A research group is building a comparison table for a set of peptide analogs. For one analog, a published report gives an inhibition constant from a radioligand competition experiment on cell membranes. A second report, using a fluorescence polarization kit and purified receptor, gives a value roughly ten times weaker. Nobody made an arithmetic mistake. The two numbers came from different experiments that happen to share a name, and a table that places them side by side without saying so is comparing unlike things.
This article sets out what the common binding formats actually measure, where each one is prone to error, and what a reader should look for before trusting an affinity figure for a peptide.
First principle: occupancy is not activity
A binding experiment reports whether, and how tightly, a ligand occupies a receptor. It says nothing about what the receptor does once occupied. A full agonist, a neutral antagonist and an inverse agonist can all show similar affinity for the same site. Telling them apart needs a functional readout, such as a second-messenger or reporter assay. Binding data and functional data answer different questions, and a compound characterized only by affinity has been characterized only in part.
Direct and indirect designs
Saturation binding
In a saturation experiment, a fixed amount of receptor is exposed to increasing concentrations of a labeled ligand until binding plateaus. Fitting the curve gives two values: the equilibrium dissociation constant (Kd) of the labeled ligand and the total number of binding sites (Bmax). The limitation is practical. The compound of interest has to exist in labeled form, and making a labeled version of every analog in a series is rarely realistic.
Competition binding
Competition is the workhorse. A labeled reference ligand is held at one concentration while unlabeled test compound is titrated in to displace it. The midpoint of the displacement curve is an IC50. To turn that into an inhibition constant (Ki), analysts usually apply the Cheng–Prusoff relationship, Ki = IC50 / (1 + [L]/Kd), which needs the concentration and the Kd of the labeled reference ligand.
That conversion rests on assumptions: a single binding site, simple competitive displacement, and a system at equilibrium. Allosteric interaction, multiple site populations or ligand depletion all violate them, and the resulting Ki is then a number without a clean meaning. Reporting an IC50 alone, without the reference ligand details, leaves the reader unable to make that conversion.
There is also a quieter dependency. The concentration axis of a competition curve is only as good as the stock solution behind it. If a peptide stock was prepared from the gross weight of a lyophilized powder that also contains counter-ions and water, every concentration on the curve is overstated by the same factor, and so is the IC50. The distinction is explained in peptide purity versus net peptide content.
Comparing the three common platforms
| Platform | Signal source | Main strength | Main weakness for peptides |
|---|---|---|---|
| Radioligand | Radioactive isotope on the ligand | Label is tiny and rarely alters binding; long-standing reference method | Requires licensed handling of radioactivity and a separation step (usually filtration) to remove free ligand |
| Fluorescence polarization | Change in rotational speed of a fluorescent ligand when bound | Homogeneous, no separation, easy to automate | Needs a large size difference between ligand and binding partner; a bulky fluorophore can change how a short peptide binds |
| Surface plasmon resonance | Mass accumulating on a sensor surface in real time | Gives association and dissociation rates, not only an equilibrium value | One partner must be immobilized, which can block or distort the binding surface |
The fluorescence polarization row deserves emphasis for peptide work. The method depends on a small labeled molecule tumbling more slowly once it is attached to something much larger. When the peptide ligand is itself fairly large, or the binding partner is a small domain, the polarization change is small and the assay window narrows. And a fluorophore can be a substantial fraction of a short peptide’s total size, so the labeled tracer may not bind like the parent sequence.
Specific binding is a subtraction
Labeled ligand sticks to receptor, to other membrane components, to filters and to plastic. Specific binding is therefore defined operationally: total binding minus the binding that remains when an excess of unlabeled competitor occupies all the true sites. Because that residual non-specific binding depends on filter material, blocking protein and wash conditions, it is a property of the protocol, not of the molecule.
When non-specific binding is a large share of the total, the specific signal becomes a small difference between two large numbers, and its uncertainty expands accordingly. Peptides, and cationic peptides in particular, adsorb readily to surfaces, which makes this a common problem. A useful rule when reading a paper is to look for the fraction of total binding that was specific. If it is below about half, the derived constants deserve caution.
Time: the assumption that is rarely checked
Equilibrium constants assume equilibrium. The time needed to reach it is governed mainly by the dissociation rate, and a ligand that leaves the receptor slowly can take hours to equilibrate. If the incubation ends early, the measured affinity is weaker than the true one, and the error is largest for the tightest binders. That compresses the top of a ranking and can reverse the order of close analogs. The check is simple in principle: show that measured binding no longer changes when incubation is extended.
The same kinetic point explains why an equilibrium constant can hide important differences. Kd is the ratio of the off-rate to the on-rate. Two analogs can share a Kd and still differ greatly in residence time, and a single number from a competition assay cannot reveal which rate a structural change affected. Real-time methods such as surface plasmon resonance exist largely to separate those two rates.
A reading checklist for any published affinity
- Format: saturation, competition or kinetic? Is the value a Kd, a Ki or a raw IC50?
- Receptor source: membranes, intact cells or purified protein? Each changes conformation, access and degradation.
- Buffer: pH, ionic strength, divalent cations and, for G protein-coupled receptors, guanine nucleotides all shift apparent affinity.
- Specific fraction: what share of total binding was specific?
- Equilibrium evidence: was incubation time shown to be sufficient, or merely stated?
- Ligand quality: how was the test peptide characterized, and how was its concentration established? A chromatographic purity figure, read as described in how to read an HPLC chromatogram, is one part of that answer.
- Selectivity scope: which receptor subtypes were tested, and which were not?
The last item matters because selectivity claims inherit every limitation above. “Selective” almost always means selective among the subtypes that were tested, measured under one set of conditions. A careful report lists both.
Questions
Can a Ki from a radioligand assay be compared with one from fluorescence polarization?
Only with care. Different receptor preparations, buffers, tracers and incubation times all move the result. Values are most comparable when measured side by side in one laboratory under one protocol.
Why is IC50 not the same as affinity?
IC50 depends on the concentration and affinity of the labeled ligand used in that specific experiment. Converting it to Ki corrects for those, but only when the assumptions of simple competition at one site hold.
What does surface plasmon resonance add?
It reports association and dissociation rates in real time, so it can distinguish compounds that share an equilibrium constant but differ in how long they stay bound.
Does a binding result show that a peptide is active?
No. Binding shows occupancy. Whether that occupancy activates, blocks or reverses receptor signaling needs a separate functional measurement.
Research use only. All products supplied by Battle Born Peptides are laboratory reference materials for in-vitro research and analytical use by qualified professionals. They are not drugs, foods, dietary supplements, cosmetics or medical devices; they are not approved by the FDA or any other regulator for use in humans or animals; and they are not intended to diagnose, treat, cure, mitigate or prevent any disease, or to affect the structure or any function of the body of humans or animals. Nothing in this article is preparation, handling or dosing guidance. See our full research-use terms.