Problems with amplification using nebnext ultra

[Andrew.G]

Hi everyone, not sure if this has been answered yet.
So I am trying to prepare a library for transposon mapping in a yeast using the nebnext ultra kit. I fragmented the DNA into what should be an average of 350bp in length. While I ran through the procedure I took 1ul out at after size selection after ligation and 1 ul after amplification for a BioA run. I got some strange results.

It seems like my size selection was a little off possibly due to my fragmentation not working the way I thought it did.
As you can see in both samples I had fragments + linkers from 200 to 500bp. indicating a fragment range around 50 to 350, which I may be able to further optimize.

The part that really confused me however is the post PCR BioA. In both samples I had what looked like very little amplification. however I see that I have two huge peaks in both samples one at ~126 and one at ~140. The PCR was run with 10 cycles as the starting material was quite low. I was wondering if anyone had any idea as to what either of these to peaks were. I thought at first they were adaptor dimers but any evidence of them are absent in my post ligation BioA.
Sample 3 and 5 are paired as post ligation and post PCR respectively.
4 and 6 are paired as post ligation and post PCR respectively.
samples 1 and 2 were unrelated to this problem and the other samples are not mine.

Any help would be much appreciated and I can clarify any questions about the procedure or samples if necessary,
Andrew

[nucacidhunter]

120 and 140 bp peaks seems to be amplified products from 41-45 peaks (evident in trace, 3 and 4, most likely excess adapters) that has been carried over from size selection. My guess is that adapters were not diluted. Post PCR clean up should remove those peaks.

[Andrew.G]

The post PCR samples were cleaned up using ampure beads before being run on the BioA. Would you then recommend a second cleanup, and for further runs a dilution of the adaptors for the ligation step?

[nucacidhunter]

I wonder what was the input quantity and if you know the mass of DNA template (size selected adapter ligated DNA) that you used for PCR and also the amplicon. You should have a lot more output than input for the PCR. For input below 50 ng double size selection has not been recommended.

From your traces it is difficult to know the input and output. It would be useful if your traces were marked by the range or just one peak covering the area so I would have a good estimate of input and output.

To get rid of those 120 bp peak I would suggest two 0.9x AMPure clean up, but I think some library prep steps has been sub-optimal that has resulted in those 120 bp peaks.

[Andrew.G]

Do you think a gel purification might work. I have tried one before cutting out everything from 200 to 700bp in length after the ligation step. Even then I still got the peaks at 126 and 139. Could the high concentration of the adaptors in the gel prevented some from migrating fully and cause them to get "stuck" with the higher weight fragments? Also would having a larger fragmentation size and subsequent selecting for a larger size help reduce excess adaptors from being pulled out?


As for inputs and outputs
for fragmentation the input should have been ~500ng and then the whole tube was used for the ligation. After ligation and size selection qubit indicated that about 15ng was the input for the PCR. Post PCR qubit indicated either 80 to 150 ng of output.

Could the USER enzyme also be a problem as if it does not cut correctly then there should be no amplification. Is there a way I could test for this?
Andrew

[nucacidhunter]

One of advantages of the kit you are using is low tendency for adapter-dimer formation. Your PCR yield for 500 ng is also very low indicating sub-optimal reaction(s). Considering using 15 ng input for PCR the ligation efficiency must have been below 10%.

Gel cut will reduce the dimers but because of low conversion rates the library diversity wil suffer. So it is better to optimise the reactions. USER enzyme could be the problem as well as earlier reactions. As for testing USER enzyme NEB tech support might be able to help.

[Andrew.G]

As a followup and as a more general question would increasing the insert size help reduce any possible extra adaptor from going through the size selection?
Andrew

[nucacidhunter]

Yes it would. You can be more aggressive with clean up reducing bead ratio and cleaning up 2x.

[Andrew.G]

Thank you for all of your advice. I will try to work them into my experimental design. Do you have any other general advice in regards to sample preparation?

[nucacidhunter]

If you are doing shotgun sequencing, it has been covered in product user guide. I have prepared libraries using that kit successfully with 1 ng input. If you are doing something else please post it for specific considerations for that application.