[Leeenaaa]

Dear all,

What's your opinion about this shearing (for ChIP-seq with Illumina)? My settings are: suspension cell-lines, 10 min PFA crosslinking 1%, 5 min of quenching with glycine, 30 cycles of shearing (Bioruptor, 30 s on/ 30 s off). Is this shearing good enough for sequencing? I have calculated that approx 54 % of my DNA is at the correct size while approx 43 % is too large. Is it possible to get one peak of 100-500 bp with crosslinked material?

[ALive]

Hi Everyone,

Has anyone ever done a dot blot to check crosslinking and sonication times? While I'm working out the western blot, I thought I'd dot blot sonicated chromatin since I have all of the reagents. Any thoughts?

Leeenaaa: I've read that large fragments can negatively influence the ChIP experiment (see Methods Mol Biol. 2011;791:265-86), but I don't know empirically since I haven't been able to get ChIP to work at all :-)

Thanks in advance,

Ali

[ETHANol]

Leeenaaa,
If your shearing is 'good enough' depends entirely on your epitope. There isn't a one-size-fits-all criteria for ChIP. I would recommend that you do your ChIP under few different x-linking/sonication conditions and chose the conditions that gives you the best enrichment at a known locus where your protein binds. If you do not have a know locus where your protein binds, I would go ahead with a pilot experiment using your current ChIP. Hopefully, that identifies works at least a little and you can optimize from there. With multiplexing, the cost of sequencing a single ChIP and input isn't that much. (And ALive's comments are worth taking into account)

ALive, I'd say just do a western blot of reverse x-linked material. It works really well and you will see the effect of soniciation on your epitope. Western dot blot is a little sketchy with out any size information associated with your signal. Of course you will see the effect of x-linking on your epitope better but I still think the result will be too hard to interpret.

[Hamid]

Dear Alive,

You are certainly destroying your epitope during the shearing process. As a result you are only able to IP and enrich for large fragments that contain intact epitope. A very easy way for you to analyze will be to do a western with the same antibody you are using for your IP. Simply run 15-30ul of your sheared and unsheared sample on a SDS PAGE, transfer to a membrane and do the western analysis. My suggestion would be to also do a time course of your shearing, and load aliquots of those on the SDS PAGE as well.
Dot blotting is not a good idea since most antibodies will have background bands. That will interfere with the analysis of the epitope integrity with the protein of interest. On a western since you are resolving the proteins first, you can just look at the bands that represent the mass of your protein of interest.
I am certain that the western will show you how much of the epitope you are losing during the uncontrolled shearing process.

Thank you

hamid

[ALive]

Hamid,

Should I boil the chromatin (sheared and unsheared) before running on the SDS PAGE gel? I've read contrasting opinions about this...

Thanks,

Ali

[tamara.tjitrowirjo]

size discrepancy between agarose and bioanalyzer

Hi all,

Just found this threat today. So I have been having problems with bioanalyzer as well. I run my sheared samples (probe sonication) on agarose gel to check the shearing since the bioanalyzer always shows very large fragments. Anyways, I just ran my ChIP samples for NGS sample prep, the last step after PCR amplification was gel extraction where I cut the gel and excise DNA between 200-350bp. I purified the DNA using MinElute gel extraction kit from Qiagen, measured the DNA using Qubit (between 0.5-1.5 ng/ul) and run bioanalyzer high sensitivity chip. The bioanalyzer showed totally different size, between 300-3000 bp. There is no way that my samples contain such large DNA fragments as they were freshly gel extracted.

Contacted Agilent already and have not gotten any reply. My conclusion is not to trust bioanalyzer for ChIP samples. Unfortunately I cannot explain why it shows wrong sizing. I've used bioanalyzer for DNA-seq sample prep and they have shown correct sizing (proven by the seq read length). So I know that generally bioanalyzer works well. Both my DNA-seq and my ChiP-seq samples are DNA with Illumina adapter in EB buffer. But somehow they differ. I will run Caliper tomorrow to check sizing again, will update soon.

[tamara.tjitrowirjo]

Caliper results

Hello...

I ran the Caliper (similar technique as bioanalyzer). The result is similar to bioanalyzer result. Meaning that both Caliper and bioanalyzer shows that my ChIP-DNA sample prep are very large between 1000-50000bp even though my agarose gel showed size of 200-350bp.

I still do not trust the bioanalyzer or Caliper since they both use the same technique. Good old school agarose method should be more straight forward.

Any comments?? Anyone that has done the seq part and have proof of how large the fragments are?

Thanks!

[pmiguel]

What method did you use to un-crosslink your DNA after IP?

One possibility is that your DNA is single stranded for some reason. With gel electrophoresis single stranded molecules run faster than double stranded molecules of the same length. But for BioAnalyzer chips the opposite is true. See this thread for details.

Also, obviously if your DNA is still complexed with protein, you would expect aberrant migration during electrophoresis. And, again, BioAnalyzer and Gel Electrophoresis might produce different results under those circumstances.

--
Phillip

[tamara.tjitrowirjo]

Hi Phillip,

Thanks for your input. I reverse cross-link the DNA by adding (Tris, EDTA and SDS) and incubating at 65C overnight. I then proceed with RNase A and Proteinase K digestion, phenol/chloroform purification and then dissolve the DNA in TE buffer. So they should not contain any RNA or protein anymore. About the single strand? I am not sure how one strand can be lost resulting in single strand DNA. I can understand that they might be denatured at one point (perhaps during reverse cross-linking) but then they should come back together again. Especially since one of the step in the sample prep was performed at 20C.

