1. Advanced Medicinal Chemistry
Lecture 3:
Molecular Interactions and Drug Potency
Barrie Martin
AstraZeneca R&D Charnwood

2. Dose-Response Curves 100 Enzyme Inhibitors (competitive):
Measure inhibition at differing concentrations of ‘drug’. %
Response %
Inhibition 50
IC50 - The inhibitor concentration that causes a 50% reduction in intrinsic enzyme activity
IC50=85nM
EC50=85nM
10nM
30nM
100nM
300nM
1mM
pIC50 = - log10(IC50)
IC50 1M = pIC50 6.0
IC50 1nM = pIC50 9.0
[Inhibitor]
[Agonist]
Agonists: Measure % Response vs Agonist concentration
EC50 - The agonist concentration that causes 50% of the maximum response. pEC50 = - log10(EC50)
Competitive enzyme inhibitors compete with the natural substrate for the active site. As the concentration of inhibitor increases, the rate of reaction decreases and the % inhibition increases.
IC50 typically used as a measure of inhibitor affinity, concentration at which the 50% enzyme inhibition is achieved. Commonly use a pIC50 scale (negative log of the IC50).
IC50 of 1nM = M, log IC50 = -9, pIC50 = 9
For agonists, measure % of maximal response against increasing concentrations of agonist. EC50 (concentration at which 50% maximal response is achieved) and pEC50 used.
Antagonists – shift the agonist dose-response curve to the right (i.e. greater concentration of agonist required to achieve response).
Antagonists: Situation more complex. Antagonists displace the agonist dose-response curve rightwards – most accurate measure of potency (pA2) requires measurement of agonist binding at multiple concentrations of antagonist
For a drug, typically target affinity values of pIC50  8 (<10 nM concentration) 2

3. iNOS - An AZ Charnwood Discovery Project
Active Site,
Haem & Inhibitor
Nitric Oxide Synthases – catalyse production of NO from arginine in the body – implicated in inflammatory conditions e.g. rheumatoid arthritis
iNOS (inducible Nitric Oxide Synthase) project carried out at AZ Charnwood.
Catalyses formation of NO from arginine – thought to be implicated in inflammatory conditions (iNOS overexpressed in a number of diseases)
Enzyme inhibitor program run and a series of dihydroaminoquinazoline compounds identified, AZ is a representative example. AZ
pIC50 7.5
A potent, selective iNOS inhibitor 3

4. How Do Drugs Bind to Enzymes & Receptors?
Drugs bind to particular sites on enzymes and receptors. In the case of an enzyme,
this will often be the active site. Receptors have binding pockets formed between
transmembrane helixes where drugs usually bind (not always the agonist’s binding site).
GLU E
PHE F
These sites are comprised of a variety of amino acid residues which give rise to a specific 3-D shape and molecular features:
Charges: CO2- , NH3+, =NH-+
Polar groups: OH, C=O, CONH
Hydrophobic groups: Ph, Alkyl, SMe
SER S
In enzymes, reaction centres are also present:
Asp-His-Ser in esterases
SH in some proteases
Metal ions (CYP-450, iNOS).
Drugs bind to particular sites on enzymes and receptors. For competitive enzyme inhibitors, this will be the active site. For GPCRs, antagonists often bind in binding pockets formed between the transmembrane helixes.
These sites are comprised of a varietyof amino acid residues which giverise to a specific 3D shape and possess molecular features which may be charged (e.g. aspartic acid), polar neutral (e.g. serine) or lipophilic (e.g. valine, phenylalanine).
With enzymes, reaction centres are also present. Some examples of groups involved in catalysing the enzyme reaction may be available for binding to an inhibitor:
Asp-His-Ser in esterases
SH in some proteases
Metal ions (CYP-450, iNOS)
Small molecules bind to these sites by a combination of shape complementarity and energetically favourable interactions.
Small molecules bind to these pockets by a combination of:
Shape complementarity
Energetically favourable interactions
Haem group – iNOS, CYP-450 4

