Module 3: Fundamentals of Ligand-Receptor Interactions
Analyze the energetic drivers behind ligand binding. Explore Gibbs free energy, calculate potential energy curves, and inspect the structural basis of the hydrophobic effect.
Learning outcomes
- Decompose binding free energy into enthalpic and entropic contributions.
- Identify the major non-covalent interactions and their characteristic geometries.
- Explain the hydrophobic effect and why desolvation can dominate an affinity change.
- Use ligand efficiency and LLE to compare hits of different sizes and lipophilicities.
- Name the four ways the pairwise-contact picture of binding breaks down.
1. Molecular Recognition Models
How do a drug molecule and its target protein recognize each other in a crowded biological environment? Lock-and-key and induced fit are useful cartoons. In conformational selection, a ligand binds a state already populated by the unbound receptor. Selection and induced rearrangement can both contribute; endpoint structures alone do not determine the kinetic route:
Lock-and-Key Model
Formulated by Emil Fischer in 1894. Suggests the receptor and ligand possess complementary, rigid geometries. It explains high specificity but fails to account for structural plasticity.
Induced-Fit Model
Proposed by Daniel Koshland in 1958. Proposes that ligand binding triggers conformational rearrangements in the receptor pocket to optimize contacts, matching thermodynamic realities.
2. Thermodynamics of Binding
The affinity of a ligand for its receptor is defined by the change in Gibbs Free Energy (ΔG) upon complex formation. A more negative ΔG corresponds to tighter binding (higher affinity):
The standard binding free energy is related to the equilibrium dissociation constant (K_d). The logarithm requires a dimensionless ratio to the standard concentration C°. Kd describes equilibrium affinity; Ki is a model-dependent inhibition constant and IC50 is an assay-dependent half-maximal concentration, so they are not interchangeable:
Where R is the gas constant and T is the absolute temperature. Because of this logarithmic relationship, small, linear changes in Gibbs Free Energy translate into exponential improvements in binding affinity:
| ΔG (at 298 K) | Affinity Constant (K_d) | Qualitative Description |
|---|---|---|
| -4.1 kcal/mol | 1.0 mM (Millimolar) | Very weak binding, typical of small fragments. |
| -8.2 kcal/mol | 1.0 µM (Micromolar) | Moderate binding, typical of high-throughput screen hits. |
| -12.3 kcal/mol | 1.0 nM (Nanomolar) | Tightly bound drug candidate. |
| -16.4 kcal/mol | 1.0 pM (Picomolar) | Exceptional affinity binding, rare and highly optimized. |
Enthalpy (ΔH)
Refers to the release of heat resulting from the formation of specific, directional non-covalent contacts (hydrogen bonds, ionic pairs, van der Waals, and halogen bonds) between the ligand and target.
Entropy (-TΔS)
Represents the change in disorder. Complexation restricts ligand and side-chain rotation, costing conformational entropy. However, this penalty is offset by the hydrophobic effect: water displacement from hydrophobic surfaces into bulk.
Hydrogen-bond distance and angle
Leave the donor–acceptor distance at 2.9 Å and bend the angle. A hydrogen bond that looks right in distance can still be weak, because the attraction is directional. The curve is a model potential, not a measured enthalpy.
180° is linear. The attractive term is scaled by cos² of the bend, now 100% of a linear bond.
Geometry schematic, not to scale.
-4.00kcal/mol
The curve is a 12-6 well whose attraction is scaled by cos² of the bend away from 180°. It omits protonation, the competing hydrogen bond to water, and the rest of the electrostatic field. A negative value here is not an enthalpy.
3. From Affinity to Design Efficiency
Potency is essential, but it does not show what a molecule had to become to achieve that potency. Medicinal chemists use efficiency metrics and thermodynamic measurements to compare compounds while tracking size, lipophilicity, and binding mechanism.
| Measure | What it adds | Important limit |
|---|---|---|
| Ligand efficiency (LE) | Relates binding free energy to non-hydrogen atom count, helping compare differently sized hits. | Size normalization has known biases; compare related series and use the same affinity endpoint. |
| Lipophilic ligand efficiency (LLE or LipE) | pActivity - logD asks whether potency rises faster than lipophilicity. | State the assay endpoint, pH, and lipophilicity measure. The octanol-water reference is useful, not a universal measure of specificity. |
| Isothermal titration calorimetry (ITC) | A titration can estimate affinity, stoichiometry, and binding enthalpy; entropy is inferred from the free-energy relationship. | Buffer ionization, proton transfer, concentration accuracy, and coupled conformational changes can affect the observed heat. |
Enthalpy-entropy compensation
Two ligands can have similar affinity with different enthalpic and entropic profiles. Solvent reorganization, protonation, and conformational change couple the terms, so an apparently favorable enthalpy does not automatically identify a better lead.
Water is part of the mechanism
Displacing an unfavorable water can help binding, while disrupting a stable bridging network can hurt it. Inspect water networks and receptor state instead of treating every buried water release as an automatic gain.
4. Types of Non-Covalent Interactions
Each interaction below is a physical phenomenon first and a line of arithmetic second. Module 4 shows how a force field turns them into computable terms — hydrogen bonds and salt bridges fall out of the Coulomb term, dispersion and steric clash out of the Lennard-Jones term — and every docking score (Module 6) and free-energy estimate (Module 10) you meet later is built from exactly these contributions.
