Estimating MABEL from Receptor Occupancy

Why Estimate MABEL From Receptor Occupancy?

For a drug whose intended action begins with binding a receptor, the fraction of receptors occupied can provide a mechanistic link between concentration and pharmacology. A low-occupancy concentration can therefore inform a minimum anticipated biological effect level (MABEL) and, with a human pharmacokinetic prediction, a candidate starting dose. This is an estimate of a concentration associated with a chosen occupancy threshold—not a guarantee that biological effects are absent below it or present above it.

The calculation relies on the receptor’s binding affinity and on the free drug concentration at the receptor. Receptor abundance has a specific role: when free drug concentration is known and drug is in large excess over receptor, abundance does not change the fraction occupied. But when binding consumes a meaningful share of available drug, abundance affects how much total drug is needed to achieve that free concentration. Keeping these two cases distinct is essential.

Definition: Receptor occupancy, \(f_{\mathrm{occ}}\), is the fraction of the relevant receptor population bound by drug. For a single binding site, \(f_{\mathrm{occ}}=[DR]/R_T\), where \([DR]\) is the concentration of drug-bound receptor and \(R_T\) is the total concentration of accessible receptor. Occupancy as a percentage is \(100f_{\mathrm{occ}}\%\).

Equilibrium Occupancy and the Concentration Target

For a single drug binding reversibly to one receptor site, assume binding has reached equilibrium, receptors are equivalent and accessible, and there is no competing ligand or cooperative binding. Let \(C_u\) be the free (unbound) drug concentration in the receptor compartment and \(K_D\) the equilibrium dissociation constant, expressed in the same concentration units. Under these assumptions, the mass-action definition of affinity is \(K_D=C_u[R]/[DR]\), where \([R]\) is unbound receptor concentration. Since \(R_T=[R]+[DR]\), the occupancy fraction is:

f_{\mathrm{occ}}=\frac{[DR]}{R_T} =\frac{C_u}{C_u+K_D}.

Solving for the free concentration that gives a specified occupancy \(f\) yields:

C_u=K_D\frac{f}{1-f}.

Thus, for a low target occupancy such as 10%, the required free concentration is \(K_D/9\), not \(0.10K_D\). For small \(f\), \(f/(1-f)\) is approximately \(f\), so \(C_u\) is approximately \(fK_D\); this approximation is only appropriate when the resulting difference is acceptable for the intended decision.

This equation specifies free concentration. If the predicted human pharmacokinetic profile provides unbound concentration at the effect site, it can be compared directly with \(C_u\). Total concentration in plasma or a dosing solution is not interchangeable with free concentration at a receptor. Any conversion between them requires an appropriate binding or distribution model.

Worked Example: Free Concentration for 10% Occupancy

Suppose a hypothetical drug has \(K_D=18\ \mathrm{nM}\), and a 10% occupancy threshold is selected for the initial mechanistic estimate. Assume one-site equilibrium binding and that the free concentration at the receptor is known.

Using \(f=0.10\):

C_u=K_D\frac{f}{1-f} =18\ \mathrm{nM}\times\frac{0.10}{0.90} =2.00\ \mathrm{nM}.

The predicted free concentration associated with 10% occupancy is \(2.00\ \mathrm{nM}\). Checking in the forward equation gives \(2/(2+18)=0.100\). This calculation does not establish that 10% occupancy is a biologically safe threshold; that choice must be justified using the mechanism and available pharmacology.

When Receptor Abundance Changes the Required Concentration

The occupancy equation uses free drug concentration. If free concentration is specified independently, \(R_T\) cancels from the fraction occupied: doubling receptor abundance doubles the number of bound receptors, but does not change the fraction occupied at the same \(C_u\). Receptor abundance matters when the input or prediction is total drug concentration and binding is large enough to deplete free drug.

For a single compartment containing only free and receptor-bound drug, total drug concentration is \(C_T=C_u+[DR]\). Because \([DR]=R_T f_{\mathrm{occ}}\), a target occupancy \(f\) requires:

C_T=C_u+R_Tf =K_D\frac{f}{1-f}+R_Tf.

This mass balance assumes drug and receptor concentrations refer to the same compartment and that other drug-binding pools are negligible. It is not a general conversion from plasma total concentration to tissue concentration. If \(R_Tf\) is small relative to \(C_u\), depletion is negligible and \(C_T\) is close to \(C_u\). If it is not small, using the free-concentration target as though it were the required total concentration can substantially understate the amount needed in that compartment.

Worked Example: Receptor Abundance and Drug Depletion

In a hypothetical receptor compartment, let \(K_D=12\ \mathrm{nM}\), \(R_T=24\ \mathrm{nM}\), and the desired occupancy be 5%. Assume the one-site equilibrium and single-compartment mass-balance assumptions above hold.

First calculate the required free concentration:

C_u=12\ \mathrm{nM}\times\frac{0.05}{0.95} =0.6316\ \mathrm{nM}.

At 5% occupancy, the bound drug concentration is \(R_Tf=24\ \mathrm{nM}\times0.05=1.20\ \mathrm{nM}\). Therefore:

C_T=C_u+R_Tf =0.6316\ \mathrm{nM}+1.20\ \mathrm{nM} =1.832\ \mathrm{nM}.

Treating total concentration as equal to free concentration would suggest \(0.6316\ \mathrm{nM}\), only about 34.5% of the total concentration required by this mass balance. Here receptor-bound drug is not negligible relative to free drug, so receptor abundance materially changes the total-concentration estimate. The calculation applies only to this hypothetical compartment; it does not itself predict plasma exposure or disposition.

