Ki, IC50 and EC50: What Binding and Potency Figures Actually Measure
Nearly every claim made about a research peptide’s selectivity rests on one of three numbers. They sit side by side in papers, in review tables and in supplier copy, and they get treated as interchangeable measures of how strong something is. They are not. They measure different quantities under different conditions, and only one of them is a property of the molecule and its target alone.
This is the reading-the-literature companion to our note on peptide nomenclature: what the numbers beside the names actually mean.
Kd and Ki: affinity
The dissociation constant, Kd, describes how tightly a molecule and its binding site hold on to each other at equilibrium. It carries units of concentration, and the value is the concentration at which half the binding sites are occupied. Lower means tighter: a Kd of 1 nM is a hundredfold tighter than 100 nM.
Most published peptide affinity data comes from competition binding rather than direct measurement. A radiolabeled or fluorescent reference ligand is bound to the receptor, the test compound is titrated in, and the concentration displacing half the reference is recorded. That raw value is an IC50, and it depends on how much reference ligand was used and on that ligand’s own affinity.
Ki is the same IC50 corrected for both of those, via the Cheng–Prusoff relationship. The correction is what makes Ki portable: two Ki values for the same receptor can be compared meaningfully, while two competition IC50 values from different assay setups cannot.
The rule when reading a table: a column headed Ki can be compared across sources. A column headed IC50 in a binding context can only be compared within one assay.
EC50: functional potency
EC50 is a different kind of measurement entirely — the concentration producing half the maximal response in a functional assay, whether that assay reads second-messenger accumulation, reporter output, calcium flux or receptor internalization.
Because it reports response rather than occupancy, EC50 absorbs everything between binding and readout: receptor expression level in that cell line, coupling efficiency, amplification through the cascade, incubation time. A receptor expressed at high density in a recombinant line can give a full response from a small fraction of occupied sites, pushing EC50 well below Kd. The same compound in a low-expression system can show an EC50 above its Kd.
So EC50 is a property of the compound in that assay, not of the compound. An EC50 quoted without the cell system and the readout is a number with no referent.
Why the distinction changes conclusions
Selectivity claims are where this bites hardest. “Selective for one receptor subtype” is nearly always a ratio of two numbers, and the ratio only means anything if both numbers came from the same kind of measurement.
A compound can bind two subtypes with similar affinity and activate them with very different potency, because the subtypes couple to different pathways with different amplification. Run the other way, a compound can look selective on affinity and behave non-selectively in a functional assay. Both patterns appear in the melanocortin literature, which is why subtype selectivity in that family has to be read rather than lifted from a headline figure.
The same caution applies to incretin receptor agonists, where relative activity at two or three receptors is the entire point of the design and the reported ratios move depending on which assay produced them. See incretin receptor pharmacology.
Efficacy is a separate axis
Potency is where the concentration–response curve sits along the concentration axis. Efficacy is how high it climbs. A compound can be highly potent and only partially efficacious — occupying the receptor at low concentration yet never producing a full response however much is added.
Two compounds with identical EC50 values can therefore behave quite differently, and a table listing only EC50 cannot tell them apart. The maximal response, usually given as a percentage of a reference agonist’s, is the missing column. A partial agonist and a full agonist are different tools even when their potencies match.
Biased agonism, briefly
A receptor can signal through more than one downstream pathway, and a ligand can favor one over another. Where that happens a single EC50 is not just incomplete, it is pathway-specific. Two papers reporting different EC50 values for one peptide at one receptor may both be right and simply be reading different outputs.
It is one reason apparently contradictory literature often is not, and a good reason to record which readout a number came from at the moment you write it down.
Reading a number properly
When a figure turns up without context, these are the questions that make it usable:
- Affinity (Kd, Ki) or function (EC50)? A binding IC50 and a functional EC50 are not the same measurement.
- What system — recombinant line, native tissue, which species’ receptor? Sequence differences between species move affinity.
- What readout, over what incubation?
- Is the maximal response reported next to the potency?
- Is the comparison being made between numbers of the same type, measured the same way?
Where a supplier or a review states a selectivity ratio without answering those, the ratio is a claim rather than a measurement. Where the underlying paper answers them, the number is usable — and usually more interesting than the summary built on top of it.
Two further notes bear on how these figures behave for peptides specifically: half-life and analog modification, since how long a compound persists in an assay affects what a fixed-incubation EC50 reports, and cyclization, which changes affinity by constraining conformation.
ExoLabz supplies compounds for laboratory research use only. Nothing on this page is medical advice or a suggestion of human or veterinary use.
