Detection of incorporated iododeoxyuridine in colonies by immunoperoxidase staining: a novel method to measure the proportion of cycling colony-forming cells.

In vitro suicide by tritiated thymidine (3H-TdR), hydroxyurea (HU), or cytosine arabinoside (Ara-C) is assumed to reflect the proportion of colony-forming cells in S-phase at the time of exposure. However, these techniques are not always accurate. Nonradioactive iododeoxyuridine (IdUrd) is incorporated into DNA during S-phase and can be detected by monoclonal antibodies. In the present study, a new IdUrd application was developed to investigate the kinetics of hematopoietic progenitor cells. After incubation with IdUrd, colony-forming cells were cultured in semisolid assay. An immunoperoxidase staining protocol was developed to detect IdUrd in cells of colonies in agar. Colony-forming cells in S-phase during the IdUrd exposure were postulated to give rise to IdUrd+ colonies, whereas non-S-phase cells would generate IdUrd- colonies. Toxicity, sensitivity, and IdUrd inactivation studies indicated that progenitor cells could safely be pulse-labeled for 2 hours with 40 microM IdUrd, whereas prolonged labeling with 1 microM IdUrd was at least feasible for 5 days. Molt-4 cells and normal bone marrow cells were used to compare IdUrd pulse-labeling with 3H-TdR suicide. Part of the Molt-4 cells were enriched for G1- and S-phase cells by counterflow centrifugation. The bone marrow cells were either unstimulated or stimulated with growth factors. As a result, the accuracy of both techniques could be tested in populations with different quantities of S-phase cells. Wide confidence intervals of the suicide technique contrasted with the small confidence intervals obtained with IdUrd pulse-labeling. For instance, the fraction of Molt-4 cells with 27.8% S-phase cells contained 17.7% (confidence interval -8.2 to 43.6%) clonogenic cells in S-phase when determined with 3H-TdR suicide. Of this fraction, the percentage of clonogenic cells in S-phase was 30.6% with a confidence interval of 25.5 to 36.2% when determined with IdUrd pulse-labeling. In our hands, the IdUrd pulse-labeling was more accurate than the 3H-TdR suicide technique. Thus far, kinetic studies of progenitors have been limited to the determination of the fraction of S-phase cells by suicide techniques. By prolonged IdUrd labeling, it is now possible to determine the proliferating fraction of progenitor cells.