The Organic Chemistry of Drug
Design and Drug Action
Chapter 3
Receptors
Receptors
1878 Langley
Study of antagonistic action of alkaloids on cat
salivary flow suggests the compounds interacted
with some substance in the nerve endings
Receptors
1897 Ehrlich
Side chain theory - Cells have side chains that contain
groups that bind to toxins - termed receptors
1906 Langley
Studying antagonistic effects of curare on nicotine
stimulation of skeletal muscle
Concluded receptive substance that received stimulus,
and by transmitting it, caused muscle contraction
Two fundamental characteristics of a receptor:
 Recognition capacity - binding
 Amplification - initiation of response
Integral proteins embedded in phospholipid
bilayer of membranes
Figure 2.26
Drug-Receptor Interactions
Pharmacodynamics
SCHEME 3.1
Equilibrium between a drug, a receptor, and a drug–receptor complex
Kd =
[drug][receptor]
[drug-receptor complex]
(3.1)
Driving force for drug-receptor interaction - low
energy state of drug-receptor complex (binding
energy)
Kd - measure of affinity to receptor (a dissociation constant)
Forces Involved in Drug-Receptor Complex
Molecular surfaces must be close and complementary
G° = -RTlnKeq
(3.2)
Decrease in G° of ~ 5.5 kcal/mol changes binding equilibrium
from 1% in drug-receptor complex to 99% in drug-receptor
complex
Forces in drug-receptor complex generally weak and noncovalent
(reversible)
Ionic Interaction
Basic groups, e.g., His, Lys, Arg (cationic)
Acidic groups, e.g., Asp, Glu (anionic)
FIGURE 3.1 Example of an electrostatic (ionic) interaction. Wavy line represents the
receptor cavity.
Figure 3.1
G° ≈ -5 kcal/mol
Ion-Dipole and Dipole-Dipole Interactions
FIGURE 3.2 Examples of ion–dipole and dipole–dipole interactions. Wavy line represents
the receptor cavity.
Figure 3.2
G° ≈ -1 to -7 kcal/mol
Hydrogen Bonding
Type of dipole-dipole interaction between H on X-H (X is an
electronegative atom) and N, O, or F
FIGURE 3.3 Examples of hydrogen bonds. Wavy line represents the receptor cavity.
Figure 3.3
G° ≈ -3 to -5 kcal/mol
Intramolecular hydrogen bonding
FIGURE 3.4 Two examples (A and B) of how intramolecular hydrogen bonding can mimic a
bioisosteric heterocycle.
a-helix
3.5 is an example of an α-helix in a protein—Copyright 2007 from Molecular Biology of the Cell, Fifth Edition by
Alberts, et al. Reproduced by permission of Garland Science/Taylor & Francis LLC.
-sheet
3.6 is an example of a β-sheet in a protein—Copyright 2007 from Molecular Biology of the Cell, Fifth Edition by
Alberts, et al. Reproduced by permission of Garland Science/Taylor & Francis LLC.
DNA
3.7 is an example of a double helix in DNA—Copyright 2007 from Molecular Biology of the Cell, Fifth Edition by
Alberts, et al. Reproduced by permission of Garland Science/Taylor & Francis LLC.
Charge-Transfer Complexes
(molecular dipole-dipole interaction)
chlorothalonilfungicide
acceptor
donor
FIGURE 3.6 Example of a charge-transfer interaction. Wavy line represents the receptor
cavity.
Figure 3.6
G° ≈ -1 to -7 kcal/mol
Hydrophobic “Interactions”
Increase in entropy of H2O molecules decreases
free energy. Therefore the complex is stabilized.
FIGURE 3.7 Formation of hydrophobic interactions. From Korolkovas, A. (1970). Essentials of
Molecular Pharmacology, p. 172. Wiley, New York. This material is reproduced with permission
of John Wiley & Sons, Inc. and by permission of Kopple, K. D. 1966. Peptides and Amino Acids.
Addison-Wesley, Reading, MA.
Hydrophobic Interaction
butamben - topical anesthetic
G° ≈ -0.7 kcal/mol
per CH2/CH2 interaction
FIGURE 3.8 Example of hydrophobic interactions. The wavy line represents the receptor
cavity.
