Tutorials › Pharmacometrics › Receptor Binding and Pharmacology Models
Pharmacology · PK/PD Foundations

Receptor Binding and Pharmacology Models

Learn how mathematical models describe drug-receptor binding, affinity, receptor occupancy, efficacy, concentration-effect relationships, and the mechanisms that connect molecular interactions to pharmacologic response.

Intermediate Pharmacology Receptor Models PK/PD Modeling
01 · The big picture

1. What Is Receptor Binding?

Receptor binding describes the interaction between a drug, ligand, or endogenous signaling molecule and a molecular target such as a receptor. Binding models provide a quantitative way to describe how much target is occupied at a given ligand concentration and how strongly the ligand interacts with that target.

The simplest representation is a reversible interaction between a free ligand and an unoccupied receptor:

\[ D + R \rightleftharpoons DR \]

Here, \(D\) represents free drug or ligand, \(R\) represents unoccupied receptor, and \(DR\) represents the drug-receptor complex.

Drug D Receptor binding site occupancy Effect E Binding can be modeled separately from the downstream pharmacologic response

A receptor model can describe the molecular binding step and, when combined with a transduction model, the relationship between receptor engagement and observed pharmacologic effect.

Core idea: receptor binding and pharmacologic response are related but distinct. A drug can bind a receptor strongly without producing the same magnitude of response as another ligand, because binding affinity and functional efficacy describe different properties.
02 · What the models ask

2. What Questions Do Receptor Models Help Answer?

Receptor pharmacology models can address several different questions, depending on the experimental design and the level of mechanistic detail.

Question Concept What it describes
How strongly does a ligand bind? Affinity The tendency of a ligand to interact with its receptor
What fraction of receptors is occupied? Receptor occupancy The proportion of available receptor sites associated with ligand
How much response can the ligand produce? Efficacy The functional ability of a ligand to activate or otherwise influence the receptor system
What concentration produces a specified response? EC50 The concentration associated with 50% of the modeled maximal response in an appropriate concentration-effect model
Can a ligand inhibit another ligand? Antagonism Competition or other mechanisms that reduce the response to an agonist
Does receptor occupancy directly determine effect? Transduction model The relationship between receptor engagement and downstream response

These concepts should not be treated as interchangeable. In particular, Kd and EC50 arise from different concepts: \(K_d\) characterizes a binding equilibrium in a specified model, whereas EC50 characterizes a concentration-effect relationship.

03 · Binding equilibrium

3. The Basic Receptor-Binding Model

For a simple one-to-one reversible interaction, the binding reaction can be written as:

\[ D + R \underset{k_{\mathrm{off}}}{\overset{k_{\mathrm{on}}}{\rightleftharpoons}} DR \]

The forward reaction is governed by the association rate constant \(k_{\mathrm{on}}\), while the reverse reaction is governed by the dissociation rate constant \(k_{\mathrm{off}}\).

At equilibrium, the dissociation constant is defined as:

\[ K_d=\frac{k_{\mathrm{off}}}{k_{\mathrm{on}}} \]

The units of \(K_d\) are concentration units. Under the assumptions of the simple one-site equilibrium model, a lower \(K_d\) corresponds to greater binding affinity.

Important distinction: \(K_d\) is a property of the modeled binding equilibrium. It should not automatically be interpreted as the concentration that produces 50% of a clinical or functional response.
04 · Receptor occupancy

4. Receptor Occupancy

The simplest receptor occupancy model assumes one class of independent binding sites and rapid equilibrium. The fraction of receptors occupied by drug is then:

\[ f_{\mathrm{occ}}=\frac{C}{K_d+C} \]

where \(C\) is the free ligand concentration.

This equation has several useful properties. When \(C\ll K_d\), occupancy is low. When \(C=K_d\), the predicted occupancy is 50%. As \(C\) becomes much larger than \(K_d\), occupancy approaches 100% under the model.

100% 50% 0 Ligand concentration Receptor occupancy Kd 50%

For the simple one-site equilibrium model, \(K_d\) is the concentration associated with 50% receptor occupancy.

