Receptor Binding: Agonists and Antagonists

Why agonists and antagonists matter beyond lock-and-key

Every pharmacology textbook opens with the lock-and-key analogy. A receptor is the lock; a ligand is the key. It's a useful picture, but anyone who has worked with PK models will tell you it flatters the actual biology. Receptors aren't static locks. They're dynamic proteins that flex, shift conformation, and sometimes do things even when nothing is bound to them. The distinction between agonists and antagonists lives right at that complexity, and understanding it properly changes how you read almost any discussion of cannabinoids, peptides, or plant-derived compounds.

So let's be precise.

The short version is that an agonist activates a receptor and an antagonist blocks it. The longer version is that "activates" comes in degrees, "blocks" can happen in more than one way, and some ligands push a receptor below where it sits on its own. Each of those cases has a name, and the names carry real information. If you want a general reference for the vocabulary, the Wikipedia entry on agonists is a reasonable starting point, and the sources at the end of this piece go deeper.

What a receptor actually is

Receptors are proteins, usually membrane-spanning and sometimes intracellular, that detect specific chemical signals and convert them into a cellular response. That conversion is called signal transduction. The response could be a change in ion flux, activation of an enzyme cascade, altered gene transcription, or something else entirely depending on the receptor family.

G protein-coupled receptors (GPCRs) are the largest family and the one most relevant here. The CB1 receptor and CB2 receptor are both GPCRs, as are the GLP-1 receptors targeted by certain GLP-1 receptor agonist peptides. When a ligand binds a GPCR, it stabilises a particular protein conformation that either activates or fails to activate the associated G protein. That's the crux of everything below. There's a broader overview of the family on the Wikipedia page for G protein-coupled receptors.

Think of a receptor less as a fixed lock and more as a protein that samples several shapes. Some shapes signal; some don't. A ligand doesn't so much open the lock as shift the odds towards one shape or another. Once you hold that picture, the categories that follow stop feeling like arbitrary labels and start reading as descriptions of which shapes a ligand favours. For the cannabinoid-specific side of this, the library's piece on phytocannabinoid-receptor binding and the one on cannabinoid receptor signalling cover what happens downstream.

Full and partial agonists: how much of the signal comes through

Full agonists: turning the signal all the way up

A full agonist binds a receptor and produces the maximum possible response that receptor can generate, what pharmacologists call the Emax. The endogenous cannabinoid anandamide acts as a partial agonist at CB1, but tetrahydrocannabinol (THC) is also a partial agonist there, not a full one, which surprises a lot of people the first time they read it. Truly full agonism at CB1 is seen with some synthetic cannabinoids, which is part of why their pharmacological profiles differ so dramatically from plant-derived compounds. The library's comparison of endogenous, phytogenic and synthetic cannabinoids lays out that split in more detail.

Full agonism requires two things: high affinity (the molecule binds readily) and high efficacy (once bound, it reliably triggers the conformational change). A molecule can have excellent affinity and low efficacy. That distinction matters enormously in drug design, because binding alone tells you nothing about what the receptor does next.

Partial agonists: a response, but not the ceiling

Partial agonists bind the same site as full agonists but stabilise the active conformation less completely. They produce a submaximal response even at 100% receptor occupancy. This is not a weakness; it's often a feature. A partial agonist in the presence of a full agonist will actually reduce the net response, because it competes for binding without producing the same signal strength. Partial agonists can therefore behave like functional antagonists in certain tissue contexts.

That last point trips people up. Whether a partial agonist looks like an agonist or an antagonist depends partly on what else is around. With no other ligand present, it produces some signal. Alongside a full agonist, it can pull the overall signal down. Tissues with plenty of spare receptors can also make a partial agonist look nearly full, while tissues with few receptors expose its limits.

THC's partial agonism at the CB1 receptor is a textbook example of this nuance. Cannabigerol (CBG) and THCV have also been examined for partial or antagonist-like activity at CB receptors depending on concentration, which is part of what makes the pharmacology of the endocannabinoid system interesting to model. A common mistake, and one I'd flag for any reader, is to assume THCV is simply a lower-potency agonist. The receptor pharmacology is more context-dependent than that.

Antagonists and inverse agonists: blocking versus pushing below baseline

Competitive and non-competitive antagonists

A competitive antagonist binds the same orthosteric site as the agonist but produces no intrinsic response. It simply blocks access. The key word is competitive: increase the agonist concentration enough and you can outcompete the antagonist. The relationship between agonist and antagonist concentration is predictable and described by the Schild equation, one of the more elegant bits of classical pharmacology, honestly.

Non-competitive antagonists bind at a different site (an allosteric site) and reduce the receptor's ability to respond even if the agonist is still bound. You can't overcome non-competitive antagonism by adding more agonist. Cannabidiol (CBD) is thought to exert some of its effects at CB1 through a negative allosteric modulator mechanism; it doesn't block the orthosteric site but alters receptor conformation in a way that reduces THC's efficacy. This is mechanistically distinct from simple competitive antagonism and probably contributes to the entourage effect discussion, though I'd argue that literature is still genuinely underpowered for the claims often made about it.

A practical way to tell the two apart in a data set is to look at what happens to the agonist's maximum response. Competitive blockade shifts the curve without lowering the ceiling, because enough agonist wins the site back. Non-competitive blockade lowers the ceiling itself. Papers that report only one point on the curve can't distinguish them, which is worth checking before accepting a mechanistic claim.

