Why a Binding Equilibrium Reduction?
In target-mediated drug disposition (TMDD), free drug binds free target to form a drug–target complex. If reversible binding adjusts much faster than drug and target concentrations change, explicitly integrating the short binding transient may be unnecessary. The complex can instead be calculated from an equilibrium relation while slower processes—such as drug elimination, target turnover, and complex internalization—continue to change the amounts of drug and target.
This is a timescale approximation, not a claim that the entire system is at steady state. In particular, free drug and target may change substantially during the simulation. The key question is whether reversible binding remains close to equilibrium as those slower changes occur.
Model and Equilibrium Closure
Consider a well-mixed system with concentrations \(D\), \(R\), and \(C\). Free drug is supplied at rate \(u(t)\) and eliminated with first-order rate constant \(k_{\mathrm{el}}\). Free target is synthesized at rate \(k_{\mathrm{syn}}\) and degraded with first-order rate constant \(k_{\mathrm{deg}}\). Complex dissociates at rate \(k_{\mathrm{off}}C\), forms at rate \(k_{\mathrm{on}}DR\), and is internalized at rate \(k_{\mathrm{int}}C\). The mass-balance equations are
The reversible binding reaction alone is \(D+R\rightleftharpoons C\). At equilibrium, its forward and reverse fluxes balance: \(k_{\mathrm{on}}DR=k_{\mathrm{off}}C\). Rearranging gives
If \(k_{\mathrm{on}}\) is measured in concentration\(^{-1}\)time\(^{-1}\) and \(k_{\mathrm{off}}\) in time\(^{-1}\), then \(K_d\) has concentration units. It is the free-drug concentration at which half of the target is bound only in the special case where the free drug concentration is held at \(K_d\) and the binding system is at equilibrium: \(C/R=D/K_d=1\).
A consistent reduced model tracks total drug \(A=D+C\) and total target \(T=R+C\). Adding the corresponding full-model equations cancels the reversible binding fluxes and gives
The equilibrium approximation closes these balances by using \(A=D+C\), \(T=R+C\), and \(C=DR/K_d\). For specified \(A\) and \(T\), the physically admissible solution has \(0\leq C\leq \min(A,T)\); then \(D=A-C\) and \(R=T-C\). This total-pool formulation accounts for complex accumulation and loss while avoiding the inconsistency of setting the binding flux to zero in the free-drug equation and then ignoring the changing amount of drug bound in \(C\).
Worked Example: Calculating the Equilibrium Dissociation Constant
Suppose hypothetical binding measurements give \(k_{\mathrm{on}}=0.20\ \mathrm{nM}^{-1}\mathrm{h}^{-1}\) and \(k_{\mathrm{off}}=0.10\ \mathrm{h}^{-1}\). Calculate \(K_d\), then find the equilibrium complex concentration when \(D=2.0\ \mathrm{nM}\) and \(R=3.0\ \mathrm{nM}\).
Calculate \(K_d\). Applying its definition,
Calculate \(C\). Under the equilibrium closure,
This result is consistent if the specified \(D\) and \(R\) are free concentrations: the corresponding total target concentration is \(R+C=15\ \mathrm{nM}\), and total drug is \(D+C=14\ \mathrm{nM}\). A complex concentration larger than either free concentration is not itself a contradiction; it must, however, respect the total-pool constraints.
Derivation from the Fast-Binding Limit
The equilibrium relation follows from the fast reversible part of the complex equation. To see the timescale explicitly, temporarily hold free \(D\) and \(R\) fixed and ignore internalization. The complex equation becomes \(dC/dt=k_{\mathrm{on}}DR-k_{\mathrm{off}}C\). Its solution approaches \(C^*=k_{\mathrm{on}}DR/k_{\mathrm{off}}=DR/K_d\), with deviations from that value shrinking exponentially at rate \(k_{\mathrm{off}}\). When free pools are allowed to respond to binding, the relaxation rate also reflects the association flux; near a binding equilibrium with no internalization, it is \(k_{\mathrm{on}}(D+R)+k_{\mathrm{off}}\).
The justification is a separation of timescales. If the relaxation time of reversible binding is much shorter than the timescale over which drug and target pools are driven away from their current values, the system spends most of its time near the binding equilibrium manifold. The approximation replaces the brief initial adjustment and small ongoing departures from that manifold with an algebraic constraint. This does not set \(dA/dt\) or \(dT/dt\) to zero.
Internalization creates an important qualification. The full complex equation has the loss term \(k_{\mathrm{int}}C\). At a stationary complex concentration it implies \(k_{\mathrm{on}}DR=(k_{\mathrm{off}}+k_{\mathrm{int}})C\), not \(k_{\mathrm{on}}DR=k_{\mathrm{off}}C\). Thus, the equilibrium closure is appropriate when dissociation is fast relative to internalization, typically \(k_{\mathrm{int}}\ll k_{\mathrm{off}}\), as well as when reversible binding is fast relative to the other changes in the system.
