It is recognised that no cell suspension, prepared from any source tissue, can be completely devoid of dead or damaged cells and fragments of cellular debris. It is also recognised that such non-viable components of a cell suspension can interfere with accurate flow-cytometric analysis, particularly where the subject of the analysis is a rare sub-population.
The choice of a strategy for preventing contamination of the data by a damaged cell component depends on the measured parameters of primary interest. Usually, a DNA dye which can not penetrate viable cell membranes is used to stain intact but damaged cells. For debris particles no longer containing DNA, the measurement of a combination of forward scatter (FSC) and side scatter (SSC) often reveals a pattern characteristic of such particles. Size alone is not a reliable indicator of debris, particularly when the viable cells range in size.
A number of dead cell staining techniques both for samples which were to remain viable and for samples which were subsequently to be fixed have been evaluated. The fluorochromes used, apart from the standard propidium iodide (PI), were the more recently introduced ethidium monoazide (EMA) and 7-amino actinomycin D (7-AAD). In cases where the instrumentation is not limiting, choice of a dead cell fluorescent marker may hinge only on whether it can be distinguished by colour from all immunofluorescent stains used in the analysis. By this criterion it was found for example, that PI and EMA can both be used in combination with orange-emitting PE but the redder immunofluorochrome Red613 requires the use of the longer wavelength 7-AAD as dead cell marker. While it was possible to confirm the efficacy of both EMA and 7-AAD for use in immediately analysed fixed cell samples, a long term increase in background fluorescence in the (initially) viable cell component was revealed as a limitation to the techniques.
Where the number of allowed parameters is limited, useful dead cell discrimination is afforded if a single detector receives both immunofluorescence and dead-cell stain emissions, and if the latter can be separated on the basis of a significantly higher intensity. It is pointed out however that some care must be taken if significant numbers of cells are in the throes of death during the analysis. In a further step towards economising on the number of parameters used only for non-viable cell discrimination, a cell sorter (FACStar+) has been modified to allow the addition of operator-selected proportions of SSC and PI signals before their introduction into the instrument's signal processing electronics. Thus, for those samples where high-SSC cells are not of interest, debris, dead cells and, depending on the limit chosen, non-lymphoid cells may be removed by gating on FSC and this combined "SSC+PI" parameter.