Inverse agonists: below the baseline

Here's where the lock-and-key metaphor completely breaks down. Some receptors have constitutive (baseline) activity: they signal at a low level even without any ligand present. An inverse agonist binds and actively suppresses that baseline activity, producing a response that goes in the opposite direction to a conventional agonist. It's not the same as an antagonist, which simply returns the receptor to baseline by blocking agonists.

CB1 constitutive activity has been demonstrated in heterologous expression systems, and several compounds, including some cannabinoid analogues, have been characterised as inverse agonists there. This is relevant to understanding why different molecules with overlapping binding sites can have such divergent physiological profiles. The label "antagonist" is also used loosely in the literature, sometimes covering molecules that are really inverse agonists, so it pays to check which one a paper actually demonstrated.

How this plays out in peptide pharmacology

Peptides are particularly interesting here because their structural flexibility allows them to adopt conformations that can selectively stabilise receptor states. A peptide that's 12 amino acids long has a lot of surface contact with a receptor's binding pocket compared to a small molecule. That contact can be tuned to favour partial agonism, biased agonism (activating one downstream pathway preferentially over another), or antagonism. The library's comparison of peptides and small-molecule drugs explains why that size difference shapes so much of the field.

Biased agonism is a relatively recent concept. A biased agonist at a GPCR might activate the G protein pathway without recruiting beta-arrestin, or vice versa. This matters because the two downstream arms can have quite different effects. The idea that you could design a peptide to activate only one arm of a receptor's signalling repertoire is genuinely one of the more exciting threads in modern peptide drug design, though separating the preclinical promise from clinical reality takes hard yards. Quantifying bias is itself difficult, which is why methods papers on the topic are worth reading before the headline claims.

The structural basis for all of this is described in detail in work from the laboratory of researchers like Roger Sunahara, whose group published landmark GPCR cryo-EM structures, but the mechanistic principles apply across the peptide field broadly. Understanding pharmacokinetics alongside receptor pharmacodynamics is where things get practically useful: a molecule with ideal receptor binding characteristics still needs to reach its target tissue at sufficient concentration and for sufficient duration. Bioavailability and first-pass metabolism shape that story significantly, especially for peptides given orally versus via subcutaneous injection.

Affinity, potency and efficacy: keeping them distinct

These three terms get muddled constantly, even in scientific communication. They're not synonyms.

• Affinity describes how tightly a molecule binds a receptor. Expressed as a dissociation constant (Kd or Ki), lower values mean tighter binding.
• Potency describes the concentration required to produce a given effect. EC50 is the concentration producing half-maximal effect. A highly potent drug works at low concentrations; potency says nothing about maximum response.
• Efficacy (or intrinsic efficacy) describes the magnitude of the response a ligand produces once bound. Full agonists have high efficacy; partial agonists have intermediate efficacy; antagonists have zero intrinsic efficacy.

A molecule can be high-affinity and low-efficacy (a competitive antagonist with tight binding, for instance). Or low-affinity and high-efficacy (produces maximal response but requires a lot of it). These combinations determine a compound's practical pharmacological profile, and conflating them leads to misreading research.

The most common slip is treating a tight binder as a strong effect. Tight binding is a statement about receptor affinity, nothing more. A second slip is comparing potency figures across assays as if they were interchangeable; EC50 values depend on the cell system, the readout and the receptor density, so they travel poorly between studies. When an endogenous ligand and a drug are compared, it helps to ask which of the three quantities is actually being reported.

Reading cannabinoid and plant medicine research with this framework

A lot of popular discussion around cannabinoids skips past receptor pharmacology entirely, which leads to claims that don't quite hold up mechanistically. When someone describes a compound as simply "binding CB1," that tells you almost nothing useful on its own. You need to know whether it's agonising, partially agonising, antagonising, or inverse-agonising, and at what concentrations.

CBD, for example, has very low binding affinity at CB1 and CB2 by classical assay. Much of its pharmacological activity appears to involve other receptor systems: TRPV1, GPR55, serotonin 5-HT1A, and the negative allosteric modulation of CB1 mentioned earlier. Reducing CBD to "it binds cannabinoid receptors" misrepresents the pharmacology. The same critical reading applies to minor cannabinoids like THCV or CBG, and to phytomedicine compounds more broadly.

Some questions worth asking of any paper or product page that leans on receptor language: Was activity measured in isolated cells or in living tissue? Which receptor, and which species? Was the effect shown across a range of concentrations, or at one? Does the claim rest on binding data alone? Pharmacological descriptions say what a compound does at a receptor in an experiment. They don't, on their own, establish a health outcome, and in Australia the Therapeutic Goods Administration regulates how therapeutic claims about products can be made. The cannabinoid receptor glossary entry is a handy companion for the terminology.

Good receptor pharmacology is the foundation of making sense of any of it. Get the agonist and antagonist framework right and the rest of the literature becomes much easier to parse.

Sources

• Receptor Pharmacology, StatPearls, NCBI Bookshelf
• Endocannabinoid Receptor Pharmacology; British Journal of Pharmacology, NCBI PMC
• Cannabidiol (CBD) products: regulation in Australia, TGA
• Biased Agonism: From Concept to Quantification, NCBI PMC

Priya Nandakumar, Pharmacology writer; peptides & PK