5. Shape Complementarity
iNOS Enzyme Inhibitor AZ
Arginine N H S C 2
Cimetidine
Histamine
H2 Receptor Antagonist
The drug must fit into the Binding Site and shape complementarity is an important feature of a drug molecule. Competitive enzyme inhibitors often bear a resemblance to the substrate, as they bind to the same Active Site. This is also true for some receptor antagonists, but not all.
The strength of an interaction depends on the complementarity of the physico-chemical properties of atoms that bind, i.e. protein surface and ligand structure.
The ‘Binding Sites’ are not totally rigid. The side chains of the amino acids that make up the pocket have some mobility. A variety of related structures can thus be accommodated by movements that change the shape of the active site. This is known as the ‘Induced Fit Hypothesis’.
The drug must fit into the Binding Site and Shape Complementarity is an important feature of a drug molecule. Competitive enzyme inhibitors often bear a resemblance to the substrate, as they bind to the same Active Site. This is also true for some receptor antagonists, but not all.
The strength of an interaction depends on the complementarity of the physico-chemical properties of atoms that bind, i.e. protein surface and ligand structure.
The ‘Binding Sites’ are not totally rigid. The side chains of the amino acids that make up the pocket have some mobility. A variety of related structures can thus be accommodated by movements that change the shape of the active site. This is known as the ‘Induced Fit Hypothesis’.
Examples:
iNOS enzyme inhibitor mimics amidine functionality present in arginine.
Cimetidine – H2 receptor antagonist (Used for the treatment of gastric ulcers, GERD). Resemblance to natural agonist histamine, but prevents production of stomach acid. 5

6. Drug-Protein Binding Energies
For a binding Equilibrium between a Protein & a Drug K
[Protein] [Drug]
[P:D] DG
Protein
Drug
Drug
Protein
Gibbs Free Energy Changes
K = [P:D]
[P] x [D]
DG=-RTlnK and DG=DH-TDS
When thinking about the energetically favourable reversible interactions between a drug and a protein, we need to think about this as an equilibrium process between free drug, protein and drug/protein complex.
This equilibrium process is described by K, the equilibrium constant, where:
K = [drug:protein complex]
[drug] x [protein]
The efficacy of a drug depends on the position of this equilibrium. The larger that K is, the more is bound & the more effective the drug will be.
However, K is also related to DG, the change in Gibbs free energy, by the equation shown, and DG = DH – TDS.
As a result the equilibrium constant K, the degree of drug: protein binding and hence any biological effect, is dependent on changes in both enthalpy (DH) and entropy (DS).
Both Enthalpy (DH) and Entropy (DS) changes affect binding strength 6

7. Drug-Protein Interactions
Bond Example kJ/mol
Van der Waal Xe…Xe, alkyl groups 2
Hydrophobic Ph…Ph (p-stacking) 5
Dipole - Dipole C=O…HN-R (d+/d-)...(d+/d-) 5
Hydrogen H2O…H2O (X-H) …(Y-R) 35
Ion - Dipole F-…H2O (+/-ve)…(d+/d-) 170
Ion - Ion H+…Cl- (+ve)…(-ve) 450
Covalent C-O 350
NB. When a drug moves from the aqueous medium into the ‘Binding Site’ it has to
break H-Bonds with water, de-solvate etc. These processes require energy, so the
net energy available for binding is only a fraction of the above bond energies.
Examples of different types of molecular interactions that can arise between drugs and proteins and some typical values.
N.B. Although electrostatic interactions can contribute considerably to binding energies, when a drug moves from the aqueous medium into the binding site it has to break H-bonds with water, de-solvate. These processes require energy so the net energy available for binding can be only a fraction of the above energies. 7

