>We have a project in which we would like to study DNA ploidy plus protein >expression of two nuclear antigens. Our intrumentation is limited to a >FACSCAN. I assume that we will start by using FITC and propidium idodide. >But what, if any, is your recommendation for a third color? Is this >feasible with the limitation of our instrument? If you can't change the filters on the FACScan, you've got a big problem. People have been able to use FITC- and PE- antibodies with PI; to do so, you need to have a lot of the antigen which will bind the PE-antibody in order to have any hope of resolving the immunofluorescence in the presence of the PI signal. In principle, 7-aminoactinomycin D would be a better DNA stain to use with FITC- and PE- antibodies; 7-AAD, unlike PI< is DNA-specific, eliminating the need for RNAse treatment, and its emission is at a much longer wavelength (> 640 nm) than PI's, matching it well to the FACScan's standard filters. Also, 7-AAD fluorescence is much less bright than PI fluorescence, making compensation easier. Now, the bad news. 7-AAD is more sensitive than almost any other DNA stain to chromatin conformation, and has been used by Stokke, Steen et al to demonstrate differences among cell types in chromatin conformation, which means that under many circumstances, one doesn't get good stoichiometric DNA staining. We have found, based on the work of Toba et al in J Immunol Methods in 1995 and on subsequent suggestions from Dr. Toba, that staining at a relatively low pH (6.0), before fixation, somewhat improves CV's with 7-AAD. Jim McSharry, at Albany Medical College, has stained cells fixed with methanol-acetone with 7-AAD and FITC- and PE- antibodies to viral antigens; 7-AAD staining is reasonably good and could probably be improved. However, in work with him, I have come to suspect that there can be substantial energy transfer between FITC bound to ab's against nuclear antigens and 7-AAD bound to DNA; this would make it necessary to compensate between green and red fluorescence, which is, as I recall, not possible on the FACScan, at least in hardware. On the other hand, even without compensation, you would probably be able to discern differences in DNA content distribution among cells bearing one or both antigens and cells bearing neither. And by the way, what kind of cells are these?...That and whether you are looking for hundreds or millions of molecules of antigen will probably be the major determinants of success or failure. -Another Howard
This archive was generated by hypermail 2b29 : Wed Apr 03 2002 - 11:49:17 EST