05 · Affinity

5. What Does Affinity Mean?

Affinity describes how strongly a ligand interacts with its binding site. In a simple equilibrium binding model, affinity is commonly summarized using \(K_d\).

The equilibrium relationship can be rearranged to express the concentration needed to achieve a particular occupancy:

\[ C=K_d\frac{f_{\mathrm{occ}}}{1-f_{\mathrm{occ}}} \]

For example, 90% occupancy corresponds to:

\[ C=9K_d \]

while 10% occupancy corresponds to:

\[ C=\frac{K_d}{9} \]

These relationships demonstrate that substantial receptor occupancy may require concentrations substantially above \(K_d\), depending on the desired occupancy.

06 · Binding kinetics

6. Association and Dissociation Kinetics

Equilibrium affinity does not describe the entire time course of binding. Two ligands can have similar equilibrium affinity while having different association and dissociation rates.

For a simple one-site model, the rate of change in receptor-ligand complex can be expressed as:

\[ \frac{d[DR]}{dt} = k_{\mathrm{on}}[D][R] - k_{\mathrm{off}}[DR] \]

The first term represents formation of the complex, while the second represents dissociation.

The dissociation rate constant is especially relevant to the time course of unbinding. Under simple first-order dissociation conditions:

\[ [DR](t)=[DR]_0e^{-k_{\mathrm{off}}t} \]

The corresponding dissociation half-life is:

\[ t_{1/2,\mathrm{off}}=\frac{\ln(2)}{k_{\mathrm{off}}} \]
Key distinction: affinity describes the equilibrium state, whereas \(k_{\mathrm{on}}\) and \(k_{\mathrm{off}}\) describe how rapidly the system approaches or leaves that state.
07 · Saturation binding

7. Saturation Binding Models

In a saturation-binding experiment, increasing concentrations of a labeled or measurable ligand are applied to a receptor preparation. Binding initially increases with concentration and then approaches a plateau as available binding sites become occupied.

A simple model is:

\[ B(C)=\frac{B_{\max}C}{K_d+C} \]

Here, \(B(C)\) is the amount of specifically bound ligand, \(B_{\max}\) represents the modeled maximum binding capacity, and \(K_d\) controls the concentration scale of the binding curve.

Parameter Interpretation
Bmax Maximum specific binding capacity under the model
Kd Concentration associated with 50% of Bmax in the simple one-site model
C Free ligand concentration
B(C) Predicted specific binding at concentration C

The model can therefore separate two concepts that are often confused: binding capacity and binding affinity.

08 · Competition

8. Competitive Binding

When two ligands compete for the same binding site, increasing the concentration of one ligand can reduce the fraction of receptors occupied by the other.

A commonly used competitive-binding relationship introduces a competitor concentration \(I\) and inhibition constant \(K_i\). Under appropriate assumptions, the apparent concentration scale of the primary ligand is shifted according to:

\[ K_{d,\mathrm{app}} = K_d \left( 1+\frac{I}{K_i} \right) \]

The resulting binding relationship can be written as:

\[ B(C,I) = \frac{B_{\max}C} {K_d\left(1+\frac{I}{K_i}\right)+C} \]

This model describes one particular mechanism—competitive interaction at a shared site. More complex receptor systems may require multiple binding sites, allosteric models, or other mechanisms.

Modeling principle: a change in the observed binding curve does not by itself identify the molecular mechanism. Mechanistic interpretation depends on the experimental design and assumptions of the model.
09 · Binding versus response

9. Why Receptor Occupancy Does Not Always Equal Effect

One of the most important concepts in pharmacology modeling is that receptor occupancy and pharmacologic effect are not necessarily identical.

A simple occupancy model might assume:

\[ E=E_{\max}f_{\mathrm{occ}} \]

which gives:

\[ E= E_{\max}\frac{C}{K_d+C} \]

Under this particular model, the concentration producing half-maximal effect is numerically related to \(K_d\). However, real pharmacologic systems can contain signal amplification, spare receptors, receptor reserve, desensitization, downstream nonlinearities, and other processes.

