NUCLEAR DYES AND CYTOPLASMIC STAINING
Growing interest for in vivo cell staining is prompted by the development of new fluorescent probes with better properties in terms of fluorescence yield, specificity, and lower cytotoxicity. However, the interaction between dyes and living organelles is still poorly understood, and very often staining affinities in vivo are different from those observed in fixed cells. The best-known case is that of acridine orange (AO) which, in fixed cells, stains nucleic acids in red and green, according to their single-or double-stranded configuration, whereas in living cells it stains acidic cytoplasmic granules in red and green depending on their internal pH (4). Less-known cases concern pyronin B (PB) and ethidium bromide (EB).
In a recent report (3), PB was found to differentially stain normal and apoptotic cells in vivo. Following a comparison with other nuclear dyes, PB, considered in every respect to be a nuclear dye, proved as efficient in detecting apoptosis as LDS-751 and SYTO-16. We do not dispute the unquestionable evidence of the data, but wonder about the nature of PB-stained structures. Indeed, the work was exclusively based on flow cytometric measurements, ignoring any morphological control. Our observations indicate that PB, used in vivo, selectively stains mitochondria, as does its analogue pyronin Y (1). We never found PB fluorescence inside nuclei of living cells, even in apoptotic and necrotic cells and using high dye concentrations.
In an analogous investigation (2), apoptotic cells were associated with a level of EB fluorescence intermediate between the low fluorescence of healthy cells and the strong fluorescence typical of dead cells. Data were obtained in vivo, and proved by the fact that EB was