Patrick, D.; Percival, M. D.; Perrier, H.; Prasit, P.; Rodger,
202. Hochgesang, G. P., Jr.; Rowlinson, S. W.; Marnett, L. J. Tyrosine-385 I.; Tagari, P.; Therien, M.; Vickers, P.; Visco, D.; Wang, Z.; Webb,
is critical for acetylation of cyclooxygenase-2 by aspirin. J. Am. J.; Wong, E.; Xu, L.-J.; Young, R. N.; Zamboni, R.; Riendeau, D.
Chem. Soc. 2000, 122, 6514–6515. Rofecoxib (Vioxx, MK-0966; 4-(4’-methylsulfonylphenyl)-3-phe-
203. (a) Raz, A.; Wyche, A.; Siegel, N.; Needleman, P. Regulation of fibro- nyl-2-(5H)-furanone): a potent and orally active cyclooxygenase-2
blast cyclooxygenase synthesis by interleukin-1. J. Biol. Chem. 1988, inhibitor. Pharmacological and biochemical profiles. J. Pharmacol.
263, 3022. (b) Masferrer, J. L.; Zweifel, B. S.; Seibert, K.; Needleman, Exp. Ther. 1999, 290, 551–560.
P. Selective regulation of cellular cyclooxygenase by dexamethasone 210. Talley, J. J.; Brown, D. L.; Carter, J. S.; Graneto, M. J.; Koboldt, C.
and endotoxin in mice. J. Clin. Invest. 1990, 86, 1375–1379. M.; Masferrer, J. L.; Perkins, W. E.; Rogers, R. S.; Shaffer, A. F.;
204. (a) Xie, W.; Chipman, J. G.; Robertson, D. L.; Erikson, R. L.; Zhang, Y. Y.; Zweifel, B. S.; Seibert, K. 4-[5-methyl-3-phenylisox-
Simmons, D. L. Expression of a mitogen-responsive gene encoding azol-4-yl]- benzenesulfonamide, valdecoxib: A potent and selective
prostaglandin synthase is regulated by mRNA splicing. Proc. Natl. inhibitor of COX-2. J. Med. Chem. 2000, 43, 775–777.
Acad. Sci. U.S.A. 1991, 88, 2692–2696. (b) Kujubu, D. A.; Fletcher, 211. Rubin, R.; How the Vioxx debacle happen? USA Today, October
B. S.; Varnum, C. R.; Lim, W.; Herschman, H. TIS10, a phorbol ester 12, 2004, http://www.usatoday.com/news/health/2004-10-12-vioxx-
tumor promoter-inducible mRNA from Swiss 3T3 cells, encodes cover_x.htm.
a novel prostaglandin synthase/cyclooxygenase homolog. J. Biol. 212. Egan, K. M.; Lawson, J. A.; Fries, S.; Koller, B.; Rader, D. J.; Smyth,
Chem. 1991, 266, 12866–2872. E. M.; FitzGerald, G. A. COX-2-derived prostacyclin confers athero-
205. (a) Tally, J. J. Selective Inhibitors of cyclooxygenase-2. Exp. Opin. protection on female mice. Science 2004, 306, 1954–1957.
Ther. Pat. 1997, 7, 55–62. (b) Bjorkman, D. J. Nonsteroidal anti- 213. Dogné, J.-M.; Supuran, C. T.; Pratico, D. Adverse cardiovascular
inflammatory drug-induced gastrointestinal injury. Am. J. Med. effects of the coxibs. J. Med. Chem. 2005, 48, 2251–2257.
1996, 101(Suppl. A), 25S–32S. (c)Seibert, K.; Zhang, Y.; Leahy, K.; 214. Piomelli, D.; Giuffrida, A.; Calignano, A.; Rodriguez de Fonseca, F.
Hauser, S.; Masferrer, J.; Perkins, W.; Lee, L.; Isakson, P. Pharmaco- The endocannabinoid system as a target for therapeutic drugs. Trends
logical and biochemical demonstration of the role of cyclooxygenase Pharmacol. Sci. 2000, 21, 218–224.
2 in inflammation and pain. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 215. Duggan, K. C.; Hermanson, D. J.; Musse, J.; Prusakiewics, J. J.;
12013–12017. Scheib, J. L.; Carter, B. Bannerjee, S.; Oates, J. A.; Marnett, L. H.
