Adventures in DNA extraction

[Joseph C. Bagshaw]

Mornin' folks,

I've got a curious phenomenon to report for your amusement and
perhaps discussion.  First I have to describe my current method
for isolating genomic DNA.  I work with shrimp, and after a lot of
frustration I developed a tedious but effective protocol that
produced digestible DNA of adequate quality.  A few months ago
one of my grad students was preparing total RNA from shrimp
using the Chomczynski and Sacchi method as described in the 
Big Red Book.  Noting how rapid the procedure was, I tried back
extracting the organic phase with uncalibrated Tris to get out 
the DNA.  This is an obvious idea that has been published at least 
twice in Biotechniques.  I got DNA, but it was rather fragmented.
Next I added 30 mM Tris, pH 8.8 to the C & S lysis buffer, and
extracted with buffered phenol and chloroform.  I got the best
genomic DNA I've ever seen from shrimp, good yield, unfragmented,
and colorless.  I've subsequently used this protocol to extract
genomic DNA from Limulus sperm (no challenge), whole terrestrial
and marine isopods, and little fishies.  Haven't tried any plants 
yet, but I'll bet it works.

Now here's the funny part.  Through a series of blunders I discovered
that substituting either sodium acetate  or sodium chloride for 
the sodium citrate in the lysis buffer buggers up the protocol.
I then get a massive precipitate upon ethanol precipitation, most of
which will not dissolve in TE.  It also inhibits recovery of DNA from
the pellet, although the quality is not reduced.  The stuff looks
like glycogen, which of course like to co-partiion with DNA and
precipitates with ethanol.  The presence of citrate, but not acetate
or chloride, causes the glycogen (or whatever it is) to either
partition into the interface or not precipitate with ethanol.  
Can anyone explain this?

BTW, the addition of Tris to the lysis buffer described in 
Current Protocols, 4.2.4, make no difference.  Just use buffered
phenol, and don't acidify the extract.  If the RNA you get with
the DNA is a problem, just RNase it away.  Try it, you'll 
like it.

-- 
********************  HAVE GENES, WILL TRAVEL  ********************
Joe Bagshaw, Worcester Polytechnic Institute
jbagshaw at wpi.wpi.edu
Roadkill on the information superhighway.