Over-shearing? - (Jul/17/2009 )

Hi All -

It looks like probably the most common problem I'm reading about in this forum is that everyone's samples are not sheared enough, yet I seem to be having the opposite problem. Every time I run the assay under a host of different sonication conditions, I seem to get fragments of ~100 bp. I'm barely sonicating my samples (which are from frozen mouse liver biopsies, btw) - I'm doing 1 round of 15x1 second pulses on a Branson Sonifier 350 (Power 1, Duty 20%) in an ice bucket. I've tried both 1.5 and 2.0 mL eppendorf tubes, it doesn't seem to make a difference.

I know something fishy is going on because when I do no sonication whatsoever, I get no DNA, yet increasing sonication (to a point) only leads to more DNA on my gel, but all this DNA is of roughly the same size.

(The protocol I'm using is the "Fast Chip" protocol most recently described by Nelson et al. in Methods Mol Biol. 2009;567:45-57.)

Also, the only success I've had with getting DNA at around 500 bp is by doing a wash with 1% SDS, but I've heard that SDS interferes with antibody binding? Will using SDS completely ruin my sheared chromatin samples for a later IP?

Any thoughts on either issue? Am I making a completely obvious mistake? Or do you think my lab just has the world's most powerful sonicator?

If anyone has any suggestions, I am all ears!

Thanks,

Eric

-awesomedisplayname-

awesomedisplayname on Jul 17 2009, 02:31 PM said:

Ok, let me ask a stupid question right up front. The samples are crosslinked, right? You know that your formaldehyde hasn't gone completely bad (i.e. it's not 10 years old)? Yeah, stupid question but I had to ask.

Second, try treating your chromatin with RNase to see if that 100bp stuff is RNA or DNA. It seems to me that you wouldn't be able to shear smaller than mononucleosomes which will be closer to 200bp.

You are reverse crosslinking and using proteinase K before running on the gel, correct? The quickest way is to dilute the equivalent of 10^6 cells of chromatin in 100ul chelex. Add 20ug proteinase K and digest at 55C for 15 minutes. Then boil for 10 minutes, cool, and run 10ul of the supernatant on your gel (using this amount of chromatin I've found that the smear is easiest to see when running on wells that except no more than 20ul of sample).

Good luck,
Joel

-KPDE-

Joel - thanks for the reply. First off, yup, I am definitely cross linking (1% for 20 mins) and my formaldehyde isn't terribly old (bought in January 09, opened in June 09). And yes, I'm reverse crosslinking with 10% chelex/proteinase k. I haven't tried the RNase treatment yet, but if it turns out that I'm getting RNA, whats my next step?

Thanks again,
Eric

-awesomedisplayname-

awesomedisplayname on Jul 21 2009, 09:28 AM said:

If it turns out the band you were looking at is RNA then I would suggest looking for the DNA smear by playing with the contrast when looking at your gel (it's likely that the DNA smear is much dimmer than the RNA band and you are missing it when you adjust contrast to the RNA band). You could also try loading more sheared chromatin into your well. My guess is that it is there but you are just having a difficult time seeing it.

Joel

-KPDE-

I use Sonics VibraCell VCX 130 and 3 mm step probe. I always get around 150 bp fragments. Please see my picture. The interesting thing is there are one large fragment part larger than 1 kb, and one small fragment part around 150 bp. Does anyone encounter this problem?

-rqliang-

Hi there -

When I had this problem, I followed Joel's advice and found that it was in fact RNA and not DNA that was present around 150-200 bp (test this by treating an aliquot of your "DNA" with RNase for an hour prior to running it on your gel - and thank you again for your help Joel btw!). After a few tries, I realized that I wasn't getting good enough lysis of the nuclei - which I fixed basically by sonicating it at a way higher power/duty cycle and using the right detergent (Igepal and NOT Tergitol). I also throw my samples in liquid nitrogen and do some freeze thaws to help split open the nucleus prior to sonication.

So, why don't you do the RNase treatment to find out if you're getting RNA and not DNA? And if it turns out that you are just getting RNA, then try increasing your sonication time/power and do some freeze/thaws to help split open the nucleus.

rqliang on Aug 19 2009, 05:17 PM said:

-awesomedisplayname-

I ran the gel after I treated my sample with RNase. Anyway I will try. Thank you very much.

Hi there -

When I had this problem, I followed Joel's advice and found that it was in fact RNA and not DNA that was present around 150-200 bp (test this by treating an aliquot of your "DNA" with RNase for an hour prior to running it on your gel - and thank you again for your help Joel btw!). After a few tries, I realized that I wasn't getting good enough lysis of the nuclei - which I fixed basically by sonicating it at a way higher power/duty cycle and using the right detergent (Igepal and NOT Tergitol). I also throw my samples in liquid nitrogen and do some freeze thaws to help split open the nucleus prior to sonication.

So, why don't you do the RNase treatment to find out if you're getting RNA and not DNA? And if it turns out that you are just getting RNA, then try increasing your sonication time/power and do some freeze/thaws to help split open the nucleus.

-rqliang-