ChIP PCR question - (Feb/22/2011 )

Thanks for replying. You guys are like a wealth of knowledge! I actually did prepare an IP using unconjugated beads (no Ab) but i didn't use it to run the pcr because i thought it wasn't a great control. Now that you guys mention, i should use it to run as well to see how dirty the background is using those beads, and if they give bright bands too, i'll be able to convince my prof to change to magnetic beads.

chabraha, can you explain further about subtracting out the %Input for the mock IP from the %Input of the specific IP?

Also, i have another question about inputs. Sorry for having to many questions. Is there a difference in saving an aliquot right after sonication as compared to saving the aliquot after preclearing as input? The reason i'm asking is because i've seen several protcols and some saves the input right after sonication, while others saves after preclearing. I actually do have several TFs that i'm interested to look at but right now i'm only doing one because the remaining Abs against the other TFs are of different hosts and i don't have the respective IgG isotype to preclear yet (i'm correct to say that Ab from different animals require the respective IgG isotype to preclear right?) So in the future when i'm ready to look at all the TFs, saving an aliquot right after sonication would mean that i will only have 1 input for as many IPs i'm doing since they all come from the same tube. However, if i need to save the input aliquots after preclearing, different IPs will require different inputs because they are precleared using different isotype Ab right? So which one is correct, or is both acceptable?

I would take chromatin aliquots after clearing, since you may actually loose target in the pre-clear step......it is best to use the appropriate isotype for the species you are looking at but I doubt many people would say its a requirement......you work with what you have available. As for the %Input calaculation......this is my way of controlling for DNA that appears either A) non-specifically binds to your beads/IgG; or is contamination, I know that alot of people think their labs are free of DNA contamination but when you work in a viral lab like mine that stuff is everywhere and difficult to get rid of completely. So to ensure my data reflects true enrichment I calculate the %Input for my specific IP and then calculate the %Input for my mock IP. Next, I subtract the mock IP value from the specific IP value. This gives me an IgG corrected %Input for my specific IP. I do this calculation for each loci looked at. So then I divide the each corrected experimental loci looked at by the corrected %Input of my control region.

First of all, for what it's worth I never pre-cleared my samples. I just have an Ab tube and an IgG tube. For the input I would typically mix one of my tubes (chromatin + buffer) then remove 1% (usually from the IgG tube) and use that as my 1% input. Since I use 2ug of chromatin, it would be impossible to actually pipette 1% of the used chromatin directly from my samples.

As far as the calculations, I do it a little differently than Chabraha. I do not use my %input from my mock samples for any quantification. This is because, at least for me, my Ct values for my mock samples typically have a lot of variability associated with them using qPCR. I typically just use them as a qualitative control, i.e, I just check to make sure my Ab enrichment is higher than my mock. For my calculations, to check for actual enrichment I calculate %input for my Ab at my target or experimental site and compare that to the %input I get at my negative control site. For the size of signal I get it can actually be difficult to get a good negative control region, because if you pick something that doesn't come up with a reasonable Ct on your qPCR, it gets very noisy. You've probably seen me recommend using a retrotransposable element (LINE1) as a negative control because it exists as a very high copy number in the genome (at least the mouse genome that I was using...should be good for the human genome too I imagine) at is actively repressed, so it will contain markers of transcriptional repression (and not be very likely to bind many TFs). The other thing I always try and run is a positive control primer set in which the TF has been shown to bind. This can get tricky when the literature is a bit sparse in your particular system, but it's worth a good check.

MM

Thank you guys again. just to have some form of reference, can I roughly know what's the approximate ct you get generally for your 1%input, your target primer with sample ip, your neg ctrl primer with sample ip, and also those primers with mock ip? I just want to have a feel of what kind of results it may be, like do 1%input usually give better ct than target primer with sample ip? Or how many more ct is mock sample results and etc.

Mm, why is it that neg primer that don't give reasonable ct on qpcr will make your result noisy? Shouldn't neg primers, in theory, give like the biggest/ undetermined ct? I designed a neg primer pair in the intergenic region far from my promoter and did see very much lower binding there as compared to my region of interest. So thats a good neg ctrl right? And yup i've already bought positive ctrl primers based on other published articles

Sorry guys, i'm facing another problem. I've been trying out normal PCR on my samples and inputs and i'm getting a lot of variability in the results. Using the same samples and primers, sometimes i get bands, and sometimes not. Sometimes they are bright, and other times faint. Both the inputs and the IP samples give the problem. I've tried changing every reagents i have, including re-diluting the primers, and doing a 2nd cleanup for my samples, and i still get the same problem. Its just impossible to get a proper set of data with all the positive ctrls primers out. I've had a lot of experience running PCR and have never faced such issue. If you can think of any other possibility, can you please enlighten me. Thank you.

how many PCR cycles are you running?

35. I did a titration of the cycle number b4, from 25 to 45.

If you run a dilution series of you Inputs do you get any bands?

I did not do serial dilution. I have my sample eluted in 100ul. Then I use 1ul to run pcr in 25ul rxn. Sometimes they work, sometimes not. How would serial dilution help me?

After hearing how much you elute in and how much you use in your PCR I don't think that dilutions would help you. I was just thinking that maybe sometime your PCR template concentration may be out of sync with the primer concentration. Anyways, as for a suggestion to your problem.......maybe your reverse cross-link/ProK treatment is not working efficiently. You can check this by running an agarose gel just like you check your shearing except for one lane use a sample where the cross-link reversal method has been omitted.