8. Electrostatic Interactions
These result from the attraction between molecules bearing opposite
electronic charges.
Strong ionic interactions can contribute very strongly to binding.
Proteins contain both CO2- and NH3+ residues and these may be present
at the binding site to interact with oppositely charged groups on the drug.
AZ iNOS Inhibitor N H O F +
GLU
Neuraminidase Inhibitor (Antiviral GSK) O H R N A G - +
Examples of electrostatic interactions
iNOS inhibitor:
Attraction between basic dihydroaminoquinazoline group (positively charged) and glutamic acid residue (negatively charged) present in active site.
Neuraminidase inhibitor:
Interaction between negatively charged acidic group on drug and positively charged arginine side-chain.
The energies involved in a ‘salt bridge’ can be in the order of >30 kJ/mol
This can lead to increase in observed binding of >106 fold 8

9. Neuraminidase Inhibitor Charge re-inforced H-Bond
Hydrogen Bonding Interactions
A hydrogen bond results when a hydrogen is shared between two electronegative atoms
The Donor provides the H, while the Acceptor provides an electron pair
D-X-H….Y-A
e.g. R-O-H…..O=C O H N O H O N H O O O H H O R O O H O R H H O G L U O H H N N
Neuraminidase Inhibitor
Charge re-inforced H-Bond N N + N H
Examples of Hydrogen bonding interactions
iNOS inhibitor:
Hydrogen bonds formed between tyrosine O-H and amide carbonyl O, and between N-H and water molecule present in active site.
Neuraminidase inhibitor:
Charge re-inforced hydrogen bond formed between two O-H groups and glutamate residue. F N H H
AZ iNOS complex
Amide to Tyrosine H-Bond 9

10. Hydrophobic Interactions
Drugs, in general, are hydrophobic molecules
The ‘Binding Sites’ of proteins are also hydrophobic in character
Thus a mutual attraction can result (like attracts like).
What drives this attraction?
Enthalpy gains may result from van der Waals bonding:
Between Alkyl, Aryl, Halogen groups
p-p Stacking is an important type of this
Entropy gains are achieved when water molecules are displaced
from ‘active site’, and return to a more random (high S) state.
Many compounds synthesised during the drug discovery process are typically hydrophobic molecules. Binding sites of proteins are often also hydrophobic in character and hydrophobic interactions can form a significant contribution to drug-protein interactions
Enthalpy contributions arise from van der Waals bonding between e.g. alkyl, aryl groups and from p-p interactions between aromatic rings.
Entropy gains are achieved when water molecules are displaced from the active site and return to a more random (high entropy state – i.e. DS increases)
Each -(CH2)- group can contribute >1 kJ/mol towards binding
Each -Ph ring can contribute >2 kJ/mol towards binding
These effects are additive and hence Hydrophobic Bonding
can make a very high contribution to binding 10

11. Hydrophobic Bonding : D Entropy
When a hydrophobic drug
is placed into water, the
structure of the water around
the drug is more ordered.
This allows the H2O-H2O
H-bonds to be maintained.
This leads to lower entropy
and is not favoured.
Water molecules are
in a highly disordered
state. Each molecule
maximises H-Bonds
to other molecules
of water.
Hydrophobic effect 1 11

12. Hydrophobic Bonding : D Entropy
Hydrophobic interaction between protein and drug is favoured by entropy gains:
Bulk water returns to less ordered state
Water molecules may be expelled from being bound in active site.
In addition enthalpy gains due to new bonds may also be favourable (e.g. van der Waals interactions)
Hydrophobic effect 2 12

13. Probing Hydrophobicity
in Drug Discovery
New iNOS lead identified: R =Me, small lipophilic substituent iNOS pIC50 7.8
Aim: Probe lipophilic pocket – what else could we put there?
How would we make it?
Probing Hydrophobicity in drug discovery – iNOS example
A new iNOS lead is discovered where the R group is a methyl. The compound has good potency pIC50 = 7.8
Methyl = small, lipophilic substituent
We decided to explore this region further – what else could we put there? 13

14. Effect of Hydrophobicity on Activity
Binding into Lipophilic pocket of iNOS
R cLogP IC50 mM Me Et CF
Thiophene
Phenyl
2-Me-Thiophene
cLogP
7.6
7.8 8
8.2
8.4
8.6 1
1.2
1.4
1.6
1.8 2
2.2
2.4
2.6
iNOS_pIC50
Effect of hydrophobicity on activity – iNOS example
Too big to fit in pocket optimally
(Shape complementarity) 14

