A Pitfall of Propidium Iodide Staining in Fluorescence-Activated Cell Sorting Cell Cycle Analysis? Free

To the Editor:

We read with great interest the recent article by Fischer et al. (1) on uracil-DNA glycosylase and its expression modulated by fluorodeoxyuridine (FdUrd). As a particular observation, the authors provide evidence that FdUrd-induced uracil-DNA glycosylase degradation occurred in late G₁-early S transition instead of late S-G₂-M phase as was observed previously in normally cycling cells (2). This was shown by fluorescence-activated cell sorting (FACS) analysis using propidium iodide staining, which is a most common assay of cell cycle analysis.

In our opinion, FACS analysis of cell cycling after propidium iodide staining alone contains a major drawback by the fact that early and late S-phases are superposed by the G₁ and G₂-M phases. Under normal cycling circumstances though, only a minor fraction of S-phase cells might be hidden by the G₁ and G₂-M peaks and can be quite reliably extrapolated by conventional cell cycle analysis software. However, when using cell cycle–modifying chemicals such as FdUrd, simple propidium iodide staining might be less adequate for such analyses.

We and others have observed repeatedly that FdUrd was able to induce a cell cycling block early in S phase (3–5). This was shown by double staining FACS analysis using iododeoxyuridine (IdUrd) and anti-IdUrd fluorescence staining combined with DNA staining by propidium iodide. In this analysis, all S-phase cells are stained based on incorporation of IdUrd into DNA and appear in a window different from that of G₁ and G₂-M phases, the latter being colored by propidium iodide staining alone. We thus showed in three different glioblastoma cell lines and an ovarian cancer line that cells accumulated in early S phase after FdUrd treatment (3, 4). Similar results have been reported for HT29 cells (5). The fact that these cells were ready to incorporate IdUrd indicates to us that they had already passed the primary “restriction” G₁ checkpoint.

Concerning the article by Fischer et al. (1), we therefore think that the cell cycle analysis shown in Fig. 2 is not reliable and possibly resulted in a major overestimation of G₁ phase and a correlated underestimation of the S-phase cells. This led to the particular difficulty of defining the phase of uracil-DNA glycosylase degradation, which, according to their discussion, remains hypothetically in late G₁-early S phase, possibly associated with a checkpoint mechanism. We would postulate that with a more adequate cell cycle analysis, the cell cycle arrest induced by FdUrd could have been better defined, would possibly be attributed to the early S phase, and would therefore better fit to their previous observation of uracil-DNA glycosylase expression. In our opinion, their results observed after aphidicolin G₁-S boundary block (1) might well fit to the latter interpretation. Could the authors possibly have been misled by a pitfall of cell cycle FACS analysis after simple propidium iodide staining?