1. What Is Binding Kinetics?
Binding kinetics describes how quickly a drug associates with and dissociates from its molecular target. Whereas pharmacokinetics describes how drug concentration changes over time, binding kinetics describes how target binding changes in response to that concentration.
For a simple reversible interaction between a drug \(C\) and a target \(R\), the binding process can be represented as:
Here, \(C\) is free drug, \(R\) is unbound target, and \(CR\) is the drug-target complex. The forward rate constant \(k_{\mathrm{on}}\) describes association, while \(k_{\mathrm{off}}\) describes dissociation.
A mechanistic PK/PD model can explicitly represent the intermediate binding process between drug concentration and pharmacologic effect.
2. Association and Dissociation
For a simple reversible binding interaction, two processes occur simultaneously. Drug molecules associate with available target sites, and drug-target complexes dissociate back into free drug and free target.
Association
Association is the formation of the drug-target complex:
The association rate depends on the concentrations of free drug and free target and on the association rate constant \(k_{\mathrm{on}}\). Under mass-action kinetics:
where \(C\) and \(R\) denote the concentrations of free drug and free target, respectively.
Dissociation
Dissociation is the reverse process:
The dissociation rate is proportional to the concentration of the existing complex:
The net rate of formation of the complex is therefore:
More explicitly, using \(C\), \(R\), and \(CR\) for the respective concentrations:
This differential equation is the basic building block of a mechanistic binding model.
3. What Do \(k_{\mathrm{on}}\) and \(k_{\mathrm{off}}\) Mean?
The two kinetic constants describe different aspects of the binding interaction.
| Parameter | Name | Interpretation |
|---|---|---|
| \(k_{\mathrm{on}}\) | Association rate constant | Describes the rate of complex formation per unit free drug and free target concentration. |
| \(k_{\mathrm{off}}\) | Dissociation rate constant | Describes the fractional rate at which the drug-target complex dissociates. |
| \(K_D\) | Equilibrium dissociation constant | Characterizes binding affinity under the equilibrium assumptions of the simple reversible model. |
| \(1/k_{\mathrm{off}}\) | Dissociation time scale | Provides a characteristic time scale for complex dissociation when dissociation follows first-order kinetics. |
The units of \(k_{\mathrm{on}}\) depend on the concentration unit used, commonly \(\mathrm{M^{-1}time^{-1}}\). The dissociation rate constant \(k_{\mathrm{off}}\) has units of inverse time, such as \(\mathrm{s^{-1}}\) or \(\mathrm{h^{-1}}\).
A useful distinction is that \(k_{\mathrm{on}}\) and \(k_{\mathrm{off}}\) are kinetic parameters. They describe how rapidly binding changes. Affinity, represented by \(K_D\), summarizes the equilibrium relationship.
4. From Binding Kinetics to \(K_D\)
For the simple reversible binding model, the equilibrium dissociation constant is related to the kinetic rate constants by:
This relationship follows by setting the net rate of complex formation to zero at equilibrium:
Rearranging gives:
Thus, \(K_D\) is determined by both association and dissociation kinetics. A small \(K_D\) corresponds to higher affinity in this model, but the same \(K_D\) can arise from different combinations of \(k_{\mathrm{on}}\) and \(k_{\mathrm{off}}\).
5. Binding Kinetics and Receptor Occupancy
One common way to connect binding to pharmacologic effect is through receptor occupancy. Define the fractional occupancy as the fraction of target sites present as drug-target complex:
where \(R_{\mathrm{tot}}\) is total target concentration:
Under equilibrium conditions for a simple one-to-one interaction, fractional occupancy can be expressed as:
This has the familiar hyperbolic form of a receptor-binding relationship. When free drug concentration is much lower than \(K_D\), occupancy is relatively low. When free drug concentration approaches or exceeds \(K_D\), occupancy increases toward saturation.
For a simple one-to-one equilibrium binding model, \(K_D\) is the free drug concentration associated with 50% target occupancy.
The equilibrium relationship is useful, but it does not capture the transient behavior that occurs when drug concentrations change rapidly. That is where explicit binding kinetics become important.
6. Why Binding Kinetics Matter in a PK/PD Model
Suppose plasma or effect-site drug concentration changes rapidly after a dose. If receptor binding is effectively instantaneous relative to the PK time scale, an equilibrium occupancy model may be adequate.
