I think you should not worry about losing too much DNA with column purification. I used phenol:chloroform extraction followed by ethanol precipitation in the past to get higher yields. However, in my experience using columns (doesn't really matter which company, we have used Quiagen and Zymo colums equally successfully) is far faster, easier and you do not have contamination issues. Forget Nanodrop for quantification, most of the cases the DNA concentration is way too low. Even with Qubit HS kit the concentration can be below the detection limit (especially in IPs with transcription factors). If the amount of IP-d DNA is a problem, try finding a library preparation kit that uses less dsDNA for input. For example, I have used Nugen Ovation Ultralow DR kit with good results. The protocol is simple and only takes about 5h to complete. The recommended minimum amount of starting material is 1ng, however, it has never failed to give me enough material for sequencing even in cases where I could not determine concentration with Qubit. Of course if you have the material, using more DNA and less PCR cycles is always preferable. Finally, if possible, always check enrichment before library preparation - put crap in get crap out.

By my experience, the DNA concentration from ChIP-seq is always low. I understood that you get 0.5-1 ng/ul (nanodrop). You need to know what is the min amount of DNA that you need to send your samples to sequence, because depend on the system (Ilumina...) with 10ng (total) could be enough.

I am more worried about the purity an the the quantity as I know I would require only aroung 10ng of total DNA. But the purity is very low. I mentioned the 260/230 ratio to be low (0.5-1.0) which represents contamination from some salts, EDTA, Phenol, carbohydrates etc... I want to know if anyone has faced such issues and how they purify their DNA after ChIP..

Need to know the type and scale of your experiment, how many cells, what antibidy (transcription factor, histone modifications) and how much total DNA you end up with. Nanodrop in general is nowhere new sensitive enough for ChIP-DNA quantification at least for transcription factors and at low concentration the 260/280/230 ratios are highly inaccurate. Qubit high sensitivity works, but might use up a lot of your DNA.

We routinely use QIAGEN clean up (heating buffer to 55C prior to elution improves yield greatly), but have now started using SPRI beads. We never look at purity measures for our ChIP-DNA...if you give us a bit more info can advise further

I use le kit "PCR purification kit" from Qiagen. Normally, it gives a very good quality of DNA. Don't forget treatting your samples with RNase after reverse crosslink. I check my samples with Qubit High Sensivity.

Ethanol precipitation is your friend. After elution from the beads, de-cross-linking and proteinase K treatment and inactivation, add equal amound of phenol-chloroform-isoamyl alcohol mix. Shake for 30 seconds, spin at max speed for 5 min at RT and collect aqueous phase. Add 0.1 volume of sodium acetate (3 M, pH 5.2), 3 volumes of 95-100% Ethanol and (optional) 1ul of Glycogen (1 mg/ml). Precipitate on dry ice or at -80 for at least an hour. Spin at maximum speed for 30 min at 4C. Wash pellet (do not resuspend) in 1 ml of ice cold 70% Ethanol. Air dry and resuspend in as much nuclease-free water as necessary. Can't guarantee you'll get a lot of DNA, but it'd be more than from a column and just as (if not more) pure.

I never check the purity of my DNA after ChIP. I usually purify using the ZymoResearch ChIP clean and concentrator kit, try it out and see if it improves your DNA purity. It's advantage is that you can elute in as little as 6ul for increased concentration.

We also use Qiagen colums for purifying ChIP DNA. And it works quite fine. Perhaps try to dry the columns (air dry 5-10 min) before eluting. For conc measurement Qbit is the best. Adding NaAcetae improves the purification efficiency.

I do not know your experimetn setup, but if you can afford to have that many material, then perform ChIP in triplicates and at the end pool all in one column for purification. This gives high conc of DNA.

In my experience measuring DNA concentration after ChIP does not help. You should do a PCR with primers amplifying one of your expected products to check that you have enriched fragments (compare with the imput). Then proceed with library preparation which normally involve amplification... then measure with Qubit your DNA and purity... Good luck

According to my experience, the amount of ChIP-DNA depends on the quantity and variety of cells and antibodies used for ChIP, and the input DNA should be 100-600bp after sonication. Usually, 2μg antibody pulls down 10-20ng DNA for H3K27ac in 5×10^6 cells. To test the quality of ChIP-DNA, ChIP-PCR should be performed before the library construction (10ng ChIP-DNA is enough) with positive and negative primers. Good luck to your experiments.

In addition, nanodrop should not be used to test the concentration (too low) of ChIP-DNA, it is inaccurate as the DNA concentration < 10ng/μl. Qubit is recommended.

By my side I use the iPure kit from Diagenode for the elution step as well as the DNA purification. It work very well and you could adjust the elution buffer volume to increase your concentration.

