A number of factors could contribute to the outcome of your experiment. To start with, your lysis procedure may not be releasing enough protein in soluble form. I would try RIPA buffer for lysis. If the antibody is well-characterized and works in IP, it should be ok for your purpose. You need to run total cell lysate as a control to assess whether protein of interest is expressed in cells and detectable using the antibody.

I won't sonicate the lysate for the same reason why you choose to use mild strength of lysis buffer. Another buffer you could try is Pierce IP Lysis Buffer from Thermo Scientific (25 mM Tris•HCl pH 7.4, 150 mM NaCl, 1% NP-40, 1 mM EDTA, 5% glycerol).

If the endogenous protein expression is low in HeLa (you should confirm it as mentioned by Choubey), have you considered to repeat the experiment with overexpression approach? In that case, you can use tag Ab which is generally very good at immunoprecipitate ectopic tagged protein. Some antibodies are good for detection but not immunoprecipitate protein. You should check the data sheet. If you get a positive result, only then you confirm the interaction at endogenous level.

Did you WB for the protein that you immunoprecipitate to confirm an enrichment/successful immunoprecipitation compared to inputs? This will answer you hypothesis whether your Ab isotype affect the binding capability with A/G beads. 

If you protein is indeed co-immunoprecipitated, you might be able to detect it with longer exposure duration and higher sensitivity substrate reagent.

-We already use an overexpression approach with the protein of interest tagged by the FLAG.

-The Ab used has been used previously to immunoprecipitate the same protein and it worked wel.

- In the input (proteinaceous fraction before doing the ip) we've observed both the protein of interest that we want to precipitate.

Thank you guys for your precious suggestions!! 

Have you ever tried other lysis buffer (with triton X-100 for instance?). Does you Ab work for IP? Have you tried with G prot/beads (for IgG1 it is the best : https://www.neb.com/tools-and-resources/selection-charts/affinity-of-protein-ag-for-igg-types-from-different-species). Does the analysis work under overexpression conditions?

I usually did a pre cleared step (lysate + beads first and after lysatefor 1 or 2 hours) and then :  Ab overnight in rotating apparatus at 4deg C, and then + Gprot/beads for 1 or 2 hours).

Good luck

Alice

Thank you for the advices Alice. I was thinking that the problem could be the buffer used. Regarding other options I would say that the Ab has been already tested in IP and it worked, About changing beads I know that could be a better choice but at the moment the Co-IP is not as important as it may be thought so we won't spend for other beads. Regarding the way to make the initial lysate, do you think that sonicating the extract could be detrimental to the proteinaceous fraction? 

Use RIPA Lysis Buffer

Incubate with Abs for 1 hr at 37C or overnight 4C

If possible try to use rabbit Poly clonal anti sera . (Since it doesn't ve isotype variation)

Hi. To be honest, I never used the sonication process... so I don't know. I worked on an endothelial cadherin (massively expressed in cells), so I used like 1% triton in my lysis buffer (+ NaCl, NP40 + HEPES + protease inhibitor in the buffer if I well remember). Because I worked on adherent cells, they grew on a 10cm culture dish, I put 300 ul of my cold lysis buffer (and make sure it covers all the dish), leave 20 min on ice, scrap the dish, centrifuge 14000 rpm for 20 min, retrieve the supernatant and work on it.

You got a lot of excellent suggestions.  We use IP buffer as suggested by Kai Ling Lialngand homogenize by freeze and thaw.  We freeze the cells at -80C just by putting the cells in the freezer for 15-20 min when the volumes are small (