ChIP basics - Newbie ChIP questions (Jun/15/2009 )

I am an analytical chemist with a freshman-level understanding of molecular biology. I have been given the task of developing ChIP as a useful method on a campus where it has never been done before. Thus, I have many questions. I am hopeful, though, that a suitable guide can help me.

1. Can you think of a good, useful introductory source for ChIP? I understand the theory, but I need something more nuts-and-bolts (ie, what controls are necessary and WHY).

2. It seems that most of you (on this board) use qPCR for ChIP analysis. I do not have this capability. Does this shoot down all possibility of useful ChIP results?

3. I have (naively) attempted ChIP a couple times now and get beautiful amplification of all samples, including my no primary antibody control. Is there a typical place I should start troubleshooting?

4. Anyone have a good protocol using the sheep anti-mouse (or rabbit) Dynabeads?

Thank you very much.
Mark

If you need to know nuts and bolts of chip, I am not aware of any one stop shop, but you can read many protocols found online. In addition, product instructions such as those from Active Motif and Millipore usually contain comprehensive information about experimental design.

Ideally you want to see differences on regular PCR. If there are only subtle changes, qPCR is an option.

Have you optimized your sonication conditions before you went that far?

pcrman on Jun 15 2009, 11:24 PM said:

Sadly, I don't have the equipment for qPCR. Is there a way to completely eliminate background, or is this just not possible? Currently, my no primary Ab control has similar intensity to both experimental and positive controls.

pcrman on Jun 15 2009, 11:24 PM said:

I think so. Most DNA shows up in the 300-500bp range, though there is a lighter smear outside that range (150-1000bp). Does that sound reasonable, or should all DNA be in a smaller range?

Thanks for your help!

MVW on Jun 15 2009, 12:23 PM said:

Seems to me there are two likely reasons you got a similar signal in your IP and mock IP. Either you ran too many cycles for PCR (thus you hit the plateau phase for both your IP and mock) and/or you used too much chromatin in your IP, which increases your background non-specific binding.

Here's some things you need to make sure you are doing:

1) Optimize sonication so that you get an average fragment size of ~0.2-1kb. To determine the size you need to run on 1% agarose gel AFTER reversing the crosslinking and treating with proteinase K. You can look around here for info on optimizing sonication.

2) You need a positive control antibody, one which has been used extensively for ChIP and which can give you stereotypical results at a few different loci in the genome. For instance. the 4H8 clone for the CTD of Pol II or an antibody to trimethyl lysine 4 of histone H3 (H3K4m3) work well. As a negative control (for the mock IP) I've found that there isn't much difference between using no antibody at all or using pre-immune total IgG.

3) You can do your analysis with regular PCR as long as the differences you are looking at are not small and you take the time to run your amplification at several different cycle numbers (e.g. 28, 30, 32, and 34 cycles) to make sure you are in the linear range for the primers. Even easier than this, you can do serial 4X dilutions of your samples to accomplish the same. If you standardize your chromatin input you may only need to do this optimization once.

4) For your PCR you need primers to a positive control region (i.e. a region where you should get a high signal for your IP). For Pol II CTD and H3K4m3, a region near (within 100kb of) the transcription start site of a gene transcribed in the cells or tissues you are using works well. You will also need primers for a negative control region (a region where you should get a very low or background signal for your IP). For both Pol II CTD and H3K4m3, a region 5-10kb away from any gene (an intergenic region) works well. You could also use the rhodopsin gene as a negative control as it's transcription is restricted to only cells in the retina.

5) For the best signal to noise ratio (i.e. best IP to mock IP signal ratio) you should optimize the amount of chromatin input. If you can estimate the number of cells used to generate your chromatin then a good place to start is an amount of chromatin equivalent to 5 X 10^5 to 1 X 10^6 cells. If you want to optimize from there to reduce your background (mock signal) then you can "titrate" the amount of chromatin you use (i.e. doing a pilot ChIP with a few different amounts of chromatin and finding which has the largest difference between the IP and the mock IP but still has good amplification for the IP --> 35 cycles or less).


If you are looking for a simplified ChIP method I can plug the method we developed, Fast ChIP (Nelson et al., 2006. Nature Protocols 1(1) 179-185). If you can't access this journal, let me know and I can send you a PDF. It is much simpler than the traditional method at least as far as steps after washing the beads. You can use whatever means for IPing your chromatin as you have used previously with the main difference in the method being after the washing steps of the IP.

Briefly:

After the last wash of the beads add 100ul of 10% chelex suspension directly to the beads. 10% chelex is 1g chelex-100 (BioRad cat#142-1253) in 10ml of DI water; the cleaner the better with nuclease free water being the best.

Add 20ug of proteinase K (i.e. 1ul of 20ug/ul). Mix briefly.

Incubate at 50-55C for 15 minutes

Incubate at 100C for 10 minutes. This can be done in a boiling water bath. Be sure to use boilproof tubes so the caps don't pop open. Tube clamps can also be used.

Centrifuge in a refrigerated microfuge for a minute or two, or cool on ice for a few minutes, followed by centrifuging to spin down the condensation

Transfer the supernatant (about 80ul) to a new tube.

Add 120ul cleanest DI H2O to the beads/slurry, mix briefly, centrifuge to pellet the beads, remove and pool the supernatant with the 80ul supernatant from the previous step

You're done; run PCR on the supernatant

This method has been very reliable for us and many other laboratories and has been incorporated into a few ChIP kits. Send me a message if you have any questions.

Joel