> >I am working with a group that has transfected a cell line with a GFP >mutant GFPS65T. >I then sort out the top 15% or so of the brightest FITC labeled cells, I'm assuming you mean the green fluorescent GFP labeled cells > out >of the right arm of the gaussian (no second peak is distinct for the >positives, but rather what we see is a continuum of increasingly brighter >cells). Everything looks great for the sort (I check a sample under the >fluorescent scope, number of sorted events are accurate, etc.) until I run >a portion of the sorted sample back through to check for purity. I get >about an 80:20 ratio of positive to negative cells, which fall out in 2 >distinct peaks, scatter is still perfectly intact. > When cell populations have broad scatter distributions, scatter is fairly reliable for discriminating cells from small debris, but unreliable for discriminating single cells from aggregates. In my experience, based on adding Hoechst dyes to samples, there can be a substantial number of aggregates in a sample when you wouldn't expect them to be there. They won't be eliminated by coincidence circuitry. An aggregate will be brighter than the average cell, and the upper tail of your fluorescence distribution is likely to contain some aggregates, some of which will be composed of a highly fluorescent cell and a dimmer one. If they disaggregate between the time you sort and the time you resort, this will give you a distribution with bright and dim peaks. I'd say this is the most likely explanation of your results. -Howard
This archive was generated by hypermail 2b29 : Wed Apr 03 2002 - 11:49:34 EST