π-π-Interactions
FIGURE 3.9
Example of π–π stacking. The wavy line represents the receptor cavity.
Cation-π-interactions
FIGURE 3.10 Example of a cation–π interaction. The wavy line represents the receptor
cavity.
Halogen bonding
FIGURE 3.11 Example of halogen bonding. A compound bound into phosphodiesterase 5.
The wavy line represents the enzyme cavity.
Van der Waals (London Dispersion) Forces
As molecules approach, temporary dipoles in one
molecule induce opposite dipoles in another;
therefore, producing an intermolecular attraction
G° ≈ -0.5 kcal/mol per
CH2/CH2 interaction
Dibucaine - local anesthetic
FIGURE 3.12 Example of potential multiple drug–receptor interactions. The van der Waals
interactions are excluded.
Dose-Response Curve
Use any measure
of response
(LD50, ED50, etc.)
Means of measuring
drug-receptor interactions
FIGURE 3.13 Effect of increasing the concentration of a neurotransmitter (ACh) on muscle
contraction. The Kd is measured as the concentration of neurotransmitter that gives 50% of the
maximal activity.
Full Agonist
FIGURE 3.14 Dose–response curve for a full agonist (W).
Antagonists
Competitive
Antagonist
NT + R
NT R
X
X R
Noncompetitive
Antagonist
Different
binding sites
FIGURE 3.15 (A) Dose-response curve for an antagonist (X); (B) effect of a
competitive antagonist (X) on the response of a neurotransmitter (acetylcholine;
ACh); (C) effect of varying concentration of a competitive antagonist X in the
presence of a fixed, maximally effective concentration of agonist (ACh); and (D)
effect of various concentrations of a noncompetitive antagonist (X’) on the response
of the neurotransmitter (ACh).
Partial Agonist
low [neurotransmitter]
added
agonist effect
high [neurotransmitter]
added
antagonist effect
FIGURE 3.16 (A) Dose–response curve for a partial agonist (Y); (B) effect of a low
concentration of neurotransmitter on the response of a partial agonist (Y); and (C) effect of a
high concentration of neurotransmitter on the response of a partial agonist (Y). In (C), the
concentration of the neurotransmitter (a,b,c) is c > b > a.
Inverse Agonists
full inverse agonist
Addition of an agonist or antagonist
to an inverse agonist (a, b, c are
increasing concentrations of agonist
added)
partial inverse agonist
FIGURE 3.17 (A) Dose–response curve for a full inverse agonist (Z); (B) effect of a competitive
antagonist on the response of a full inverse agonist (a, b, and c represent increasing
concentrations of the added antagonist or natural ligand to Z); and (C) dose–response curve for
a partial inverse agonist (Z′).
 To effect a certain response of a receptor, design an
agonist
 To block a particular response of a natural ligand of
a receptor, design an antagonist
 To produce the opposite effect of the natural
ligand, design an inverse agonist
Table 3.1
Agonists - often
structural similarity
Antagonists - little
structural similarity
How can agonists and antagonists bind to same
site and one show response, other not?
agonist
antagonist
enantiomer
• All naturally-occurring chemicals in the body are agonists
• Most xenobiotics are antagonists
• Drugs that bind to multiple receptors  side effects
Two stages of drug-receptor interactions:
1) complexation with receptor
affinity
5 different drugs
a = 1 full
agonist
a < 1 partial
agonists
2) initiation of response
(Stephenson)
efficacy
intrinsic activity (Ariëns)
All are full
agonists
FIGURE 3.19 Theoretical dose–response curves illustrate (A) drugs with equal affinities and
different efficacies (the top compound is a full agonist, and the others are partial agonists) and
(B) drugs with equal efficacies (all full agonists) but different affinities.
Affinity and efficacy are uncoupled: a compound
can have great affinity but poor efficacy (and vice
versa).
A compound can be an agonist for one receptor
and an antagonist or inverse agonist for another
receptor.
A full or partial agonist displays positive efficacy.
An antagonist displays zero efficacy.
