In those old chemical kinetics courses it was explicitly or
implicitly clear that [I] refers to I(free), not I(total).
The way assays are usually run, [E]<<K(I), so to a good
approximation the amount of bound ligand is negligible.
The way we set up crystallization, [E] is often mM
while K(I) is uM or below, and you had better consider
concentration AND total amount. You will never get
decent occupancy if you add 1 uM ligand to 1 mM protein,
even if the K(I) is 10 nM! On the other hand if you
ignore K(I) and add stoichiometric ligand, dissociation
of a few time K(I) to satisfy the free concentration
will not hurt occupancy significantly.
If you add 2x stoichiometric to
be safe, you may also fill some very low affinity
(100 uM) site. Use stoichiometric pus a few x Kd,
or solve quadratic equation to see how much is
really required.
Marius Schmidt wrote:
think about your old chemical kinetics courses.
what counts is concentration and not amount.
M
Dear Marius and others,
here I would like to comment: The problem with soaking is often not
so much the concentration of the ligand, but the amount of ligand
needed to fill all binding sites in the crystal. A typical crystal
contains about 20 mM protein so one has to add the equivalent amount
of ligand. On the other hand, the concentration UNBOUND inhibitor,
asuming a 1:1 protein-ligand complex, need only to be in the order of
10-100 times the Kd (90-99% occupancy). Ways to overcome this is to
soak in a large volume of mother liquor (containing a large amount of
ligand) or to add solid ligand to the mother liquor and let the
ligand slowly dissolve and diffuse into the crystal.
Herman
