Competitive Antagonism and Parallel Dose-Response Shifts

Why Competitive Antagonism Shifts a Curve

An agonist concentration-response curve describes how response changes as agonist concentration increases. A reversible competitive antagonist can reduce the response produced by a given agonist concentration without permanently removing receptors or changing the agonist’s action once it occupies a receptor. At equilibrium, increasing agonist concentration can displace the antagonist. The characteristic prediction is therefore a rightward shift: more agonist is needed to reach a given response, while the same maximum response remains achievable.

This prediction follows from receptor occupancy only under defined conditions. In particular, the antagonist and agonist must compete reversibly for the same mutually exclusive binding site, and their free concentrations must be known and maintained at equilibrium. The derivation below connects these assumptions to a concentration ratio, then explains how that ratio appears on a functional response curve.

Definition: Let $[A]$ and $[B]$ denote the free agonist and antagonist concentrations at the receptor, and let $K_A$ and $K_B$ denote their equilibrium dissociation constants. For a one-site receptor with mutually exclusive binding, the fraction occupied by agonist is $$ \theta_A=\frac{[A]/K_A}{1+[A]/K_A+[B]/K_B}. $$ The dose ratio here means the ratio of agonist concentrations required to produce a specified matched point on the curves, with antagonist divided by without antagonist. Under reversible competitive equilibrium, the ratio for restoring the same agonist occupancy is $1+[B]/K_B$.

Deriving the Occupancy Dose Ratio

Write $x=[A]/K_A$ for normalized agonist concentration and $b=[B]/K_B$ for normalized antagonist concentration. The agonist occupancy is $\theta_A=x/(1+x+b)$. The fraction of receptors not occupied by agonist is $1-\theta_A=(1+b)/(1+x+b)$, so the occupancy odds are

\frac{\theta_A}{1-\theta_A}=\frac{x}{1+b}.

Without antagonist, $b=0$, and the same occupancy odds equal $x_0=[A]_0/K_A$. To restore the occupancy reached at $[A]_0$ without antagonist, the agonist concentration in the presence of antagonist must satisfy

\frac{[A]_B/K_A}{1+[B]/K_B}=\frac{[A]_0}{K_A}, \qquad [A]_B=[A]_0\left(1+\frac{[B]}{K_B}\right).

Thus the occupancy dose ratio is

DR_{\mathrm{occ}}=\frac{[A]_B}{[A]_0}=1+\frac{[B]}{K_B}.

This is a dimensionless ratio. It is one when no antagonist is present and increases as the antagonist concentration rises relative to its dissociation constant. For example, $[B]=K_B$ gives an occupancy dose ratio of two: twice the agonist concentration restores any specified agonist occupancy.

Worked Example: Restoring a Specified Occupancy

Suppose a hypothetical agonist has $K_A=10$ nM and a reversible competitive antagonist has $K_B=15$ nM. The free antagonist concentration is $[B]=30$ nM. Find the agonist concentrations required for 60% agonist occupancy with and without antagonist.

Without antagonist, set $\theta_A=0.60$ in $\theta_A=[A]_0/(K_A+[A]_0)$. Rearranging gives $[A]_0=K_A\theta_A/(1-\theta_A)$. Substitution yields

[A]_0=10\ \mathrm{nM}\times\frac{0.60}{0.40}=15\ \mathrm{nM}.

The antagonist dose ratio is $1+[B]/K_B=1+30/15=3$. Therefore $[A]_B=3\times15=45$ nM. Checking directly, $[A]_B/K_A=4.5$ and $[B]/K_B=2$, so

\theta_A=\frac{4.5}{1+4.5+2}=0.600.

The calculation agrees with the target occupancy. All concentrations are free concentrations at equilibrium; substituting administered dose or total plasma concentration for receptor-site concentration would require additional justification.