A full or partial inverse agonist displays negative
efficacy.
Drug-Receptor Theories
Occupancy Theory (1926)
Intensity of pharmacological effect is directly
proportional to number of receptors occupied
Does not rationalize how two drugs can occupy
the same receptor and act differently
Rate Theory (1961)
Activation of receptors is proportional to the
total number of encounters of a drug with its
receptor per unit time.
Does not rationalize why different types of
compounds exhibit the characteristics they do.
Induced Fit Theory (1958)
• Agonist induces
conformational change response
• Antagonist does not induce
conformational change - no
response
• Partial agonist induces
partial conformational
change - partial response
FIGURE 3.20 Schematic of the induced-fit theory. Koshland, Jr., D. E., and Neet, K. E., Annu.
Rev. Biochem., Vol. 37, 1968. Annual Review of Biochemistry by Annual Reviews. Reproduced
with permission of Annual Reviews via Copyright Clearance Center, 2013.
Macromolecular Perturbation Theory
 Two types of conformational perturbation (Belleau)
 Specific conformational perturbation allows molecule
to induce a response
 Nonspecific conformational perturbation does not
result in a response
 How to explain an inverse agonist?
Activation-Aggregation Theory
Monad, Wyman, Changeux (1965)
Karlin (1967)
Receptor is always in a state of dynamic equilibrium
between activated form (Ro) and inactive form (To).
Ro
biological
response
To
no biological
response
Agonists shift equilibrium to Ro
Antagonists shift equilibrium to To
Partial agonists bind to both Ro and To
Binding sites in Ro and To may be different, accounting
for structural differences in agonists vs. antagonists
Two-state (Multi-state) Receptor Model
R and R* are in equilibrium
(equilibrium constant L), which
defines the basal activity of the
receptor.
Full agonists bind only to R*
Partial agonists bind preferentially
to R*
Full inverse agonists bind only to R
Partial inverse agonists bind preferentially to R
Antagonists have equal affinities for both R and R* (no effect on
basal activity)
In the multi-state model there is more than one R state to account
for variable agonist and inverse agonist behavior for the same
receptor type.
Drug and Receptor Chirality
Drug-Receptor Complexes
Receptors are chiral (all L-amino acids)
Racemic mixture forms two diastereomeric
complexes
[Drug]R + [Drug]S + [Receptor]S
[Drug]R [Receptor]S + [Drug]S [Receptor]S
Have different energies and stabilities
Topographical and Stereochemical
Considerations
Spatial arrangement of atoms
Common structural feature of antihistamines
(antagonists of H1 receptor)
Pharmacophore - parts of the drug that interact
with the receptor and cause a response
Figure 3.22
CH-O, N-, CH-
2 or 3 carbons
Chiral antihistamine
Kd for enantiomers are different
- two diastereomers are formed
(S)-(+)-isomer 200x more potent
than (R)-(-)More potent isomer - eutomer
Less potent isomer - distomer
Ratio of potencies of enantiomers - eudismic ratio
High eudismic ratio when antagonist has stereogenic
center in pharmacophore
Distomer is really an impurity
(“isomeric ballast”)
May contribute to side effects and/or toxicity
O
H
N
O
O
O
H
N
O
O
N
N
H
H
O
O
3.13
(R)-(+)-thalidomide
sedative/hypnotic
(S)-(-)-thalidomide
teratogen
Enantiomers of ketamine
S-ketamine is several fold more
potent than R-ketamine
Prilocaine, a local anesthetic
Both enantiomers are active, but only one is toxic
Drugs useful as mixtures of enantiomers
Both are local anesthetics, Diuretic, but one enantiomer causes
But l-form is vasoconstrictor uric acidretention, the other inhibits it
Enantiomers can have different activities
S-enantiomer: NSAID
R-enantiomer: Reduces bone loss in periodontal disease
Enantiomers can have different activities
dextropropoxyphene (Darvon®)
analgesic
levopropoxyphene (Novrad®)
antitussive (anticough)
Enantiomers can have opposite activities
barbiturate
S-(+)convulsive
(actually inverse agonist)
R-(-)narcotic
One enantiomer may antagonize the other with
no overall effect observed.