Worked Example: Comparing Equilibrium and QSS Closures
Use the hypothetical rates \(k_{\mathrm{on}}=0.20\ \mathrm{nM}^{-1}\mathrm{h}^{-1}\), \(k_{\mathrm{off}}=0.10\ \mathrm{h}^{-1}\), and \(k_{\mathrm{int}}=0.0050\ \mathrm{h}^{-1}\), with free concentrations \(D=2.0\ \mathrm{nM}\) and \(R=3.0\ \mathrm{nM}\). Compare the equilibrium complex with the quasi-steady-state (QSS) complex.
The equilibrium value from the rapid reversible binding assumption is
In the QSS approximation, the complex equation is set to zero while retaining internalization:
Substitution gives \(K_{\mathrm{SS}}=(0.10+0.0050)/0.20=0.525\ \mathrm{nM}\), and
The QSS value is \(11.43/12.0=0.9525\) times the equilibrium value, about \(4.75\%\) lower. In general, at fixed free \(D\) and \(R\), \(C_{\mathrm{QSS}}/C_{\mathrm{eq}}=k_{\mathrm{off}}/(k_{\mathrm{off}}+k_{\mathrm{int}})\). The small difference here is consistent with \(k_{\mathrm{int}}/k_{\mathrm{off}}=0.050\), so internalization is relatively slow compared with dissociation.
Timescale Conditions and the QSS Distinction
Both reductions eliminate the explicit fast complex transient, but their algebraic assumptions differ. The equilibrium reduction treats reversible binding as equilibrated and uses \(K_d=k_{\mathrm{off}}/k_{\mathrm{on}}\). QSS sets the net rate of change of complex approximately to zero, including its internalization loss, and uses \(K_{\mathrm{SS}}=(k_{\mathrm{off}}+k_{\mathrm{int}})/k_{\mathrm{on}}\). QSS can therefore accommodate internalization that is not negligible relative to dissociation, provided complex adjustment is still fast compared with the slower evolution of the pools.
| Feature | Rapid-binding equilibrium | QSS |
|---|---|---|
| Complex relation | \(C=DR/K_d\) | \(C=DR/K_{\mathrm{SS}}\) |
| Constant | \(K_d=k_{\mathrm{off}}/k_{\mathrm{on}}\) | \(K_{\mathrm{SS}}=(k_{\mathrm{off}}+k_{\mathrm{int}})/k_{\mathrm{on}}\) |
| Role of internalization | Small relative to dissociation for the binding-equilibrium closure | Included in the complex-loss balance |
| Core requirement | Reversible binding equilibrates rapidly and remains near equilibrium | Complex adjusts rapidly enough that its accumulation transient is negligible |
A useful local timescale check is to compare binding relaxation with the processes that change free drug or target. With concentrations approximately fixed over the short binding transient, a characteristic reversible relaxation rate is \(k_{\mathrm{on}}(D+R)+k_{\mathrm{off}}\); its reciprocal is a local binding timescale. If internalization is included in the transient complex dynamics, the corresponding local rate has an additional \(k_{\mathrm{int}}\) contribution. These estimates are guides, not universal guarantees: changing concentrations, input schedules, or model compartments can alter the relevant timescales.
Worked Example: Checking Timescale Separation
Continue with the hypothetical rates \(k_{\mathrm{on}}=0.20\ \mathrm{nM}^{-1}\mathrm{h}^{-1}\), \(k_{\mathrm{off}}=0.10\ \mathrm{h}^{-1}\), and \(k_{\mathrm{int}}=0.0050\ \mathrm{h}^{-1}\), at \(D=2.0\ \mathrm{nM}\) and \(R=3.0\ \mathrm{nM}\). Suppose \(k_{\mathrm{el}}=0.020\ \mathrm{h}^{-1}\) and \(k_{\mathrm{deg}}=0.030\ \mathrm{h}^{-1}\). Assess whether the rates make an equilibrium reduction plausible.
Using the reversible relaxation estimate,
The corresponding binding timescale is \(1/1.10=0.909\ \mathrm{h}\). By comparison, the characteristic drug-elimination and free-target-degradation times are \(1/k_{\mathrm{el}}=50.0\ \mathrm{h}\) and \(1/k_{\mathrm{deg}}=33.3\ \mathrm{h}\). Also, \(k_{\mathrm{int}}/k_{\mathrm{off}}=0.050\). At these concentrations and rates, binding is much faster than the stated turnover and elimination processes, and internalization is small relative to dissociation. Those comparisons support the equilibrium approximation locally. They do not prove it will hold at every concentration or during every dosing transient.