But if association or dissociation is slow, receptor occupancy can lag behind drug concentration. In that situation, using an instantaneous equilibrium relationship may miss an important component of the observed pharmacology.
The distinction can be represented conceptually as:
In an equilibrium model, occupancy is calculated directly from the current concentration:
In a kinetic binding model, occupancy is generated dynamically from the differential equation for the drug-target complex:
The second formulation allows the binding state at time \(t\) to depend on its previous history.
7. Binding Time Scales
Binding kinetics introduce characteristic time scales that can interact with PK and PD time scales.
For dissociation following first-order kinetics, a characteristic dissociation half-life is:
A small \(k_{\mathrm{off}}\) therefore corresponds to slow dissociation and a longer residence time of the drug-target complex.
Association is more dependent on the concentrations of the interacting species. Under conditions where free target remains approximately constant, the characteristic association rate can be related to the sum of association and dissociation terms:
The corresponding relaxation time scale is approximately:
This provides a useful way to compare binding kinetics with other processes in a PK/PD model.
| Process | Representative time scale | Potential modeling consequence |
|---|---|---|
| Rapid binding | Much shorter than PK changes | Equilibrium occupancy may be a reasonable approximation. |
| Intermediate binding | Comparable with PK changes | Transient occupancy may visibly lag concentration. |
| Slow dissociation | Long relative to concentration decline | Target engagement can persist after free drug falls. |
| Slow downstream signaling | Long relative to binding | Effect may persist even after occupancy has changed. |
The relevant comparison is therefore not simply whether binding is "fast" or "slow." The important question is how the binding time scale compares with the time scales of drug disposition and downstream pharmacology.
8. Connecting a PK Model to Binding
A mechanistic PK/PD model can place a binding model downstream of a pharmacokinetic model.
For example, suppose a one-compartment IV bolus model gives:
This concentration becomes the input to the binding model:
The resulting target occupancy is then:
The complete sequence is therefore:
This structure is especially useful when the pharmacologic effect cannot be adequately described as an instantaneous function of plasma concentration.
9. From Target Occupancy to Pharmacologic Effect
Target occupancy does not necessarily equal pharmacologic effect. A PD model is needed to describe how target engagement is translated into the measured endpoint.
A simple proportional relationship might be:
where \(E_0\) is baseline response and \(E_{\mathrm{max}}\) represents the maximum response attributable to the modeled occupancy pathway.
A more flexible transduction relationship could use a Hill function:
Here, \(\gamma\) controls the steepness of the occupancy-effect relationship.
10. Worked Example: A Simple Binding Kinetic Model
Consider a hypothetical drug-target interaction with:
- \(k_{\mathrm{on}}=0.10\ \mathrm{nM^{-1}h^{-1}}\)
- \(k_{\mathrm{off}}=0.020\ \mathrm{h^{-1}}\)
- Total target concentration \(R_{\mathrm{tot}}=10\ \mathrm{nM}\)
- Free drug concentration \(C=2\ \mathrm{nM}\)
Step 1: Calculate \(K_D\)
Step 2: Calculate equilibrium occupancy
Under the equilibrium assumption:
Thus, the predicted equilibrium occupancy is approximately 90.9%.
Step 3: Calculate the amount of bound target at equilibrium
Step 4: Calculate the characteristic dissociation half-life
This example illustrates an important point. The equilibrium affinity is strong, but the dissociation time scale is also long. Consequently, target engagement may persist for many hours after free drug concentration begins to decline.
Step 5: Compare with the equilibrium assumption
If drug concentration changes much faster than the approximately 34.7-hour dissociation time scale, instantaneous equilibrium may not be a good representation of the transient occupancy profile. A kinetic binding model would retain the history of the target-bound state.
11. Residence Time and Duration of Target Engagement
The dissociation rate constant provides a simple measure of how long a drug-target complex persists. A commonly used characteristic quantity is the mean residence time:
For the worked example:
The residence time is distinct from systemic PK half-life. A drug can have a relatively short plasma half-life while maintaining target engagement because the drug-target complex dissociates slowly.
