Are EC50, Kd, and IC50 values for MHC binding assays comparable?

Question

[alex.rubinsteyn]

I'm working on predicting the binding of peptides to MHC molecules using assay data from IEDB. Experimental binding results come with measurements in a variety of units, the most common being IC50 (concentration of the query peptide which inhibits 50% of a reference peptide binding), as well nM concentrations for Kd (dissociation constant) and EC50. Though the underlying assays are different, it seems like these concentration values sometimes get used interchangeably. Is this kosher? Will a peptide which has an IC50 concentration of 500nM also achieve a similar EC50?

Answer 1

[Devon Ryan]

Just because a compound/peptide binds a protein/complex doesn't mean it'll activate it...that's how inhibitors work. So the answer is no. A peptide with an IC50 of 500nM may not even have an EC50. It may also help to consider partial agonists/antagonists when thinking about this.

Edit: I'll add that Kd describes binding. IC50 and EC50 describe activity. They're related concepts, but you can't trivially convert between them.

[linus]

Can you define the two more precisely?

[Devon Ryan]

Binding and activity? You can put a little kid in the driver's seat of a car (binding) but you probably wouldn't expect the kid to drive (little to no agonistic activity/high EC50) and you would then yourself be blocked from driving the car (high antagonistic activity/low IC50). These are general pharmacology terms that you should probably learn if you want to understand literature about protein binding (an intro pharmacology class can be quite useful).

[alex.rubinsteyn]

My impression is that neither the IC50 nor EC50 values given for MHC binding assays directly describe the activity of the MHC molecule itself. In the case of IC50, you're inhibiting the binding of some other reference peptide. I can't find a clear description of what activity gets measured for EC50 values but I think it's related to triggering cytokine release.

[Devon Ryan]

Using a downstream readout like cytokine release is pretty typical since it's easier to assay than actual binding. Measuring direct activity is only common for things like enzymes and ion channels, since they're amenable to more direct measurement.