Consequently, the concentration-effect relationship may not have the same parameters as the underlying binding relationship.

10 · Functional pharmacology

10. Agonists, Partial Agonists, and Antagonists

Receptor pharmacology distinguishes ligands not only by whether they bind, but also by what they do after binding.

Ligand type Binding Functional description
Agonist Binds to the target Can activate the receptor system and produce a response
Partial agonist Binds to the target Produces a lower maximal response than a full agonist under the same modeled system
Antagonist Binds to the target Can reduce agonist-mediated response without producing the same receptor activation
Inverse agonist Binds to a receptor system with constitutive activity Can reduce activity below the baseline level defined by the model

These functional classifications depend on the receptor system and assay conditions. They cannot be inferred solely from a binding-affinity measurement.

11 · Concentration-effect models

11. The Emax Model

A widely used pharmacology model relates drug concentration to effect using the Emax equation:

\[ E(C) = E_0+ \frac{E_{\max}C} {EC_{50}+C} \]

Here, \(E_0\) is the baseline effect, \(E_{\max}\) is the maximum additional effect represented by the model, and \(EC_{50}\) is the concentration associated with half of the modeled maximum effect.

A Hill-type extension introduces a slope parameter:

\[ E(C) = E_0+ \frac{E_{\max}C^n} {EC_{50}^n+C^n} \]

where \(n\) is the Hill coefficient.

When \(n=1\), the equation reduces to the standard Emax form. Values of \(n\) different from 1 allow the concentration-effect curve to change its steepness.

12 · Concentration-effect curves

12. Reading a Concentration-Effect Curve

Emax 50% EC50 Concentration Effect

The Emax model summarizes the baseline, maximum modeled effect, and concentration scale of the response.

13 · Potency and efficacy

13. Potency Is Not the Same as Efficacy

Potency describes the concentration or dose required to produce a specified effect, whereas efficacy describes the magnitude of effect that the drug can produce within the relevant system and model.

In an Emax model, \(EC_{50}\) is commonly used as a measure of potency, while \(E_{\max}\) describes the maximum modeled effect.

Concept Typical model quantity Interpretation
Affinity Kd Binding property in a specified receptor-binding model
Potency EC50 Concentration scale of a functional response
Efficacy Emax Magnitude of the maximum modeled response
Receptor capacity Bmax Maximum binding capacity in a specified binding assay/model
Do not collapse these concepts: a ligand can have high binding affinity without having high functional efficacy, and potency can be affected by system-level properties beyond receptor affinity alone.
14 · Receptor reserve

14. Spare Receptors and Signal Amplification

In some biological systems, maximal response can occur without occupancy of every available receptor. This phenomenon is often described using the concept of spare receptors or receptor reserve.

The underlying idea is that the receptor-to-effect pathway can amplify the signal. As a result, a relatively small amount of receptor activation can produce a substantial functional response.

This provides one explanation for why an observed \(EC_{50}\) can be lower than a binding \(K_d\) under some experimental conditions.

The relationship should not be interpreted as a universal rule. Receptor reserve is a system property and can depend on receptor density, downstream signaling, assay conditions, and the particular model used.

15 · Mechanistic pharmacology

15. From Binding to Pharmacologic Response

A mechanistic pharmacology model can be constructed as a sequence of linked processes rather than treating concentration and effect as a single empirical relationship.

\[ C(t) \rightarrow \text{Receptor binding} \rightarrow \text{Receptor occupancy} \rightarrow \text{Signal transduction} \rightarrow E(t) \]

For example, a PK model may first predict the free concentration \(C(t)\). That concentration can drive receptor binding, which produces an occupancy signal. The occupancy can then drive a downstream response model.

This approach is useful when the biological mechanism is scientifically important or when a simple direct Emax relationship does not adequately describe the observed behavior.

