Slow-Binding Inhibition Versus Rapid-Equilibrium Inhibition

Why Binding Time Changes the Activity Profile

A reversible inhibitor can produce a delayed pharmacodynamic response even when the inhibitor concentration changes immediately. The reason is that enzyme molecules need time to associate with inhibitor, and bound molecules need time to dissociate. A rapid-equilibrium model treats this adjustment as instantaneous; a slow-binding model represents it explicitly. The distinction matters when the measured activity is collected before binding has equilibrated, or when inhibitor concentration changes on a timescale comparable to binding.

The model below describes a single, reversible binding step. It isolates binding delay from changes in enzyme abundance, inhibitor depletion, and other mechanisms. Activity is taken to be proportional to the unbound enzyme fraction. This is appropriate when inhibitor-bound enzyme is inactive and the binding rates are constant under the assay conditions. If substrate affects inhibitor binding, the rates or model must account for that dependence; they should not automatically be treated as universal constants.

Definition: In the one-step reversible slow-binding model, free enzyme $E$ associates with inhibitor $I$ to form inactive complex $EI$, which can dissociate back to $E$ and $I$: $$ E+I \underset{k_{\mathrm{off}}}{\overset{k_{\mathrm{on}}}{\rightleftharpoons}} EI. $$ Here $k_{\mathrm{on}}$ is the association rate constant, with units of concentration$^{-1}$ time$^{-1}$, and $k_{\mathrm{off}}$ is the dissociation rate constant, with units of time$^{-1}$. The dissociation constant is $K_D=k_{\mathrm{off}}/k_{\mathrm{on}}$, with units of concentration.

Occupancy and Activity Equations

Let $E_T$ be the total enzyme concentration, assumed constant, and let $p(t)=[EI](t)/E_T$ be the fraction of enzyme bound to inhibitor. Then the unbound fraction is $1-p(t)$. Assume that free inhibitor concentration $[I](t)$ is known, that it is not materially depleted by enzyme binding, and that the binding event is the only process changing this fraction. Association removes unbound enzyme at rate $k_{\mathrm{on}}[I](t)(1-p)$; dissociation releases bound enzyme at rate $k_{\mathrm{off}}p$. Thus:

\frac{dp}{dt}=k_{\mathrm{on}}[I](t)(1-p)-k_{\mathrm{off}}p =k_{\mathrm{on}}[I](t)-\big(k_{\mathrm{on}}[I](t)+k_{\mathrm{off}}\big)p.

The terms in this equation have units of inverse time because $p$ is dimensionless. Define the instantaneous relaxation rate $\lambda(t)=k_{\mathrm{on}}[I](t)+k_{\mathrm{off}}$, also in inverse time. If the inhibitor-bound enzyme is inactive and activity per unbound enzyme is unchanged, relative activity is $r(t)=v(t)/v_0=1-p(t)$, where $v_0$ is activity without inhibitor under the same assay conditions.

Key result: At constant inhibitor concentration, the equilibrium bound fraction is $p_{\mathrm{eq}}=[I]/(K_D+[I])$, and the rate of approach to it is $\lambda=k_{\mathrm{on}}[I]+k_{\mathrm{off}}$. For an initial bound fraction $p(0)=p_0$, $$ p(t)=p_{\mathrm{eq}}+(p_0-p_{\mathrm{eq}})e^{-\lambda t}. $$ The binding half-time is $t_{1/2}=\ln(2)/\lambda$. It describes the time for the difference from equilibrium to halve, not necessarily the time for activity itself to halve.

Deriving the Constant-Concentration Profile

When $[I]$ is constant, the occupancy equation becomes $dp/dt=\lambda(p_{\mathrm{eq}}-p)$. To see the solution, subtract the equilibrium value and let $q(t)=p(t)-p_{\mathrm{eq}}$. Since $p_{\mathrm{eq}}$ is constant, $dq/dt=-\lambda q$. Separating variables gives $dq/q=-\lambda\,dt$; integrating from time zero to $t$ gives $\ln(q(t)/q(0))=-\lambda t$. Therefore $q(t)=q(0)e^{-\lambda t}$, which yields the stated profile. The exponential approaches equilibrium without overshooting it under this one-step model.