Hydrogen Bonds
Formed between a hydrogen atom covalently bound to an electronegative atom (Donor: O-H, N-H) and another electronegative atom with lone pairs (Acceptor: O, N). Strong, highly directional, with typical optimal donor-acceptor distances of 2.7–3.2 Å and bond angles close to 180°.
Halogen Bonds (σ-Hole Interactions)
An interaction between an electronegative atom and the electropositive region on the tip of a halogen atom (Cl, Br, or I) bound to carbon. This positive region, known as the σ-hole, renders halogen bonds highly directional.
Electrostatic Interactions
Salt bridges formed between oppositely charged functional groups (e.g. protonated amine on ligand and carboxylate side-chain of Aspartate or Glutamate on receptor). These interactions are long-range (V ∝ 1/r).
Cation-π Interactions
A special ion-dipole interaction between a cation (e.g. a protonated amine, Lys-NH₃⁺, or Arg guanidinium) and the electron-rich π face of an aromatic ring. Because the effect depends on ring electron density, electron-donating substituents (e.g. -NH₂) strengthen it while electron-withdrawing groups (e.g. -CN) weaken it. This explains why electron-rich Trp and Tyr engage in cation-π contacts far more often than Phe.
Van der Waals / London Dispersion & π-π Stacking
Weak (~0.5–1 kcal/mol each), short-range forces from transient, induced dipoles between all atoms in close contact. Individually negligible, but summed over a well-packed binding pocket they contribute substantially to affinity — this is the energetic basis of shape complementarity. π-π stacking between aromatic rings (parallel-displaced or T-shaped, ~3.5–4.0 Å) is a directional special case.
Interactive Playground: What Burial Is Worth
Bury nonpolar surface, then add one polar group that has no partner. The hydrophobic estimate is the old rule of thumb, about −25 cal/mol per Ų (−0.025 kcal/mol/Ų). An unmatched polar group costs more than that gain. The readout is a free-energy estimate, not a measured −TΔS: at room temperature the hydrophobic effect is often discussed as entropy, while many protein–ligand ITC profiles are enthalpic.
At 60 Ų the rule of thumb gives -1.5 kcal/mol. Turn on the unmatched carbonyl and watch that gain disappear.
Interactive Playground: Benzamidine in Trypsin
Benzamidine binds the S1 pocket of trypsin. The amidinium faces Asp189, the ring fills the hydrophobic well, and a reported Ki is about 20 µM. Converting it to roughly −6.4 kcal/mol assumes an equilibrium inhibition model in which Ki approximates Kd; this is not a directly measured binding free energy. Build the same complex as a ledger. Switch desolvation off to see the nanomolar answer a pairwise contact sum invents.
Benzamidine itself is rigid. Each extra rotor is charged 0.8 kcal/mol, a scoring-function estimate, not an ITC entropy.
Equivalent Kd if this ledger were a real standard free energy: 24 µM.
With desolvation included, the same contacts give -6.3 kcal/mol (24 µM). The contact-only sum is -11.3 kcal/mol (5.1 nM). Benzamidine’s measured Ki is about 20 µM, close to the solvent-aware ledger and far from the contact-only number.
Interaction magnitudes are teaching numbers, chosen so the solvent-aware benzamidine case lands near experiment. They are not a force-field result, and the equivalent Kd is only a translation of the ledger through ΔG = RT ln(Kd).
5. Where This Picture Breaks Down
The ledger you just moved is a sum of pairwise contacts, with solvent written in or left out. That picture is useful enough to design against, and wrong in four specific ways that resurface throughout the course.
Interactions are not additive
Adding a hydrogen bond worth 2 kcal/mol to a ligand rarely buys 2 kcal/mol of affinity. The new contact may cost desolvation, restrict a rotatable bond, or strain the pose. Non-additivity is precisely why medicinal chemistry still requires synthesis rather than arithmetic — and why scoring functions (Module 6) fail in the way they do.
The pocket is not rigid
Induced fit above is a cartoon of a much larger effect: side chains rotate, loops close, and some pockets do not exist until a ligand arrives. Anything computed on one fixed structure inherits that structure's assumptions (Modules 6 and 10).
Water is a participant, not a background
Every contact you form must first break a contact with water, and a few ordered waters in a pocket can be worth more than a whole substituent. Desolvation is the single most commonly underestimated term in this module.
Affinity is not the objective
ΔG tells you how tightly a ligand binds its target — not whether it is selective, absorbed, metabolically stable, or safe. Ligand efficiency and LLE exist to keep potency honest, and Modules 12 and 15 supply the constraints that ultimately decide whether a tight binder becomes a drug.
Every one of these is a physical shortcoming of the pairwise model, and each is the reason a later technique exists. Module 4 turns these same interactions into computable energy terms; keep the four caveats in mind, because they explain most of what goes wrong afterwards.
Knowledge check
Research decision checkpoint
A matched pair gains a hydrogen bond but loses affinity. What would you measure next?
Your output: Build a solvent-aware energy ledger and convert a Kd to standard binding free energy with units.
Compare your reasoning
Consider desolvation, pKa, strain, and conformational populations. A contact count cannot distinguish these explanations. IC50 is an assay endpoint, not automatically Kd.