Mechanistic modeling principle: each additional biological layer introduces additional parameters and assumptions. More mechanistic detail can be valuable, but only when the available data can support the added complexity.
16 · Dynamic receptor models

16. When Binding Is Not at Equilibrium

The simple occupancy equation assumes equilibrium. That assumption may be inadequate when drug concentration changes rapidly, binding is slow, or receptor turnover and trafficking are important.

A dynamic receptor model can explicitly track receptor states. For example:

\[ \frac{dR}{dt} = -k_{\mathrm{on}}CR +k_{\mathrm{off}}DR \]

and:

\[ \frac{dDR}{dt} = k_{\mathrm{on}}CR -k_{\mathrm{off}}DR \]

These equations allow receptor occupancy to lag behind changes in drug concentration.

This distinction can become important when a drug has rapid PK changes but relatively slow receptor association or dissociation. In such cases, the effect may reflect the history of exposure rather than concentration at one instant.

17 · Receptor turnover

17. Receptor Turnover and Regulation

Receptors themselves can be synthesized, degraded, internalized, recycled, or otherwise regulated. When receptor abundance changes over time, a constant-receptor model may no longer be sufficient.

A simple turnover model might be written:

\[ \frac{dR_{\mathrm{tot}}}{dt} = k_{\mathrm{in}} - k_{\mathrm{out}}R_{\mathrm{tot}} \]

where \(R_{\mathrm{tot}}\) is total receptor abundance, \(k_{\mathrm{in}}\) represents receptor production, and \(k_{\mathrm{out}}\) represents first-order loss.

Drug exposure can then be linked to receptor turnover or activity through an additional model. This creates a bridge between receptor pharmacology and time-dependent pharmacodynamic models.

Such models can help represent delayed effects, tolerance, desensitization, and recovery when those mechanisms are supported by the experimental data.

18 · Worked example

18. Worked Example: Receptor Occupancy

Suppose a drug has a modeled receptor-binding constant of \(K_d=10\) nM. What receptor occupancy is predicted at 2 nM, 10 nM, and 90 nM?

Step 1: Occupancy at 2 nM

\[ f_{\mathrm{occ}} = \frac{2}{10+2} = \frac{2}{12} \approx0.167 \]

The predicted occupancy is approximately 16.7%.

Step 2: Occupancy at 10 nM

\[ f_{\mathrm{occ}} = \frac{10}{10+10} = 0.50 \]

The predicted occupancy is 50%, as expected because the concentration equals \(K_d\) under the simple one-site model.

Step 3: Occupancy at 90 nM

\[ f_{\mathrm{occ}} = \frac{90}{10+90} = 0.90 \]

The predicted occupancy is 90%.

Step 4: Interpret the result

Concentration Calculation Predicted occupancy
2 nM \(2/(10+2)\) 16.7%
10 nM \(10/(10+10)\) 50.0%
90 nM \(90/(10+90)\) 90.0%

The example illustrates the nonlinear relationship between concentration and occupancy. Increasing concentration from 2 to 10 nM increases occupancy substantially, but progressively larger concentrations are required to move occupancy closer to complete saturation.

19 · Binding versus effect

19. Worked Example: Why Kd and EC50 Can Differ

Suppose the same drug has \(K_d=10\) nM but an observed \(EC_{50}=2\) nM in a functional assay.

These two values should not automatically be treated as contradictory. They describe different quantities:

Quantity Value What it describes
Kd 10 nM Binding equilibrium in the specified receptor model
EC50 2 nM Concentration scale for the functional response

A downstream signaling system may amplify receptor activation so that a relatively small fraction of occupied receptors produces a substantial response. Alternatively, differences in assay conditions, receptor expression, endogenous signaling, or model structure can contribute to the observed relationship.

Interpretation: comparing \(K_d\) and \(EC_{50}\) can be informative, but the comparison should be made within the context of the specific receptor system, assay, and model.
20 · Model hierarchy

20. A Hierarchy of Pharmacology Models

Receptor pharmacology can be represented at several levels of complexity. The appropriate model depends on the scientific question and the information contained in the data.

