generally, once I have identified -LTH +Bgal transformants, I select for
loss of pAS2 by nutritional supplementation (grow on rich media, then on
SD -L), followed by a screen for loss of Bgal signal. -Bgal -L colonies
are then subcloned twice to ensure a homogenous isolate containing
(hopefully) a single flavor of pACT, pACT is isolated, sequenced, and
retransformed. This works in 9 out of ten cases (i.e. most, but not all,
pACT plasmids trigger the two-hybrid system upon retransformation).
You should be able to weed out false positives due to DNA binding
proteins encoded by pACT, as well as (extremely rare) chromosomal
mutations affecting transcriptional control, by screening for loss of
Bgal (as above).
If you have >1000 -LTH +Bgal transformants to screen, I'll be happy to
send you my protocol (basically, you do everything in 96 well plates,
and a robot certainly, helps, too).
jliphard at reed.edujtcl3 at cam.ac.uk
Div. of Virology
University of Cambridge