Cell-cycle studies by multiparametric automatic scanning of topographically preserved cells in culture

Abstract

The authors have developed a new methodology for characterizing in situ the cell cycle of human mammary epithelial cell lines. Using a SAMBA 200 cell image processor (scanning cytometry), 15 densitometric and textural parameters were computed on each Feulgen-stained nucleus. Parameters computed from the grey level cooccurrence and run-length section matrices allowed assessment of the chromatin pattern. Multiparametric analysis of data defined: 1) the relative position of each cell; 2) the relative positions of groups of cells, each group corresponding to a definite phase of the cell cycle; and 3) the function of these parameters best separating these phases. Files then were constructed for each phase: G0/G1, S, G2/ and M. Using these three files as a reference to classify cells, it was possible to ascertain the phase of the cell cycle for each cell of a population. The MDA AG human cell line synchronized by mitotic selection was used as a model to develop this method. The criteria used to assign cells to G0/G1, S, or G2 was DNA content. Classification in M phase was achieved by visual identification of mitotic cells. This method was checked on unsynchronized MDA AG and then applied to other human cell lines (MDA MB231, MCF-7, T47D C111). Comparison of results obtained by scanning cytometry and flow cytometry showed the proportion of cells assigned to G0/G1, S, and G2/M by the two methods to be similar. This new method removes some of the limitations of flow cytometry by 1) allowing visual verification of each cell analyzed; 2) lowering the number of cells required for study; 3) discriminating between G2 and M; and 4) preserving cell topography.