Roger,
Very low values of Kd (nM) may mean that you have a good chance of finding both the proteins (or protein-nucleic acid) together if you get crystals of the mixture. I therefore think that the measurements are useful in that sense. However, low Kd does not necessarily mean that you get the complex to crystallize. It is possible that one of the components packs better, therefore would crystallize on its own. Also, there could be cases where the proteins have very low affinity for each other (Kd in 100's of micromolar to millimolar range), yet would readily crystallize as a complex. If you want to know some specific cases in this category, look for examples of some ubiquitin receptors co-crystallized with ubiquitin.
Gel filtration of the mixture is not the only answer. You could also mix A and B with one of the components in slight excess (say B is 1.5-fold molar excess than A) and then set up trays without any chromatography. I have spoken to a few crystallographers working on protein-protein complex and they all seem to agree that the the mixing approach works.
I am not sure if anyone has studied the relationship between thermodynamics of protein-protein or protein-nucleic acid complex and their crystallizability. I would like to know, if you find any.
Chitta
----- Original Message -----
From: "Roger Pickman" <[log in to unmask]>
To: [log in to unmask]
Sent: Friday, December 7, 2012 10:11:15 AM
Subject: [ccp4bb] Binding constants/kinetics for crystallisation
Dear all - is there a rule of thumb for favourable values of Kd, kon and koff of protein-protein or protein-dna complexes for protein crystallisation? Are these measurements useful in crystallisation, or should one just put it down a gel filtration column, hope for a complex and not worry? If anyone can point toward a reference, i'd appreciate it.
Roger
|