I recently did chromatin extraction from brain nuclei samples. Since the individual sample amount is too small, I pooled 10 samples together (~40 mg). When I used the Epigentek Chromatin Extraction Kit, it is suggested on their protocol to shear at 2*20s and there was detectable dsDNA concentration right after chromatin extraction. Following their instructions for reverse cross-linking to get input DNA (pre-mixed solution from the kit with 2.5 uL proteinase K, 15 min incubation at 65 C then 10 min at 95 C), there was no band showing up and the dsDNA concentration after this input DNA extraction was also very low (~10 ng/uL on average and the loading amount on the gel was ~80 ng). I then tried with the homemade recipe for chromatin extraction (freshly made 1% formaldehyde cross-linking for 10 min rotation at RT, then quenched with 0.125M glycine for 5 min rotation at RT; lysis [0.01M pH8.0 Tris-HCl, 0.01M NaCl, 0.2% NP-40 with protease inhibitor cocktail] using needle (referred to the protocol from ThermoFisher website); extraction [1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS and 0.004% sodium azide in 1x PBS] and sonicate for 10 min (30s on, 30s off on ice). The input DNA was then reversed cross-linked with 6 uL 5M NaCl, 2 uL RNase A and 2 uL proteinase K at 65 C for 4-6 h, phenol-chloroform-isoamyl extracted with ethanol precipitation at -20 overnight and eluted in water. There was no detectable dsDNA right after chromatin extraction but the concentration for input DNA was over 1000 ng/uL but the ratios were only about 1.4 as when using the kit. Using about 120 ng DNA on a 1.2% gel, there was no band at all. I am very confused by these results. Have any of you confronted with such scenario before? Any suggestions? Thank you!
Hi Yang,
I would suggest you to try some other cross-linker and also try one round of chromatin prep without cross-linking. That will make sure that your protocol or reagents are working fine and there is no problem with that. I have seen cases where cross-linking itself creates problems in some way. As far as the concentration is concerned, how are you quantifying your DNA? Is it using Nanodrop? If yes, then I would say Nanodrop is not an accurate estimation of your DNA concentration, even when you have good 260/280 ratios. So, you could try Qubit for the quantification. It will give you better estimation.
Best,
Vikas
Hi Vikas,
Thank you very much for your kind suggestions! That is also what I am thinking to do a non-cross-link control, but if this is the case, what other cross-linker instead of 1% formaldehyde would you suggest? Also, I indeed measured DNA concentration by Nanodrop, and I also read several suggestions talking about Qubit system. However, we do not have that in our lab. Is there any other common approaches as a substitution? I also doubt about the DNA concentration, but concerned more about the band on the gel since this is really a basic approach to detect the sheared DNA fragments aside from concentration measurement. Do you think the concentration is actually too low to be detected by the gel? If so, why even the non-sheared control did not show a big bright gDNA cluster at the top?
Thank you so much!
Yang
Is it possible that you either over or under fragmented your dsDNA samples? Regardless of what a kit tells you to do, different methods and different instruments are going to give wildly different shearing efficiencies of chromatin. You really should run a test experiment with different times of sonication, if that's the method you are using (i.e. 1x10min, 2x10min, 1x20min, etc., etc.). If you over fragmented, it is possible that the DNA ran right through the gel or that the fragments are too small for gel detection. If you under fragmented, I would imagine you should still see a band unless there is a column step involved, which might make you lose DNA fragments that are too large. In my personal experience, *almost* all ChIP problems can be traced back to bad chromatin shearing, so it is essential you get that optimized to have a successful protocol.
I also agree with Vikas that a Qubit is basically essential for quantification of dsDNA for the purposes of ChIP. You can also try using some of the fluorescence based quantitations like PicoGreen from Promega, but I don't know how well those work. Nanodrop is almost certainly not going to help you out for this experiment.
As a final note, if input DNA you are reading is 260/280 of 1.4, it is very hard to trust that you are actually quantifying DNA, which could account for why loading so much DNA according to your quantification isn't giving you bands. Have you considered using a DNA column to purify the DNA instead of an ethanol precipitation? In my experience, those are more straightforward and yield more consistent results.
Best,
Jim
Thank you James Ropa for your kind reply! I indeed tried the sonication/shearing under several different conditions (sonicate frequency and duration) with a non-shear control, and all the previous DNA extraction did not include a column step, but none of these ever gave a fragmented/smear DNA band on the gel, nor a bright gDNA cluster. I agree that a ratio of 1.4 is definitely not trustworthy and that is why I doubt if no band on the gel is purely an indication of no successful extraction of chromatin, which results in little undetectable amount of DNA after reverse cross-linking. I will give the column DNA purification a try after reverse cross-linking next time.
For the fluorescent-based DNA quantification, I just want to know if my DNA concentration is really that low, does it necessarily mean that my chromatin concentration is also low? If so, then would such quality and quantity of chromatin be good enough for the downstream immunoprecipitation?
Thank you very much!
When you de-crosslink the DNA at 95C for 10 minutes, is will most likely be completely denatured and converted to single stranded DNA. This would explain why there is no visible band on the gel. Try to lower the temperature for reverse crosslinking to 80C. You may have to extend the reaction time with the lower temperature.
Yang Xiao, I hope you have found some solution for your problem and you were able to rectify it. For cross-linking, people have tried some other cross-linker, but always with formaldehyde. I am not sure how it helps but you can give it a try. You can find some answers for that here: https://www.researchgate.net/post/Does_anybody_know_of_an_alternative_to_formaldehyde_formaline_in_Chip-Seq_experiments
For the quantification purpose, if you don't have Qubit, you can also simply do a PCR to check if there is any DNA in your preparation. If you want to quantify the amount, you can do qPCR which will give you a fair idea of DNA amount. You can check the website of the product that you use for qPCR for more details on this. I am sure all of them have some guidelines. However, it does not tell anything about the quality of your sample. So, I will also check the samples using Nanodrop for quality.
Hope it helps.
Best,
Vikas
- Badges
- Science topic