Comparison of cellular and nuclear flow cytometric techniques for discriminating apoptotic subpopulations.

We compared cellular flow cytometric methods employing carboxyfluorescein (CF), Höechst 33342, and Höechst 33258 with a nuclear method in their ability to discriminate apoptotic subpopulations in rat thymocyte cultures exposed to dexamethasone or tributyltin. In the nuclear technique, apoptotic cells appeared as a single population containing reduced DNA content, while in the cellular techniques, apoptotic cells appeared as two or more subpopulations exhibiting increased fluorescence. Of these subpopulations, early apoptotic cells [which excluded propidium iodide (PI), indicating maintenance of membrane integrity] exhibited higher fluorescence but the same level of axial light loss (i.e., size/refractive index) as viable cells; late apoptotic and dead cells (which incorporated PI) exhibited decreased axial light loss. However, while the Höechst dyes allowed discrimination of late apoptotic from dead cells, CF did not. In comparing sensitivity to staining conditions, Höechst 33258 fluorescence was the most stable over time, Höechst 33342 the least, and CF fluorescence not only varied with time, but with tri-n-butyltin methoxide concentration. Comparison of single-parameter analyses revealed that axial light loss was sensitive only to late apoptotic changes; nuclear fluorescence was a better indicator of apoptotic subpopulations, but still underestimated the total percentage of affected cells, and Höechst 33342 distinguished early apoptotic cells as those with elevated fluorescence. Early apoptotic cells stained with Höechst 33258 also exhibited increased fluorescence, but could not be distinguished from late apoptotic and dead cells without a second parameter. These findings indicate that of the methods investigated, the method of choice for detecting apoptosis depends on the goal of analysis: Höechst 33258 was best for discriminating apoptotic subpopulations, and CF was best for assessing alterations of membrane fluidity. For single-parameter analyses, Höechst 33258 was best for determining the total percentage of affected cells, while Höechst 33342 could be used to determine the percentage in early apoptosis.