Model Primary purpose Typical quantities
One-site binding Describe simple receptor-ligand equilibrium Kd, Bmax
Binding kinetics Describe association and dissociation over time kon, koff
Competitive binding Describe competition for a binding site Ki, Kd
Emax model Describe concentration-effect relationships E0, Emax, EC50
Hill model Allow flexible concentration-effect slope EC50, Hill coefficient
Mechanistic receptor model Connect concentration, binding, signaling, and response Multiple kinetic and transduction parameters
Receptor turnover model Describe changing receptor abundance Production and loss parameters

The progression from simple to mechanistic models should not be viewed as a requirement to always use the most complex model. A simpler model may be more appropriate when the data cannot identify additional mechanisms.

21 · PK → receptor → PD

21. Connecting PK to Receptor Pharmacology

Receptor models become especially useful when they are connected to pharmacokinetics. PK describes how drug concentration changes over time, while receptor models describe how that concentration interacts with a molecular target.

\[ \text{Dose} \rightarrow C(t) \rightarrow \text{Receptor binding} \rightarrow \text{Occupancy} \rightarrow \text{Effect} \]

A simple integrated model might use a PK-generated concentration \(C(t)\) in the occupancy equation:

\[ f_{\mathrm{occ}}(t) = \frac{C(t)} {K_d+C(t)} \]

A downstream response model could then use occupancy as its driver:

\[ E(t) = E_0+ E_{\max}f_{\mathrm{occ}}(t) \]

This is one possible mechanistic PK/PD structure. More complex models can include delayed receptor binding, receptor turnover, signal transduction, active and inactive receptor states, tolerance, or indirect-response mechanisms.

22 · From data to model

22. How Are Receptor Models Estimated From Data?

The model-estimation process depends on the type of experimental data. Binding experiments, functional assays, biomarker studies, and clinical PK/PD datasets provide different types of information.

  1. Define the scientific question. Determine whether the goal is to estimate affinity, receptor capacity, binding kinetics, functional potency, efficacy, or a mechanistic pathway.
  2. Identify the observable. Examples include bound ligand, free ligand, receptor occupancy, concentration, biomarker response, or clinical effect.
  3. Choose the structural model. Start with a model that represents the biological hypothesis at an appropriate level of complexity.
  4. Specify the observation model. Measured binding or response contains experimental variability and may require an appropriate residual-error structure.
  5. Estimate parameters. Parameters such as \(K_d\), \(B_{\max}\), \(k_{\mathrm{on}}\), \(k_{\mathrm{off}}\), \(EC_{50}\), and \(E_{\max}\) can be estimated using appropriate statistical methods.
  6. Evaluate model adequacy. Inspect residuals, fitted curves, parameter plausibility, uncertainty, sensitivity, and alternative model structures.
  7. Use the model for inference or prediction. Predictions should remain conditional on the model assumptions and the information contained in the data.
Important: a curve that visually fits the data well does not by itself establish the underlying molecular mechanism. Different models can sometimes produce similar observable curves.
23 · Identifiability

23. Why Identifiability Matters

Mechanistic pharmacology models can contain many parameters. A central statistical question is whether the available data contain enough information to estimate those parameters reliably.

For example, a dataset might provide enough information to estimate a concentration-effect curve but not enough information to separately identify every parameter in a detailed receptor-transduction model.

Parameters can become difficult to distinguish when:

  • the concentration range does not cover the relevant part of the curve;
  • there are too few observations near informative regions of the response;
  • multiple parameters influence the observations in similar ways;
  • binding and downstream signaling processes occur on time scales that cannot be resolved by the sampling schedule;
  • the experiment does not contain sufficient perturbation to distinguish competing mechanisms.

This is why experimental design is an important part of mechanistic pharmacology modeling. The best model cannot recover information that the experiment did not measure.

24 · Interpretation

24. What Receptor Models Do Not Tell Us Automatically

Receptor models are useful mathematical representations, but their parameters require interpretation within the experimental and biological context.

