Protocol Online logo
Top : New Forum Archives (2009-): : ChIP and Next Generation Sequencing

Antibody validation for ChIP - high backgound after knockdown - (Jun/21/2013 )

Hello everybody!

I have got a huge problem with my ChIP assay. I get a ChIP signal after precipitation with an antibody for a transcription factor in a promoter region of a gene. The problem I am facing at the moment is, that I am not sure anymore, whether this antibody is actually specifically recognising this transcription factor, or wether it pulls down something else.
Here is what I have tried so far to validate the antibody:
- When I overexpress the transcription factor, I do not get a stronger ChIP signal. Is this because free protein, that is not bound to DNA blocks the antibody? Or is my antibody not recognising my protein of interest? Or is the DNA "saturated" with my protein and does not bind any more of the overexpressed version?
- When I knock down the protein, I get a stronger signal in all the DNA regions tested. The negative control region shows about 10fold more enrichment, my region of interest shows about 2fold more enrichment. I am sure that there was no problem with washing this sample, as I performed a big assay with over 20 samples and ALL others are okay. Does this indicate, that the antibody binds unspecifically to other proteins? And if so, can I rely on my ChIP results? Does it only bind unspecifically when the specific protein is gone?
- The antibody recognises lots of bands in western blot.
- Neither overexpression nor knockdown of the protein influence the expression of the potential target gene.

Okay, I actually think, that I can't rely on this antibody. But how do I validate a new antibody? I do not have a positive control gene, as nobody ever worked on this factor in my experimental setting. And more impotantly, how can I prove to my supervisor that the antibody I am currently using is not working?

Thanks a lot,



Have you looked at your RNA to determine what the actual % knockdown is?


I don't get knockdown on RNA level, but I get about 50% knockdown on protein level. I have stable, inducible knockdown clones and I check the knockdown by western blot for every experiment I do. So the protein is definitely down.


This antibody doesn’t seem working well for your ChIP and WB. To validate an antibody for ChIP, I normally do as the followings:

1) Use IgG control to set the baseline for background. If the ChIP signal of the antibody is lower than that of IgG, I would not consider the antibody works. However, since some antibodies generate much higher non-specific signals than IgG control does, only comparing the ChIP signal strength from your antibody and IgG might be misleading. I would recommend to compare the ChIP signals of different regions, preferring different regions along the same gene. You may design primers at promoter, coding and 3’ end of the gene and test their signal strength. If the ratio of signal/noise is more than 5, I would be comfortable to consider that antibody is good for ChIP and the binding is specific.

2) Check If the ChIP signal from the specific region changes with treatment, e.g. activation of pathway or knocking down. I don’t think overexpression is a good idea, since it is difficult to control the protein expression level and not physiological relevant.

3) In your case, you may try IP with your antibody from cell lysates, then blot with your antibody. If you can’t get enriched and more specific band, it suggests that this antibody is not working well for IP and, in most cases, the antibody won’t work well for ChIP either. Also you can try using your knocking down cells to check if the IP signal changes.

Hope this helps,


Why not find genes that your TF regulates (in the literature) and test those promoters. Also try native ChIP instead of xChIP the cross-linking may be blocking the epitope or a complex your protein of interest is in may be blocking the epitope.............especially if its a monoclonal.............I've found that if your antibody looks like garbage in immunofluorescence then it will look like garbage in your ChIP.