immunoaffinity chromatography-choose of matrix and functional group? - (Oct/09/2007 )

Thanks The Bearer, Almasy for answer.
Other question. . .
We would like to do immuno-affinity purification (using HPLC) of protein Mw = 50kDa (the protein is in triton x-100 insoluble fraction after centrifugation, 18000xg). We have polyclonal IgG rabbit antibody (produced for us at a company). This antibody is able to detect very nicely recombinant protein and probably also natural one. Unfortunately, we are not able to do classical (A/G agarose beads) immuno-precipitation with this antibody. We would like to buy/use NHS-activated sepharose (Amersham Biosciences) that is recommended for such a purpose.
1. Does have any sense to try to do immuno-affinity chromatography with this antibody if normal IP does not work?
2. Does anyone have any experience with the NHS-activated sepharose (or other) (is it working well, are there any problems etc.)?
3. How/in what buffer to dissolve (perfectly) protein for immuno-affinity chromatography, if the protein is (in) the triton x-100 insoluble (fraction)?
4. NHS-activated sepharose is designed for covalent coupling of ligands containing primary amino groups - if it will be coupled to primary amino group of (our) antibody, will not be the affinity (functionality) of the antibody to (our) protein restricted? Should not we use preferably any other medium - for example, EAH sepharose (Amersham Biosciences) that couple ligands containing carboxyl groups?

Thanks for suggestion!
vic

Coupling through primary amines (e.g. NHS-activated resin) is the most common, but it will result in a random orientation of antibodies and, in my experience, at least a 50% reduction in antibody activity (amines may be at or near the antigen binding site). The same will probably be true for a carboxyl-reactive support. I would recommend a hydrazide support, such as CarboLink, that reacts with carbohydrates if you are concerned about preserving antibody activity. This method will bind the antibody to the resin through glycosylation sites on the Fc portion of the antibody.

there is a support, available from pierce (but hard to find in their online catalog), that uses protein a or g (they have both) so that the antibody orients so that all bound antibodies are available to bind antigen. you also crosslink the antibody to the ligand so that you can release the antigen and leave the antibody behind for reuse.

to answer to question 1 :

yes it can make a sense, if your antibody recognizes the antigen. IP doesn't work, maybe because the antibody is not well recognized by protein A or G?