Enantiomers can have opposite activities
(+)-isomer: Narcotic agonist analgesic
(-)-isomer: Narcotic antagonist
Enantiomers can have opposite activities
R-enantiomer: Serotonin agonist at 5HT-1a
S-enantiomer: Serotonin antagonist at 5HT-1a
Stereospecificity of one compound
can vary for different receptors
Eudismic ratio (-/+) is 800
(+) - 3.24 butaclamol - antipsychotic
(-) is almost inactive
Eudismic ratio (+/-) is 1250 for D2-dopaminergic, 160
for D1-dopaminergic, and 73 for a-adrenergic receptors
Hybrid drugs - different
therapeutic activities
Antagonist of -adrenergic
propranolol (X = NH
receptor (-blocker) triggers vasodilation
antihypertensive
Eudismic ratio (-/+) is 100
But propanolol also is a local anesthetic for
which eudismic ratio is 1
)
Pseudo-hybrid drug - multiple isomeric forms
involved in biological activity
FIGURE 3.23 Four stereoisomers of labetalol
labetalol - antihypertensive
R,R- mostly -blocker (eutomer for -adrenergic block)
S,R- mostly a-blocker (eutomer for a-adrenergic block)
S,S- and R,S- almost inactive (isomeric ballast)
Epinephrine, a natural hybrid drug
Racemates as Drugs
 90% of -blockers, antiepileptics, and oral
anticoagulants on drug market are racemates
 50% of antihistamines, anticholinergics, and local
anesthetics on drug market are racemates
 In general, 30% of drugs are sold as racemates
Racemic switch - a drug that is already sold as
a racemate is patented and sold as a single
enantiomer (the eutomer)
Omeprazole, a chiral switch
RS, Prilosec, now generic
S-enantiomer, Nexium
Single enantiomer drugs are expected to have
lower side effects
Antiasthma drug albuterol binds to 2-adrenergic
receptors, leading to bronchodilation
The (R)-(-)-isomer is solely responsible for
effects; the (S)-(+)-isomer causes pulse rate
increases, tremors, and decreased blood glucose
and potassium levels
Sometimes, it is better to use the racemate than
one isomer. In the case of the antihypertensive
drug nebivolol, the (+)-isomer is a -blocker;
the (-)-isomer causes vasodilation by a different
mechanism. Therefore, it is sold as a racemate
to take advantage of both vasodilating
pathways.
Prozac is the racemic drug. The R-enantiomer
showed cardiotoxicity so the chiral switch failed
Verapamil is used as a racemate
S-enantiomer is an antihypertensive
R-enantiomer inhibits resistance of cancer cells
Receptor Interaction
Figure 3.24
Enantiomers cannot be distinguished with only
two binding sites.
Three-point attachment concept
Figure 3.25
Receptor needs at least three points of interaction
to distinguish enantiomers.
Unnatural enantiomers of natural products may
have useful activities
Both of these are more active than the natural enantiomers!
Diastereomers
The antihistamine activity of (E)-triprolidine
(3.36a) is 1000-fold greater than the (Z)-isomer
(3.36b).
Diastereomers
 The antipsychotic activity of 3.37a is 12 times
more than 3.37b
Diastereomers
 Diethylstilbestrol (3.38a) is a much more potent
estrogen than the Z-isomer (3.38b)
Conformational Isomers
 Pharmacophore is defined by a particular conformation
of a molecule (the bioactive conformation)
 The conformer that binds need not be the lowest energy
conformer
 Binding energy can overcome the barrier to formation of
a higher energy conformer
Note that the bioactive
conformation bound to
the peroxisome
proliferator activated
receptor gamma
(PPAR) is not the
lower energy extended
conformation.
Figure 3.26
If the lead has low potency, it may be because of
the low population of the active conformer. If the
bioactive conformer is high in energy, the Kd will
appear high (poor affinity) because the
population of the ideal conformer is low.
SCHEME 3.2 Cyclohexane conformations. a, chair (substituent equatorial); b, half-chair; c, boat; d, half-chair; e,
chair (substituent axial).
