Sonicate for >60min: What did I do wrong?? - ChIP sonication disaster (Nov/18/2010 )

HI All,

I am running into sonication problem back and forth and still find it very difficult to get the optimal size (200-500bp) of fragments for ChIP.

This is what I do: I cross-link cells on a 15 cm dish with 1.1% formaldehyde (directly added to culture medium drop-wise) at RT for 10 min; quench with 1.25M Glycine for 5 min RT; remove medium and wash with PBS 3 times (this may take up to about 15-20min depending on how many dishes that I'm dealing with. I'm not sure if this actually contributes to cross-linking?); then scrape cells off, spin down and lyse in 2 mL IP buffer for sonication (this should be below 10^7 cells/mL).

The cells i'm dealing with right now specifically are MCF7 (breast cancer cells, tend to clump on dishes) but i found similar situation on other cells that grow as monolayer. So after I collected the cells, I took 500 uL lysate and sonicated with Branson 250 microtip sonifier at 50% output to sonicate for 15s with 1 min rest in between for 5, 10 and 15 min of sonication time (which is 25, 50 and 75 min with resting time). I ran the gel and yes the 15 min sonication gave desired fragment size. And then my problem is how to sonicate a total of 2 mL lysates, I can either do 4x 75 min or try to sonicate it together, which was what I did. So i sonicated it for 3x 15 min (that is 3x 75 min total time!) and didn't get the size I want.

I asked myself, have I seen anyone sonicating for a full day? The answer is No; there ought to be other way to do this! and then I asked myself, how did other people do it? Did I do something really wrong? The answer is I have no clue; and so I was wondering, where else can I ask. So i turned here. Please help!

I appreciate all your kind replies in advance! This will make my dissertation!

Very Anxiously,

Shan

Maybe use a lysis buffer containing 1%SDS and just dilute in 0.1% SDS containing dilution buffer. I use 600ul SDS lysis buffer per 1X10^7 cells and found that this works for me......in fact using MEFs I could never get frags below 1-2kb using the IP Lysis Buffer.

I agree with chabraha,

adding SDS to the lysis can help a lot. I have similar problems with sonication, mine being the complete removal of genomic DNA. I always seem to have some left in the sample no matter how long i sonicate for. I am about to try using a brief and mild MNase digestion before sonication to see if this will help, it may also help in your case.

Ben

Thanks chabraha and Ben! I will definitely try that. thinking of directly adding 1% SDS into IP buffer then dilute with IP buffer that doesn't contain SDS. Has this worked for you? Or shall I just lyse the cells in 1% SDS? Since I have been using IP buffer for other cells, I'm trying to change as little as possible. As a side question, can you help me why SDS should help with soication? and what is MNase...?

Thanks a lot!!

Shan

Hi Shan,

I have some concerns over the use of SDS in my sonications, however I fear this may be the only way forward for me as my supervisor is refusing to buy and MNase for some unkown reason. MNase is micrococcal nuclease and cleaves DNA inbetween nucleosomes. It is great for making preps of these and is often used in nucleosome mapping experiments. It can be used in extremely mild contitions to break up genomic DNA prior to sonication. Using MNase alone is not recommended for ChIP as it can introduce some sequence bias into the results (or so Ive been warned).

Ben

Also I don't think MNase works well at all on cross-linked chromatin

@chabraha, really? This is contrary to what I have heard, however I am testing it now. Luckily I dont actually want it to work too well, just break the chromatin up mildly and finish by sonication.

I was under the impression that MNase wasn't compatible with FA but I have never used both so I can't say with certainty. Let me know how your results turn out. I'm curious as to the differences between FA alone and in combo with MNase.

Hi there, so for Shan's experiment and Chabraha's general information:

MNase digestion does work very well following cross linking. I tested using an undiluted (10U/ul) Vs diluted 1:10 (1U/ul) stock on my tough to sonicate chromatin. The first I incubated for 1min at 37C and the second for 6min at 37C. I have attached the gel that I ran after reversing the cross linking. The first lane is the 10U/ul sample and the second the 1U/ul. As you can see the undiluted sample has worked extremely well giving a very good smera for ChIP-qPCR. However as I have mentioned above it has been reported to generate a sequence bias when performing ChIP-seq and so I am now attempting a mild MNase digest (using 1U/ul) followed by a round of sonication. This looks like it will yeild the desired results

Ben

Thanks Ben! and Chabraha for keeping this post alive. Very much appreciated

So unfortunately, i can't see the gel image (only my problem?) but i was wondering if that means Ben you can get the desired fragment size without any sonication? I don't know much about MNase but if that's an enzyme, doesn't it cut at a recognition site? Will it constantly cut at a point, where possibly is my sequence of interest and hence generate false negative?

As a side question, could you explain why you don't like SDS in your ChIP assay? I'm very intrigued as it seems pretty standard in many ChIP protocols.

Thanks!

Shan