Nothing it's unexplainable, we have just to find the explanation. Firstly, why do you say that the reaction is exothermic? Have you taken into account the heat of dilution?

Thank you very much. I presume the reaction is exothermic because toward the end of the titration, I started to observe the heat presumably from binding of the ligand and the protein. I don't have other evidence to indicate that. As for the heat of dilution, I already did the control (by titrating ligand to buffer without protein) and the heat of dilution is very small (~0.02 ucal/sec).  

Hi,

what is the ratio of lingad to protein at the end of the titration? Might the binding process a cooperative process? I would suggest at the moment, to increase the ligand concentration in the syringe, and observe what happens at  higher ligand concentrations.

Greets,

Leonardo

Thank you for your suggestion. As for the ratio, here I used 300 uM for ligand and 20 uM for protein, so the ratio is 15:1. Could you clarify on the concept of cooperative process since actually I am not familiar with that. 

Cooperativity: Protein has multiple binding sites. Binding of first ligand is weak, but once bound enhances affinity at another or the other site(s). Leads to a sigmoid binding curve. https://en.wikipedia.org/wiki/Cooperativity 

Another _potential_ explanation might be that your ligand is titrating something out in the buffer with no heat change. Only when that is exhausted does the heat-producing (protein-binding) reaction produce any heat. Does the pH of the sample stay constant? If not it may be that binding only occurs when the protonation state of protein and/or ligand reaches an accommodating value.

On initial examination of your data, without knowing anything else about your experiment, my observations are:

-Overall heat change per injection  is detectible but appears small, also there is observable noise in the baseline. Which instrument are you using for ITC experiments?   You may need to use conditions to increase binding heat to get your ITC data- increased concentrations, increased injection volumes (if possible), and/or different experimental conditions (temperature, buffer. pH, etc)

-Jaime was questioning your comment about exothermic binding.  Whenever I see ITC data where the heats for the first few injections are smaller than the last injections, I would fit the binding isotherm data (molar ratio vs delta H) data by assuming heat of dilution is at the end, and correct by that value.  When you do that the binding isotherm at the beginning will have a value above  zero kcal/mole, and be endothermic binding heat.  I would also like to see the actual control ITC titration - are the heats/injection similar to the initial injections, or the final injections, or some other value?

-I also do not fully observe the binding is saturated at the end,  which makes the results hard to interpret.  So Leonardo's suggestion of increasing  ligand concentration is good to see if the binding is saturated by the end, plus the possibility of more than one binding event.  If you cannot increase ligand concentration, you can also complete injection of one syringe-ful, leave the material in the iTC cell, fill up the syringe and continue the injections.  I can help you with that experiment if you want.

-How did you check protein and ligand concentrations? What buffer are you using for ITC?

Feel free to email or message me directly.  Verna, MicroCal applications scientist, Malvern Instruments

Thank you very much for all of this helpful information. From what I understand, cooperative binding is certainly possible. I actually have to check the pH of the sample after the titration to see if it is still the same as when I began. I also agree with the suggestion on increasing the heat signal. We actually tried different ligand with the same protein and observed the saturation and large heat signal. It is when we started this ligand that we observed this unfamiliar result. 

Thank you so much again. Once I obtain more result, I will update you all. 

In addition to all of the preceding suggestions, I would consider one more detail.  Since you indicated that the ligand dilution heats are small, I presume that the large exothermic heats observed towards the end of the experiment originate from the protein sample. Did you dialyze this protein batch extensively and use the final dialysate to prepare your ligand?  The protein preparation may contain salt or other solute(s) that contribute large exothermic dilution heats which effectively offset an apparent endothermic titration profile.  When the protein is dialyzed extensively, these large dilution heats are removed and you should be able to measure the actual binding heats in the absence of such artifacts.  The reason that you probably did not observe this phenomenon with other ligands might be attributed to larger exothermic binding heats that overcome buffer imbalances.  On a final note, I agree with Leonardo and Verna who suggested that you probably need to increase the ligand and protein concentrations as a means of improving overall signal to noise.  Good luck with your ITC experiments.

Thank you for your suggestion. The idea of apparent endothermic titration profile sounds like a good suggestion. I will go back to see how well I prepared my protein sample and follow what you suggest. Thank you so much again.

what have you interpreted from this graph I want to know?

Hi Udom, I read all comments about your ITC experiment and especially those one about cooperativity. I'm actually studying a possible protein protein interaction with ITC and the heat profile is very similar to yours. Could you tell me if your binding was really affected by coopeativity? Did you find the answer at the end? Thanks