206. Khanna, I. K.; Weier, R. M.; Yu, Y.; Collins, P. W.; Miyashiro, J. (R)-Profens are substrate-selective inhibitors of endocannabinoid
M.; Koboldt, C. M.; Veenhuizen, A. W.; Currie, J. L.; Seibert, K.; oxygenation by COX-2. Nature Chem. Biol. 2011, 7, 803–809.
Isakson, P. C. 1,2-Diarylpyrroles as potent and selective inhibitors of 216. (a) Chandrasekharan, N. V.; Dai, H.; Turepu Roos, K. L.; Evanson,
cyclooxygenase-2. J. Med. Chem. 1997, 40, 1619–1633. N. K.; Tomsik, J.; Elton, T. S.; Simmons, D. L. COX-3, a cyclooxy-
207. Khanna, I. K.; Weier, R. M.; Yu, Y.; Xu, X. D.; Koszyk, F. J.; Collins, genase-1 variant inhibited by acetaminophen and other analgesic/
P. W.; Koboldt, C. M.; Veenhuizen, A. W.; Perkins, W. E.; Casler, J. antipyretic drugs: cloning, structure, and expression. Proc. Natl. Acad.
J.; Masferrer, J. L.; Zhang, Y. Y.; Gregory, S. A.; Seibert, K.; Isak- Sci. U.S.A. 2002, 99, 13926–13931. (b) Botting, R. M. Mechanism of
son, P. C. 1,2-Diarylimidazoles as potent, cyclooxygenase-2 selective action of acetaminophen: Is there a cyclooxygenase 3? Clin. Infect. Dis.
and orally active antiinflammatory agents. J. Med. Chem. 1997, 40, 2000, 31(Suppl. 5), S202–S210. (c) Willoughby, D. A.; Moore, A. R.;
1634–1647. Colville-Nash, P. R. COX-1, COX-2, and COX-3 and the future treat-
208. Penning, T. D.; Talley, J. J.; Bertenshaw, S. R.; Carter, J. S.; Col- ment of chronic inflammatory disease. Lancet 2000, 355, 646–648.
lins, P. W.; Docter, S.; Graneto, M. J.; Lee, L. F.; Malecha, J. W.; 217. Matheson, A. J.; Figgitt, D. P. Rofecoxib: a review of its use in the
Miyashiro, J. M.; Rogers, R. S.; Rogier, D. J.; Yu, S. S.; Anderson, management of osteoarthritis, acute pain and rheumatoid arthritis.
G. D.; Burton, E. G.; Cogburn, J. N.; Gregory, S. A.; Koboldt, C. Drugs 2001, 61, 833–865.
M.; Perkins, W. E.; Seibert, K.; Veenhuizen, A. W.; Zhang, Y. Y.; 218. Warner, T. D.; Giuliano, F.; Vojnovic, I.; Bukasa, A.; Mitchell, J.
Isakson, P. C. Synthesis and biological evaluation of the 1,5-dia- A.; Vane, J. R. Nonsteroid drug selectivities for cyclo-oxygenase-1
rylpyrazole class of cyclooxygenase-2 inhibitors: Identification of rather than cyclo-oxygenase-2 are associated with human gastroin-
4-[5-(4-methylphenyl)-3-(trifluoromethyl)-1H-pyrazol-1-yl]ben- testinal toxicity: a full in vitro analysis. Proc. Natl. Acad. Sci. U.S.A.
zenesulfonamide (SC-58635, celecoxib). J. Med. Chem. 1997, 40, 1999, 96, 7563–7568.
1347–1365. 219. Ochi, T.; Motoyama, Y.; Goto, T. The analgesic effect profile of
209. (a) Prasit, P.; Wang, Z.; Brideau, C.; Chan, C. C.; Charleson, S.; FR122047, a selective cyclooxygenase-1 inhibitor, in chemical noci-
Cromlish, W.; Ethier, D.; Evans, J. F.; Ford-Hutchinson, A. W.; ceptive models. Eur. J. Pharmacol. 2000, 391, 49–54.
Gauthier, J. Y.; Gordon, R.; Guay, J.; Gresser, M.; Kargman, S.; 220. Buckley, M. M.; Brogden, R. N. Ketorolac. A review of its pharma-
Kennedy, B.; Leblanc, Y.; Leger, S.; Mancini, J.; O’Neill, G. P.; codynamic and pharmacokinetic properties, and therapeutic poten-
Ouellet, M.; Percival, M. D.; Perrier, H.; Riendeau, D.; Rodger, tial. Drugs 1990, 39, 86–109.
