Hello everyone!
I have looked through some past answers on this subject but am still hesitant to take this approach, as the answer seems too obvious or beyond my capacity of understanding (funny how science is like that sometimes).
The ligand I am working with is not very soluble and forms aggregates, clogging the syringe and changing the concentration. To get around this issue I thought of reversing the solutions of ligand and receptor in the ITC (called, unsurprisingly, reverse titration). I would be lowering the concentration of the ligand to 20uM in the cell, which should make it soluble, and increasing the concentration of my protein to 200uM in the syringe. As far as I know, neither the protein nor the ligand form dimers or some sort of complex within themselves, and the protein only has one binding site.
My concern lies mainly in being such a novice to ITC, so I feel like this approach is greatly overlooking some obvious factor that I am still too naive to see or understand. One past answer I read stated this approach would be equivalent to a direct (or forward 'ligand-to-receptor) titration as long as there is only one binding site, which there is. However, I am worried there are minute details or more complex issues in the changes of enthalpy I do not understand. Would this approach give me the same Kd if I did a direct titration or would I have to take some series of equations to invert the data to give me the true Kd (as if I did a direct titration)?
Thank you for your time and support - I really appreciate it.
Tom
The molecular mass of reactants is not a factor affecting the equilibrium binding equations. in addition, those equations are symmetrical regarding protein and ligand (it is the experimental design what breaks the asymmetry). Then, there is no problem for exchanging protein and ligand.
However, you have mentioned some issues to be aware of:
- protein with more than one ligand binding site
- oligomerization equilibrium aggregation propensity for any of the reactants
- high enthalpy of dilution for any of the reactants
which may limit the possibility of the reversing the set-up
Dear Sean,
If the protein and ligand are fully active, and if the concentrations are well estimated, the forward and reverse titrations are assumed to be equivalent and are modeled by the same set of equations.
Note that the Ka measured is a macroscopic parameter, which differ from the microscopic ka if there are several binding sites. In the "Origin" software provided by microcal, the apparent Ka differs from direct titration to reverse titration, due to stoichiometry. (e.g. in Origin, for a protein: ligand stoechiomery 1: 2, direct tritration gave a Ka = 2.ka and n = 2 macroscopic, and for the reverse titration the association constant provided was directly ka and n = 0.5).
For the interpretation of the data, I propose to use the NITPIC / SEDPHAT suite, which offers several binding models, adapted to your specific case of protein-ligand interaction, and a clear interpretation of thermodynamics.
However, be aware that your protein is probably not 100% active, and from batch to batch you will observe variations in the so-called “competent fraction”. This is an important point because the measured enthalpy is directly proportional to the concentration of the syringe and not to the cell concentration.
In SEDPHAT, "incompetent fraction can be corrected from batch to batch if you consistently record the same titration as a reference.
In other words, for the comparison of ligands, you should systematically titrate the same compound used as the "reference assay" to renormalize the future assays, in order to avoid some potential mismatches of the protein batches.
best whishes
jean-francois Gaucher
- Badges
- Science topic