A peptide binds a receptor with a dissociation constant of 3 nM. That tells you how tightly it holds on. It tells you nothing about whether it switches the receptor on, switches it off, quietly blocks the natural ligand, or makes the natural ligand work better. Those are separate properties measured in separate experiments, and conflating them is the most common source of confused results in peptide pharmacology. This page sorts out the vocabulary, the experiments that distinguish the cases, and why peptides are unusually well placed to act as modulators rather than as simple on-off switches.
Affinity is how well a ligand occupies the receptor. Efficacy is what the receptor does once occupied. A binding assay measures the first; only a functional assay measures the second. The practical consequence is that a binding assay cannot tell an agonist from an antagonist, because both bind. If a project has measured a KD and nothing else, it does not yet know what its compound does.
An orthosteric ligand binds where the natural ligand binds. An allosteric ligand binds somewhere else and changes what happens at the orthosteric site1. That geometric difference produces a pharmacological one that is worth being precise about.
| Class | Effect on the natural ligand | Activity alone |
|---|---|---|
| PAM | Increases its affinity and/or its efficacy1,2 | None, by definition |
| NAM | Decreases its affinity and/or its efficacy1,2 | None |
| Neutral (silent) modulator | None — but blocks other modulators from the site2 | None |
| Ago-PAM | Potentiates it, and also activates the receptor by itself3 | Yes |
| Biased modulator | Potentiates one pathway while leaving or suppressing another2,3 | Varies |
The reason this matters clinically is that a pure modulator does nothing where the natural ligand is absent. It amplifies or damps an existing signal rather than imposing one, so its effect is confined to tissues and moments where the physiology is already active. That is a fundamentally different safety profile from a receptor-wide agonist.
Probe dependence is the trap. A modulator's effect depends on which orthosteric ligand it is tested against and which signalling pathway is read out — the property is called probe dependence2. One neurotensin receptor modulator has been reported as a negative modulator on one G protein pathway, neutral on another, and an activating positive modulator for β-arrestin recruitment3, and in the same receptor a positive effect on arrestin recruitment has been described alongside a negative one on G protein signalling2. "Is it an agonist?" is not a well-formed question until the pathway and the competing ligand are named.
Biased agonism is the capacity of different agonists at the same receptor to activate different subsets of its downstream pathways2. Two ligands can bind the same site with the same affinity and produce different physiology, because they stabilise different receptor conformations. For peptide work this is consequential rather than academic: the therapeutic effect and the adverse effect of a receptor often run down different pathways, and a biased ligand is an attempt to separate them. It also means a single-readout assay can rank a series in an order that reverses when a second readout is added.
The argument for peptides as modulators rather than as mere agonists is a strong one. Many peptides and proteins were selected by evolution precisely to modulate receptors, so nature has already done a large screen; they are, as one review puts it, leads hiding in plain sight. Peptides also tend to deliver high affinity and tolerability from far fewer compounds than a small-molecule campaign requires, and where a peptide derives from a known protein interaction, structural information and a putative binding site usually come with it — which makes optimisation and site verification far easier than starting from a screening hit of unknown binding mode4.
The examples are not hypothetical. Endogenous peptide modulators include PLG-NH2, a positive modulator of the dopamine D2 receptor, and 5-HT moduline, a negative modulator of the 5-HT1B receptor. Pepcan-12, derived from a haemoglobin fragment, is a positive modulator of the cannabinoid CB2 receptor, and LVV-hemorphin-7 a positive modulator of the angiotensin II receptor. On the synthetic side, the palmitoylated pepducin PZ-128 is a negative modulator of PAR1 that reached clinical trials in coronary artery disease, and the α-conotoxin Vc1.1 is a GABA-B positive modulator that entered trials for neuropathic pain, with a head-to-tail cyclised version reported as orally available in preclinical models4.
Two of those illustrate points made elsewhere on this site: the pepducin is palmitoylated, which is what anchors it in the membrane (see lipidated peptides), and the conotoxin was cyclised head-to-tail to survive oral administration (see cyclic and constrained peptides).
For designing the peptide series itself, residue-by-residue substitution is the standard way to locate the pharmacophore — the same scanning logic described in epitope mapping and peptide scanning, and the controls in control peptides in experimental design apply directly.
| Affinity | How tightly it binds — a binding assay |
| Efficacy | What the receptor then does — a functional assay |
| Orthosteric | Binds the natural ligand's site |
| Allosteric | Binds elsewhere; changes affinity and/or efficacy of the natural ligand |
| Probe dependence | The answer depends on which ligand and which pathway you test |
Characterising a peptide ligand properly means making more than one peptide: alanine and D-amino acid scans to find the pharmacophore, truncations to find the minimum active fragment, a scrambled control, and often a cyclised or N-methylated variant to test whether constraint improves or destroys activity. LifeTein makes these as sets. Send the parent sequence and what you need varied.