Chapter | 5 Enzyme Inhibition and Inactivation 273

221. Picot, D.; Loll, P. J.; Garavito, R. M. The X-ray crystal structure of 237. Schechter, P. J.; Barlow, J. L. R.; Sjoerdsma, A. In Inhibition of Poly-
the membrane protein prostaglandin H2 synthase-1. Nature 1994, amine Metabolism. Biological Significance and Basis for New Ther-
367, 243–249. apies; McCann, P. P.; Pegg, A. E.; Sjoerdsma, A. (Eds.), Academic,
222. Kurumbail, R. G.; Stevens, A. M.; Gierse, J. K.; McDonald, J. J.; Orlando, FL, 1987, p. 345.
Stegeman, R. A.; Pak, J. Y.; Gildehaus, D.; Miyashiro, J. M.; Pen- 238. (a) Kuzoe, F. A. S. Current situation of African trypanosomiasis. Acta
ning, T. D.; Seibert, K.; Isakson, P. C.; Stallings, W. C. Structural Trop. 1993, 54, 153–162. (b) Pegg, A. E.; Shantz, L. M.; Coleman, C.
basis for selective inhibition of cyclooxygenase-2 by anti-inflamma- S. Ornithine decarboxylase as a target for chemoprevention. J. Cell
tory agents. Nature 1996, 384, 644–648. Biochem. 1995, 22, 132–138. (c) Wang, C. C. In Burger’s Medici-
223. Gierse, J. K.; McDonald, J. J.; Hauser, S. D.; Rangwala, S. H.; nal Chemistry and Drug Discovery, 5th ed., Wolff, M. E. (Ed.), John
Koboldt, C. M.; Seibert, K. A single amino acid difference between Wiley & Sons, New York, 1997; Vol. 4, p. 459.
cyclooxygenase-1 (COX-1) and -2 (COX-2) reverses the selec- 239. McCann, P. P.; Pegg, A. E. Ornithine decarboxylase as an enzyme
tivity of COX-2 specific inhibitors. J. Biol. Chem. 1996, 271, target for therapy. Pharmacol. Ther. 1992, 54, 195–215.
15810–15814. 240. Danzin, C.; Bey, P.; Schirlin, D.; Claverie, N. α -Monofluoromethyl
224. Shan, B.; Medina, J. C.; Santha, E.; Frankmoelle, W. P.; and α -difluoromethyl putrescine as ornithine decarboxylase inhibi-
Chou, T. C.; Learned, R. M.; Narbut, M. R.; Stott, D.; Wu, P. G.; tors: in vitro and in vivo biochemical properties. Biochem. Pharma-
Jaen, J. C.; Rosen, T.; Timmermans, P. B. M. W. M.; Beckman, col. 1982, 31, 3871–3878.
H. Selective, covalent modification of β-tubulin residue Cys-239 241. Selikoff, I. J.; Robitzek, E. H.; Ornstein, G. G. Toxicity of hydrazine
by T138067, an antitumor agent with in vivo efficacy against derivatives of isonicotinic acid in the chemotherapy of human tuber-
multidrug-resistant tumors. Proc. Natl. Acad. Sci. U.S.A. 1999, culosis. Quart. Bull. Seaview Hosp. 1952, 13, 17–27.
96, 5686–5691. 242. Zeller, E. A.; Barsky, J.; Fouts, J. P.; Kirchheimer, W. F.; Van Orden,
225. Medina, J. C.; Roche, D.; Shan, B.; Learned, R. M.; Frankmoelle, L. S. Influence of isonicotinic acid hydrazide and 1-isonicotinoyl-
W. P.; Clark, D. L.; Rosen, T.; Jaen, J. C. Novel halogenated sulfon- 2-isopropylhydrazine on bacterial and mammalian enzymes. Experi-
amides inhibit the growth of multidrug resistant MCF-7/ADR cancer entia 1952, 8, 349–350.
cells. Bioorg. Med. Chem. Lett. 1999, 9, 1843–1846. 243. (a) Kline, N. S. Clinical experience with iproniazid (marsilid). J. Clin.
