MNase vs sonication for ChIP - (Sep/23/2005 )

Hi everyone,

I am trying to optimize the chromatin shearing for ChIP. I was using strandard sonication with a microtip but it wasn't giving great results (maybe I was sonicating with too much power... foaming was frequent). Anyways, I wanted to try alternatives, so I tryed sonication with a cup horn and MNase digestion (micrococcal nuclease). See the attached picture for the results.

MNase digestion seems great. What you see is the sheared DNA from increasing amounts of enzyme (I can give out my protocol if anyone is interested). My main concern is that the strongest band (around 180 bp, the length of a nucleosome I guess) might be too small for ChIP, i.e. the ChIPped DNA will be too small to be amplified by PCR (unless the PCR fragment is very small). Is that why 500-1000bp is generally the target size for ChIP ? It is so easy to do such an enzymatic digestion of DNA... I don't understand why it is not used more frequently (compared to sonication, which seems to be much more tricky to set-up).

Cup-horn sonication stills needs optimization... I did 0-5-10-15-20-25-30 pulses (10 sec) and some of the DNA doesn't shear at all. It is not an overfixation problem since the same DNA can be sheared by MNase (these experiments were done in parallel).

Event though the same amount of nuclei was used for digestion and sonication, you can see that you get more DNA from sonication. This is because MNase digestion does not burst the nuclei and a lot of them are lost in the spin after digestion. Maybe I should try a few sonication pulses to burst the nuclei to recover the sheared chromatin inside... or another method to burst the nuclei (any suggestions on this issue?). Conclusions up to now: I get better sheared chromatin with MNase than sonication but lower DNA yields.

Anyways, I would really appreciate your thoughts on these results or any other suggestions... Thanks!

Controller

I agree, your mnase looks like it is shearing more cleanly...probably does have to do with the foaming thing....

there should not be a problem with using small primers for the ChIP, it should improve the probability that the amplicon remains intact after shearing... I think this is also the reason people use an average of 500bp for sonication, you shear enough to exclude background from neighboring sequences, but the pieces are still large enough that there is a good possiblity your amplicon will remain intact.... along these lines, I like the third lane of your mnase for IP...

I only wonder about the reproducability of the mnase, sonication is very reproducible as long as the amount of DNA is not hugely different... seems like mnase may be more susceptible to having variations even with the "same" treatment b/c enzyme is temperature dependent etc... if you get some data either way would love to know....

good luck!

Hi there... I'm happy to see that someone else is having trouble with a cup horn, I was beginning to think its me! I see almost the identical pattern that you see (too much big stuff).

I agree with the other poster... the third lane in your MNase gel looks ideal for ChIP. It gives you a range of DNA fragments, but all the sizes are appropriate for ChIP. If you are concerned about 180 bp being too small for your PCR, under these conditions you still have plenty of DNA that is in the 500-2000 bp range.

I would love to see your protocol for an MNase ChIP! I have tried it a few times myself and I get great DNA shearing, but I haven't yet succeeded to IP from these samples... I'm missing something in my protocol!

Good luck in your experiments!
-Melissa

Have any of you tried using hydrodynamic shearing to fractionate DNA? Pushing the DNA through a fine needle under high pressure is very effective at shearing long DNA fragments. I routinely do this 3x with a 28 ga needle and a 1 ml syringe, producing 4Kb fragments. Depending on the number of times and the pressure used, this technique can produce a range of fragment sizes. See:

Thorstenson YR, Hunicke-Smith SP, Oefner PJ, Davis RW.
An automated hydrodynamic process for controlled, unbiased DNA shearing.
Genome Res. 1998 Aug;8(8):848-55. PMID: 9724331

4 Kb is too long for ChIP because you start to get effects from neighboring chromosomal regions. I agree that enzymatic digestion is a good alternative for soninication and maybe more reproducible depending on how skilled you are at avoid bubbles, ect. during sonication. You should probably consider lysis of your cells before digestion for the enzyme to have better access at the DNA. You may have to do a purification step however to get the correct buffer conditions.

Controller!
I'm working with nChip protocol, and I'm using short amplicons (90-120 mer) with very good results. Just, You have to be sure that the amplicons are present in the input sample, after Mnase digestion. I'm trying to work at the mononucleosome level to improve the resolution of the nchip.
But I'm having some problems with the recovery of chromatin after digestion, because two bands appears in the gel at 4000 bp and 1000 bp approximately (I believe that are RNAs) and this difficult the DNA estimation before add the Mnase..
Could you send me a pdf copy of your protocol of Mnase digestion?
Thanks and good luck

Hi,

here is a ChIP protocol using MNase digestion or sonication to shear DNA.

I should mention that I think MNase digestion is ok for histone ChIPs but not for transcription factors ChIPs. I have not tried it but I guess it would be quite inefficient if the TF binds between two nucleosomes or in a nucleosome-free region.

Here is the link: Chromatin Immunoprecipitation (ChIP) Assay.

good luck!

Controller