sticky proteins in co-IP - (Aug/28/2009 )

I'm looking for protein-protein interactions in mouse brain lysate using co-IP. My problem is that all of my results are coming up positive, even my negative controls: 1) using a non-specific mouse IgG antibody on the beads, or 2) using a mouse brain lysate that is knockout for the protein that the specific antibody on the beads recognizes (knockout status was double checked on western). I don't see a positive band on the western when I only use beads and lysate (no antibody).

I've probed the western blots for 4 different proteins that should interact with the primary protein (the one the antibody on the beads recognizes). I always see the correct band (either ~170 kD or ~250 kD). There are no non-specific bands. They are cytoskeletal proteins.

I've tried three different kinds of beads - Santa Cruz agarose protein A/G plus, Invitrogen dynabeads protein G, and Sigma Sepharose protein A beads.

I've tried three different primary antibodies on the beads (recognizing the same protein, but some polyclonal, monoclonal, rabbit and mouse)

I've tried washing with buffers containing 0.2%-1% triton X, or 1% NP40, or RIPA with 1% triton X, 0.1% SDS and 1% deoxycholate, or RIPA with 1% triton X, 1% NP40, 0.1% SDS and 0.25% deoxycholate.

So it seems that the proteins I'm trying to IP are sticking non-specifically to any antibody. Any suggestions?

Have you tried preclearing with beads and isotype antibody?

How are you lysing your cells?