Welcome to the wonderful world of methylene blue. We use methylene blue as
described in numerous papers by Wickner to stain killed cells in the killer
virus assay (haloes of killed sensitive cell lawns around killer colonies
replica plated onto them). The dye stains the dead cells at the border of the
halo, but not the live cells of the colonies or lawn. However, after several
decades of looking at colonies on methylene blue plates, I have noticed that
individual colonies may differ in staining for reasons other than viability.
Strains with unstable mitochondrial genomes seem to vary more than stable
strains, suggesting one reason for variation. In the 1960-70's a student in
E.A. Bevan's lab in London wrote a thesis on Blue Mutants of yeast. The
student tried to investigate why blue variants would be found among white
strains, and found that multiple stable and unstable chromosomal and
cytoplasmic genes seemed to be responsible. Interesting, but nothing you
probably with to pursue.
I suggest that MB staining is a crude method that is good when in works, which
is not always. I hope some other responder can give us both a better method.