The equilibrium expression follows by setting $dp/dt=0$: $k_{\mathrm{on}}[I](1-p_{\mathrm{eq}})=k_{\mathrm{off}}p_{\mathrm{eq}}$. Rearranging gives $p_{\mathrm{eq}}=k_{\mathrm{on}}[I]/(k_{\mathrm{on}}[I]+k_{\mathrm{off}})$; dividing numerator and denominator by $k_{\mathrm{on}}$ gives $[I]/(K_D+[I])$. Thus $K_D$ determines the equilibrium occupancy, whereas both rate constants determine how quickly it is reached. A small $K_D$ alone does not establish rapid equilibration.

Worked Example: Activity After an Inhibitor Step

Suppose a hypothetical assay has $k_{\mathrm{on}}=0.20\ \mathrm{\mu M^{-1}\,min^{-1}}$, $k_{\mathrm{off}}=0.10\ \mathrm{min^{-1}}$, and inhibitor concentration stepped to $2.00\ \mathrm{\mu M}$. Initially, no enzyme is bound, so $p_0=0$. Calculate occupancy and relative activity after 4.00 minutes.

First, $K_D=k_{\mathrm{off}}/k_{\mathrm{on}}=(0.10\ \mathrm{min^{-1}})/(0.20\ \mathrm{\mu M^{-1}\,min^{-1}})=0.500\ \mathrm{\mu M}$. The equilibrium occupancy is $p_{\mathrm{eq}}=2.00/(0.500+2.00)=0.800$.

The relaxation rate is $\lambda=(0.20)(2.00)+0.10=0.500\ \mathrm{min^{-1}}$. Therefore, $p(4.00)=0.800+(0-0.800)e^{-(0.500)(4.00)} =0.800(1-e^{-2.00})=0.692$.

Relative activity is $r(4.00)=1-0.692=0.308$, or 30.8% of uninhibited activity. The equilibrium prediction is $1-0.800=0.200$, or 20.0%. At four minutes, the enzyme has not yet reached that lower equilibrium activity. The binding half-time is $\ln(2)/0.500=1.39$ minutes.

Changing Concentrations and the Rapid-Equilibrium Approximation

For changing inhibitor concentration, the occupancy remains governed by the differential equation, but the equilibrium target and relaxation rate both change over time. An integrating factor gives the exact solution for a specified concentration history:

p(t)=p_0\exp\left(-\int_0^t\lambda(u)\,du\right) +\int_0^t k_{\mathrm{on}}[I](s) \exp\left(-\int_s^t\lambda(u)\,du\right)ds, \qquad \lambda(u)=k_{\mathrm{on}}[I](u)+k_{\mathrm{off}}.

The rapid-equilibrium approximation replaces the dynamic bound fraction by its instantaneous equilibrium value, $p_{\mathrm{eq}}(t)=[I](t)/(K_D+[I](t))$. It is suitable only when occupancy can adjust much faster than that equilibrium value changes, over the time and concentration range of interest. A useful local justification follows by writing $q=p-p_{\mathrm{eq}}$. Differentiating gives $dq/dt=-\lambda q-dp_{\mathrm{eq}}/dt$. If $\lambda$ is large and the forcing changes slowly, $q$ is approximately $-\lambda^{-1}dp_{\mathrm{eq}}/dt$. Thus the approximate occupancy error is small when $|dp_{\mathrm{eq}}/dt|/\lambda$ is small relative to the occupancy scale or to the prespecified acceptable error.

This criterion compares quantities with compatible units: $|dp_{\mathrm{eq}}/dt|/\lambda$ is dimensionless. When $[I](t)>0$, the fractional concentration-change rate $|d\ln[I]/dt|$ can also help characterize the concentration timescale, but comparing that rate with $\lambda$ is a timescale check, not a guarantee by itself. The equilibrium occupancy may change more or less sharply than concentration, depending on its position relative to $K_D$. At $[I]=0$, a logarithmic concentration rate is undefined, so inspect the occupancy equation or evaluate the trajectory directly.

Worked Example: Washout Reveals Slow Dissociation

Consider hypothetical parameters $k_{\mathrm{on}}=0.10\ \mathrm{\mu M^{-1}\,min^{-1}}$ and $k_{\mathrm{off}}=0.10\ \mathrm{min^{-1}}$. Enzyme is exposed to $1.00\ \mathrm{\mu M}$ inhibitor for 10.0 minutes from an initially unbound state, then inhibitor is removed completely. Find the bound fraction at washout and five minutes later.

