Why Equal Affinity Does Not Mean Equal Engagement
Two ligands can have the same equilibrium affinity yet bind and leave a receptor at very different rates. At equilibrium, their fractional occupancies may be identical at a given free-ligand concentration. During a finite exposure, however, the faster ligand may approach that equilibrium sooner; after free ligand falls, the ligand with slower dissociation may remain bound longer. Thus the duration and timing of target engagement can differ even when an equilibrium affinity comparison does not distinguish the ligands.
The distinction matters for pharmacology because a biological response depends not only on how much target is occupied at equilibrium, but also on the exposure history and the relationship between occupancy and function. Kinetic differences alone do not establish greater clinical benefit or intrinsic efficacy. They identify a potential difference in engagement that must be interpreted with the response model and the exposure conditions.
For example, multiplying both \(k_{\mathrm{on}}\) and \(k_{\mathrm{off}}\) by the same positive factor leaves \(K_D\) unchanged but shortens both the characteristic association and dissociation times. A ligand with a larger \(k_{\mathrm{off}}\) therefore has a shorter residence time, even if its \(K_D\) matches that of another ligand. At a fixed \(C\), this kinetic distinction can affect how quickly occupancy is reached and how quickly it is lost.
Occupancy During Exposure and Washout
Assume a homogeneous, well-mixed system with one reversible binding site, constant receptor availability, and free ligand concentration \(C\) that is not appreciably depleted by binding. The association and dissociation steps follow mass-action kinetics. Under these assumptions, occupancy changes according to the rate of binding to unoccupied receptors minus the rate of dissociation from occupied receptors.
At constant \(C\), define \(k_{\mathrm{obs}}=k_{\mathrm{on}}C+k_{\mathrm{off}}\). Setting the rate of change to zero gives the steady-state occupancy. Starting at occupancy \(\theta_0\), solving the differential equation gives the complete time course:
The equilibrium expression depends on \(K_D\), while the approach rate depends on \(k_{\mathrm{obs}}\). After free ligand is removed so that \(C=0\), association stops and the model reduces to first-order dissociation:
Worked Example: Equal Affinity, Different Occupancy-Time Exposure
Consider two hypothetical ligands at a receptor with initially unoccupied sites. Ligand F has \(k_{\mathrm{on}}=0.060\ \mathrm{nM^{-1}h^{-1}}\) and \(k_{\mathrm{off}}=0.060\ \mathrm{h^{-1}}\). Ligand S has \(k_{\mathrm{on}}=0.012\ \mathrm{nM^{-1}h^{-1}}\) and \(k_{\mathrm{off}}=0.012\ \mathrm{h^{-1}}\). Each is applied at \(C=1.00\ \mathrm{nM}\) for 2.00 h, then free ligand is removed for 4.00 h. Assume the stated free concentration is maintained during exposure and becomes zero at washout.
Affinity and residence time. For F, \(K_D=0.060/0.060=1.00\ \mathrm{nM}\), and for S, \(K_D=0.012/0.012=1.00\ \mathrm{nM}\). Their equilibrium occupancy at \(1.00\ \mathrm{nM}\) would therefore be \(1/(1+1)=0.500\). Their residence times differ: \(1/0.060=16.67\ \mathrm{h}\) for F and \(1/0.012=83.33\ \mathrm{h}\) for S.
Occupancy at the end of exposure. Starting from zero occupancy, each ligand has \(\theta(t)=0.5(1-e^{-k_{\mathrm{obs}}t})\) during exposure, because \(C=K_D\). For F, \(k_{\mathrm{obs}}=0.060(1)+0.060=0.120\ \mathrm{h^{-1}}\), so \(\theta_F(2)=0.5(1-e^{-0.120(2)})=0.1067\). For S, \(k_{\mathrm{obs}}=0.024\ \mathrm{h^{-1}}\), so \(\theta_S(2)=0.5(1-e^{-0.024(2)})=0.02343\). F has higher occupancy at the end of this short exposure, although their equilibrium occupancies are equal.