Conversely, a drug can remain in the body for a long time while having relatively rapid target dissociation. In that situation, systemic exposure may persist without equally persistent target occupancy.
| Quantity | Primarily describes | Typical controlling parameters |
|---|---|---|
| PK half-life | Drug concentration decline | CL, V, and structural PK model |
| Binding residence time | Persistence of drug-target complex | \(k_{\mathrm{off}}\) |
| PD duration | Persistence of pharmacologic response | Binding plus downstream signaling and turnover |
12. Equilibrium Binding vs. Kinetic Binding
Both equilibrium and kinetic binding models can be useful. The appropriate choice depends on the scientific question, available data, and relative time scales.
| Feature | Equilibrium model | Kinetic model |
|---|---|---|
| Primary parameters | \(K_D\) | \(k_{\mathrm{on}}\), \(k_{\mathrm{off}}\) |
| Binding state | Calculated directly from current concentration | Obtained by solving a differential equation |
| Transient behavior | Not explicitly represented | Explicitly represented |
| Binding history | Generally ignored | Retained through the dynamic state |
| Usefulness | Useful when binding is fast relative to relevant PK/PD changes | Useful when binding kinetics themselves influence the response |
A useful modeling strategy is to begin with the simplest model that can answer the question. If an equilibrium model cannot explain observed hysteresis, delayed onset, prolonged target engagement, or other temporal features, explicit binding kinetics may provide the necessary additional structure.
13. Binding Kinetics Can Produce Hysteresis
When effect depends on a dynamic binding state rather than directly on instantaneous plasma concentration, the relationship between concentration and effect can differ during the rising and falling phases of exposure.
This phenomenon is commonly described as hysteresis. In a concentration-effect plot, the path during increasing concentration can differ from the path during decreasing concentration.
A dynamic binding or downstream-effect model can create a time-dependent concentration-effect relationship. The exact direction and shape of hysteresis depend on the model and mechanism.
Hysteresis is not itself proof of receptor binding kinetics. Similar patterns can arise from effect-site equilibration, active metabolites, indirect-response mechanisms, downstream signaling, or other delayed processes. Mechanistic interpretation therefore requires consideration of the complete PK/PD system.
14. Binding With Target Turnover
Many biological targets are not static. Receptors, enzymes, proteins, cells, and other target species can be synthesized and removed over time.
A simple target-turnover model can be written as:
At baseline steady state:
When binding is added, the model can distinguish between free and bound target. For example:
The target turnover and binding processes can then interact. A drug may alter target availability, target internalization, target synthesis, or downstream signaling in addition to simply occupying the target.
15. When Binding Kinetics Also Affect PK
In some systems, target binding is not merely downstream of PK. Binding can feed back into drug disposition itself.
This occurs when the drug-target complex is internalized, degraded, recycled, or otherwise removed from the system. Target-mediated drug disposition (TMDD) is a prominent example of this type of mechanism.
A simplified conceptual structure is:
In such a model, binding can create an additional nonlinear route of drug elimination. Consequently, the concentration profile itself can contain information about the binding and target-mediated processes.
This creates a bidirectional relationship:
This is one reason mechanistic PK/PD models can become substantially more complex than simple exposure-response models.
16. How Are Binding Parameters Estimated?
Estimating \(k_{\mathrm{on}}\) and \(k_{\mathrm{off}}\) requires data that contain information about the temporal binding process. Depending on the experimental system, measurements may come from biochemical binding assays, cellular systems, preclinical experiments, clinical biomarker measurements, or integrated PK/PD studies.
- Measure drug concentration or exposure. The concentration driving the binding process should be characterized as well as possible.
- Measure target engagement or a suitable biomarker. The observation should contain information about binding or a downstream consequence of binding.
- Specify a mechanistic model. Define association, dissociation, target turnover, and other relevant processes.
- Estimate parameters. Estimate \(k_{\mathrm{on}}\), \(k_{\mathrm{off}}\), target quantities, and other model parameters using an appropriate estimation method.
- Evaluate identifiability. Determine whether the available data can distinguish the parameters individually rather than only identifying a combination such as \(K_D\).
- Check model adequacy. Use observed-versus-predicted plots, residual diagnostics, parameter plausibility, and sensitivity analyses.
A major practical issue is that \(k_{\mathrm{on}}\) and \(k_{\mathrm{off}}\) can be difficult to identify separately if the data primarily inform equilibrium affinity. In that situation, the data may support estimation of \(K_D\) more strongly than independent estimates of both kinetic constants.
