checking sonication using chelex and agarose gel - (Nov/16/2010 )
Hi there,
I have a basic fundamental question relating to checking the size of shearing using the chelex purification method described in the fast ChIP protocol.
If the DNA purified by this method is mostly single stranded DNA will this not run differently on an agarose gel? Is it not difficult to tell the size of the shearing in samples prepared this way? Any help understanding this would be extremely well received.
Ben
I have seen a difference in a side by side comparison using Chelex vs. regular de-crosslinking and extraction. THe Chelex method usually gives me larger fragment sizes than when I use the more traditional method, but it doesn't worry me, ChIP using the Chelexed material works great for qPCR and since I did a side by side comparison I know for example that a 1-2kb Chelex smear = a 300-1000bp traditional smear. I am not exactly sure why though. Maybe you get partially hybridized DNA concatamers so your DNA length looks longer than it actually is? Or it could be that the reverse cross-linking just leaves a greater peptide content in the sample and that abrogates the DNA migration in the gel? I'm not really sure though.
OK thanks chabraha,
I am looking to do ChIP-seq, but only want to use the chelex method to speed up my quantification of shearing step. So, if it cannot acurately tell me the size of my sonicated chromatin it is pretty useless to me. From what you say I guess I'll have to try traditional and chelex in parallel and work out what size smear with chelex coresponds to the correct smear determined with o/n cross-link reversal.
Ben
the semi-fast (faster than traditional) reverse cross-link I perform may help combine speed and accuracy:
Spin sheared chromatin down at 14,000xg for 10min
Take 100ul of sheared sample & combine with 200ul 1.5X Szak's Crosslink Reversal Buffer
Incubate 1hr @ 100C
Let cool and add 20ug ProK, incubate at 57C for 1hr
Extract once with Phenol:Chloroform. Transfer 250ul to a new tube
Add 10ug glycogen, 1/10vol NaAcetate, and 2.5vol ice cold EtOH
Spin at max speed at 4C for 20min
Wash pellet with ice cold 70%EtOH
Resuspend pellet in 60ul Tris pH8
Load 40 and 20ul on a 1.2% agarose gel
Oh yeah, and when your precipitating your DNA don't put it in the freezer to help precipitation, you'll cause the SDS to come out of solution. You will get plenty of DNA by just adding the ice cold EtOH and then spinning down at 4C
Ah ok, thanks for that Chabraha,
I'll give it a try. Dont want to appear ignorant but what is Szak's crosslink reversal buffer? Do you have the recipe by any chance?
Ben
Not to worry I found the paper. Thanks for the info.
Ben
chabraha on Tue Nov 16 16:42:43 2010 said:
the semi-fast (faster than traditional) reverse cross-link I perform may help combine speed and accuracy:
Spin sheared chromatin down at 14,000xg for 10min
Take 100ul of sheared sample & combine with 200ul 1.5X Szak's Crosslink Reversal Buffer
Incubate 1hr @ 100C
Let cool and add 20ug ProK, incubate at 57C for 1hr
Extract once with Phenol:Chloroform. Transfer 250ul to a new tube
Add 10ug glycogen, 1/10vol NaAcetate, and 2.5vol ice cold EtOH
Spin at max speed at 4C for 20min
Wash pellet with ice cold 70%EtOH
Resuspend pellet in 60ul Tris pH8
Load 40 and 20ul on a 1.2% agarose gel
Now that I'm using different cell types I'm noticing the same thing you did, that boiling in chelex is insufficient for reversal of crosslinking if you want to check sizes of DNA fragments on a gel, even though it still works fine for PCR.
I just looked at the reverse crosslinking buffer in Szak's paper and I'm curious what your yield is using it. I would have expected the low pH (6.8) would cause nicking of the DNA. We were never able to get amplifiable PCR product when boiling in pHs below 8-8.5 and didn't get anywhere near 100% yield unless the pH was above 9.8. I always assumed this was because the DNA was getting badly nicked.
That's in fact very interesting. I had the same question as Ben before. Later on I ran a side-by-side comparison and what I found was chelex extracted DNA ran smaller (100-500bp) on the gel as apposed to 65C O/N (100-850bp). I can't tell if that's because of incomplete reverse X-link; although the DNA from 65C O/N did ran nicely into the gel.
chabraha on Tue Nov 16 15:39:13 2010 said:
I have seen a difference in a side by side comparison using Chelex vs. regular de-crosslinking and extraction. THe Chelex method usually gives me larger fragment sizes than when I use the more traditional method, but it doesn't worry me, ChIP using the Chelexed material works great for qPCR and since I did a side by side comparison I know for example that a 1-2kb Chelex smear = a 300-1000bp traditional smear. I am not exactly sure why though. Maybe you get partially hybridized DNA concatamers so your DNA length looks longer than it actually is? Or it could be that the reverse cross-linking just leaves a greater peptide content in the sample and that abrogates the DNA migration in the gel? I'm not really sure though.
KPDE,
I agree that the DNA is probably being nicked during using Szak's reverse crosslink buffer. I wouldn't be inclined to use that buffer on ChIPed DNA, just solely for checking the DNA fragment length. I wonder if the Beta-ME in the buffer can protect from the nicking? Nevertheless, the Chelex works great for ChIP and I will continue to use it for that but using Szaks Buffer is a nice compromise for me to get the speed and accuracy for DNA fragment analysis. I am using MEFs in my ChIPs right now.......just thought I'd throw that in for comparison since you were experiencing the Chelex issue in other cell types. As far as the DNA yield it's hard to compare since I only use around 20ul when i was using Chelex to check the DNA fragment size. But with Szaks I get plenty of DNA to check the fragment sizes.......I always load 40 and 20ul on a gel just to make sure i don't overload the gel.