AIF leaves mitochondria for the nucleus in aged yeast cells (bottom).

Yeast, as a unicellular organism, would seem to benefit most from self-preservation. But yeast altruism, in the form of apoptosis, is a new-found, if controversial, field of study. Many doubt the validity of experiments supporting programmed cell death in yeast and call for better controls. In this issue, the debate continues with two new articles that suggest that yeast cells do organize their own deaths—for the sake of their brethren.

Biologists often address how a phenomenon occurs, but seldom why. For the yeast apoptosis field, however, why is a painfully obvious question. On page 1055, Fabrizio et al. suggest a method to the madness. They show that yeast populations survive better in the long run when they initiate an early death program through superoxide.

Superoxides are produced by the everyday activities of life, but their mutational and death-inducing activities can be curtailed by superoxide dismutases (Sods). The authors find that Sods are normally down-regulated in older yeast cultures, which are surviving in nutrient-poor medium, leading to cell death. Mutants that circumvented this programmed death mechanism by maintaining high Sod activity had extended life spans. These long-lived populations, however, were unable to repopulate their culture once most of the cells died, a phenomenon known as adaptive regrowth. As a result, in competition experiments, strains that initiated early death eventually outgrew the wild type.

Early death probably allows the best-adapted cells in a population to reproduce before they are too old by using nutrients left behind by the dead yeasts. Superoxides are mutation inducers. The higher mutation rates that the authors noted in wild-type and other short-lived populations might impart surviving cells with the ability to use newly available nutrients rather than rely on the vanishing original supply. Mathematical models of adaptive regrowth based on experimental measurements of growth and mutation rates confirmed that long-lived populations are less able to survive under changing growth conditions.

The group has shown before that aging is similar among yeast and higher eukaryotes. The finding that yeast programs its own aging for altruistic reasons raises the pessimistic possibility that our own life span is best left similarly limited. The authors do not insist that superoxide-mediated death must be apoptotic, but hallmarks of mammalian apoptosis were seen, including chromatin condensation, cytosolic acidification, and extracellular exposure of phosphatidyl serine.

Another parallel with mammalian apoptosis, AIF-induced cell death, is revealed in an article by Wissing et al. (page 969). AIF is a mitochondrial protein that moves to the nucleus in response to apoptosis-inducing stimuli and then triggers DNA degradation. Wissing et al. show that similar translocation and degradation events occur in yeast apoptosis and that its loss delays age- or peroxide-induced death, accompanied by chromatin condensation and DNA fragmentation. Overexpression of yeast AIF sensitized cells to death in response to low levels of peroxide in a pathway that was partly dependent on the yeast caspase-like protein YCA1.

Many still remain skeptical, however, about the idea of yeast apoptosis. A recent report by Wysocki and Kron (J. Cell Biol. 166:311–316) suggested that some past evidence of apoptosis was artifactual and that death is independent of YCA1. Proponents of yeast apoptosis counter that Wysocki and Kron used extreme death-inducing conditions, which could lead to nonapoptotic or caspase-independent death. They say that the identification of so many yeast counterparts of mammalian apoptotic proteins, including YCA1, Cdc48, Cdc6, and the caspase regulator OMI, suggests that the pathway is ancient and conserved.

Things may be more complicated for some of the mammalian proteins implicated in apoptosis. Past evidence, for instance, suggests that the loss of AIF in mammalian cells makes the cells more, not less, susceptible to death.

If yeast do undergo apoptosis, scientists will be better able to do experiments that would be difficult in mammalian cells. Wissing et al. hope to find suppressors of AIF overexpression, for example. For now, it seems, they are still fighting to prove that apoptosis is a real phenomenon in yeast. We therefore await the next battle in this deadly war. ▪