During exposure, $K_D=0.10/0.10=1.00\ \mathrm{\mu M}$, $p_{\mathrm{eq}}=1.00/(1.00+1.00)=0.500$, and $\lambda=(0.10)(1.00)+0.10=0.200\ \mathrm{min^{-1}}$. Hence $p(10.0)=0.500(1-e^{-2.00})=0.432$. Relative activity immediately before washout is $1-0.432=0.568$.

After washout, $[I]=0$, so the association term is zero and $dp/dt=-k_{\mathrm{off}}p$. Five minutes later, $p=0.432e^{-(0.10)(5.00)}=0.262$. Relative activity is then $1-0.262=0.738$. Activity has recovered, but it has not returned to 100% because dissociation is not instantaneous. A rapid-equilibrium model predicts no occupancy immediately after complete washout, and therefore predicts immediate full recovery under these assumptions.

Worked Example: Checking Whether Equilibrium Can Track a Declining Concentration

Suppose, hypothetically, that $[I](t)=1.00e^{-0.050t}\ \mathrm{\mu M}$, with $t$ in minutes, $k_{\mathrm{on}}=0.20\ \mathrm{\mu M^{-1}\,min^{-1}}$, and $k_{\mathrm{off}}=0.10\ \mathrm{min^{-1}}$. Assess the local rapid-equilibrium approximation at $t=0$, and consider whether the same conclusion can be extended to the full decline.

Here $K_D=0.10/0.20=0.500\ \mathrm{\mu M}$, and $p_{\mathrm{eq}}(0)=1.00/(0.500+1.00)=0.667$. Since $d[I]/dt=-0.050[I]$, the initial concentration derivative is $-0.050\ \mathrm{\mu M\,min^{-1}}$. Differentiating $p_{\mathrm{eq}}=[I]/(K_D+[I])$ gives $dp_{\mathrm{eq}}/dt=K_D(d[I]/dt)/(K_D+[I])^2$. At time zero this is $(0.500)(-0.050)/(1.50)^2=-0.0111\ \mathrm{min^{-1}}$.

The initial relaxation rate is $\lambda(0)=(0.20)(1.00)+0.10=0.300\ \mathrm{min^{-1}}$. The local tracking-error estimate is therefore $|dp_{\mathrm{eq}}/dt|/\lambda=0.0111/0.300=0.0370$, about 0.037 in bound fraction. This may be acceptable for an analysis allowing several percentage points of occupancy error, but it is not negligible if tighter accuracy is required. As the concentration declines, $\lambda(t)$ approaches $k_{\mathrm{off}}=0.100\ \mathrm{min^{-1}}$, while the concentration fractional decline rate remains $0.050\ \mathrm{min^{-1}}$. The separation becomes less strong. The initial check alone does not justify rapid equilibrium throughout the entire profile; compare the trajectories or assess the criterion across the relevant interval.

Model Checks, Alternatives, and Limitations

The rapid-equilibrium and dynamic models answer different questions. The first predicts activity from current inhibitor concentration alone; the second predicts activity from concentration history and initial occupancy. A steady-state approximation based on setting $dp/dt=0$ at each instant produces the same algebraic occupancy as rapid equilibrium in this one-step model, but it is justified only when binding relaxes quickly enough to track the changing conditions. It should not be used merely because an equilibrium formula is available.

Approach	What it represents	When it is useful
Rapid equilibrium	Occupancy is set to $[I](t)/(K_D+[I](t))$ at every time.	Binding relaxation is fast relative to relevant changes and measurement times.
Dynamic reversible binding	Occupancy follows an ODE with $k_{\mathrm{on}}$ and $k_{\mathrm{off}}$.	Onset, washout, or concentration changes show appreciable binding delay.
Equilibrium data only	Describes the concentration–occupancy relationship but not the approach to it.	Useful for equilibrium potency summaries, but cannot establish binding speed.