To determine the active conformation, make
conformationally rigid analogs. The flexible lead
molecule is locked into various conformations by
adding bonds to rigidify it.
First we will use this approach to identify the
bioactive conformation of a neurotransmitter, then
a lead molecule.
Consider acetylcholine binding to muscarinic and
nicotine receptors
O
Me 3NCH2CH2OCCH3
OAc
Me 3N
acetylcholine
Four conformers of acetylcholine (just
staggered conformers)
Lowest energy
conformer
Newman
projections
Conformationally rigid analogs
All exhibited low muscarinic receptor activity, but
3.43a was most potent (0.06 times potency of ACh).
Analogues of acetylcholine
 The threo isomer (3.44) is 14 time more potent than
acetylcholine.
 The erythro isomer (3.45) is 0.036 times as potent as
acetylcholine.
To minimize the number of extra atoms, the
cyclopropane analog was made.
The (+)-trans isomer (3.46) has about the same
muscarinic activity as acetylcholine; (-)-trans isomer
1/500th potency.
Excellent support for the anti-conformer as the bioactive
conformer.
()-cis isomer (3.47) has negligible activity. Therefore,
acetylcholine binds to the muscarinic receptor in an
extended form (3.42a)
However, both the trans and cis cyclopropane
analogs are weakly active with the nicotinic
receptor for acetylcholine.
Therefore, a conformation other than the anticonformation must bind to that receptor (i.e., a
higher energy conformer).
Conformationally Rigid Analogs
in Drug Design
moderate tranquilizing activity
Maybe it is because the piperidino ring needs
to be in a higher energy conformation for good
binding.
Possible conformers of piperidino ring
R= F
O
C (CH2) 3
Conformationally Rigid Analogs
order of potency
3.51 > 3.52 > 3.50
Therefore, the less stable axial conformer binds
better than the equatorial conformer.
Lead modification should involve making analogs
in which the hydroxyl group is preferred in an
axial orientation.
Conformations of PCP
FIGURE 3.27 PCP, 3.53 and three conformationally rigid analogs of PCP
All these analogs bind poorly to the NMDA receptor, but
bind well to the σ-receptor.
Conformations of peptides
FIGURE 3.28 Use of a triazole as a conformationally rigid bioisostere to lock in an amide bond
conformation
Atropisomers
FIGURE 3.29 General example of atropisomerization
What makes atropisomers stable?
FIGURE 3.30 Example of a nonatropisomer, an unstable atropisomer, and a stable atropisomer
Telenzepine racemizes very slowly
FIGURE 3.31 Exceedingly slow isomerization of atropisomers of telenzepine (3.57)
(+) isomer is 500 times more active at muscarinic acetylcholine receptors
Atropisomers in drug optimization
A neurokinin 1
antagonist is a
lead for an
antidepressant
The active atropisomer of 3.58 is 3.59.
3.60 has two atropisomers
3.61 has only a single isomer
Avoiding atropisomers—make rotations fast
Symmetrization to avoid atropisomers
Ring Topology
chlorpromazine - tranquilizer
amitriptyline antidepressant with a
tranquilizing side effect
imipramine - pure antidepressant
Figure 3.32
bending of ring planes
annellation
angle of ring axes
torsional angle
tranquilizers - only a
mixed - a and 
antidepressants - a, , 
You must consider the 3-dimensional structures of rings.
Case History of Rational Drug Design Cimetidine
(no QSAR, computer graphics, or X-ray
crystallography)
Another action of histamine - stimulation of
gastric acid secretion
Antihistamines have no effect on H2 receptor
Nobel Prize (1988) to James Black for
antagonist discovery
H1 and H2 receptors differentiated by agonist and
antagonists
H1 receptor agonist
(no effect on H2
receptor)
H2 - receptor agonist (no
effect on H1 receptor)
H2 - receptor antagonists
would be antiulcer drugs
Bioassay used to screen compounds
Histamine was infused into anesthetized rats to
stimulate gastric acid secretion, then the pH of
the perfusate from the stomach was measured
before and after administration of the test
compound.