  • A Kd is not automatically an EC50. The quantities come from different models and experimental concepts.
  • Binding does not automatically imply functional activation. A ligand can bind without producing the same downstream response as another ligand.
  • A fitted curve does not prove a mechanism. Alternative models can sometimes provide similar fits over the observed concentration range.
  • Parameters are model-dependent. Changing the structural model can change the estimated parameters and their interpretation.
  • Assay conditions matter. Receptor expression, signaling components, ligand depletion, incubation time, and other experimental features can affect observed relationships.
  • Extrapolation requires caution. Predictions outside the observed concentration or time range can depend strongly on model assumptions.
Modeling principle: distinguish what was directly measured from what is inferred through the receptor model. A mechanistic interpretation is strongest when the experimental design contains information capable of testing that mechanism.
25 · Practical workflow

25. A Practical Receptor-Pharmacology Modeling Workflow

  1. Start with the biological question. Are you studying affinity, occupancy, kinetics, efficacy, antagonism, receptor regulation, or exposure-response?
  2. Identify the experimental system. Specify the receptor, ligand, assay, concentration range, and sampling schedule.
  3. Determine the relevant observable. Binding, occupancy, biomarker response, and clinical effect require different modeling approaches.
  4. Begin with an appropriate simple model. Use the simplest structure that can answer the scientific question.
  5. Check whether equilibrium assumptions are reasonable. If binding is time-dependent, consider an explicit kinetic model.
  6. Estimate parameters and uncertainty. Do not interpret point estimates without considering how precisely the data support them.
  7. Evaluate competing explanations. Consider whether alternative receptor or transduction models could explain the observations.
  8. Connect the receptor model to PK when appropriate. Use concentration-time profiles as the input when receptor behavior is being studied in vivo.
  9. Validate predictions. Compare model predictions with independent observations whenever possible.
  10. Use the model for simulation and decision support. Clearly separate observed evidence from model-based predictions.

26. Key Takeaways

  • Receptor-binding models describe how ligands interact with molecular targets and how receptor occupancy changes with concentration.
  • For a simple one-site equilibrium model, \(f_{\mathrm{occ}}=C/(K_d+C)\), and \(K_d\) corresponds to 50% occupancy.
  • \(K_d\) describes binding affinity in a specified model; it is not automatically equivalent to EC50.
  • Binding kinetics introduce \(k_{\mathrm{on}}\) and \(k_{\mathrm{off}}\), which describe how rapidly association and dissociation occur.
  • Saturation-binding models can estimate both binding capacity \(B_{\max}\) and affinity-related parameters such as \(K_d\).
  • Competitive-binding models describe how one ligand can alter the apparent binding behavior of another.
  • Receptor occupancy and pharmacologic effect are related but need not be identical because signaling pathways can amplify, attenuate, or otherwise transform receptor activation.
  • Emax and Hill models describe concentration-effect relationships, with EC50 representing a functional concentration scale and Emax representing the modeled maximum effect.
  • Potency, efficacy, affinity, and receptor capacity are distinct concepts and should not be used interchangeably.
  • Mechanistic receptor models can connect PK concentration-time profiles to binding, occupancy, signal transduction, and pharmacologic effect.
  • Dynamic receptor and receptor-turnover models are useful when equilibrium or constant-receptor assumptions are inadequate.
  • Model complexity should be supported by the available data. A more detailed mechanistic model is not automatically more informative if its parameters cannot be identified.
Next step

Where to Go Next

A natural progression from receptor-binding models is to study target engagement and receptor occupancy models, followed by Emax and Hill models, competitive antagonism, indirect-response models, receptor turnover, tolerance and desensitization, and mechanistic PK/PD models.

The next step is to connect these receptor concepts to observed drug concentrations over time. This provides the foundation for exposure-response modeling and for more mechanistic pharmacometric models that link dose → PK → target engagement → pharmacologic response.

← Back to Pharmacokinetics Tutorials