[Yeast] Is this a false positive one from yeast two hybrid screen?

Linfeng Huang (TSL) linfeng.huang at sainsbury-laboratory.ac.uk
Tue Sep 26 07:21:58 EST 2006


Hi Y2H experts,
I got a puzzle in my yeast two-hybrid experiment and eager for some
expert advice.
I was using a Clontech MATCHMAKER system (LexA system). My bait was
fused to the BD domain (pLexA vector).
>From the screen, I got a strong candidate (very blue on selection
medium), which is represented by several colonies. I did
re-transformation of the isolated library plasmid with my bait and it
looks fine. I also tested to swap the bait and pray vectors and it was
also fine. 
Then I did RACE for the gene and got the full length cDNA sequence.
Then the problem came: it turned out that the library plasmid we have
got contains the 5'UTR sequence (about 40AA) which is in-frame
translated with the full coding sequence. And when I put only the full
CDS in the plasmid, the interaction did not work. Then I did several UTR
replacement constructs (similar length of original UTR and flexible) -
they all did not work. But 5UTR itself or deletion of a portion of the
3'sequence within the coding region also did not interact which ruled
out the possibility that the 5UTR alone interacted with bait. It seems
the 5UTR and CDS are both indispensable for the interaction.
A recent experiment also showed although the CDS only construct did not
show interaction tested on plate but a colony lift essay (grow on SD
-Ura/Trp/His G/R) sowed blue but also in empty vector control. It might
be worthy to notice that CDS only construct also grow slower than the
library construct with CDS. How weird is it?
Since there is no evidence that this 5UTR could be translated in vivo.
Could this gene be just be a false positive/artificial interactor? Do
you have suggestion for some more Y2H experiments or controls?
Thanks a lot in advance!
Best regards!

Linfen

Linfeng.huang at bbsrc.ac.uk



More information about the Yeast mailing list