Some time ago I noted my inability to obtain "advertized yields" when
trying to extract a PCR product of 650 bp from agarose gel. I asked for
help. A number of you replied, and I have tried most of your suggestions.
Since this topic keeps coming around, as now, I offer the following:
1. I made the mistake of not noting that I was doing direct sequencing of
double stranded product as opposed to subsequent cloning. A few cloners
wrote....why worry, almost any amount will do....20-30% is plenty...don't
be greedy. My fault here as I need, maybe, 200 ng at 800 bp but 1.0 ug if
the total, double stranded PCR product in the sequencing reaction is 2500
bp. I know, I know, use shorter PCR product......I will, and do.
2. Most folks indicated that yields of 90-95% are pie-in-the-sky. Be
happy with, and plan for, 30-60% and you should be OK.
3. I have tried all of the freeze-squeeze methods offered and I just
don't have the karma. Must be something like males producing serine at
their fingertips and scaring away the fish. Everytime I touch the tube,
the DNA becomes insoluble. To those of you who do have the karma, guard
4. One technique which may have real promise is to run your DNA into
Whatman DE81 paper. The yields looked as good as #5, but cutting the slit
in the gel, getting your timing down as to how far to run the band
especially if you have multiple bands running very close together, and the
subsequent elution and recovery of very small amounts of DNA takes some
practice, at least in my hands, by doing it time after time. I don't
think I really gave it enough of a chance to offer a yes or no.
5. On a happier note, I did find success using both Elu-Quik and Qiaex
(must be the clever spelling), at the 50-60% recovery level, **BUT** only
with TAE rather than TBE. Even doubling the dissolution/binding buffer
did not bring TBE even close to the yield of TAE gels.
So, if it fits your work, use lots of a short PCR product, probably a
glassmilk binding matrix, TAE gel, and take it to the Applied Biosystems
373A which will love it.
Thanks again for the help.
Oregon State University
Corvallis, OR 97331
ricksonf at bcc.orst.edu