From Occupancy to Response

Let the response at agonist concentration $[A]$ without antagonist be $E_0([A])=g(\theta_A)$, where $g$ is the receptor-to-response relationship and is unchanged by antagonist. With a fixed antagonist concentration, the occupancy at $[A]$ equals the no-antagonist occupancy at $[A]/DR_{\mathrm{occ}}$. Consequently,

E_B([A])=E_0\left(\frac{[A]}{DR_{\mathrm{occ}}}\right).

This equation gives the curve-level meaning of a parallel rightward shift: the antagonist curve is the control curve evaluated at an agonist concentration smaller by the same multiplicative factor. If the response is a strictly increasing function of occupancy over the range of interest, a matched response identifies a unique occupancy. The concentration ratio for that response is then $DR_{\mathrm{occ}}$. This is the usual interpretation for a monotone sigmoidal concentration-response curve.

The qualification matters. If $g$ is not one-to-one, a response value can correspond to more than one occupancy. The two curves still obey the pointwise horizontal-scaling equation above, but matching a response on different branches need not give the occupancy dose ratio and may not give a unique response-level ratio. Thus the familiar matched-response interpretation assumes a monotone response curve, or explicitly pairs the points that correspond to the same occupancy.

At very high agonist concentration, $\theta_A$ approaches one for any finite antagonist concentration because the agonist can outcompete the antagonist. If the same response function applies and full agonist occupancy can produce the maximum response, the maximum is therefore unchanged. This is the sense in which reversible competitive antagonism is surmountable. It does not mean that every finite agonist concentration overcomes the antagonist, nor that a finite experiment necessarily reaches the plateau.

Worked Example: Calculating a Parallel Shift on a Response Curve

Consider a hypothetical monotone response curve described by $E=5+100[A]/(4+[A])$, where concentrations are in nM and response is in arbitrary response units. The agonist EC50 is 4 nM, the lower asymptote is 5 units, and the upper asymptote is 105 units. Suppose a competitive antagonist concentration gives $DR_{\mathrm{occ}}=3$. Find the shifted curve’s EC50 and compare responses at the relevant concentrations.

For a curve of this form, multiplying all agonist concentrations by three shifts the curve right by three while preserving its shape and asymptotes. The shifted EC50 is $3\times4=12$ nM. At 4 nM without antagonist,

E_0(4)=5+100\frac{4}{4+4}=55\ \text{units}.

At 12 nM with antagonist, the shifted curve has the same fractional position:

E_B(12)=5+100\frac{12}{12+12}=55\ \text{units}.

As a second check, at 12 nM without antagonist the response is $5+100(12/16)=80$ units; at 36 nM with antagonist it is $5+100(36/48)=80$ units. Both matched pairs have concentration ratio three. In each curve the response tends to $5+100=105$ units as agonist concentration becomes large, so the maximum is not reduced by this model of competition.

Reading Parallel Shifts in Data

In practice, a parallel shift is assessed by comparing concentration-response curves collected under comparable conditions, with and without antagonist. A useful summary is the agonist concentration required to reach a defined fraction of the response range. For monotone curves, one can compare such concentrations at more than one response level. Similar ratios across the range, together with similar lower and upper asymptotes and curve shape, are consistent with a parallel shift.

Worked Example: Comparing Ratios at Two Response Levels

In a hypothetical assay, the control curve has lower asymptote 0 units, upper asymptote 100 units, and EC50 of 5 nM, with Hill slope one. Under a fixed antagonist concentration, the curve has the same asymptotes and shape but EC50 of 15 nM. Check the concentration ratios at 20% and 80% of the response range.

For a Hill-slope-one curve, the fractional response $f$ is $f=[A]/(EC50+[A])$. Solving gives $[A]=EC50f/(1-f)$. At 20%, the control concentration is $5(0.20/0.80)=1.25$ nM, and the antagonist concentration is $15(0.20/0.80)=3.75$ nM. The ratio is $3.75/1.25=3$.