Occupancy integrated over the full 6.00 h. This integral, with units of occupancy·h, summarizes target engagement over time; it is not itself a receptor response. For an exposure of duration \(T\), the exposure-period integral is \(\theta_{\mathrm{ss}}[T-(1-e^{-k_{\mathrm{obs}}T})/k_{\mathrm{obs}}]\). For F this is \(0.5[2-(1-e^{-0.240})/0.120]=0.1109\ \mathrm{occupancy\cdot h}\). Its washout integral is \(\theta_F(2)(1-e^{-0.060(4)})/0.060=0.3794\ \mathrm{occupancy\cdot h}\). The total is \(0.4903\ \mathrm{occupancy\cdot h}\).
For S, the exposure-period integral is \(0.5[2-(1-e^{-0.048})/0.024]=0.02362\ \mathrm{occupancy\cdot h}\). Its washout integral is \(\theta_S(2)(1-e^{-0.012(4)})/0.012=0.09152\ \mathrm{occupancy\cdot h}\). The total is \(0.1151\ \mathrm{occupancy\cdot h}\). Under this particular short exposure followed by a 4 h washout, F has the larger integrated occupancy. The result does not contradict S's longer residence time: F builds occupancy more rapidly, while S dissociates more slowly. A longer exposure or washout can change the comparison.
Why the Ranking Can Change With Time
During washout, each ligand's occupancy follows a different exponential decay. A ligand that begins washout with higher occupancy can lose that advantage if its dissociation is sufficiently faster. The crossing time can be calculated directly when the two washout occupancies are known.
Worked Example: A Washout Crossover
Suppose two hypothetical ligands have equal \(K_D=1.00\ \mathrm{nM}\), with \(k_{\mathrm{off,F}}=0.240\ \mathrm{h^{-1}}\), \(k_{\mathrm{on,F}}=0.240\ \mathrm{nM^{-1}h^{-1}}\), \(k_{\mathrm{off,S}}=0.060\ \mathrm{h^{-1}}\), and \(k_{\mathrm{on,S}}=0.060\ \mathrm{nM^{-1}h^{-1}}\). For this example, the modeled occupancies at the start of washout are 0.1504 for F and 0.07539 for S. These initial conditions represent the exposure history; they are not assumed to arise from the same exposure regimen as in the preceding example.
After washout begins, \(\theta_F(t)=0.1504e^{-0.240t}\) and \(\theta_S(t)=0.07539e^{-0.060t}\), where \(t\) is time in hours after washout. Setting these equal and solving gives:
Before this time, F has higher occupancy; after it, S has higher occupancy. For instance, at 4.00 h, F has occupancy \(0.1504e^{-0.960}=0.05759\), while S has occupancy \(0.07539e^{-0.240}=0.05930\). The crossover illustrates why a single measurement time cannot fully characterize kinetic selectivity.
More generally, if two ligands have washout occupancies \(a e^{-k_Ft}\) and \(b e^{-k_St}\), with \(a>b>0\) and \(k_F>k_S\), their positive crossover time is \(t=\ln(a/b)/(k_F-k_S)\). This follows by equating the trajectories, taking logarithms, and rearranging. If the initial occupancies or dissociation-rate ordering differs, a positive crossover need not exist.
From Kinetic Differences to Functional Differences
Kinetic selectivity describes differences in target engagement over time; it is not automatically functional selectivity. To infer a functional consequence, specify how occupancy maps to a response. Earlier material on occupancy-response curves and receptor reserve establishes why the same occupancy change need not produce the same response change in every system. A response may be nonlinear, delayed, amplified, or affected by receptor regulation.
The following calculation makes the mapping explicit. Assume, for illustration, that a proximal response tracks occupancy instantaneously and follows the same occupancy-response curve for both ligands. This is a simplifying model, not a general property of receptor signaling.
Worked Example: Applying a Shared Occupancy-Response Curve
Use the end-of-exposure occupancies from the first example: \(\theta_F=0.1067\) and \(\theta_S=0.02343\). Suppose both ligands have the same hypothetical maximal response \(E_{\max}\), and normalized response is \(E/E_{\max}=\theta/(\theta_{50}+\theta)\), with \(\theta_{50}=0.200\). Assume response follows this curve without a time delay.
