All 3 are likely correct as in an ITC you measure all heat difference as a consequence of the titration. That includes what ever the buffer doing when it is removed from the protein upon interaction which could very well be different from buffer system to buffer system. Also you have a different ionic strength across your buffers which could also be a cause.

Hope this was helpful.

How large are the differences in the observed enthalpy?

From assay 1 to assay 2 you are changing buffer (and ionic concentration?), and to assay 3 you are changing buffer, pH and ionic concentration. The three factors (buffer, pH and ionic interaction) may influence the observed enthalpy.

If proton association/dissociation is coupled to complex formation, the use of buffers with different ionization enthalpies will result in a different observed enthalpy change. In addition, even small changes in pH will affect the observed enthalpy (that is why asked about the magnitude of the potential discrepancias).

If electrostatic interactions are important for complex formation, ionic concentration differences will also affect.

And finally, although not very common, certain buffers may show some specific interaction with one of the interacting molecules, thus resulting in a distorted enthalpy change due to the competition effect.

Thank you so much for the explanation. I made a lot of mistakes in sample preparation. Could you please take a look on the signal in each trials if you have any clues from it? And from each trial:

1st ITC (deltaH:5.416E4±5510 cal/mol, n=1.75, deltaS: 206 cal/mol/deg); 2nd ITC (del H: -2471±249.8, n=1.16, delta S: 19.1); 3rd ITC (1043±150.8, n=1.18, 25.5)

Have you performed control experiments injecting ligand into buffer?

Titrations show very low affinity.

Strangely, first and second titration have identical estimated enthalpy, while the range of thermal power is rather different and the peaks look endo- and exothermic, respectively. In addition, that enthalpy is huge: 54 kcal/mol.

Fittings of the isotherms (neither the isotherms) are not shown.

Actually, my lab don't have this equipment, I sent the samples to run by a service from an institute and they gave back the results via email. I am really sorry that I did not mention this. I made big mistake that my peptides were dissolved in water instead of same buffer with protein in the 1st and 2nd trial, which could reason why they gives high enthalpy. I was wrong not to use the same buffer in first two trials, so they totally can not be used, can they? The ITC is extremely more sensitive than I had thought. But I do not know whether the different buffer could explain the different endo-/exothermic trend or not.

In each trial of ITC, I did 3 separated ones that used the same protein and 3 different peptides to find which peptide have highest affinity to protein (from previous publication, 1 peptide was determined to interact and 1 was not shown to interact with protein. Therefore, I used them as the positive and negative. The other is my test peptide.)

The third trial I used the same buffer for both peptides and protein but I don't understand why the baseline seems to be noisy and the peaks are quite fluctuated. One of three tests in the 3rd trial even can not be read (as shown in the below figure), does it mean the peptide is not interact with the protein in this case?

Thank you so much for help!

If the solutions for protein and peptide are different, there might be additional heat effects which preclude relying in the binding isotherm or mask the potential interaction heat effect.

In any case, you must perform control experiments. That would help in concluding if what you observe is a real interaction or just dilution/mixing effects.

Thank you so much for the advice! I realize it is very important now when you mention the dilution/mixing effect. Do you have any suggestion about the noisy base line and fluctuated peak on 3rd trial ITC? I used the same dialysis buffer for both protein and peptides that is 0.1 M phosphate buffer (Na2HPO4, NaH2PO4), pH 7.8, NaCl 100mM. I am going to make another trial more carefully, so I really appreciate all of your valuable recommendations.

Noise is difficult to assess, considering it gets lower after injection 7

Shakir, did you just copy-paste one of my answers?