Is my ChIP result positive or negative
Posted 20 July 2011 - 07:18 AM
First, a quick description of what I am doing: I am treating a suspension cell line with 1% formaldehyde in media, isolating and washing cells, then lysing, sonicating, and diluting chromatin as described by the UpState EZ ChIP protocol. After pre-binding with protein A agarose resin, I am adding antibodies for (generally) lymphocyte specific transcription factors (Ets-1, Runx-3, others); I add about 3 ug each for about 7 million cell equivalents of chromatin. After overnight incubation, I add protein A agarose resin, wash resin as usual, elute DNA from resin, purify the DNA with Qiagen columns, and quantify the amount of DNA recovered using quantitative PCR.
Using primers specific to my promoter of interest, I see a 3.5 unit increase in Ct for Runx-3 over the non-specific IgG, and about 2 unit increase in Ct between Ets-1 and the nonspecific IgG, which is of course good.
I want to test a negative control region so I use a region downstream of the promoter, and I see about a 0 unit increase in Ct between my specific antibodies and the nonspecific IgG, which is of course good.
Another negative control region I checked is the GAPDH promoter. This promoter is ubiquitously expressed, and so I would not expect it to bind lymphocyte specific factors like Ets-1 and Runx-3. Yet, using primers specific for the GAPDH promoter I see a 3.5 unit increase in Ct for Rx-3 over the non-specific IgG and about 2 unit increase in Ct between Ets-1 and the nonspecific IgG; in other words, the same results I got with my promoter of interest.
These results may indicate that I have binding of Runx-3 and Ets-1 to my promoter of interest, and the GAPDH promoter. I find the binding of Runx-3 and Ets-1 to GAPDH very hard to believe, and am wondering if there is some kind of mistake I am making so that these are false positive results.
So I am trying to figure out if my results for my gene of interest are positive or negative.
What do you think? Any input is appreciated.
Posted 20 July 2011 - 11:11 AM
I had the same problem looking at GAPDH as a negative control region for a transcription factor - in my case Ct values (I suppose that is what you mean with 'unit') were 5 cycles lower for specific antibody vs IgG (converts to ~128fold "enrichment" of the TF in question at GAPDH). Two things helped me get rid of the problem:
- I changed from amplifying a fragment of the GAPDH promoter to a fragment from a GAPDH intron, and more importantly
- I included a centrifugation step after incubation with the primary antibody. (Spin max speed, 8min, 4°C and transfer the supernatant to a new tube with your beads). I compared two samples, where the only difference was this centrifugation step, and it made all the difference between having the false enrichment or no enrichment.
Unless you absolutely want to use that kit, changing from agarose to magnetic beads may also help reduce unspecific signal.
Posted 20 July 2011 - 11:24 AM
With my work, I already have the extra centrifugation step as you suggested. Yet one negative control region has no Ct difference while the other (GAPDH promoter) does have a 3.5 Ct difference. I would think if it was a false positive then both negative control regions would have a 3.5 Ct difference.
Posted 22 July 2011 - 10:16 AM
Posted 22 July 2011 - 10:26 AM
Thanks for the response. That is what I was thinking myself; that what is happening is that my promoter of interest and the GAPDH promoter may be sharing a "transcriptional factory", such that Ets and Runx crosslinks to RNA Pol II reading my gene of interest, which cross links another RNA Pol II reading other promoters, like GAPDH.
I have checked a downstream Intron and I do not see any Ets and Runx binding, so I think this maybe enough to rule out a false positive result. Therefore I am leaning toward thinking that my results are positive for specific binding of Ets and Runx to my gene of interest.