shaney at watson.princeton.edu (Steven A. Haney) wrote:
>URA3 expression cannot be directly selected against because of its fairly
>long phenotypic halflife ("phenotypic lag"), and because FOA selection in a
>URA3(+) strain is toxic, as opposed to growth static. Direct selection on
>FOA will only identify those cells that were URA3(-) before the
>The best solution is to cotransform the gene fragment with a small amount
>of a closed circular plasmid with a different selectable marker (eg: TRP1
>or LEU2). Select for the cotransformed plasmid marker. Then either
>replica print to FOA or check the phenotype of the transformants directly,
>if you can. I typically get about 0.1% to 1% of the original transformants
>that show a gene conversion that removes the URA3 marker.
>Steven A. Haney
>Broach Lab/Dept. of Mol. Biol.
>shaney at watson.princeton.edu
Rather than perfoming a double transformation, just include a step in
the protocol which allows the cells to "lose" the functional URA3p
which is present in the cells. (the YD growth step mentioned in one
of the other posts.)
In Jack von Borstel's lab, they found that selecting UV-induced ura3-
cells could be facilitated by holding the cells in water(or buffer)
for 24 hours; or by growing the cells in liquid SC-ura overnight.
The URA+ cells grow out, but the URA- cells stop dividing once all
functional Ura3p is gone. Direct plating on FOA is possible after the
non-selective delay, and a reasonably true estimate of frequency is
I'm just here, I don't speak for the U of A. | Buy me a drink, sometime
| And have your way with me.