Some ChIP Questions - (Mar/02/2012 )

Hi everyone,

I have started doing ChIP a few months ago and am trying to optimise some conditions. I have a few questions/concerns:

1) How necessary is the NaHCO3 in the elution buffer? I am following a protocol used by a previous student, and adapted from a "micro ChIP" paper, and in this protocol the elution/reverse cross-linking/proteinase K treatment is done all at once by shaking at 1300rpm, 68 degrees, for 2 hours. However this elution buffer consists of SDS, NaCl, Tris and EDTA - no NaHCO3.

2) What is an acceptable amount of input DNA to get back with antibody, and what CTs from qPCR are okay? The most recent ChIP I did had these numbers:Input (200 ul starting material, 22.2% of IP) concentration: 40 ng/ul in 40 ul (total 1.6 ug DNA). Running 2 ul of this undiluted has a CT of 19 for target gene.Antibody (900 ul starting material): CT of ~26 for target gene, CT of ~30 for control regionIgG control (900 ul starting material): CT of ~31 for target gene, CT of ~31 for control regionSo I have an enrichment of 64-fold over IgG for my target gene, but an enrichment of 2-fold over IgG for my control region. I assume this normalises to ~36-fold enrichment? Furthermore, when I calculate the % of input that I get with antibody, it is about 0.3% - is this okay, or very low??
3) I've been trying a ChIP with a different antibody (for diff protein). This time, I get:Antibody: CT of ~28 for target gene, CT of ~28 for control regionIgG control: CT of ~31 for target gene, CT of ~31 for control regionSo just looking at the target gene, it looks like I have ~8-fold enrichment over IgG, however the enrichment is also there for the control region. What would you suggest to try to overcome this? Would more washing help remove the non-specific binding, more antibody/less antibody? I currently do not have the LiCl wash as part of my protocol, which seems to be working okay for the previous antibody, but would doing this possibly help? I am already using magnetic beads which are blocked in BSA.
Any input would be greatly appreciated, thanks!!!

Julie-Ann

One more thing, in relation to sonication. I am sonicating C2C12 mouse myoblasts in 1.5ml, doing several rounds of 3 sec bursts at 40% amp, and this is typical of how my chromatin looks:

http://i.imgur.com/KEncc.jpg
In image 1, I've sonicated for 12 bursts and 24 bursts and taken 20 ul aliquots. I've run the crosslinked chromatin next to the reversed ones, and also done an RNase treatment on some unsonicated material. As you can see before I do any sonication (lane 2) I already start with a smear around 700bp, but this isn't RNA (still present in RNase treated). I haven't seen anyone else have this before, is it normal? I also don't really see any difference at all with RNase treatment, except for the presence of some low molecular weight stuff. After doing several rounds of sonication, I can start to see the DNA migrate out of the wells and accumulate around 3kb. When I reverse crosslink, the only difference is now I can see a smear below the original 700bp smear.

http://i.imgur.com/qviLO.jpg
Image 2 is exactly the same as 1, except the gel has been run for longer. Question is why does it now seem that my chromatin smear is so much larger?

http://i.imgur.com/3pOZh.jpg
Image 3 is another few rounds of sonication. By 42 bursts my chromatin is also 500bp-3kb (I only checked crosslink). This is what I have been going ahead with in my ChIPs.

In answer to question 1, the NaHCO3, I'm fairly sure, is there to raise the pH thus stablizing the DNA at high temp. Other reagents used for this same purpose are Tris base (at pH 10 rather than a buffer at lower pH) and chelex resin. I haven't looked at the micro ChIP protocol for a while but I would guess that the pH of the elution buffer is relatively high.

In answer to question 2, your IP looks great and the Cts are in a range I've seen with several other antibody/primer pair combinations.

For question 3, is the antibody you are using one which has been validated to work in ChIP by another lab. This is certainly not a guarantee that it will work for you but it's a good place to start. There are a number of things that can affect the efficiency of an antibody (e.g. crosslinking time and reagents or shearing buffer and shearing time; I haven't found that changing the washes is all that effective in reducing background though with a Ct that high you might try reducing the amount of antibody). You can waste a lot of time on optimization though, so you might want to look and see if there are other antibodies against the same protein, which have worked in ChIP for others, and try them with your current conditions.

Regarding your chromatin sizes, the only way to accurately determine them is to reverse crosslink before running on a gel. With the small sizes of your reverse-crosslinked chromatin I'm assuming you're using an SDS buffer? In any case, given what you showed on the gel I think your best conditions are the shortest time where you get most of the chromatin to migrate out of the well.

Thanks very much for your answers

The micro chip elution buffer is 20mM tris pH 8, 5mM EDTA and 50mM NaCl to which I add SDS and proteinase K, and incubate at 1300rpm, 68 degrees, 2 hours. So this doesn't really fit with the theory about needing a higher pH??

Good to know my CTs look okay! And getting back around 0.3% of input DNA is acceptable too?

Unfortunately it isn't a ChIP grade antibody, and there aren't any available for my protein There have been two studies published with using my antibody in ChIP, but both are only very small enrichments and the papers don't give very detailed descriptions of their methods.

Yes, I am using 1% SDS buffer to sonicate. Why is it that after reverse crosslinking, I get a very small product, but still a lot of DNA >3kb, and even still in the wells? Does this mean that the reversal wasn't 100% effective? Have you also had experience with the chromatin looking different on a gel depending on how long the gel was run for?

One more general thing, after looking online the buffer recipes given to me seem to be slightly different to what the majority of everyone else is using. For example, all of my washing buffers contain 0.01% SDS, while everyone else uses 0.1%. Also all of my wash buffers and TE contain 1mM EDTA and 16.7mM Tris ph8, yet others use 1mM EDTA and 10mM Tris for low salt/TE, and 2mM EDTA and 20mM Tris for high salt. Do you think this would have any impact on my results?

Thanks.

Also I'm still not sure how the RNase treatment managed to produce a band, not remove anything, but I guess it's not something to get too worked up about?

Hi Sel,

reply to your usual band -
I had same problem of usual band you are seeing on the gel. Looked to me too that I might be loosing some crosslinked product!

In fact, the band will not migrate at specific size as you have mentioned and it is not a DNA band as it does not have edges and also its too smiley! It looked more like histone/protein band. To be double sure its not any contamination in my buffers which is affecting nuclear isolation, I ran my buffer and also constitutes of it and found that its SDS.

Took two weeks to nail this down!

just run 10% SDS (or your stock) and check on gel to confirm what I said above.I dont think it will interfere with your ChIP as its quite a std. buffer you (and myself) are using.

Hope this helps!

FYI- your IP control and all info is very useful as I am std. ChIP and currently at pull-down step. just posted a question on the form with some pull-down queries.

Cheers!