RIPA wont do. Please used NP40 or Triton X100 but without sodium deoxycholate and SDS. RIPA buffer is not used to lyse cells for IP experiments.

RIPA wont do. Please used NP40 or Triton X100 but without sodium deoxycholate and SDS. RIPA buffer is not used to lyse cells for IP experiments.

There is good commentary available here: http://www.abcam.com/protocols/immunoprecipitation-protocol-1

Do I understand that the Ab to be used for the IP is purified with the caprylic acid method? If so indeed? Why did you perform that purification and are you certain the purified Ab is still functional?

Syed A Ali his comment is correct. RIPA and SDS should not be used in the IP-buffer.

Try to incubate your lysates with your Ab of interest first (O/N). Subsequently you add the Protein A/G beads (4 h). This may help if you have a low-yield Ab.

I never found better results in IP when the Ab was first mixed w/ the POI than when the POI was added to immobilized Ab. What I can advise you is not to incubate very long. The background signal will get higher with longer incubation times. Let me explain that. See included figure. The Ab has a high affinity for the Ag. This Ag will bind rather fast to the Ab (red dots). After a while it has bound the maximum Ag. The a-specific binding goes much slower  (blue) and has its max much later. The cummulative signal is shown as green dots.

Think rather in minutes than in hours for Ab-Ag binding! This will result in much less BG-signal. 4 hours is way too long and o/n is never my advise for Ab-reactions.

First of all, it is not true that RIPA buffer does not work in IP. Many papers report that it can work. I also did a lot of IP with RIPA. NP40 is milder and thus in some cases is the only option. For sure SDS does not work for IP, or only very rarely and at very low concentrations.

I think that if you want to use your polyclonal antibodies in IP, the best would be to adsorb them directly from the serum on Sepharose Protein A or G (no purification, after which surprisingly no more activity can be seen in your case). The fact that the antibodies work in western blot, but not in IP, suggests another possibility (although rare) that the antibodies recognize the protein only if it is denatured, but not if it is in its native conformation. It is quite strange since this generally is observed with monoclonal antibodies (recognizing a single epitope), but not with polyclonal antobodies.