226. Silverman, R.B.; Invergo, B.J. Mechanism of inactivation of Exp. Psychopathol. Quart. Rev. Psychiat. Neurol. 1958, 19(Suppl. 1),
γ-aminobutyrate aminotransferase by 4-amino-5-fluoropentanoic 72–78. (b)Zeller, E. A. (Ed.), In vitro and in vivo inhibition of amine
acid. First example of an enamine mechanism for a γ-amino acid oxidases. Ann. N. Y. Acad. Sci. 1959, 80, 583–589.
with a partition ratio of 0. Biochemistry 1986, 25, 6817–6820. 244. Ganrot, P. O.; Rosengren, E.; Gottfries, C. G. Effect of iproniazid
227. Nelson, S. D. Metabolic activation and drug toxicity. J. Med. Chem. on monoamines and monamine oxidase in human brain. Experientia
1982, 25, 753–765. 1962, 18, 260–261.
228. Schechter, P. J.; Barlow, J. L. R.; Sjoerdsma, A. In Inhibition of Poly- 245. Dostert, P. L.; Strolin Benedetti, M.; Tipton, K. F. Interactions of
amine Metabolism. Biological Significance and Basis for New Thera- monoamine oxidase with substrates and inhibitors. Med. Res. Rev.
pies. McCann, P. P.; Pegg, A. E.; Sjoerdsma, A. (Eds.), Academic, 1989, 9, 45–89.
Orlando, FL, 1987, p. 345. 246. (a) Squires, R. F.; Lassen, J. B. Pharmacological and biochemi-
229. Isaacson, E. I.; Delgado, J. N. In Burger’s Medicinal Chemistry. 4th cal properties of γ-morpholinobutyrophenone (NSD 2023), a new
ed.; Wolff, M. E. (Ed.), Wiley, New York, 1981, Part III; p. 829. monoamine oxidase inhibitor. Biochem. Pharmacol. 1968, 17,
230. Houser, W. A. In Epilepsy. A Comprehensive Textbook, Engel, J.; 369–384. (b) Squires, R. F. Additional evidence for the existence of
Pedley, T. A. (Eds.), Lippincott-Raven, Philadelphia, 1997, Vol. 1, several forms of mitochondrial monoamine oxidase in the mouse.
Section 1. Biochem. Pharmacol. 1968, 17, 1401. (c) Johnston, J. P. Some obser-
231. Nanavati, S. M.; Silverman, R. B. Design of potential anticonvulsant vations on a new inhibitor of monoamine oxidase in brain tissue. Bio-
agents: mechanistic classification of GABA aminotransferase inacti- chem. Pharmacol. 1968, 17, 1285–1297.
vators. J. Med. Chem. 1989, 32, 2413–2421. 247. (a) Palfreyman, M. G.; McDonald, I. A.; Bey, P.; Schechter, P. J.;
232. Iadarola, M. J.; Gale, K. Substantia nigra: site of anticonvulsant Sjoerdsma, A. Design and early clinical evaluation of selective
activity mediated by γ-aminobutyric acid. Science 1982, 218, inhibitors of monoamine oxidase. Prog. Neuropsychopharma-
1237–1240. col. Biol. Psychiat. 1988, 12, 967–987. (b) McDonald, I. A.; Bey,
233. Lippert, B.; Metcalf, B. W.; Jung, M. J.; Casara, P. 4-Amino-hex- P.; Palfreyman, M. G. In Design of Enzyme Inhibitors as Drugs,
5-enoic acid, a selective catalytic inhibitor of 4-aminobutyric-acid Sandler, M.; Smith, H. J. (Eds.), Oxford University Press: Oxford,
aminotransferase in mammalian brain. Eur. J. Biochem. 1977, 74, 1989; p. 227.
441–445. 248. Green, L. D.; Dawkins, K. In Burger’s Medicinal Chemistry and
234. De Biase, D.; Barra, D.; Bossa, F.; Pucci, P.; John, R. A. Chemistry Drug Discovery, 5th ed., Wolff, M. E., Ed.; Wiley: New York, 1997;
of the inactivation of 4-aminobutyrate aminotransferase by the anti- Vol. 5, p. 121.
epileptic drug vigabatrin. J. Biol. Chem. 1991, 266, 20056–20061. 249. Maeda, Y.; Ingold, K. U. Kinetic applications of electron paramag-
235. Nanavati, S. M.; Silverman, R. B. Mechanisms of inactiva- netic resonance spectroscopy. 35. The search for a dialkylaminyl
tion of γ-aminobutyric acid aminotransferase by the antiepilepsy rearrangement. Ring opening of N-cyclobutyl-N-n-propylaminyl. J.
drug γ-vinyl GABA (vigabatrin). J. Am. Chem. Soc. 1991, 113, Am. Chem. Soc. 1980, 102, 328–331.
