Discussion: ChIP PCR control - (Aug/03/2005 )
Hi, I would like to discuss PCR control in ChIP with you.
I have 9 pairs of primers which amplifies the respective candidate region, I didn't mix two pairs of primers in a single PCR tube (i.e. control and candidate) because I found sometimes there are severe primer dimers. So everytime I do the enrichment test I have to run a parallel PCR reaction with control primer.
***********************************************************
After I recovered the DNA, I always calibrate my input and "ChIPed" DNA amount so that they produce similar level of PCR product with control primer. This step has been repeated twice.
Then, I started to do the enrichment test, of course I have to do the control PCR as well. This test was done twice. The PCR results of all the 9 candidate regions were repeatable. However, in the first round of control PCR was nice (equal intensity), but in the second round, I found an obvious difference in my control PCR, If I adjust everything according to this one then every candidate got no enrichment at all. So sad......I don't have enough DNA to do the third round 
I'd like to ask, can I just treat this bad control as PCR variation?
How do you guys make control if you cannot mix two pairs of primers?
Thank you very much!
Hi bullfrog,
I curious to know how and why you have calibrated both your input and IP DNA to your control PCR. If you do this, then essentially you are normalising the concentrations of your target in both input and IP, therefore you would essentially get no enrichment.
Nick 
Sorry for such confusion 
The primer for control PCR amplifies a genomic fragment of tubulin gene, which is not regulated by our protein of interest. What I did is firstly normalize tubulin PCR product, then with adjusted template amount, I did PCR again to check the enrichment on my target genomic sequence.
I am not sure that this will be much help, it seems you are more advanced than I... but here goes....
Maybe too many rounds? How much was your first round diluted before adding to the second.. If they look the same after one round and everything is in linear range, they should look the same after two rounds right?? If you see product after one round, is it necessary to do two/three? Probably you have already controlled for these but I thought to throw them out there....
In answer to your other question, I always do seperate reactions for my negative control (like the tubulin you use), positive control, and my gene. To normalize between these different primers (different efficiencies etc.) I normalize to dilutions of the input DNA (in the range of 0.01 to 1%) that are amplified in the same master mix... This also helps me confirm that I remain within the linear range.
Question for you... What percent of input do you pull down with your antibody, is it a TF antibody?
Yes it is a transcription factor. I didn't pull down all the TF from my sample because I found protein degradation with longer incubation time, even with protease inhibitors. I think only around 50% has been pulled down.
So you get enrichment equivalent to 50% of the input?? I see about 50% bound protein on WB but routinely pull down only about 0.2-1% of input DNA is this similar to what you see?
True. I guess the ChIP DNA that I obtained is roughly 0.2% of input DNA, since I have to dilute input DNA about 500 times to get equavalent band intensity for my control (tubulin).
So you get enrichment equivalent to 50% of the input?? I see about 50% bound protein on WB but routinely pull down only about 0.2-1% of input DNA is this similar to what you see?
Do u agree on the background amplification should be better if less than 0.1% of INPUT DNA?
So you get enrichment equivalent to 50% of the input?? I see about 50% bound protein on WB but routinely pull down only about 0.2-1% of input DNA is this similar to what you see?
Do u agree on the background amplification should be better if less than 0.1% of INPUT DNA?
Yes, if your fixation, sonication, immunoprecipitation step work well, it is better if you have less input DNA recovered. Even your antibody does not work (I have such bad experiences), you wash and wash after the binding step, lots of DNA will be discarded, won't it? Therefore personally I don't care how less it is, I only care about the enrichment test result.
I am also more interested in the enrichment.
BTW,
In papers, it looks like everyone gets a completely clean background reaction. This has concerned me in the past, but based on some of the posts here I am no longer convinced that it is wrong to get a band in the negative control.
Maybe the authors(of the above-mentioned papers) could have used tricks of PCR to eliminate any visible amplimer in the negative control, I don't know if I agree with this tactic, it seems a little like manipulating data, but if your background is 0.1% and your enriched fraction is 1% then using conditions where your limit of detection by PCR is 0.2% you can do an experiment that looks like there is no background but is really due to using less sensitive assay conditions.... This can also be accomplished by manipulating the exposure time of the gel...
Anyway, does anyone have a comment on this? Is it right to optimize your assay to eliminate a visible background band??
