The analysis and interpretation of DNA distributions measured by flow cytometry

A principal use of flow cytometers is for the measurement of fluorescence distributions of cells stained with DNA specific dyes. A large amount of effort has been and is being expended currently in the analysis of these distributions for the fractions of cells in the G1, S, and G2 + M phases. Several methods of analysis have been proposed and are being used; new methods continue to be introduced. Many, if not most, of these methods differ only in the mathematical function used to represent the phases of the cell cycle and represent attempts to fit exactly distributions with known phase fractions or unusual shapes. In this paper we show that these refinements probably are not necessary because of cell staining and sampling variability. This hypothesis was tested by measuring fluorescence distributions for Chinese hamster ovary and KHT mouse sarcoma cells stained with Hoechst-33258, chromomycin A3, propidium iodide, and acriflavine. Our results show that: a) single measurements can result in phase fraction estimates that are in error by as much as 40% for G2 + M phase and 15-20% for G1 and S phases; b) different dyes can yield phase fraction estimates that differ by as much as 40% due to differences in DNA specificity; c) the shapes of fluorescence distributions and their interpretation are very dependent on the dye being used and on its binding mechanism.