15. Bioisosteres Isostere: Bioisostere:
Similarities in physicochemical props. of atoms/groups/molecules with similar electronic structures (no. and arrangement of electrons in outermost shell). Often observed with groups in the same periodic table column (Cl  Br, C  Si).
Grimm – Hydride Displacement Law (1925) - Replacement of
chemical groups by shifting one column to the right & adding H.
Bioisostere:
Simplest definition - any group replacement which improves the molecule in some way
Two different interchangeable functionalities which retain biological activity.
Bioisosteric replacements can offer improvements both in potency and other properties (e.g. metabolic stability, absorption)
Chemical isosterism – Langmuir (1919)
Similarities in physicochemical properties of atoms/groups/molecules with similar electronic structures, most often observed with atoms in the same column of the periodic table.
Grimm – Hydride Displacement (1925)
Isosteric groups formed by displacing one place to the right in the periodic table and adding H (e.g. O vs NH vs CH2)
Bioisosteres:
Interchangeable functionalities which retain biological activity
Pharmacophoric bioisosterism – different functionalities which have similar interactions with a target protein e.g. CO2H & tetrazole or acyl sulfonamide.
Template bioisosterism – replacements which do not have a specific interaction with the target but act as spacer groups which orientate the pharmacophoric elements e.g. -CH2- & -O- or -S-
-CH2 & bioisosteres
Carboxylic acid & bioisosteres
amide & bioisosteres 15

16. Invisible Bioisosteres
EGF-R 2.2 nM
EGF-R 7.5 nM
H-bonds can be directly to protein or via water molecules
Further example of bioisosterism - quinazoline to cyanoquinoline switch
EGF-R (Epidermal Growth Factor Receptor - involvement in uncontrolled cell proliferation and cancer target)
Original nitrogen group bound to threonine residue via water molecule, replacement with CH (unsubsituted quinoline) gave significant loss in activity. Postulated that replacement with nitrile would allow direct interaction with protein and maintain activity
Ref. J. Med. Chem., 2000, 43, (Inhibitors of EGF-R Kinase) 16

17. Optimising Potency
How might we improve potency further from this compound?
Develop understanding of which molecular features are important for activity – remove substituents.
Look at incorporating new groups for additional potency e.g. through lipophilic interactions, hydrogen bonds etc.
Functional group bioisosteres.
Use available structural information – e.g. crystal structures of compound bound to enzyme.
Use of modelling to design/evaluate new targets.
Develop and test hypotheses.
Identify good disconnections/robust chemistry to allow rapid synthesis of multiple analogues – build up information.
pIC50 7.5
Identify pharmacophore - the molecular framework that carries the essential features responsible for a drug’s activity
Incorporation of new chemical groups/variation of substituents to introduce additional interactions with receptor.
Where available make use of structural information to guide target design.
Identification of late stage intermediates suitable for late stage diversification key, requires good chemical disconnections & synthetic routes to targets.
N.B. Potency is one of many properties that needs to be optimised in drug discovery - need to consider absorption, metabolism, selectivity etc. 17

18. Forward Synthesis - 1 Forward Synthesis of AZ10896372
Conversion of nitrile into amidine achieved over 2 steps – nucleophilic attack of hydroxylamine on the nitrile followed by hydrogenolysis of the N-O bond of the intermediate hydroxyamidine.
Tricyclic system formed by reaction of ortho-anilinoamidine with protected ketopiperidine. 18

19. Forward Synthesis - 2
Removal of the carbamate protecting group under basic conditions gave the secondary amine.
Coupling of amine with 6-cyano-3-pyridinecarboxylic acid achieved via activation of the acid with oxalyl chloride in dichloromethane followed by reaction with the resultant acid chloride with the amine in the presence of excess base (triethylamine).
Ref. J. Med. Chem., 2003, 46, 19