From a Concentration Target to a Candidate Dose

A concentration target becomes a dose estimate only after it is linked to a predicted human concentration-time profile. The PK prediction must correspond to the intended route and regimen and to the concentration at the relevant site of action. For a first-dose calculation based on peak occupancy, a convenient input is predicted unbound \(C_{\max}\); if the pharmacological concern depends on exposure over time, the profile should instead be assessed over the relevant interval using the occupancy equation.

If the PK model predicts dose-proportional unbound \(C_{\max}\) over the dose range, the dose giving a target peak free concentration can be found by scaling from a reference dose. If concentration is nonlinear in dose, the target must be found by inverting the model prediction rather than by simple proportional scaling. Predicted PK and uncertainty in it should be treated as part of the estimate, not as exact facts.

Worked Example: Translating an Occupancy Target Into a Dose

For a hypothetical drug, let \(K_D=6.0\ \mathrm{nM}\) and choose a 10% peak-occupancy target. Suppose the predicted human unbound \(C_{\max}\) after a 1 mg dose is \(0.050\ \mathrm{nM}\) per mg, and concentration is predicted to be dose proportional in this range. Assume this is the concentration at the relevant site, or that the PK prediction appropriately maps plasma unbound concentration to that site.

The free concentration target is:

C_{u,\mathrm{target}} =6.0\ \mathrm{nM}\times\frac{0.10}{0.90} =0.6667\ \mathrm{nM}.

The dose-proportional PK prediction is \(C_{\max,u}(D)=0.050\ \mathrm{nM/mg}\times D\). Setting this equal to the target gives:

D=\frac{0.6667\ \mathrm{nM}}{0.050\ \mathrm{nM/mg}} =13.33\ \mathrm{mg}.

Thus, \(13.33\ \mathrm{mg}\) is the hypothetical dose predicted to give the selected 10% peak-occupancy target under the stated assumptions. The units cancel to mg. This estimate is not, by itself, a recommended clinical starting dose: uncertainty in affinity, human PK, effect-site exposure, and the biological relevance of 10% occupancy still requires evaluation.

Assumptions, Diagnostics, and Alternatives

Before using the estimate, check that the affinity measurement is relevant to the human receptor and the experimental conditions of interest; that \(K_D\), receptor abundance, and drug concentrations use compatible units; and that the predicted concentration is unbound and representative of the binding compartment. The equilibrium calculation is inappropriate if equilibrium is not a reasonable approximation over the relevant exposure period. Endogenous ligands, multiple receptor states, non-specific binding, or heterogeneous receptor populations can also make the simple equation inadequate.

Useful diagnostics include recalculating occupancy at the lower and upper plausible \(K_D\) values, checking whether \(R_Tf\) is small compared with \(C_u\), and verifying the concentration and dose units at every conversion. A predicted concentration far above the affinity scale for a very low occupancy target, or a total concentration target that changes substantially when receptor abundance is varied, should prompt a review of the algebra, units, and model assumptions. Where measurements are available, compare predicted occupancy with direct occupancy data or with relevant downstream pharmacology; agreement should not be assumed simply because both are linked to drug concentration.

Receptor occupancy provides a mechanistic concentration target, whereas the earlier material on in vitro potency and cellular assays uses functional assay evidence to characterize biological activity. These approaches can complement one another, but an occupancy fraction is not itself a response magnitude: receptor reserve, signaling amplification, and biological context may separate occupancy from effect. Receptor-abundance correction in the mass-balance equation is a static binding calculation; it does not replace a PK model when binding affects concentration over time.

A common mistake is to say, “10% occupancy requires a concentration equal to 10% of \(K_D\).” The correct relation is \(C_u=K_Df/(1-f)\), so 10% occupancy requires \(K_D/9\). A second mistake is to multiply the free-concentration target by receptor abundance or to ignore abundance without checking depletion. Abundance does not alter fractional occupancy at a fixed free concentration; it enters the total-concentration mass balance as \(R_Tf\). Finally, a low predicted occupancy does not prove the absence of biological effects. The selected threshold is a modeling choice that should be justified by the mechanism and interpreted alongside other evidence.

Key takeaway: Under one-site equilibrium binding, target occupancy \(f\) corresponds to \(C_u=K_Df/(1-f)\). Receptor abundance contributes to total concentration only when bound drug is included in the same-compartment mass balance. A dose estimate then depends on predicted human exposure at the relevant site and on the validity of the PK and binding assumptions.

Check Your Understanding

Try each question, then compare with the solutions.

For \(K_D=9\ \mathrm{nM}\), what free concentration gives 20% occupancy?
At fixed free drug concentration, does doubling receptor abundance change fractional occupancy under the one-site model? Explain briefly.
For \(K_D=4\ \mathrm{nM}\), \(R_T=10\ \mathrm{nM}\), and target occupancy \(f=0.10\), calculate the free and total concentrations in a single compartment.
A PK model predicts unbound \(C_{\max}=0.030\ \mathrm{nM/mg}\) after dose \(D\). If the required free concentration is \(0.60\ \mathrm{nM}\) and dose proportionality holds, calculate \(D\).

Solutions. (1) \(C_u=9\times0.20/0.80=2.25\ \mathrm{nM}\). (2) No. At fixed \(C_u\), \(f_{\mathrm{occ}}=C_u/(C_u+K_D)\) contains no \(R_T\); the absolute bound-receptor concentration changes, but the fraction does not. (3) \(C_u=4\times0.10/0.90=0.4444\ \mathrm{nM}\). Bound drug is \(R_Tf=10\times0.10=1.00\ \mathrm{nM}\), so \(C_T=0.4444+1.00=1.444\ \mathrm{nM}\). (4) \(D=0.60\ \mathrm{nM}/(0.030\ \mathrm{nM/mg})=20.0\ \mathrm{mg}\).