For F, \(E_F/E_{\max}=0.1067/(0.200+0.1067)=0.3479\). For S, \(E_S/E_{\max}=0.02343/(0.200+0.02343)=0.1049\). Under this stipulated model and at this time point, F produces the larger predicted response because it has higher occupancy. The calculation does not show that F has greater intrinsic efficacy: the response mapping and \(E_{\max}\) were assumed identical. A different exposure duration can alter the occupancy ordering, and a signaling delay or receptor turnover can alter the response ordering.
Assumptions, Diagnostics, and Alternatives
The simple kinetic comparison is most useful when the system is close to the one-site, well-mixed model. Check whether free concentration is known and maintained, whether binding materially depletes ligand, whether receptor availability changes during the experiment, and whether internalization, recycling, or endogenous competitors affect occupancy. If these processes matter, the apparent association or dissociation pattern may not be explained by \(k_{\mathrm{on}}\) and \(k_{\mathrm{off}}\) alone.
A practical diagnostic is to compare the model-predicted time course with observations at multiple times during both exposure and washout. Systematic early or late residuals, a non-exponential washout pattern, or estimates that change markedly with the sampled interval can indicate that the assumed single binding process is inadequate or that concentrations are not controlled as assumed. A fitted \(K_D\) alone cannot diagnose these issues: different rate pairs can share the same ratio.
The main alternatives answer different questions. An equilibrium occupancy calculation is appropriate when concentrations and binding have had enough time to approach equilibrium; it is concise but omits the path to equilibrium. A time-dependent occupancy model uses changing \(C\) in the differential equation and is appropriate for finite or fluctuating exposures, but requires kinetic parameters and a concentration history. An occupancy-time integral compares cumulative engagement over a chosen interval, but does not by itself represent a biological response. A functional model adds a specified occupancy-to-response relationship and, if needed, signaling or receptor-state dynamics; it is more informative for function but introduces further assumptions and parameters.
Avoid treating residence time as a universal measure of drug quality. A long residence time can preserve engagement after free concentration falls, but it can also delay reversibility. The duration of exposure, the target's biological role, the response kinetics, and off-target binding all matter. Kinetic differences support a mechanistic hypothesis about timing; they do not alone establish therapeutic selectivity or clinical benefit.
Check Your Understanding
Work through each question, then compare your reasoning with the solutions.
- Two ligands have \(k_{\mathrm{on}}=0.080\ \mathrm{nM^{-1}h^{-1}}\), \(k_{\mathrm{off}}=0.040\ \mathrm{h^{-1}}\) for the first, and \(k_{\mathrm{on}}=0.020\ \mathrm{nM^{-1}h^{-1}}\), \(k_{\mathrm{off}}=0.010\ \mathrm{h^{-1}}\) for the second. Compare their \(K_D\) values and residence times.
- For a ligand with \(k_{\mathrm{off}}=0.060\ \mathrm{h^{-1}}\), calculate the dissociation half-life.
- At constant \(C=K_D\), starting from \(\theta=0\), why can two equal-affinity ligands have different occupancy after a short exposure?
- Does a larger occupancy-time integral necessarily imply a larger measured functional response? State one reason.
Solutions. (1) Both have \(K_D=0.040/0.080=0.500\ \mathrm{nM}\) and \(0.010/0.020=0.500\ \mathrm{nM}\). Their residence times are \(1/0.040=25.0\ \mathrm{h}\) and \(1/0.010=100\ \mathrm{h}\), respectively. (2) The half-life is \(\ln(2)/0.060=11.55\ \mathrm{h}\). (3) Their steady-state occupancies at \(C=K_D\) are both 0.500, but their \(k_{\mathrm{obs}}=k_{\mathrm{on}}C+k_{\mathrm{off}}\) values differ, so they approach that occupancy at different rates. (4) No. The occupancy-to-response relationship may be nonlinear, delayed, or altered by receptor reserve or signaling dynamics, so an occupancy-time integral alone does not determine the measured response.