the plasmid has a tendency to recombine out the transposase in bacteria thanks to the homologies between the homology arms and the repair template.
Two recombinations are possible depending on which arm does the recombination:
In both case the plasmid will lack the transposase and the begining of ADE2.
None of the recombined plasmid will generate ADE+ progeny.
To diagnose the amount of recombination in your plasmid prep, digestions with EcoRV and PvuII is advisable:
Here are examples of preps with little or lots of recombination products, digested with EcoRV
Here are examples of preps with little or lots of recombination products, digested with PvuII
Of course, the plasmid can also recombine in S cerevisiae. To avoid starting a screen with the wrong clone, streak several transformants (10-20) directly from the transformation plate (-Ura) onto -Ade +Dex. A recombined clone will generate zero ADE+ colonies, while a good clone will generate a small number (2-10) of spontaneously ADE+ colonies by homologous recombination (see below).
Make sure that you streak a good amount of cells on -Ade, since the spontaneous homologous recombination is rare.
Also streak on -Ura to propagate the clone.
Conversely, if your streak is full of colonies (>100) on -Ade, it means that your clone is full of spontaneously recombined ADE+ cells. You don't want to use this one either.
Continue with good clones, starting from the -Ura streak (see below, clone #3 has been used further. Any but #1 would do).