gDNA purification - anything better than alcohol precipitation? (Jan/29/2009 )

I need to produce LARGE amounts of gDNA for whole-genome sequencing from very small organisms. My extraction protocol works rather well (about 1.5 microGrams DNA per miliGram tissue), but I've lost a LOT of DNA during precipitation (more than 40%!). I use the "standard" Sodium Acetate - Ethanol protocol, no glycerol (should I?), >7 hours at -20°, then 20 minutes centrifuge at >14K G, 6°C; then twice wash with EtOH 70%, resuspend in TE. The quantities I'm obtaining (prior to precipitation) are about 6 microGrams per tube.

I've searched quite a lot in the Internet/Books/etc, but found no substantial improvements, other than using glycerol, but this seems to be mainly to be able to see the pellet. Is it so?

I start with quite pure DNA (A 280/260 about 1.8-1.9), in TE. The concentration is around 0.03 microGrams per microLiter, usually 200 microLiters.

I need to get maximum 100 microLiters, at 0.1 microGrams per microLiter.


I would GREATLY appreciate if you could suggest a better protocol/kit to concentrate DNA.

Thanks!

I do not know what kind of tissue you are working with, but I get higher yields of DNA when using only Isopropanol to precipitate the DNA (then wash with 70% EtOH)....

Or you try a Whole genome amplification protocol (like MDA) to amplify your gDNA. I have worked with the repli-G Kit of Quiagen and got good results....starting with the DNA of app. 100 cells, after using the Kit I had app. 50 µg of DNA at the end.

gebirgsziege on Jan 29 2009, 03:06 PM said:

I'm using very "watery" tissue from small worms, they're probably 90% water, so my numbers are probably quite underestimated.

gebirgsziege on Jan 29 2009, 03:06 PM said:

By "using only Isopropanol" I guess you mean after your lysis and extraction... or do you just wash the DNA out of the tissue with Isopropanol? You must have a LOT of tissue to start with... I have around 2 miliGrams.

gebirgsziege on Jan 29 2009, 03:06 PM said:

That is an option which I'd rather use after trying more efficient precipitations... Despite what QIAGEN suggests, I'd be afraid of amplification biases. Worth trying at a later stage, nevertheless, thanks!

So I am starting with few cells (less than 100) INSIDE of woody tree roots. So this IS a biased sample (polyphenols, sugar, thick cell walls). If your DNA is well purified, MDA does a great job. If you do not like kits, you can buy a suitable enzyme and set up the reaction yourself.


I think all cells (besides resting stages) consist of 90% + water, so this is nothing to be worried about. And two mg of pure tissue is quite a lot of material to start with in my eyes.

Isoprop instead of Sodium Acetate - Ethanol mixture.

ditaviz on Jan 29 2009, 08:46 AM said:

glycogen is used as a coprecipitant. it will not only make the pellet visible but will also assist in improving yield.

are you centrifuging at 14k xg or 14k rpm? 14k rpm gives more than 14k xg and is probably better to use for collecting precipitated dna.

mdfenko on Jan 29 2009, 06:03 PM said:

ok, i'll include it.

mdfenko on Jan 29 2009, 06:03 PM said:

actually, i'm centrifuging at 16.2k x g first, 20 minutes (out of the freezer, still in 100% EtOH + NaAcetate), then at 15.4k x g, 5 minutes for washing with 70% EtOH. The lower speed has been a recommendation for which I haven't found any arguments.

Right now I'm repeating the extraction (a modification of the DNeasy Tissue Kit from QIAGEN), which I plan to elute with water and then concentrate by evaporation. I know the yield is lower when eluting with water, but I doubt that it is 40% lower. I'll comment on the results tomorrow. And include the glycogen, too, thanks!

ditaviz on Jan 29 2009, 06:46 AM said:

I don't think you would lose 40% DNA the way you do the precipitation (>7 hours at -20°, then 20 minutes centrifuge at >14K G, 6°C). For ug of DNA, vortex mix with ethanol, spin immediately at >12000xg for 10min, you should recover 100% of the DNA. More likely after your repeated the ethanol precipitation, you got rid of some small RNAs, oligos, and nucleotides etc. (did you treat with RNase during the extraction?). You could load eq. amount of sample pre- and post-ethanol precipitation in a gel to see if there is any DNA loss. Seeing is believing, UV spec often tells you the wrong story.

AquaPlasmid on Jan 29 2009, 11:41 PM said:

According to some different sources (too many to remember which right now) one does lose DNA during Ethanol precipitation. I did treat with RNase, and the gel looked quite clean, rather indistinguishable between pre and post. I'm extracting with a QIAGEN kit, with which I've always had quite good results (regarding cleanliness).

Next time I'll save the gel digitally... it would have been nice to have your opinion.

Just to be sure: the co-precipitant is glycogen, not glycerol (you said glycerol in your first post). I use Pellet Paint (Novagen) which is blue dyed glycogen. You might also look at the Zymo genomic prep kits, which worked well for us in 96 well plate format. What are you doing with the DNA? How pure do you need it? How much do you need?

phage434 on Jan 30 2009, 05:21 PM said:

yes, I noticed my mistake (the late hours are staring to show...), thanks!

phage434 on Jan 30 2009, 05:21 PM said:

I need extra-pure DNA (will have to ask about the glycogen) for 454 sequencing. I need several micrograms, around 5-6 per tube.