Nathalie CUEILLE <73373.1544 at CompuServe.COM> wrote:
>> Hello everybody
>> I am a french PhD student (CNRS, Montpellier, France) seeking for
> some help
> with my saccharomyces yeast. (e-mail at 73373.1544 at compuserve.com>> I have screened a Xenopus cDNA bank (dihybrid technique) to find
> to my bait
>> I have found 9 clones giving the following phenotype His+/Lac+
>> After segregation of the clones followed by amplification of the
> cDNA (PCR)
> and transformation in E.coli HB101 to eliminate the bait I tried
> to test
> the specificity of the interaction by backtransforming these
> with my bait or with RAS.
> But I lost in this backtransformation the phenotype Lac+/His+ in
> both cases
> though my witness reacts well
> what can I do to clear up this mystery?
> How can I explain the loss of this phenotype?
> How can I test the specificity? thank you -- Nathalie CUEILLE
If I understand your message correctly, what you found is that after
you had identified 9 colonies that were His+/LacZ+ in your screen,
you purified the plasmid from each of the 9 and then attempted to test
specificity by retransforming them into naive yeast containing your
original bait or your bait fused to Ras. The result of this was that
none of the 9 repeated the His+/LacZ+ phenotype you had originally
obtained. (I'm not sure what you mean by the witness behaving well).
If the above summary is true: I can tell you from experience in a
different two hybrid system that it is possible to get yeast strain
mutants that simultaneously activate the two reporter genes being used.
What these are is unclear - however these are particularly common if
your bait has ANY background ability to activate transcription, however
weak...and in one case, we went through the trouble to drive out the
original bait by propagation on medium not selecting for it, and
retransformed in a different bait, and found that it was activating
more than it should, too. My guess is that these would be mutations
in transcriptional activation core components, giving a general increase
to expression from your target genes. Stan Fields has a Biotechniques
article from 1994 peripherally related to this type of thing
Another more remote possibility is that your original 9 positives
contained 2 separate library plasmids (sometimes happens with libraries
based on 2u), and that the one with the positive real phenotype is
selectively less reisolated than some bystander. I think this is
unlikely with 9/9...but it has been seen to happen on occasion.
If I were you, I might return to the original positives, repeat plasmid
isolation and retransformation: and if it happens again, figure you
have a background problem. Good luck!