Ethanol precipitation following sucrose gradient - problem - (Jul/21/2006 )

I have been having problems with ethanol precipitation of some DNA I am preparing to make a genomic library with - I get <10% recovery! (Usually when I EtOH ppte I get 80% recovery)
The DNA I am trying to EtOH ppte is the 9-23kb fragments from a partial MboI digest of some genomic DNA which was separated on a sucrose gradient. I removed the sucrose from my fractions of interest by dialysis using 10 000MWCO tubing (using 4litres TE 1X O/N changing the buffer once after 4 hours at 4oC).
I have now repeated this procedure 3 times, and have the same loss of DNA each time. Obviously I do not want to make a library such DNA as I may be selecting against certain fragments.
Please, can anyone help me?!

I couldn't find a recommendation of a particular sized MWCO tubing to use. I simply used 10kDa as that's what the biochemists in our department have, and Maniatis didn't state anything in particular for dialysis with DNA.
I thought 10kDa would be ok as I figured the smallest fragments would have a challenge to get through this pore size as each base pair is roughly 600Da. On reflection after your comment, I guess the long strands can wiggle their way out as they are not necessarily globular.
I did try to do precipitation without prior dialysis and got a low yield (Maniatis suggests lowering the concentration of sucrose to <10% and then apparently it should not interfere with pelleting - however I found that I only got <10% recovery with this - much lower than the 80% I would usually expect). Then I figured it wouldn't be challenging to do dialysis, so why not get rid of the sucrose... The rest is history.

Thanks very much for your comment - I'll certainly try different tubing.
Cheers!!

That sounds good for proteins - however I am working with DNA. My smallest fragments are 9kb. Since one bp is roughly 600Da the smallest molecules should be 9000x600=5 400 000Da (or 5400 kDa). So in theory it should have been ok. And yet I was still losing my molecules of DNA though 10kDa MWCO dialysis tubing! I think this is to do with the fact that DNA is linear, and can wiggle out anyway. So I have ordered some 1kDa MWCO tubing which I hope will work better.

I am making a genomic DNA library. I need to do the sucrose gradient to isolate 9-23kb fragments from my MboI partial digest for ligating into Lambda phage. I want only to add inserts which are the correct size to be accommodated into the ligation reaction so as to have a high titre when I make the library, and thus good representation of the genome. Stratagene (who manufacture the kit I am using) recommend size fractionation of the insert DNA to minimise cloning of multiple inserts.

PS if anyone knows the exact MWCO tubing I should be using to retain 9-23kb DNA I would appreciate your input!!

You could try a fast-desalting column to get rid of the sucrose. If you do it in a reasonable amount of salt, you should reduce the amount of non-specific binding to the column. Then try your EtOH precipitation again.

This reply is a bit late but. I do a similar technique in the lab but we use Dialysis Cassetts from PIERCE there called Slide-A-lyzer cassettes. They come in different MW cut off points. I rarely loose DNA after a sucrose gradient/ dialysis but the times I have it has allways been during dialysis. Another thing you can try is Amicon Ultra filter devices for spin dialysis. you put your fractions of intrest into the tube and centrifuge it down to ~1ml sample then add a few ml of your buffer to dillute the sucrose. once you do this about 3 times the sucrose should be allmost all filtered out. I have never lost DNA doing spin dialysis.

Good luck
RWG