9341–9349. 250. Silverman, R. B. Mechanism of inactivation of monoamine
236. Pegg, A. E. Polyamine metabolism and its importance in neoplas- oxidase by trans-2-phenylcyclopropylamine and the structure
tic growth and as a target for chemotherapy. Cancer Res. 1988, 48, of the enzyme-inactivator adduct. J. Biol. Chem. 1983, 258,
759–754. 14766–14769.
274 The Organic Chemistry of Drug Design and Drug Action

251. Paech, C.; Salach, J. I.; Singer, T. P. Suicide inactivation of mono- 264. Markey, S. P.; Johannessen, J. N.; Chiueh, C. C.; Burns, R. S.;
amine oxidase by trans-phenylcyclopropylamine. J. Biol. Chem. Herkenham, M. A. Intraneuronal generation of a pyridinium metabo-
1980, 255, 2700. lite may cause drug-induced parkinsonism. Nature 1984, 311, 464.
252. In MAO-B Inhibitor Selegiline (R-(-)-Deprenyl). Riederer, P.; Przun- 265. Javitch, J. A.; D’Amato, R. J.; Strittmater, S. M.; Snyder, S. H. Par-
tek, H. (Eds.), Springer-Verlag: Wein, 1987. kinsonism-inducing neurotoxin, N-methyl-4-phenyl-1,2,3,6-tetrahy-
253. Weinreb, O.; Amit, T.; Riederer, P.; Youdim, M. B. H.; Mandel, S. A. dropyridine: uptake of the metabolite N-methyl-4-phenylpyridine by
Neuroprotective profile of the multitarget drug rasagiline in Parkin- dopamine neurons explains selective toxicity. Proc. Natl. Acad. Sci.
son’s disease. Int. Rev. Neurobiol. 2011, 100, 127–149. U.S.A. 1985, 82, 2173–2177.
254. Davis, G. C.; Williams, A. C.; Markey, S. P.; Ebert, M. H.; Caine, E. 266. (a) Priyadarshi, A.; Khuder, S. A.; Schaub, E. A.; Priyadarshi, S. S.
D.; Reichert, C. M.; Kopin, I. J. Chronic parkinsonism secondary to Environmental Risk Factors and Parkinson’s Disease: A Metaanaly-
intravenous injection of meperidine analogs. Psychiat. Res. 1979, 1, sis. Environ. Res. 2001, 86, 122–127. (b) Priyadarshi, A.; Khuder, S.
249–254. A.; Schaub, E. A.; Shrivastava, S. A meta-analysis of Parkinson’s dis-
255. Langston, J. W.; Ballard, P.; Tetrud, J. W.; Irwin, I. Chronic Parkin- ease and exposure to pesticides. Neurotoxicology 2000, 21, 435–440.
sonism in humans due to a product of meperidine-analog synthesis. (c) Le Couteur, D. G.; McLean, A. J.; Taylor, M. C.; Woodham, B.
Science (Washington, D.C.) 1983, 219, 979–980. L.; Board, P. G. A meta-analysis of Parkinson’s disease and exposure
256. (a) Burns, R. S.; Chiueh, C. C.; Markey, S. P.; Ebert, M. H.; Jaco- to pesticides. Biomed. Pharmacother. 1999, 53, 122–130.
bowitz, D. M.; Kopin, I. J. A primate model of parkinsonism: Selec- 267. Maycock, A. L.; Abeles, R. H.; Salach, J. I.; Singer, T. P. The struc-
tive destruction of dopaminergic neurons in the pars compacta of the ture of the covalent adduct formed by the interaction of 3-dimeth-
substantia nigra by N-methyl-4-phenyl-1,2,3,6-tetrahydropyridine. ylamino-1-propyne and the flavine of mitochondrial amine oxidase.
Proc. Natl. Acad. Sci. U.S.A. 1983, 80, 4546–4550. (b) Langston, Biochemistry 1976, 15, 114–25.
J. W.; Forno, L. S.; Robert, C. J.; Irwin, I. Selective nigral toxicity 268. Hubálek, F.; Binda, C.; Li, M.; Herzig, Y.; Sterling, J.; Youdim, M.
after systemic administration of 1-methyl-4-phenyl-1,2,5,6-tetrahy- B.H.; Mattevi, A.; Edmondson, D.E. Inactivation of purified human
dropyridine (MPTP) in the squirrel monkey. Brain Res. 1984, 292, recombinant monoamine oxidase A and B by rasagiline and its ana-
390–394. logues. J. Med. Chem. 2004, 47, 1760–1766.
