A problem with the sort of GFP transfected cells on the FACS DiVa

Dear Flowers,

I would like to ask your input / help with a problem, we recently faced 
twice trying to sort GFP transfected cells (once mouse bone marrow and 
once adherent cultured human cell lines.

Running the samples we found only very few and very weekly fluorescing 
cells. But whenever we stopped the sample flow during acquisition, the 
last "burst" of cells, showed up in the region, where we normally found 
strongly positive cells. This effect was always identical (when we 
switched on and off every few seconds or stopped directly after a 30 
minutes run). With the bone marrow cells backgating showed a scatter 
pattern as we expected.

The same phenomenon was seen, when we used the sample boost during 
acquisition: The first "burst" of the boosted cells seemed to be 
strongly positive. Then the pattern went back to normal of course with 
the low reolution of a boosted sample.

With a normal run at different sample speeds we hardly found any 
positive cells (well, the transfected population was supposed to be 
rare). But one could believe that at lower sample rates the percentage 
of positives grew slightly. The latter is to take with a grain of salt, 
since we talk about very low numbers.

Neither a non transfected sample nor CaliBrite nor AccuDrop beads showed 
this phenomenon.

We tried the following to either improve the detection of the positive 
cells or to get rid of the artefact depending on how you look at it:

-	   Working at different sheath pressures (12, 30, 40 and 60 PSI)

-	   Trying several nozzles, some of them with different diameters 
(70 and 100 um)

-	   Realigning with beads, while slightly changing the laser 
focus in order to broaden the spot and compensating the intensity loss 
with more laser power.

-	   Vigorous flushing of the sample line with bleach etc. before 
starting the "strangely" behaving sample didn't stop the effect.

On the other hand we sort quite regularly (I am tempted to say ad 
nauseam) GFP positive cells (who in the community doesn't ?) and we 
sorted similar samples for the same "clients" without this problem 
before. The samples were without any treatment apart from the 
preparation and were kept in FACS Wash. We run the machine with PBS and 
flush it daily after the last sort.

Since I am quite puzzled with this phenomenon and do not have any 
explanation at hand I would be glad, if anyone in the community has one.

Any ideas or help in order to explain this phenomenon or even better to 
overcome the trouble would be greatly appreciated.

Thank you in advance and best regards


Jochen