Viable cell sorting, based on flow cytometric analysis of DNA content and cell volume, was used to evaluate the cycle position and survival potential of Adriamycin (AdR)-treated or 1-β-d-arabinofuranosylcytosine (ara-C)-treated CHO cells. Drug-treated cells initially stained with the vital, DNA-specific fluorochrome, Hoechst 33342, were analyzed for DNA content and volume, and sorting “windows” were established for subsequent sorting of duplicate unstained cell samples based only on cell volume. Another portion of the cell sample was fixed in ethanol, and stained with three fluorochromes for correlated flow cytometric analysis of DNA, RNA, and protein. Similarities in the viable cell volume distributions and the protein content distributions of the ethanol-fixed samples provided a means for indirectly determining the DNA and RNA contents of the sorted cells. Three regions (S, L, and I) were selected in the cell volume distributions corresponding to the range of near normal cell size (S), larger than normal cell size (I), and the extremely large cells (L). Adriamycin-treated or ara-C-treated cells sorted from the S region had survival values, respectively, 46 times and 7 times greater than the abnormally large cells in region L. Cells from the S region also respectively survived 14-fold (AdR-treated) and 7-fold (ara-C-treated) greater than the cells sorted from the I regions. RNA content levels for cells within the L region were three times and two times greater, respectively, than the AdR-treated and Ara-C-treated subpopulations in the S regions. Survival of subpopulations of G2-arrested, AdR-treated cells (I and L regions) was better correlated with relative abnormality in cell size than with position in the cell cycle. In addition to providing further support for the validity of the “balanced growth hypothesis,” the results of this study suggest that two-parameter DNA content and cell volume measurements would be extremely useful for providing general guidelines for judging the effectiveness of therapy, especially in clinical diagnoses where cell sorting is impractical or impossible. From these analyses the frequency and cycle position of cells resistant to therapy can be estimated. Such information would be particularly useful for rapidly detecting drug-resistant cells and design of subsequent therapeutic regimens.


Supported by the Division of Research, NIH Grant RR01315, and the United States Department of Energy.

This content is only available via PDF.