Vertical Agarose gels, help

John Brunstein brunstei at
Tue Jun 4 15:55:10 EST 1996

OK  :-))  Been there, seen those problems, and solved them!  Here's how 
I handle them (this is for 1.5% agarose, 1% glycerol gels 1.5mm thick, 
about 20x20cm).
	(1)  Slippage of the gel from between the plates (boy, is it ever a 
bummer to see your hard-won bandshift samples slide right out of the 
plates and into a sodden little lump an hour into the gel run!)  Before 
pouring the gel, I run a bit of sequencing tape (the yellow electrical 
stuff) across the bottom of the gel just inside the spacer, in a sort 
of 'U' shape...(I'll attempt a bad ASCII art rendition below).  Leave 
these on for the run, as the gel starts to slide it hits them and holds 
the gel in place..but since most of the bottom is unobstructed, current 
flow is not interfered with
									|		|											|		|
									|		|					GEL 		|		|<---- Spacer
									|		|											|		|
									|		|T									T|<-----  Tape runs down, around bottom, and 		
									|		|T									T|		|					partly up far side

	(2)  Removal of comb.  Another really frustrating problem...but easily 
solved with a bit of practice.  The problem is that as you remove the 
comb, a vacuum is created in the well and the gel collapses into it, 
causing tears.  Solution:  I use a thin spatula to pry at the glas 
plates as I ease the comb out little by'll have to move the 
spatula around a bit to ensure that all wells get air into them as you 
work the comb out, so don't rush it...but after a few wrecked gels 
you'll get the hang of it.

	Hope these answers solve your problems, they took me a few weeks of 
trial and error to settle on as the final system.  I use these gels 
routinely for bandshifts (my DNA is too large for acrylamide systems), 
and I no longer have any problems with preparing them. There is a bit of 
a problem with getting the wells really clean, bits of agarose tend to 
stay stick in there despite my best attempts to cut them out, and these 
can make the results messier than one might like..but the sytem works.

More information about the Methods mailing list