ELISA for protein-protein binding assay

I am tring to determine the specific binding activity
of purified protein which binds to another glycoprotein.

For some protein which can bind to small molecules,
the radio-labelled ligand is avilable, and the specific
binding activity can be monitored from CPM value after
free-bound separation.

My question is if ELISA is useful to determine
specifc binding activity between protein and protien.
Problem is it seems to be hard to label with radio to
protein, and it is not easy to separate big protein
from also big bound protein.

I could not find 'specific binding activity' value
such as pmole/mg for protein protein interaction
system using ELISA. Also SPR can determine kinetics value
but not for specific activity value.

H.R.T., MIT

takayama wrote:


If you are prepared to work with radioactivity, labeling proteins with
iodone-125 is no big deal, check the Pierce catalog for kits to do that.
Disposable gelfiltration columns (e.g. Pharmacia) work well to separate
the labeled protein from free iodine. If you do not have access to a
gamma-counter you can count 125-I in a beta-counter as well, if you add
a solution of 10% ZnCl2 to the szintillator. The gamma-radiation removes
Compton-electrons from the Zn-ions, these are counted. Overall counting
efficiency for 125-I is about 80%.

Note however that the setup would not be ELISA- (enzyme linked
immunsorbent assay) but a RIA-like (radio-immuno assay). The whole point
of ELISA is to avoid the costs and burocratic hassle involved in the use
of radioisotopes by using enzymes as label. Conjugating a protein with
an enzyme is no big deal either, several chemistries are available. A
search in Meth. Enzymol. is appropriate here, there is also a good book
by Kessler.

Either case to detect interactions you'd incubate the first, unlabeled
protein with your ELISA-plate, which is made from polystyrene. The
protein attaches by hydrophobic interaction (thus this incubation
happens in the _absence_ of carrier protein, which needs to be present
in all following steps). There are also activated plates available where
the protein is bound covalently, rather than by hydrophobic interaction.
After washing away unbound protein you incubate with the second protein.
The protein may carry the label itself (direct assay), or you can use a
labeled antibody against the second protein to detect and quantify
binding (indirect assay). For quantitation you need to include standard
dilutions of your second protein, this is of course easier in a direct
assay.

Since only a few percent of the first protein bind to ELISA-plates I
find it more useful to dot-blot it onto a PVDF-membrane, which is >98%
efficient. This increases your sensitivity by 2 orders of magnitude.