I am trying to optimize my control titration for ITC so that I can ensure my ligand to protein traces are reliable.
My small molecule is stored in DMSO. I have been very careful to avoid buffer mismatch. In short I make a 2x stock solution of my compound in 20% DMSO diluted with protein dialysis buffer, and a 2x stock solution of 20% DMSO (without compound) diluted with protein dialysis buffer. The 2x compound stock is added to an equal volume of protein dialysis buffer (species in syringe) and the 2x stock 20% DMSO in protein dialysis buffer without compound is added to an equal volume of protein dialysis buffer (for control) or 2x protein in dialysis buffer (species in sample cell). This results in a final 10% DMSO and 1x ligand/protein solutions or for my control 1x ligand and buffer (for control experiment). This is also all done with locked pipettes.
Upon titrating compound into buffer I am seeing a significant heat transfer but it is not constant or high with a slight linear decrease, which I know is indicative of buffer mismatch.
Is it possible that this compound into buffer just produces a large heat release intrinsically?
I have attached the raw data below.
This looks like your compound self-associates in solution and, when injected into the cell, it dissociates resulting in decreasing heat effects.
Buffer mismatch, as you said, usually results in more similar injection heat effects with little variation along the experiment.
Adrian Velazquez-Campoy Is it possible to do a ligand into protein titration with a subtraction of the above data to get an accurate Kd?
If the ligand-into-buffer thermogram is significantly different to the ligand-into-protein thermogram, then, as a first approximation, you may subtract the first from the second thermogram. This will provide you the net heat effect "mostly" coming from the interaction. However, this is an approximation since the ligand self-dissociation will be coupled to the interaction with the protein and will result in an apparent interaction affinity and enthalpy that do not correspond to the intrinsic interaction parameters. However, if you are just concerned about demonstrating there is interaction, that is fine, and, at least, you have some approximate estimation for the interaction parameters.
A more appropriate alternative would be determining the self-association parameters from the ligand-into-buffer thermogram, and then develop a model in which ligand dissociation is coupled to the protein binding. Using the self-association parameters as known parameters, you may determine the protein binding parameters.
- Badges
- Science topic
- Similar topics
- Engineering
- Materials Engineering
- Powders