17. Designing Data to Inform Binding Kinetics
The sampling design should reflect the time scales that the model is intended to estimate.
| Objective | Data feature that can help |
|---|---|
| Estimate rapid association | Early observations after a change in drug concentration. |
| Estimate slow dissociation | Observations extending sufficiently beyond the concentration change to capture complex decay. |
| Estimate equilibrium affinity | Measurements spanning concentrations around the relevant \(K_D\) range. |
| Distinguish kinetic models | Sampling that captures transient rather than only equilibrium states. |
| Characterize target turnover | Measurements over a time window long enough to observe target recovery or depletion. |
Sampling too sparsely can make a dynamic binding process appear instantaneous. Conversely, very dense sampling at times that contain little additional information may add observations without substantially improving parameter identifiability.
The experimental design should therefore be based on expected biological and kinetic time scales rather than on a generic sampling schedule.
18. A Hierarchy of Binding Models
Binding can be represented at different levels of mechanistic detail.
| Model | Representation | Potential use |
|---|---|---|
| Equilibrium occupancy | \(\mathrm{Occ}=C/(K_D+C)\) | Simple exposure-occupancy relationships when binding is effectively rapid. |
| Reversible kinetic binding | \(C+R\rightleftharpoons CR\) | Explicit association and dissociation dynamics. |
| Binding + target turnover | Binding coupled to synthesis and degradation | Targets whose abundance changes over time. |
| Binding + effect model | Occupancy drives downstream response | Mechanistic exposure-response modeling. |
| TMDD model | Binding coupled to drug disposition | Systems in which target interaction contributes materially to drug clearance. |
| Multi-state binding model | Multiple bound or conformational states | Systems with more than one kinetically relevant target state. |
The appropriate level of complexity depends on the question. Adding parameters that are not supported by the data can make a model harder to estimate and interpret without providing additional useful information.
19. Quasi-Equilibrium and Quasi-Steady-State Approximations
Full kinetic binding models can sometimes be simplified when their time scales are sufficiently fast relative to the processes of primary interest.
Under a quasi-equilibrium approximation, binding is treated as effectively equilibrated at each relevant drug concentration:
A related quasi-steady-state approach can arise when the complex formation rate rapidly approaches a dynamic steady state relative to slower processes. The resulting relationship may involve a parameter combination rather than the equilibrium \(K_D\) alone.
These approximations can be extremely useful, but they should be treated as model assumptions rather than universal biological laws.
20. Binding Kinetics vs. Effect-Site Equilibration
A delayed pharmacologic effect does not automatically imply slow receptor binding.
An alternative model introduces a hypothetical effect compartment that equilibrates with plasma concentration:
The pharmacodynamic effect is then modeled as a function of \(C_e\), rather than directly from \(C_p\).
Both an effect-compartment model and a kinetic binding model can generate delayed effects, but they represent different mechanisms.
| Model | Dynamic state | Mechanistic interpretation |
|---|---|---|
| Effect compartment | \(C_e\) | Empirical or semimechanistic delay between plasma and effect site. |
| Kinetic binding | \([CR]\) | Explicit drug-target association and dissociation. |
| Target turnover | \(R_{\mathrm{tot}}\) | Changing target abundance. |
| Indirect response | \(E\) or response mediator | Drug modifies production or loss of the response. |
Model selection should therefore be guided by available mechanistic evidence and by whether the data can distinguish the proposed mechanisms.
21. What Binding Models Do Not Tell Us Automatically
Binding models can be highly informative, but their parameters require careful interpretation.
- \(K_D\) is not the same as clinical potency. Cellular signaling, receptor density, tissue distribution, and downstream biology can all influence the concentration-effect relationship.
- \(k_{\mathrm{off}}\) is not automatically the duration of clinical effect. Downstream turnover and other processes can prolong or shorten pharmacologic effects.
- Observed plasma concentration may not equal the concentration at the target site. Tissue distribution and permeability can introduce additional delays.
- Binding may involve multiple target states. A single \(k_{\mathrm{on}}\)-\(k_{\mathrm{off}}\) pair may be an approximation.
- Equilibrium assumptions may fail during rapidly changing exposure. The binding state can lag behind concentration.
- Parameter identifiability depends on the data. A complex mechanistic model does not guarantee that every parameter is estimable.
- Model fit does not establish mechanism by itself. Distinct mechanisms can sometimes produce similar observable profiles.
