In article <Ct60L9.G7G at UQuebec.CA>, Benoit_Hebert at IAF.UQUEBEC.CA () wrote:
> Reinhard Kunze (kunze at gen1.genetik.uni-koeln.de) wrote:
> > Is there everyone having experience in the use of degenerated oligonucleotides?
> > My plan is to order a 60 bases long oligonucleotide for site-directed
> > mutagenesis, which should have four wobble-positions. Could you please send
> > me your advice, in which ratios
>>> First, think about having your mutations near the center of your oligo. By the
> size of your oligo, I figure you'll be using the Kunkel method. I have more
> experience with PCR mutagenesis.
>> Obviously, the more degenerate your oligos, the greater number of different
> oligos you'll be using, and the lower the concentration of each population of
> oligo. If you can avoid degeneracy and can pre-select your intended mutations,
> the easier this experiment will be.
I regularly use PCR to mutate our favourite protein (SRF) and have on
occasions put mismatches within 6bp of the 3'end. In one cockup I used the
wrong template which meant i had a mismatch 2bp from the end, I still got a
good yield of product (of course 27 of the other 30 bp did match). We
always make the 3' end a G or a C, no idea how important this is as I've
never done otherwise.
For cassette mutagenesis I made a 55bp oligo with 12 positions randomized
(ie all 4 bases can occur) the last randomized position was 25bp from the
3' end. It was used in PCR with a T7 17mer using the cycles
the reactions used 10pmol of each primer and 100ng of template in a 100ul
rxn, dNTPs at 0.2mM, Mg2+ 1.5mM.
My randomized oligo had a Cla site 7bp from the 5' end and the region PCRed
included an NcoI site. Thus the PCR rxn was phenoled, EtOH prec, resus,
digested and gel purified. The resulting library was amplified with
electrozap and hey presto a whole library of SRF mutants which were
randomized at 4 amino acids (possible number of mutant proteins =160,000,
if you don't count the stops). Its screening the little buggers that is the
problem. In my case it was the DNA binding domain and we screened then
using a yeast one hybrid screen.
The library appears to give a good representation of the 4 bases at the
randomized positions, ie we didn't detect any bias.