Cell fusion to study nuclear-cytoplasmic interactions in endothelial cell apoptosis.

Studies examining the regulation of nuclear rearrangements during apoptosis have led to conflicting results. Cytoplasmic control of nuclear events has been strongly suggested by cell-free experimental systems. In contrast, strict cytoplasmic control cannot account for the results of fibroblast-thymocyte fusion experiments in which dexamethasone induction of polykaryons led only to thymocyte nuclear apoptosis. Unresolved by these fusion studies was whether fibroblast nuclei were indifferent to heterologous cytoplasmic signals. Our objective was to resolve this discrepancy using cell fusion in a homologous system. Our strategy was to fuse endothelial cells with high levels of susceptibility to the induction of apoptosis (log phase cells arrested in G1 for 48 hours by isoleucine deprivation) with those manifesting low levels of susceptibility (serum-deprived, G0). Resultant fused and unfused cells were induced to undergo apoptosis by incubation with tumor necrosis factor-alpha and cycloheximide. Depending on the parental cell of origin, between 14 and 30% of dikaryons contained one apoptotic and one intact nucleus, indicating that strict cytoplasmic control was not occurring. In accord with this, the total frequency of nuclear apoptosis was unchanged after fusion. However, the distribution of apoptotic nuclei revealed a pronounced cytoplasmic influence, with a two- to fivefold increase in coordinate nuclear behavior. This pattern of nuclear apoptosis was consistent with a model of control in which both the state of nuclear susceptibility to apoptosis and expression of cytoplasmic pro-apoptotic regulators determined whether nuclear apoptosis would eventuate.