low total/input DNA levels in ChIP work-up - (Nov/19/2009 )

Hi all,
I am beginning to work-up the fast ChIP protocol but am having difficulty at one of the first steps - perhaps this doesn't bode well?
I took an aliquot of sheared chromatin before the IP step and extracted the DNA using the chelex protocol given in the Nature Protocols paper for Fast ChIP. When I ran this on a gel, I saw a bright low MW band, and a very faint smear of very high MW. After RNAse treatment the low MW band disappeared, as expected, but there is very little DNA on the gel, if any (if I squint there may be something there, but it could be my imagination).
I used a relatively confluent plate, so should have approximately the 10 million cells/1ml recommended in the protocol. I had reasonably sized cell and nuclear pellets, and small but visible DNA pellets after extraction. Could I have not re suspended my DNA sufficiently in the Chelex?

Can anyone give me some ideas as to where my loss of DNA may have occurred?
Many thanks...

labchef on Nov 19 2009, 12:49 AM said:

Hi,

Sorry about the difficulty. When I initially wrote up Fast ChIP I included the EtOH precipitation step in the input DNA extraction because I was worried that the various inhibitors (protease and phosphatase) would be inhibitory to PCR. It turns out that this is not an issue and you can dispense with the precipitation step entirely. It is likely that after EtOH precipitation you are having difficulty redissolving the pellet which is difficult to tell with all of the chelex in the tube. In any case, you can take 10µl or so of your chromatin and add it to 100µl 10% chelex plus 20µg proteinase K and do the digestion and boiling steps directly, no precipitation necessary. It has worked well in my hands as well as a few others I've taught the method to.

Hope this helps and let me know if you have any other questions.

Joel

Thanks for the tip!
I re-extracted, omitting the EtOH precipitation, and got slightly better yield and minimal RNA. There is still very little DNA coming out though, which makes me concerned I will miss low-abundance fragments after IP. How concentrated can you make the initial re-suspension of pellet in IP buffer without incurring problems? My sonication step isn't optimal; could this also be contributing to the aparrent low levels of DNA?
Also, the proteinase K treatment is described as optional. What factors would make this more or less necessary?
Thanks again...

labchef on Nov 19 2009, 10:15 PM said:

I'm just curious, when you centrifuge the chromatin right after sonication, what is the size of the pellet you get with respect to the original cell pellet. I typically get a pellet which is less than 10% in size of the original pellet. Anything more than this is likely cutting into your yield and requires more sonication. I would suggest sonicating in volumes of 500µl or less by dividing up your 1ml resuspension into 2 or 3 aliquots (wouldn't suggest going below 300µl or so with most microtips, as foaming starts to become a problem). If you feel that you're needing to sonicate for too long to decrease the size of the residual pellet you can reduce the amount of cells you use per volume of buffer, since decreasing the viscosity of the cell suspension increases the efficiency of sonication. I know it sounds counterintuitive to dilute your cells further, since you're trying to increase your DNA yield per volume of chromatin, but if you're not sonicating efficiently then you're losing yield.

As for the prot K treatment, this is definitely necessary for determining fragment size on a gel but for PCR it may not always be necessary. I can usually get amplification with most primers if I don't include the proteinase K step. However, there are some runs where it seems like the digestion makes a difference (I didn't control everything perfectly so it could have been that other factors were involved). Also, in the original Fast ChIP paper in NAR we showed that, while the digestion had little effect on the enrichment of the well expressed egr-1 gene, it made a large increase in the enrichment of the silent b-globin gene. Since the proteinase K digestion doesn't take very long and involves almost zero labor, I always include it.

Let me know how it goes,
Joel

Hmmm. My pellet after sonication is definitely much smaller; hard to estimate but 10% of the original sounds about right.
However, whilst I am definitely seeing a DNA smear it is very very faint, and even after eight rounds of 15 1 sec pulses of sonication it extends all the way up the gel. I suppose I'm primarily questioning just how faint the bands should be? I have proteinase K treated so doubt this is the source of the problem. could my DNA be degrading?
And I thought these would be the simple steps of the ChIP!

labchef on Nov 25 2009, 03:44 AM said:

I wouldn't worry about the smear being faint. I always noticed a much brighter RNA smear than DNA smear. And also, since your DNA is stretched out over the length of the lane it's not as concentrated and thus, not as bright. The bigger problem is that you have so much high molecular weight chromatin even after prot K digestion and crosslink reversal. You might try a few different dilutions of your cell pellet and see if diluting results in more efficient shearing.

With regards to sonicating being the easiest step in ChIP ... it isn't. The rest of the assay is pretty easy as long as you have an antibody that works . In fact, your incompletely sonicated chromatin would probably work pretty well except that the spacial resolution wouldn't be so great (you'll get longer tails on the peaks of binding of your factor).

By the way, I highly suggest using a "positive control antibody" the first time you do ChIP. That would be an antibody which is known to work well and gives you a preditable pattern, like the Pol II CTD antibody 4H8 (Santa Cruz sc-47701; Abcam Ab5408) or H3K4m3 (Abcam ab8580). When using these antibodies, primers to the transcription start site of an active gene vs. primers to an intergenic region should give clear differences in enrichment.

FYI, H3K4m3 (Abcam ab8580) is not currently available from Abcam.

Mushin on Nov 26 2009, 07:53 AM said:

They're still out of it? Wow. Are you aware of a good replacement?