I think you should not worry about losing too much DNA with column purification. I used phenol:chloroform extraction followed by ethanol precipitation in the past to get higher yields. However, in my experience using columns (doesn't really matter which company, we have used Quiagen and Zymo colums equally successfully) is far faster, easier and you do not have contamination issues. Forget Nanodrop for quantification, most of the cases the DNA concentration is way too low. Even with Qubit HS kit the concentration can be below the detection limit (especially in IPs with transcription factors). If the amount of IP-d DNA is a problem, try finding a library preparation kit that uses less dsDNA for input. For example, I have used Nugen Ovation Ultralow DR kit with good results. The protocol is simple and only takes about 5h to complete. The recommended minimum amount of starting material is 1ng, however, it has never failed to give me enough material for sequencing even in cases where I could not determine concentration with Qubit. Of course if you have the material, using more DNA and less PCR cycles is always preferable. Finally, if possible, always check enrichment before library preparation - put crap in get crap out.

You have to consider that some treatments during isolation of ChIP-ed DNA may damage and denature it to single strand, which is not good if you want to sequence it. For ChIP-Seq I use the iDeal Chip-Seq and library prep kit from Diagenode, the bead purification kit (iPure) is included. Then I measure the concentration with Qubit, using the high sensitivity kit. Then I prepare library with the iDeal library prep, and then check again the concentration with Qubit and analyze the amplicons size with the Bioanalyzer high sensitivity DNA chip. In the iDeal library prep kit you can make a size selection, and isolate the ChIPed DNA with specific size, e.g. 200 bp fragments.

We had same problems as you mentioned in the post. So, we designed a Tini DNA/RNA spin column for recovering and concentrating trace amount of DNA or RNA for Chip and LCM samples. The elution volume for this tiny column is as small as 5ul.

Since Tini columns are also silica membrane based (like DNA/RNA mini columns in the Qiagens kits), all the buffers for Tini spin columns are compatible with the ones for Mini columns. So you can buy bulk Tini columns ($33) and use all the solutions in Qiagen kits. We also have Chip-DNA cleanup kit with Tini DNA spin columns ($55/50preps). If you want to make your own solutions in the kits, please contact [email protected] for the solution recipes in DNA miniprep, DNA gel extraction and PCR cleanup kits. Hope this helps. Good luck!

Tini column: http://www.enzymax.net/columns/tini_spin_DNA_column.htm

DNA purification Kits: http://www.enzymax.net/columns/DNA_purification.htm

Make sure you do the entire procedure in a cold room ~16oC.

Qiagen columns, Zymo or Phenol/ Chloroform should be fine. Since there are couple of purification steps during library prep you do not need to worry about Phenol/ Chloroform contaminiation while it has high ChIP DNA yields. However, your nanodrop quantification, including ratios might be off (Qubit gives you a better picture there). Also, if you are interested in your DNA quality, try out some predicted targets by qPCR.

Tonis Org, you can start library preparation with only 50pg of DNA using the microplex library prep. kit from diagenode but the complexity of your library will be too low as a result of the high number of PCR cycles. Getting a library is a much smaller burden compared to the significance of the sequencing data. With a library of low complexity you will miss the unrepresented binding sites and so the results will be unbiased. I am interested to know if you have successfully made libraries from DNA undetected on HS Qubit and could make biological sense of the sequencing data.

Someone may have already wrote this, but do you rnaseA treat you samples prior to decrosslinking? I'd suggest to add this step if you don't already. I work with very small amounts of chromatin for chip and I get good results with phenol chloroform extraction followed by ethanol ppt overnight. I use glycogen to help pull down my DNA, and it works well, as you can see a "pellet" after you spins, so it's easier not to lose your precious material. If you have access to a bioanalyzer, it is the most accurate way of assessing your sample.

I agree with Tonis statement. Mostly, the Nanodrop shows poor quality of chip DNA and undetectable levels  in specific and sensitive Qubit systems. In our experience, these samples on  PCR analysis shows sufficient enriched DNA. However, I'm sceptical to take this further for library construction. I would consider pooling of Chip DNA prepared with good number of cells before the purification steps to improve the DNA yield. Make sure the final immunoprecipitation buffer contains no more than 0.1% SDS.  I find iDEAL chip protocol supplement buffers to perform chip with high volume cell lysates. This could also help in working with low abundant proteins.

Qubit dsDNA HS kit can detect concentrations down to 10pg/ul, so for example if you elute in 10 ul of buffer your final total amount has to be less than 100pg in order for it to be undetectable with Qubit. 50pg DNA is about 16 full haploid mouse genomes. If the average fragment size in ChIP is about 300bp, there should be about 10 million unique fragments. Let's assume that 10% of the fragments are IP specific, making it 1 million. If there are 5000 TF binding sites in the genome there are theoretically 200 fragments per binding site. So in theory, there should still be enough complexity to get a good library. This is of course theory. In reality, the mapping quality will decrease and the number of PCR duplicates will increase. Nevertheless, although the data won't be ideal it is still absolutely possible to get biologically meaningful data with such a low amount of starting material.