Lead Discovery
Histamine analogs synthesized at Smith, Kline,
and French (now GlaxoSmithKline)
Took four years and 200 compounds
3.75 was very weakly active (actually, partial
agonist)
Na-guanylhistamine
Isosteric replacement
Isothiourea 3.76 is more potent
than the cyclic analogue 3.77
imidazole retained
for recognition
homolog
not + charged
Had weak antagonistic activity without
stimulatory activity.
Homologation
further homologation
R = CH3
burimamide
purely competitive antagonist
for H2 receptor
Tested in humans - poor oral activity
Could be pharmacokinetics or pharmacodynamics
Consider pharmacodynamics
Imidazole ring can exist in 3 forms
FIGURE 3.33 Three principal forms of 5-substituted imidazoles at physiological pH
Thioureido group can exist as 4 conformers
Side chain can
be in many
conformations
Maybe only a
small fraction
in the
bioactive form
FIGURE 3.34 Four conformers of the thioureido group
To increase potency of burimamide
Compare population of the imidazole form in
burimamide at physiological pH to that in
histamine.
Hammett Study of Electronic Effect of Side Chain
favored for
R = e- -withdrawing
favored for
R = e- -donating
pKa of imidazole = 6.80
pKa of imidazole in histamine = 5.90
Therefore, side chain is e- -withdrawing, favoring 3.80a.
pKa of imidazole in burimamide = 7.25
Therefore, side chain is e- -donating, favoring 3.80c.
Need to make side chain e- -withdrawing.
Isosteric replacement to lower the
pKa of the imidazole
A second way to increase
population of 3.80a is to
put an e- -donating group
at 4-position.
thiaburimamide (R = H)
pKa of imidazole in
thiaburimamide = 6.25
thiaburimamide is 3 times
more potent than
burimamide
metiamide (3.82, R =
CH3)
pKa of imidazole in
metiamide = 6.80
8-9 times more potent
than burimamide
Oxaburimamide is less potent than burimamide, even
though O is more electronegative than S
Conformationally-restricted analog forms by
intramolecular H-bonding.
Does not occur with thiaburimamide.
Metiamide (3.82)tested in 700 patients with
duodenal ulcers - very effective.
However, side effect in a few cases
(granulocytopenia).
Thought the side effect was caused by the thiourea
group.
Isosteric replacement (X = O, X = NH) is 20 times
less potent.
When X = NH, basic
To lower basicity, add e- -withdrawing group
X = N-CN (cimetidine)
(pKa -0.4)
X = N-NO2
(pKa -0.9)
Both are comparable to metiamide in potency but
without the side effect.
Linear free energy relationship between
potency and lipophilicity
cimetidine
FIGURE 3.35 Linear free energy relationship between H2 receptor antagonist activity (pA2) and the partition coefficient.
Reprinted with Permission of Elsevier. This article was published in Pharmacology of Histamine Receptors, Ganellin, C. R., and
Parsons, M. E. (1982), p. 83, Wright-PSG, Bristol.
A cyclic analogue is less active
Other H2 receptor antagonists made using
cimetidine as the lead
ranitidine (Glaxo)
(no imidazole at all)
nizatidine (Eli Lilly)
famotidine (Yamanouchi)
Case history #2: Suvorexant
Insomnia is a serious health problem
Orexin A and B are neuropeptides that regulate sleep
Orexins bind to a GPCR
Orexin antagonists could be sleep aids
Merck identified a lead compound (3.89) from
high throughput screening
Modification of the aromatic rings gave 3.90, 3.91, and finally 3.92
3.92 has low bioavailability and undergoes rapid
metabolism
SCHEME 3.3
Oxidative metabolism of the 1,4-diazepane ring of 3.92
Further optimization
Methylation gave 3.95, which is resist to metabolism, but has low
bioavailability
Fluorination and removal of a methyl group gave 3.96, which has
better bioavailability
Adding a benzoxazole in 3.97 reduces metabolism further, but has
lower potency
Adding a chlorine increases potency, resulting in 3.98 (Suvorexant)
An alternative orexin antagonist
More potent than suvorexant in vivo