Agonist, antagonist—or something in between?

Binding is only the first question

A ligand is a molecule that can bind to a receptor. Affinity describes the tendency of the two to form a complex under specified conditions. An equilibrium dissociation constant summarises binding once binding and unbinding have reached a balance. Association and dissociation rates answer a different question: how quickly binding and unbinding occur.

Efficacy describes the response a ligand can produce after binding, relative to a reference response in a specified system.

Potency describes the concentration or dose needed to produce a defined effect; EC50 is a concentration producing half of a specified maximal effect, while ED50 is a dose-based measure. Potency is influenced by affinity and efficacy. Other influences include the number of receptors, signal amplification, exposure time and how the test (assay) is performed. A lower concentration needed for an effect does not make a medicine better, and efficacy in a cell assay is not the same as therapeutic efficacy in people.21

Full agonists and partial agonists

A full agonist can produce the system’s reference maximum response under the conditions of the experiment. A partial agonist can activate the same receptor but reaches a lower maximum even when the receptor is occupied at high concentrations. The word partial does not mean half as effective in every person or at every dose. In the presence of a full agonist, a partial agonist can compete for receptor occupancy and lower the combined response, so it can look agonist-like in one context and antagonist-like in another.12

Neutral, competitive and non-competitive antagonism

A neutral antagonist has affinity for a receptor but, in the system being measured, does not measurably increase or decrease baseline activity. A reversible competitive antagonist can reduce agonist occupancy; in a concentration–response experiment this commonly shifts the agonist curve to the right, while the original maximum may remain reachable if enough agonist is present. A non-competitive antagonist can reduce signalling through another site or mechanism. This can lower the maximum response observed, so increasing the agonist may not fully restore the response. A reduced maximum alone does not identify the molecular mechanism: receptor reserve, slow dissociation or downstream interference can also affect the observed curve.1328

Inverse agonism and response context

Some receptors show constitutive activity: measurable signalling even without an external ligand. An inverse agonist preferentially stabilises a less-active receptor state and lowers that baseline activity; a neutral antagonist occupies the receptor without changing baseline activity in the assay. Receptor density, coupling proteins, downstream amplification, timing and the signalling pathway measured can change the result. A ligand may therefore be an agonist in one assay and a weak agonist or antagonist in another. Selectivity is relative, not the same as specificity.1345

One receptor can signal along different routes

A receptor can connect to more than one signalling pathway. In a particular cell and experiment, one agonist may favour one measured pathway relative to another agonist used as a reference; researchers call this ligand bias or functional selectivity. The comparison depends on the reference, the pathways, the timing and the sensitivity of the assays. Differences in receptor abundance or cellular machinery can also change the pattern. A laboratory claim of pathway bias does not by itself show that a medicine has fewer harms or better outcomes in people: those claims require clinical evidence.95

A simple concentration–response experiment

Imagine a tissue in which a reference agonist produces a maximum response of 100 units. Compound A reaches 100 units at 10 nanomolar; compound B reaches 60 units even at a much higher concentration; compound C produces no response by itself but makes the reference agonist require a higher concentration. In that experiment, A behaves as a full agonist, B as a partial agonist and C as an antagonist whose pattern is compatible with competitive antagonism. A rightward shift alone does not prove that C competes at the same receptor site; binding and additional experiments are needed to test that mechanism. The example shows how the labels arise from a comparison. It does not establish human dosing, durability, adverse effects, tissue distribution or benefit on a patient-important outcome.248

What a receptor assay cannot answer alone

A binding experiment can show that a molecule interacts with a target, but it cannot by itself show that the target interaction changes a symptom or prevents an illness. A cellular signal can be downstream of several receptors, and a biomarker can move without improving how a person feels or functions. Even a well-controlled human trial may answer only its prespecified population, dose, comparator and follow-up period. Keep the levels separate: binding, cellular response, tissue or circuit response, and patient-important outcome. Move from one level to the next only when evidence connects them.467

Why concentration and dose are different

Concentration is the amount of a substance in a measured volume; dose is the amount administered. A dose is translated into concentration by absorption, distribution, metabolism, clearance and time. Two studies can use the same dose but produce different exposure, while a high concentration in a dish may never occur in human tissue. Route, formulation, sampling time and the population studied therefore belong beside every potency claim. The concentration–response curve describes the tested system, not a personal dosing rule.26

A mechanism claim is a starting point

When a headline says that a medicine targets a receptor, ask which receptor and subtype were studied, whether the result was binding, a cellular signal, a tissue response or an outcome in people, what reference ligand and endpoint were used, and whether benefits and harms were measured over a meaningful period. A target explains a research question; it does not establish that a treatment works, is approved or is safe for a particular person.246