22. A Practical Binding-Kinetics PK/PD Workflow
- Define the scientific question. Is the objective to characterize affinity, residence time, target engagement, pharmacologic effect, or drug disposition?
- Characterize the PK input. Determine which concentration should drive the binding process and whether plasma concentration is an adequate surrogate.
- Identify the target process. Define target concentration, binding sites, turnover, and relevant downstream processes.
- Start with the simplest plausible binding model. An equilibrium model may be sufficient when binding is rapid.
- Add kinetic binding when needed. Estimate \(k_{\mathrm{on}}\) and \(k_{\mathrm{off}}\) when transient binding behavior is supported by the data.
- Connect occupancy to effect. Specify an appropriate PD or transduction model.
- Consider target turnover. Include synthesis, degradation, internalization, or recycling when supported by the biology.
- Consider PK feedback. If binding contributes to elimination, evaluate whether a TMDD or related model is necessary.
- Assess identifiability and diagnostics. Determine which parameters are actually supported by the available observations.
- Use the model for simulation and prediction. Clearly distinguish measured data from model-based predictions.
23. Key Takeaways
- Binding kinetics describes how rapidly a drug associates with and dissociates from its molecular target.
- The basic reversible binding model is \(C+R\rightleftharpoons CR\), governed by \(k_{\mathrm{on}}\) and \(k_{\mathrm{off}}\).
- For the simple equilibrium model, affinity is characterized by \(K_D=k_{\mathrm{off}}/k_{\mathrm{on}}\).
- Two compounds can have the same \(K_D\) but substantially different association and dissociation kinetics.
- Receptor occupancy can be calculated dynamically from a kinetic binding model rather than assumed to instantaneously equilibrate with concentration.
- The dissociation half-life \(0.693/k_{\mathrm{off}}\) provides a characteristic time scale for loss of drug-target binding.
- Slow dissociation can produce persistent target engagement even when systemic drug concentration is declining.
- Drug concentration, target occupancy, and pharmacologic effect are distinct quantities and can have different time courses.
- Binding kinetics can contribute to hysteresis between concentration and effect, although other mechanisms can produce similar patterns.
- Target turnover can introduce additional delays and persistence beyond the binding process itself.
- When target binding contributes to drug elimination, binding and PK become coupled, as in target-mediated drug disposition.
- Equilibrium, kinetic, turnover, and TMDD models represent different levels of mechanistic detail and should be selected according to the scientific question and available data.
- Identifiability is critical: having \(k_{\mathrm{on}}\) and \(k_{\mathrm{off}}\) in a model does not mean that both can be estimated precisely from a particular dataset.
Where to Go Next
A natural progression is to study target-mediated drug disposition (TMDD), where the binding process is coupled directly to drug elimination. From there, related topics include nonlinear PK from target binding, receptor occupancy models, indirect-response models, translational PK/PD, and quantitative systems pharmacology.
The next tutorial can build directly on the equations introduced here by showing how \(k_{\mathrm{on}}\), \(k_{\mathrm{off}}\), target concentration, internalization, and systemic PK combine to produce nonlinear drug disposition in a TMDD model.
References
| Reference | Relevance |
|---|---|
| Weiss, M. & Mettetal, J. (general pharmacometric modeling literature) | Provides broader context for mechanistic PK/PD modeling and dynamic drug-effect relationships. |
| Copeland, R. A. The Pharmacological Revolution: How New Drug Targets and Mechanisms Are Transforming Drug Discovery. | Provides background on drug-target interactions, affinity, kinetics, and residence time. |
| Hocht, C. et al. and related receptor-occupancy literature | Provides examples of linking receptor binding and occupancy to pharmacologic effects. |
| Jusko, W. J. and related PK/PD modeling literature | Provides foundational context for mechanistic pharmacokinetic-pharmacodynamic modeling and indirect-effect relationships. |
| Meibohm, B. & Derendorf, H. Basic Concepts in Pharmacokinetics and Pharmacodynamics. | Provides foundational treatment of PK/PD relationships, concentration-effect models, and pharmacodynamic time courses. |
The equations presented in this tutorial describe standard mass-action models for a simple reversible one-to-one binding interaction. More complex biological systems may require multiple binding sites, cooperative binding, receptor turnover, internalization, active metabolites, or additional downstream signaling states.