Fig. 3. Analysis of cycle progression by time lapse microscopy. Cells were plated out in a grid, photographed as described in Materials and Methods, then fixed and stained at 18 days. (A,B) Cardiomyocytes may divide more than once. (A) A field where a cardiomyocyte divided twice and another did not divide. The number at the top right of each picture shows the number of days elapsed since the cells were plated. A schematic drawing of the field is shown on the top left part of the figure. Note that the cell on the left at 8 days gives rise to 2 daughter cells at 10 days. Each of the daughter cells divides once more at 12 days. All the progeny cells stain positive for MyHC (lower right), but the staining is less strong and organised (red arrow) than in the non-dividing cell (white arrow). (B) Tree diagrams (one per starting cell) for the proliferation of cardiomyocytes which divided more than once giving rise to mononucleate progeny. Cells that divided only once (n=17) are not represented. Pedigrees were constructed from the time-lapse analysis of three independent cultures and they illustrate the diversity observed. All the cells were followed for 18 days (to simplify the scheme, lines for each cell stop in the last division observed during the 18 days). The arrows indicate clones which continued to divide; using this experimental set-up it was impossible to trace those divisions as accurately as all the others represented. The cross indicates that a cell died. (C) Example of a field where a cardiomyocyte underwent acytokinetic mitosis, thus becoming multinucleate. (D) Same experiment as in B. Pedigrees of cardiomyocytes that gave rise to multinucleate progeny. Cells that underwent only one cycle are not represented. Note that the mitotic division of a multinucleate cell can have several different outcomes. Scale bars: 50 µm.