ChIP-seq library prep failed, NEBNext ULTRA DNA kit

[DinaHer]

Hi everyone,

I encountered an issue with histone modification ChIPseq library prep using the NEBNext Ultra DNA lib prep kit. Hope anyone here could help to troubleshoot this.

1. I sheared my DNA using a Covaris S220 sonicator, the fragment size after sonication was 300-400 bp.
2. Immunoprecipitation was done using unmodified H3 (Abcam, H3K4me3 (active motif), and H3K27me3 (active motif) antibodies.
3. I got a low yield of ChIP-enriched DNA from the H3K*me3, with a concentration range from 2-5 ng.
4. I then decided to use 2 ng of DNA for preparing the library. I optimized the PCR cycles prior to the library preparation (using the input DNA) and found that 11 cycles are suitable for 2 ng of starting material, i.e. library final conc. around 10-20 ng/uL, with average fragment size 600-700 bp.

When I did the lib prep using the ChIP-enriched DNA, I only got good results from the input and unmodified H3, with average fragment size at around 600 bp and concentration of 10-20 ng/uL. The H3K*me3 libraries concentration was very low, i.e. 0.08 - 0.4 ng/uL. I dilute my library to a final concentration of 30 uL.

I did the same experiment previously with a lower amount of starting material and using different library prep kit and got the exact same pattern: no problem with input and unmodified H3 samples and a lower concentration of H3K*me3 libraries.

Any ideas about why this could have happened, or how to troubleshoot it would be greatly appreciated.

PS: I use the ChIP technique to study the developmental process in a non-model plant. It is a bit of a challenge for me to check the enrichment with PCR due to minimal information regarding potential target loci/ gene-specific involves in the developmental process of this plant.


Cheers,
Dina

[KB*]

Hi,

Sorry, it is a bit late reply. I hope you solved your problem.

I used NEBNext Ultra DNA lib prep kit and it worked OK for me. I guess you might not have enough DNA to start with. Did you measure your chiped DNA yield?

At what point you did your size selection?

I had a very limited amount of DNA from my chip - estimated previously. I knew that DNS fragments distribution is not optimal. So, I skipped the first size selection, but just removed adapters. Also, when I got my final chiped DNA samples I did not have extra for measuring DNA concentration and quality check. So, I blindly prepared libraries with everything I had.

What else u can do. After each PCR cycle (at lest 3 for the best of my memory), you take a sample for tapestation and Qubit. The will allow you to assess primers amount left and to know if you have to do one more PCR cycle.

Hope this helps

[DinaHer]

Hi KB*,
Thank you for your reply.

Yes, I measured the DNA with Qubit before preparing the library. I then took 2 ng DNA each from the input, unmodified histone H3, and the modified histone H3 for preparing the library.

The supplier suggests following an optimized protocol for ChIP experiment (attached the protocol). It uses 0.9X volume of beads to clean the adaptors without size selection. Did you use the same amount of beads in your library preparation?

Thanks,
Dina

[KB*]

Hi, Dina,

yes, I used 98 uL (0.8X), just as it is described in the kit manual.

I am a bit concerned about your fragments length. Are not they a bit longer than it is needed? E.g. 600bp. When I was preping my library, I did not know what platform I would use, so I did not know what would be the optimal fragments size. Later on it became a problem as I needed longer inserts. When you know what platform you are going to use, you will know the optimal inserts size and so your library fragments size. This will determine your last size selection step (there is a table in the manual kit for the 1st size selection). You shall have a similar table for the beads your are going to use - to decide what fragments to leave for sequencing.
Well, some people would recommend to sonicate your chiped DNA to get the range of fragments allowing you not to remove a chunk of longer fragments.
Hope this helps,
K