I understand that yeast integrating plasmids (YIps) must be linearized in order to promote homology-directed recombination. However I don't quite understand where.
In the end, the external regions of the linear constructs must be homologous to the genomic target; but quite often in a naked YIp backbone the only region that has homology to the host genome is the resistance marker... and cutting it would defeat the whole purpose of having a selection marker, would it not?
A basic example (from the Bartel Lab, 2009): - The region 500-2000 is homologous to some weird, unrelated species. No homology with yeast genome. - An insert would go to the multiple cloning site (XbaI, NotI etc) - AmpR and ori are just there for the bacterial step
So all that's left is URA3, which one could cut with, say, NcoI... but then the yeast will get a negative phenotype (loss of uracile metabolism, by disruption of the biosynthetic gene), which is annoying to select for - and could be the result of a spontaneous KO mutation (unlikely but not impossible).
In the precise case of URA3, one could select for the loss of the gene with 5-Fluoroorotic acid, but other markers like LEU2 and HIS3 don't have such a straightforward loss-of-function selection strategy, as far as I know (?)
So, is there a way to select for YIp integration based on gain-of-function phenotypes in this example? or would one have to incorporate one in the transgene, like a fluorescent protein?