Differentiation of mitotic melanoma cells from G2 cells and their isolation by use of 5-bromo-2'-deoxyuridine and propidium iodide

This report describes a method by which mitotic cells were isolated from nonsynchronized Cloudman melanoma cells that had been pulse labeled with 5-bromo-2'-deoxyuridine (BrdUrd) and double-stained with a fluoresceinated monoclonal antibody to BrdUrd and with propidium iodide (PI). In initial experiments, melanoma cells were first pulse labeled with BrdUrd, treated with prostaglandin E1 (PGE1 10 micrograms/m1) or vehicle (0.1% ethanol) for up to 24 hours, then stained with anti-BrdUrd and PI. PGE1-treated cells monitored at 3-hour intervals were observed to migrate from S phase to G2 phase, then, enigmatically, back into the late S phase region of the distribution. In other experiments, cells treated with PGE1 were pulse labeled with BrdUrd at the end of the treatment period and harvested. In these experiments, there was a small, discrete subpopulation of cells within the late S phase region of the DNA distribution that was negative for anti-BrdUrd. This subpopulation of cells was sorted and examined by light microscopy. We observed that 95% of these BrdUrd-negative "S phase" cells were mitotic cells. Since mitotic cells and G2 cells have equivalent amounts of DNA, the reduced red fluorescence exhibited by these cells may be due to a greater sensitivity to denaturation, which has been described for DNA of mitotic cells, and would account for the phenomenon of cells appearing to move "backwards" in the cell cycle. This report indicates that although the BrdUrd/PI method can further define the cell cycle into four compartments, it can also lead to over-estimation of S phase cells in kinetic studies because of contaminating mitotic cells.