Solving with Total Concentrations
The equilibrium closure is often most useful when the model tracks totals. Since \(D=A-C\) and \(R=T-C\), substitution into \(C=DR/K_d\) gives one equation for the physically admissible complex concentration:
Worked Example: Finding Complex from Total Drug and Target
Suppose hypothetical total concentrations are \(A=10.0\ \mathrm{nM}\) and \(T=4.0\ \mathrm{nM}\), with \(K_d=2.0\ \mathrm{nM}\). Find the equilibrium free concentrations and complex concentration.
Substitution into the total-pool relation gives \(2C=(10-C)(4-C)\). Expanding and collecting terms,
The roots are \(C=(16\pm\sqrt{96})/2\), or approximately \(3.101\ \mathrm{nM}\) and \(12.90\ \mathrm{nM}\). The second root is impossible because it exceeds the total target \(T=4.0\ \mathrm{nM}\). Thus \(C=3.101\ \mathrm{nM}\), \(D=A-C=6.899\ \mathrm{nM}\), and \(R=T-C=0.899\ \mathrm{nM}\). Checking the closure, \(DR/K_d=(6.899)(0.899)/2.0\approx3.10\ \mathrm{nM}\), consistent with the calculated complex, subject to rounding.
Diagnostics, Limitations, and Common Mistakes
The approximation should be assessed over the concentrations and time periods relevant to the application, not just at one nominal operating point. Compare local binding relaxation with the fastest meaningful changes in free drug and target, including abrupt input changes. A short transient after a dose or a rapid change in target abundance can violate the assumption even if binding is fast relative to baseline turnover.
When a full TMDD model is available, compare its simulated free and total concentrations with the reduced model across relevant doses, parameter ranges, and initial conditions. Persistent differences in complex or free-drug trajectories, especially near rapid transitions, indicate that the reduced closure may not capture the needed dynamics. Numerical instability or negative reconstructed free concentrations can also reveal an incorrect implementation; the physical constraint \(C\leq\min(A,T)\) must be respected.
- Using \(K_d\) when internalization is substantial. Equating association and dissociation while ignoring a comparable complex sink can understate the impact of complex loss. The QSS constant includes \(k_{\mathrm{int}}\) in its denominator when that loss is retained.
- Calling the whole system equilibrated. The binding relation can hold while total drug and target are changing. Do not set their time derivatives to zero unless a separate steady-state assumption is justified.
- Confusing totals with free concentrations. The relation \(C=DR/K_d\) uses free \(D\) and \(R\), not total \(A\) and \(T\). Use the mass balances to reconstruct free pools.
- Checking only one timescale. Fast association alone is insufficient if dissociation is slow, internalization competes with dissociation, or dosing and pool changes occur on similarly short timescales.
The full TMDD differential equations are preferable when the binding transient itself matters, when binding and internalization timescales are not separated, or when parameters and conditions vary enough to make the approximation unreliable. QSS is a distinct alternative when complex accumulation is negligible but internalization cannot be ignored in the complex balance. Neither reduction is automatically valid merely because it is algebraically convenient.
Check Your Understanding
Answer each question, then compare with the solutions below.
- If \(k_{\mathrm{on}}=0.50\ \mathrm{nM}^{-1}\mathrm{h}^{-1}\) and \(k_{\mathrm{off}}=0.20\ \mathrm{h}^{-1}\), what is \(K_d\)?
- For \(K_d=1.0\ \mathrm{nM}\), \(D=2.0\ \mathrm{nM}\), and \(R=0.50\ \mathrm{nM}\), calculate equilibrium \(C\).
- With \(k_{\mathrm{off}}=0.10\ \mathrm{h}^{-1}\) and \(k_{\mathrm{int}}=0.10\ \mathrm{h}^{-1}\), is the equilibrium closure justified by the condition that internalization be small relative to dissociation? What is \(K_{\mathrm{SS}}/K_d\)?
- Why can a model use an equilibrium binding relation while total drug is changing over time?
Solutions. 1. \(K_d=0.20/0.50=0.400\ \mathrm{nM}\). 2. \(C=DR/K_d=(2.0)(0.50)/1.0=1.0\ \mathrm{nM}\). 3. No: \(k_{\mathrm{int}}/k_{\mathrm{off}}=1\), so internalization is not small relative to dissociation. Here \(K_d=0.10/k_{\mathrm{on}}\) and \(K_{\mathrm{SS}}=0.20/k_{\mathrm{on}}\), hence \(K_{\mathrm{SS}}/K_d=2\). 4. The equilibrium relation approximates only the fast reversible binding process; total drug can still change through input, elimination, and complex internalization.