Useful diagnostics include plotting observed activity against both current inhibitor concentration and exposure history. A systematic lag, or different activity at the same concentration on the rising and falling portions of a profile, is consistent with dynamic binding, though it does not by itself prove that mechanism. Compare observed and predicted onset and washout, examine residuals over time, and check whether estimated rates predict the measured half-times. Estimates of $k_{\mathrm{on}}$ and $k_{\mathrm{off}}$ can be weakly determined if observations cover only equilibrium or only a narrow time range; data from both changing exposure and recovery can be more informative.

This one-step model is limited when enzyme concentration is not constant, free inhibitor is depleted, the inhibitor has multiple binding steps, bound enzyme retains partial activity, or binding rates vary with substrate or assay conditions. In those cases, the model structure or observation equation must be expanded. A rapid-equilibrium approximation can also fail near abrupt concentration changes even when the average exposure changes slowly, because the relevant comparison is local. The model should be judged against the sampling times and accuracy required for the intended PK/PD application.

Common Mistake

Wrong reasoning: “The inhibitor has a low $K_D$, so binding must be rapid.” This confuses equilibrium affinity with kinetic speed.
Correction: $K_D=k_{\mathrm{off}}/k_{\mathrm{on}}$ fixes the equilibrium ratio, but the relaxation rate at fixed concentration is $k_{\mathrm{on}}[I]+k_{\mathrm{off}}$. Distinct rate pairs can have the same $K_D$ and very different approach times.
Wrong reasoning: Directly compare $\lambda$, in inverse time, with $d[I]/dt$, in concentration per time.
Correction: Compare rates with matching dimensions, such as $\lambda$ with $|d\ln[I]/dt|$ when concentration is positive, or assess whether occupancy can track its changing equilibrium value using $|dp_{\mathrm{eq}}/dt|/\lambda$.
Wrong reasoning: Assume that activity returns immediately to baseline when inhibitor concentration becomes zero.
Correction: In the reversible model, bound enzyme remains bound until dissociation; after washout, its fraction decays as $e^{-k_{\mathrm{off}}t}$.

Key takeaway: Reversible slow binding is a history-dependent process. The rapid-equilibrium model is a limiting approximation that is defensible only when the enzyme-bound fraction adjusts quickly enough to track its changing equilibrium value.

Check Your Understanding

Answer each question, then compare with the brief solutions below.

For $k_{\mathrm{on}}=0.050\ \mathrm{\mu M^{-1}\,min^{-1}}$, $k_{\mathrm{off}}=0.20\ \mathrm{min^{-1}}$, and $[I]=4.00\ \mathrm{\mu M}$, calculate $K_D$, $p_{\mathrm{eq}}$, and $\lambda$.
At constant inhibitor concentration, what happens to $p(t)$ if $p_0>p_{\mathrm{eq}}$?
After complete washout, which parameter controls the occupancy recovery half-time?
Why are equilibrium measurements alone insufficient to establish that rapid equilibrium applies during changing exposure?

Solutions. (1) $K_D=0.20/0.050=4.00\ \mathrm{\mu M}$; $p_{\mathrm{eq}}=4.00/(4.00+4.00)=0.500$; $\lambda=(0.050)(4.00)+0.20=0.400\ \mathrm{min^{-1}}$. (2) Occupancy declines exponentially toward $p_{\mathrm{eq}}$, because $p(t)-p_{\mathrm{eq}}=(p_0-p_{\mathrm{eq}})e^{-\lambda t}$ remains positive and shrinks toward zero. (3) $k_{\mathrm{off}}$, since $p(t)=p(0)e^{-k_{\mathrm{off}}t}$ when $[I]=0$, so the half-time is $\ln(2)/k_{\mathrm{off}}$. (4) Equilibrium data describe occupancy at equilibrium but do not reveal how fast equilibrium is reached; time-course data or independently justified kinetic rates are needed to assess tracking during changing exposure.

Approach	What it represents	When it is useful
Rapid equilibrium	Occupancy is set to \([I](t)/(K_D+[I](t))\) at every time.	Binding relaxation is fast relative to relevant changes and measurement times.
Dynamic reversible binding	Occupancy follows an ODE with \(k_{\mathrm{on}}\) and \(k_{\mathrm{off}}\).	Onset, washout, or concentration changes show appreciable binding delay.
Equilibrium data only	Describes the concentration–occupancy relationship but not the approach to it.	Useful for equilibrium potency summaries, but cannot establish binding speed.