fragment isolation w/o glass milk?
brunstei at UNIXG.UBC.CA
Mon Aug 8 13:28:42 EST 1994
Actually, someone (sorry, but I don't remember who) posted the
'easier version' that you are looking for on this newsgroup just a few
months ago. I took his protocal and modified it very slightly myself
and have had good luck. It goes as follows:
(1) Silanize a batch of glass wool (the 'soft' grade is easier to
work with than the 'stiff' grade, but both work OK) by putting in a
beaker, adding a few ml of silane (anybody out there tried Rain-X for
this?) and slop it about to get the glass wool all wet. I then put the
beaker in a 60C oven for a few hours to dry the wool (original post
suggested vacume dessicator overnight, but I'm too impatient).
(2) Cut the tops off a batch of small (0.5 or 0.6ml
microcentrifuge tubes) and use a 22 gauge needle to poke holes through
(3) Take a small pinch of the silanized glass wool and squash it
into the small tube. I use the plunger from a 1cc syringe with the
bottom rubber part broken off to stomp the wool down. You want ~4-5mm
height of the wool.
(4) Cut the cap off a large (1.5 ml) microcentrifuge tube, put
the prepared small tube into the big one (it should catch on the rim,
such that there is ~100 ul free volume at the bottom of the assembly).
(5) Now the easy part---just cut your band out of the gel (as
small a slice as possible), put the slice into the premade assembly
(don't worry about crushing, etc; just pop it in) and spin 12000 x g for
(6) Your fragment will be in the aqueous phase in the bottom of
the big tube in 20-100 ul of solution. For a band of any reasonable
strength, I usually find the best results for cloning is just to use a
few ul of this directly in the ligation. For some reason, this works
better in my hands than doing a ppt. of the extract and resuspending in a
smaller volume, but then again maybe I'm just a klutz.
I have used this successfully for a few months now (~40
isolations?) on bands from 80bp to 9kb, with gels of between 0.9 and 2.0%
Hope this answers your questions, and my thanks to the original
poster whose protocal this is based on.
More information about the Methods