At 80%, the corresponding concentrations are $5(0.80/0.20)=20$ nM and $15(0.80/0.20)=60$ nM. Their ratio is $60/20=3$. The equal ratios at these two response levels and the preserved asymptotes are consistent with a threefold parallel rightward shift. This pattern is evidence compatible with competitive antagonism under the stated assumptions; it does not by itself prove that the receptor-level assumptions hold.

Assumptions, Diagnostics, and Alternatives

The occupancy derivation relies on a one-site equilibrium model in which agonist and antagonist bind mutually exclusively and reversibly. It also assumes that free concentrations at the receptor are maintained, the antagonist has no agonist effect of its own in the response measure, and antagonist exposure does not alter receptor number or the downstream relationship between occupancy and response. To interpret a functional dose ratio as the occupancy ratio, the response curve should be monotone over the matched range, or the comparison should explicitly use the corresponding same-occupancy points.

Several features of the data can challenge this interpretation. A visibly lower upper plateau, a changed lower baseline, or a different slope may indicate that a simple horizontal shift does not describe the observations. A plateau that was not reached is not evidence of a reduced maximum: the tested agonist range may simply be too low. Conversely, similar apparent maxima over a narrow range do not establish that the true maxima are equal. Check whether the concentration range covers both plateaus and whether the curves were measured under comparable conditions. Consider uncertainty in fitted parameters rather than treating a visual resemblance as conclusive.

A reduced maximum or a nonparallel change is not the simple prediction derived here. Such a pattern can arise from other pharmacological mechanisms, assay limitations, changing receptor or response properties, or model misspecification. The main alternatives are not interchangeable: competitive antagonism predicts a surmountable, multiplicative shift under its assumptions, whereas mechanisms that limit the attainable response require a different functional interpretation. In particular, irreversible antagonism and insurmountable antagonism are distinct topics; they should not be inferred merely because an experiment did not reach a high enough agonist concentration.

Common Mistake: Matching a Response Without Checking the Curve

A common but overbroad claim is: “The response function is unchanged, so any matched response has the occupancy dose ratio.” An unchanged response function alone is insufficient if it is non-injective. For example, suppose $g(\theta_A)=\theta_A(1-\theta_A)$. Occupancies 0.2 and 0.8 both produce response 0.16. With $[B]/K_B=1$, an agonist concentration of $0.25K_A$ without antagonist gives occupancy 0.2. An agonist concentration of $8K_A$ with antagonist gives occupancy 0.8, because $\theta_A=8/(1+8+1)=0.8$. These responses match, but the concentration ratio is $8/0.25=32$, not two. The correction is to use matched occupancy, or to require a one-to-one response function over the relevant range when interpreting matched-response ratios. For the usual strictly increasing concentration-response curve, that condition makes the matched-response interpretation valid.

Key takeaway: Under reversible, mutually exclusive equilibrium competition, restoring a specified agonist occupancy requires multiplying agonist concentration by $1+[B]/K_B$. With an unchanged, monotone response relationship, this appears as a parallel rightward shift with an unchanged attainable maximum. Curve shape and experimental range must be checked before applying that interpretation.

Check Your Understanding

Use the assumptions stated in this tutorial. Solutions follow the questions.

If $[B]=2K_B$, what is the dose ratio for restoring the same agonist occupancy?
With $K_A=8$ nM, what agonist concentration gives 50% occupancy without antagonist? If $[B]=K_B$, what concentration restores that occupancy?
Why does an unchanged response function not always guarantee that every matched response has the occupancy dose ratio?
A curve measured with antagonist has a lower observed plateau, but the agonist concentrations tested do not approach saturation. Does this establish a reduced maximum response?

Solutions. (1) $DR_{\mathrm{occ}}=1+2=3$. (2) Without antagonist, 50% occupancy requires $[A]_0=K_A=8$ nM; with $[B]/K_B=1$, the ratio is two, so $[A]_B=16$ nM. (3) A non-injective response function can map different occupancies to the same response, so a response match may pair different branches; matching the same occupancy, or using a strictly increasing response function, avoids this ambiguity. (4) No. The tested range may not reach the plateau, so the true maximum is not identified by those observations alone.