Last year's top-cited paper, from Xiaodong Wang's team from University of Texas Southwestern Medical Center at Dallas, is at #8 in the current Top Ten, while 1997's runner-up, from Donald D. Newmeyer and his colleagues at the La Jolla Institute for Allergy and Immunology, entered the list at #7. Both teams demonstrated, in quite similar ways, that apoptosis requires the release of cytochrome c from mitochondria, and that Bcl-2, a protein that can rescue cells from apoptosis, does so by keeping cytochrome c within the mitochondrial membranes. Another step in the cell's elaborate swan song has thus been identified.

   The natural history of cell death, in mammalian cells, includes a movement of cytochrome c from inside the mitochondria to the cytosol, in which all the cell's organelles are suspended. It starts within an hour of the signals that trigger cell death, and generally accompanies cell death. But is this process a cause or an effect of apoptosis?

   Wang's team homed in on Bcl-2, a protein that sits on the outside of mitochondria and seems to act upstream of yet another apoptosis messenger, CPP-32 (see Science Watch, 7[6]:8, November/December 1996). CPP-32 is a member of the family of enzymes now called caspases, which amplify the apoptosis signal and whose activation seems to lead irrevocably to cell death. Wang and his colleagues transfected human cells with additional copies of the gene for Bcl-2, overexpression of which protects cells from apoptosis, and triggered apoptosis. Cells with additional Bcl-2 showed no CPP-32 activity, and all their cytochrome c stayed within the mitochondria. The cells did not die. Cells transfected only with a marker leaked cytochrome c and died within a matter of hours.

   In paper #7, Newmeyer and principal collaborator Douglas R. Green used a similar approach in a different system. Xenopus frog eggs provided the cellular machinery, which responded to additional Bcl-2 by blocking the release of cytochrome c and halting the customary sequence of apoptosis. Adding extra cytochrome c to the preparation, however, started cell death up again, and no amount of Bcl-2 could actually reverse the effects of cytochrome c once it had leaked out of the mitochondria.

   What is happening? Both teams point out that one member of the Bcl family is structurally related to bacterial proteins that form ion channels across membranes; Bcl-2 most likely has a similar structure, which would imply that it too can open channels across the mitochondrial membrane. These channels do not directly allow cytochrome c out of the mitochondria, but they clearly alter the membrane properties in such a way that the cytochrome c does escape. Once in the cytosol, the cytochrome then interacts with CPP-32 or one of the other caspases to further enhance its activity.

   In late 1996 Science Watch posed the question "Is the activation of CPP-32 the final step off the edge for the cell, or is rescue possible?" No answer is available yet, but it does look as if cells can be saved before CPP-32 kicks into action. In mice, Fas is a cell-surface antigen that is part of the apoptosis signal chain. Exposing mice to an anti-Fas antibody causes massive apoptosis in liver cells. Two groups have shown that caspase inhibitors prevent the Fas-induced apoptosis of liver cells (see I. Rodriguez, et al., J. Exp. Med., 184, 2067-72, 1996; and N. Rouquet, et al., Curr. Biol., 6:1192-5, 1996). And Virginie Joulin, a senior researcher with the French national research organization INSERM, has gone on to show that transgenic mice that make the human Bcl-2 product in their liver cells are also protected.