Have you sequenced your samples? I was wondering about the whole sizing problems. If anyone have sequenced it then we would know for sure what the size is.

[pmiguel]

The other strand would still be there, but after dissociating from its complement strand, chances of it ever finding that partner again are low.

I have not experienced exactly the issue you describe, I am just speculating as to possible causes.

Have you already ligated adapters onto your fragments? If so, you could do a little enrichment PCR (not too much, though) and that would recreate the complement strands. What library prep method are you using? TruSeq? There is a phenomenon where ligase remains attached to TruSeq ligation products resulting in abnormally long looking amplicons. Could be that.

--
Phillip

[shanshanzhou]

Hi All,

Thanks for many useful tips. I have a very similar problem with Alive. And I just tried some westerns according to Hamid’s suggestion ( this is the first time I did western but I found a labmate help me out). I think it’s because of the samples are crosslinked, instead of getting a nice single band, I got many bands from very high weight to low weight. I did a time course of shearing using the following conditions with TC12x12 tube:
Duty cycle: 5% Cycles: 4, 6, 8, 10, 12 min
Intensity: 4 Temperature: 4C
Cycles per burst: 200 Power mode: Frequency sweeping
Cycle time: 60 sec Degassing mode: Continuous

I took ~15ul chromatin extract from each sample, boiled and loaded on SDS PAGE. After staining, I don’t really see much difference between these samples other than slightly denser stain on few low molecular weight band as shearing time getting longer. Just like what Alive asked, if high energy of shearing and 1% SDS can disrupt epitope, then can what we see on a SDS western with boiled sample be a good indicator to what we need?

Regardless of what I got from western I did another set of ChIP assay with two antibodies on 8min and 12 min samples (based on bioanalyzer, 4min is still not well sheared). Both samples gave results of fragments ranging from 500bp to 7000bp.
Is my Covaris setting is too harsh that once the chromatin got sheared the epitope got distroyed? Or there are other steps that actually affecting the epitope , or maybe in its not related to epitope at all?

By the way, I use Drosophila whole body tissue, grinded with liquid nitrogen. And antibodies are ordered based on the test results from ModEncode paper.

PLEASE HELP!!!

Best,
Shanshan

[ETHANol]

Hi Shanshan,
You have to reverse x-link your protein before the Western.

Your question are pretty general and vague and I have no idea what kind of protocol you are following so it's hard to give you any kind of answer. I've posted pretty much everything I know about the Covaris and x-linking on my blog ( http://ethanomics.wordpress.com/ ) including, how to prepare your samples for western. Maybe it helps.

[Hamid]

Hi Alive,

We do the following:

1. Prepare fresh sample buffer by adding 50 uL 2-mercaptoethanol to 950 uL of 2x Laemmli SDS sample buffer.
2. Dilute lysates 1:1 with the fresh prepared 2X sample buffer and incubate at 60°C for 10 minutes, and then load on to the gel.

NOTE: Heating the samples to 95°C in Tris buffers (e.g. Laemlli buffer) induces Asp-Pro cleavages of proteins. We have observed this to occur in less than 10 minutes of heating at ≥80°C.

Thank you

Hamid

[Hamid]

Hi shanshanzhou,

I am currently working on getting a protocol optimized for drosophila. Our current protocols, reagents, and settings are optimized for mammalian cells only. Since you are most likely not isolating nuclei prior to shearing, there will be quite a bit of cell debris that will interfere with the acoustic processing.

I would also like to make another comment regarding fragment sizes after IP being larger than fragment analysis on a bioanalyzer after shearing. Please note that the amount of open chromatin varies from cell line to cell line, and from organism to organism. when you shear chromatin there will be quite a bit of the fragments generated from areas not bound or loosely bound by histones. These fragments will shear easily and when analyzed right after chromatin shearing on a bioanalyzer, it will be impossible to distinguish from fragments obtained from sheared chromatin with bioanalyzer.
If your Bioanalyzer after IP is giving you larger fragments, it is likely due to:
1. destruction of the epitope by over-shearing. Over-fixation will certainly contribute to that.
2. Not shearing long enough such that you base your chromatin shearing efficiency based on the open chromatin fragments. This typically occurs in with bioanalyzer traces similar to what was posted by Leenaa where very distinct double peaks are present in the bioanalyzer.

Thank you

Hamid

[shanshanzhou]

Hi Hamid and ETHANol,

Thanks so much for the quick reply!!
ETHANol, I looked at your protocol and compared with mine. After the homogenization steps, they are generally the same except slightly difference of the buffer recipes (by the way, it's a very well written protocol, have seen so many badly written ones~~), but I will probably try to repeat your protocol again, just in case some small step may have big impact.

Hamid, I've seen several your posts on shearing conditions, and they are very useful! I adjusted my conditions according to your suggestions, using 1% fresh formaldehyde with no methanol and fix samples for 5 min at room temperature. The above shearing settings are actually suggested by one of your technician I contacted half year ago. I really try hard to avoid over-shearing. See attachment for bioanalyzer result for 4, 8 and 12 min shearing. I apologize that the test solutions are a little be over diluted, but you still could see 4 min is not fully sheared yet, kinda like Leenaa's double peaks, not as extreme, but 8 and 12 min seems well sheared (let me know if there is any signature of the curve you think is a bad sign). And the ChIP results from both 8 and 12 mins are huge fragments I don't know what else I can do to try, the only thing may worth to try is to do a chip on 6 min sample, but I don't know if there will be difference. Please let me know what do you think!!

Best,
Shanshan