257. Heikkila, R. E.; Hess, A.; Duvoisin, R. C. Dopaminergic neurotoxic- 269. Binda, C.; Hubálek, F.; Li, M.; Herzig, Y.; Sterling, J.; Edmondson,
ity of 1-methyl-4-phenyl-1,2,5,6-tetrahydropyridine in mice. Science D. E.; Mattevi, A. Crystal structures of monoamine oxidase B in
(Washington, D.C.) 1984, 224, 1451–1453. complex with four inhibitors of the N-propargylaminoindan class.
258. Hornykiewicz, O. Aging and neurotoxins as causative factors in idio- J. Med. Chem. 2004, 47, 1767–1774.
pathic Parkinson’s disease - a critical analysis of the neurochemical 270. Chen, J.J.; Swope, D.M.; Dashtipour, K. Comprehensive review of
evidence. Prog. Neuropsychopharmacol. Biol. Psychiat. 1989, 13, rasagiline, a second-generation monoamine oxidase inhibitor, for the
319–328. treatment of Parkinson’s disease. Clin. Ther. 2007, 29, 1825–1849.
259. Tanner, C. M. The role of environmental toxins in the etiology of 271. Bar-Am, O.; Amit, T.; Sagi, Y.;Y, M.B.H. Contrasting neuroprotective
Parkinson’s disease. Trends Neurosci. 1989, 12, 49–54. and neurotoxic actions of respective metabolites of anti-Parkinson
260. Markey, S. P.; Schmuff, N. R. The pharmacology of the Parkinsonian drugs rasagiline and selegiline. Neurosci. Lett. 2004, 355, 169–172.
syndrome producing neurotoxin MPTP (1-methyl-4-phenyl-1,2,3,6- 272. Müller, K.; Faeh, C; Diederich, F. Fluorine in pharmaceuticals: look-
tetrahydropyridine) and structurally related compounds. Med. Res. ing beyond intuition. Science 2007, 317, 1881–1886 and supplemen-
Rev. 1986, 6, 389–429. tary material.
261. (a) Langston, W. B.; Irwin, I.; Langston, E. B.; Forno, L. S. Par- 273. Cohen, S. S. Nature of thymineless death. Ann. N. Y. Acad. Sci. 1971,
gyline prevents MPTP-induced Parkinsonism in primates. Science 186, 292–301.
(Washington, D.C.) 1984, 225, 1480–1482. (b) Heikkila, R. E.; 274. Mukherjee, K. L.; Heidelberger, C. Fluorinated pyrimidines. IX. Deg-
Manzino, L.; Cabbat, F. S.; Duvoisin, R. C. Protection against the radation of 5-fluorouracil-6-C14. J. Biol. Chem. 1960, 235, 433–437.
dopaminergic neurotoxicity of 1-methyl-4-phenyl-1,2,5,6-tetrahy- 275. (a) Douglas, K. T. The thymidylate synthesis cycle and anticancer
dropyridine by monoamine oxidase inhibitors. Nature 1984, 311, drugs. Med. Res. Rev. 1987, 4, 441–475. (b) Benkovic, S. J. On the
467–469. mechanism of action of folate- and biopterin-requiring enzymes.
262. Langston, J. W. In Factor, S. A.; Weiner, W. J. Parkinson’s Disease. Annu. Rev. Biochem. 1980, 49, 227–251.
Diagnosis and Clinical Management. Demos Medical Publishing, 276. (a) Santi, D. V.; McHenry, C. S.; Raines, R. T.; Ivanetich, K. M. Kinet-
New York, 2002, Chap. 30. ics and thermodynamics of the interaction of 5-fluoro-2’-deoxyuri-
263. Chiba, K.; Trevor, A.; Castagnoli, N. Jr. Metabolism of the neuro- dylate with thymidylate synthase. Biochemistry 1987, 26, 8606–8613.
toxic tertiary amine, MPTP, by brain monoamine oxidase. Biochem. (b) Silverman, R. B. Mechanism-Based Enzyme Inactivation: Chemis-
Biophys. Res. Commun. 1984, 120, 574–578. try and Enzymology, CRC, Boca Raton, FL, 1988; Vol. 1, p. 59.