Genomic DNA fragmentation

[FredericRaymond]

Hello,

We will soon do our first sequencing run on our new HiSeq 1000 system. In the past, we analyzed NGS data, but we are yet to prepare a library ourselves. I am currently looking into fragmentation methods, mostly to fragment bacterial DNA for metagenomics studies. We have no Covaris system available, so we are not able to follow the protocol suggested by Illumina. However, some collegues have told me of another fragmentation method they used for hybridization. Basically, DNA to fragment is mixed with small glass beads in a 2 ml tube. The tube is then immersed in a sonicator bath for the appropriate time to obtain the right fragment distribution. I am told this protocol allows to generate a homogenous fragment distribution around 400-500 nt.

I was wondering if this protocol could produce fragments suitable for library preparation (if the illumina end-repair protocol is used afterwards) for illumina thru seq sample prep protocol.

Thank you,

Frédéric Raymond
Centre de recherche en infectiologie
Université Laval, Québec, Canada

[Heisman]

400-500 bp should be fine, although I'm not completely familiar with TruSeq. I'd be surprised if it wasn't fine, though.

[pmiguel]

TruSeq uses the canonical blunt/phosphorylate (presumably with T4 poly/T4PNK) followed by A-tailing (klenow, I presume) "end polishing" methodology. So, as long as the beads don't convert most of the DNA to single-stranded, etc, it should work.

Another option would be nebulization. (I think this is actually called for in the TruSeq protocol.) This just requires a N2 tank, an appropriate pressure regulator and consumable nebulizers you can obtain from Illumina.

--
Phillip

[lterhune]

We have been using the Diagenode Bioruptor routinely to obtain 400-500bp fragments. We previously used Fragmentase (NEB) before receiving our sonicator, which is a good alternative if you don't want to buy a new instrument. Both work fine, although the sonicator gives more reliable results. In our experience, the Fragmentase varied more with different DNA samples.

[coffee52]

hi Phillip, I want to know how to test the single-stranded DNA after fragmentation. do you have any good methods? Thanks!

[Liting]

We have been using the Diagenode Bioruptor routinely to obtain 400-500bp fragments. We previously used Fragmentase (NEB) before receiving our sonicator, which is a good alternative if you don't want to buy a new instrument. Both work fine, although the sonicator gives more reliable results. In our experience, the Fragmentase varied more with different DNA samples.

hi Iterhune,we recently used Bioruptor to creat genomic DNA fragments about 300-400bp,but it doesn't concentrate. Can you give me some advice that I can get a concentrated fragments.The Bioruptor in our lab is Diogenode UCD 600.Thank you very much!

[pmiguel]

hi Phillip, I want to know how to test the single-stranded DNA after fragmentation. do you have any good methods? Thanks!

Probably the easiest way would be to check your sample concentration with a fluorimeter using a ds-specific fluor (eg, picogreen) before and after fragmentation. If the amount of dsDNA detected in this way decreases dramatically, then you would suspect there might be an issue.

However, there are caveats to the method. First, we have never tried it. DNA libraries seem pretty easy to make--no reason to look. Second, if your fragmentation method converts a large percentage of your sample to mono or oligo nucleotides, then this will also be detected by pre/post fluorimetry. So it would not distinguish between the two possibilities.

--
Phillip

[rnal]

We also have the problem that the fragments it doesn't concentrate after Bioruptor. we get fragments about 200-800bp after Diagenode Bioruptor UCD 200.

[coffee52]

Probably the easiest way would be to check your sample concentration with a fluorimeter using a ds-specific fluor (eg, picogreen) before and after fragmentation. If the amount of dsDNA detected in this way decreases dramatically, then you would suspect there might be an issue.

However, there are caveats to the method. First, we have never tried it. DNA libraries seem pretty easy to make--no reason to look. Second, if your fragmentation method converts a large percentage of your sample to mono or oligo nucleotides, then this will also be detected by pre/post fluorimetry. So it would not distinguish between the two possibilities.

--
Phillip

I see. Thank you very much!