From the IC50, you can say the Kd is lower than 0.6 μM.

Although the optimal concentrations in ITC depend on Kd and ΔH, there is a major constraint: the amount of heat effect measured per injection.

As a rule of thumb: 10 μM in the cell and 100 μM in the syringe, for a 1:1 interaction. You may change these numbers (by scaling or expanding) depending on the enthalpy and the stoichiometry of binding.

Thank you very much Prof. Adrian Velazquez-Campoy for your advice.

If the Kd for a receptor is 0.6 μM then to get good confidence on your two state quadratic fit, your receptor concentration should be less than or equal to 0.6 μM. But as Adrian mentioned, the sensitivity of ITC may not be that low. Try to see if 0.6 or 1 μM gives you a good signal, that way you don't waste protein.

Archishman Ghosh, that reasoning is fine for many binding techniques. However, for ITC is not convenient whenever possible.

In ITC, for optimal determination of the association constant and the binding enthalpy, the product of the association constant Ka and the protein concentration in the cell PT must be much larger than 1:

c = Ka [PT]

5 < c < 500

That is equivalent to:

5 < [PT] / Kd < 500

Therefore, in ITC we usually work at protein concentrations much higher than the dissociation constant. See Wiseman et al. Anal. Biochem. 1989.

This is related to the fact in ITC we do not work with cumulative heat effects, but with individual injection heat effects. In most binding techniques the cumulative signal is monitored and analyzed.

Adrian Velazquez-Campoy Ah! I see, I was coming from a purely model based standpoint. We use this criterion of [Pt] =< Kd for fluorescence spectroscopy and microscale thermophoresis both of which, as you mentioned, deal with cumulative signal buildup. Thanks for the clarification.