fast Chip protocol..high MW bands and crabby 260/230 ratio.. - (Aug/05/2010 )

Hi there,

I am still relatively new in this forum so bear with me if i don't describe my problems very well...

I am starting some ChIP experiments lately and trying to follow the fast ChIP protocol using Chelex extraction but i am getting funny results. There were different combination of treatments that I used including #1 sheared DNA + EtOH precipitation then 10% Chelex extraction by boiling at 100C for 10 min, #2 sheared DNA + O/N reverse cross-link, EtOH precipitation + 10% Chelex extraction as stated, #3 #1 + prot. K treatment and #4 #2 + prot. K treatment. When I ran these samples on a gel, I saw a smear of from top to bottom of the gel in #1 and #2 ; after Prot. K treatment, the very top of the smear disappeared but #3 gave smears from 12K to the bottom and #4 gave smears from 1.6k to the bottom . These samples were all from one single tube of sonicated fragments. Does it mean my Chelex boil did not reverse cross-link well ?

I also nanodrop my samples and my 260/230 ratios were under 1 with or without prot. K. puzzled...

Does anyone have idea why I'm seeing what I saw?

Thanks

Shan

Shan on Thu Aug 5 18:23:39 2010 said:

Just out of curiosity, when do the proteinase K digestions happen in each of the conditions. I'm just curious if the digestion happens in the presence of chelex in one condition but not the other.

I wouldn't worry about the 260/280 ratio. Chelex doesn't actually remove anything from your DNA samples other than multivalent cations like Ca2+ and Mg2+ (that, and the raising of the pH above 10, is the main function of chelex in the assay). There should still be a lot of short peptides and amino acids left from the prot k digestion which will throw off your 260/280.

Finally, the EtOH extraction isn't necessary. We put it in the protocol because we were worried about some of the contaminants (protease and phosphatase inhibitors, etc.) causing problems with PCR, since there is no DNA cleanup step in the protocol. I have gotten fairly reliable dilution curves (even including a 1X point) in PCR without doing the EtOH cleanup on input samples.

Joel

KPDE on Sat Aug 7 04:00:41 2010 said:

Thanks Joel! It's great to have the very original author for advice, wow~

I guess my description wasn't very accurate. In fact, I did the chelex extraction on both the non-decorsslinked and decrosslinked samples. After running a gel with those, I directly did prot. K treatment to both of them and run another gel. I think maybe the prot. K digest the high MW bands (possibly DNA with protein??) so they disappeared. But what's troubling to me is that the one only with chelex runs higher than the one with decrosslink and chelex (they both smeared from top to bottom of the gel before prot.K treatment). I don't know how that happened

In addition, my 260/280 raiots were kind of reasonable, around 1.7-1.8 but my 260/230 ratios were horrible with a really high peak at 230. I am concerned if that contributes to any of the results that I have on the gel and would interfere with later PCRs...

And thanks for the advice on the EtOH clean-up, that's really good to know. I was actually thinking if I could use a 20% Chelex and add equal volume to extract so i can have concentrated samples for loading on gel for optimizing conditions. But I'm not sure if the chelex concentration is really crucial? Any suggestion on this?

I really appreciate any advice that you will offer

Thanks!!

KPDE on Sat Aug 7 04:00:41 2010 said:

Oops, I did the prot.K treatment after I boiled the samples for the first chelex extraction (EtOH precipitate-chelex-boil-load gel(#1 and #2)-add prot.K-45C 1h-boil-load gel(#3 & #4)).

Shan

Shan on Sat Aug 7 19:48:18 2010 said:

Sorry, should have read your post a little more closely. I was going to suspect the proteinase K digestion but it looks like, both for the sample with previous 65C incubation and the one without, that they were boiled in chelex after, followed by prot K, so there should have been no difference in efficiency of digestion. Instead it looks like you may not be getting complete reversal of crosslinking with the boiling in chelex alone. Any reversal that you do get may be sufficient for PCR however (just not looking at your fragments on a gel). I would try both samples (the one with 65C o/n and the one without) in PCR with the primers you want to use.

As for the 260/230 ratios, the thing that might be throwing those off could be breakdown products of PMSF, which has a phenyl group. Just a guess anyway. In my case, I don't find significant inhibition of PCR when running both the straight extract or ChIPed DNA so I would try running PCR before worrying about it.

Hope things work out but let me know if they don't.

KPDE on Mon Aug 9 17:51:24 2010 said:

Thanks KPDE!!

Well it was Friday so 280 or 230 should not be much difference

Back onto the ChIP, I was suspecting my samples were not decrosslinked well too but I am not sure if this is in general how it is or it was a unique situation? How can I improve reverse cross-linking efficiency using chelex? I guess I am asking more of how chelex works... but it is a great idea to compare the two samples at the PCR level. But I am really concerned the resolution of the pull-down given the large fragment.

With that said, I am also wondering what fragment sizes do you use (200-500bp?) if somehow the fragment size after chelex extract do not reflect the real sizes of the fragments? I was trying to optimize my sonication conditions, but I think I was getting similar fragment sizes for 4, 6 and 8 rounds of 15 1s sonication.

Thanks again and I really appreciate your answer!

Shan

Shan on Mon Aug 9 21:17:43 2010 said:

It appears that chelex works by protecting the DNA at high temps (both by raising the pH to 10 or higher and chelating Mg2+ which can catalyze autolysis of DNA). The high temp (100C) in the protocol is used to speed up the reversal of crosslinking. Chelex can be replaced by 25mM tris (pH 10), 1mM EDTA.

With regards to fragments sizes, it seems like your fragments are small enough (as 65C reversal of crosslinking gives a low MW smear). The fragment sizes seen in the sample boiled in chelex without the prior de-crosslinking may not be the true sizes, and thus won't effect your resolution.

Cheers,
Joel

KPDE on Tue Aug 10 08:41:30 2010 said:

Thanks Joel!!

I repeated the experiment omitting the EtOH precipitation step and it helped the recovery yield of DNA! Pretty happy about that.

As far as the reverse cross-linking goes, I am still getting something similar (with chelex alone, the smear was from 12K to bottom while with O/N 65C followed by chelex to boil and inactivate Prot.K, the smear was from 1.6K to bottom. ) So from what you recommended, it seems OK to proceed? I still don't quite get how Chelex reverses cross-linking but does not show the right size....because some protein fragments still hang on the DNA or the sizes got messed up during the boiling and renaturing?... In that case, what sizes of DNA that you use for your ChIP? I'm very worried that I'm going to get not repeatable or uninterpretable data...

Thanks...

Shan

I'm a little late for this post but I was wondering if anyone had an explanation for the aberrant running size of DNA using the Chelex reverse cross-link method? I also have run into a few strange diagnostic gels when i check my sheared chromatin. First, (as per munkeyspunk) I see a low MW smear (not RNA as I RNase treat my samples) in the range from 75-200 and then another high MW band in the range of 1-2kb. munkeyspunk seemed to correct this problem by performing a revers cross-link step before the ProK treatment.........but he did not state whether this was the chelex method or not. Secondly the apparent inefficiency of the Chelex method at reversing cross-links discussed in this thread has me worried that I may have sufficiently sheared chromatin but not be aware of it. Is there advice on how to ensure using Chelex that you reverse cross-links efficiently? If not what are the downstream effects on PCR.........should i be concerned about the incomplete reversal of cross-links for my PCR? Or for that matter the difference in DNA appearance after Chelex (not a good smear, but rather two smear)? Any advice is greatly appreciated as I am concerned with reproducibility of results.