I am going to do affinity purification of a polyclonal antibody from rabbit, but the antigen (100KD, with GST tag) is insoluble. Does anyone has experience in purify the antibody with insoluble antigen?
I have a suggestion, since insoluble means it can be pelleted down. In this case, you can mix your insoluble antigen with the rabbit diluted serum. You can then centrifuge to pelleted down your insoluble antigen which could contain the specific antibody.
I would suggest another way. You can precipitate all IgG from your rabbit antiserum with 45% Ammonium Sulphate, then collect the pellet by centrifugation, dissolve it a in any buffer with pH higher than 6.0, and load this solution on a column with MEP-HyperCel resin (Pall Corporation, Cat # 12035-010), elute the bound antibodies with any low-salt buffer around pH 4.0, and neutralize the solution up to pH 7.0-7.5. In this case you will get a total IgG fraction (not just your antigen-specific antibodies), but it will be good enough to make an immunoaffinity resin. We did it many times in our lab and it always worked well.
The first answer will work but will not be efficient because the antigen protein molecules will be clumped together in large particles (relative to the size of a protein molecule) so the vast majority of binding sites will be inaccessible to antibody. A better option is to run the protein on preparative SDS-PAGE and blot to nitrocellulose or PVDF. Stain a narrow strip of the blot to locate your protein band and then cut out the band for use as an affinity matrix. You elute bound antibody with low pH (in my lab we use citrate) just as for chromatographic purification of antibody by antigen affinity. For the SDS-PAGE it's better to run the protein with disulfides NOT reduced to increase the chances you'll have some conformationally correct antigenic determinants for the antibody to bind.