James King-Holmes/Science Photo Library

Predicting natural death is generally impossible, save for those who study Caenorhabditis elegans: They know the precise moment that 131 cells, and only those 131 cells, are programmed to die. The timing and location of cell death is identical during the development of every tiny C. elegans worm. Nobel laureates John Sulston and Robert Horvitz discovered these cellular suicides in 1976 when they mapped the fate of C. elegans' 1,090 cells.

"It really did jump out," recalls Sulston. "We had no idea that we were going to be able to see [cell death], and that it would be one of the cell fates, being absolutely determined by the origin of the cell." That map, which is still the only complete cartography of a whole organism, spawned the field of programmed cell-death research, known as apoptosis, and highlights how important the regulation of cell death is to normal life.

DEADLY ORIGINS The researchers knew where and when the cells died, but they wanted to know how. Horvitz, who discovered numerous genes dedicated to killing cells, was a visionary, says Michael Hengartner, University of Zurich. "He was prescient at a time when nobody believed [cell death] was relevant or important."

C. elegans research owes much to fellow Nobel laureate Sydney Brenner, who selected the nematode in the early 1960s as an ideal model organism that would tie genes directly to development and behavior. Brenner liked the worm's simple nervous system and easy genetics. Worms are easy to grow, have a short life cycle, are hermaphrodites, and are transparent, allowing for easy histological study. The worm took researchers along paths that were as unpredictable as they were important; the conservation of apoptotic pathways across evolution has ramifications beyond a worm's fate.

Courtesy of Maxim Pharmaceuticals

LIFE SAVER: Inhibiting the caspase cascade prevents cell death, or apoptosis.

The worm's versatile genetics led to the isolation of cell death mutants and the discovery of genes that control the core death pathway. Working with Sulston, Ed Hedgecock, now at Johns Hopkins University, characterized the first death mutant, called ced-1, that accumulates cellular detritus (dead cell bodies) resulting from defective phagocytosis. Shortly thereafter, in 1986, Hillary Ellis visually screened compound ced-1 mutants; she was looking for worms in which cell death was blocked and no corpses were evident. This led her to the "executioner," which is encoded by the worm gene ced-3. This gene belongs to the caspase family, a group of conserved aspartate-directed cysteine proteases that act on a range of substrates. The combined actions of an activator and an inhibitor carefully regulate this enzyme. The activator Ced-4 (Apaf) induces Ced-3 activity, while the upstream ced-9 gene prevents death. Another killer gene, egl-1, acts by inhibiting ced-9 function.

The apoptosis machinery is remarkably conserved in nature; the three core elements (executioner, activator, and inhibitor) are found in species from worms to human. Mammals have more than a dozen caspases and many ced-9-like genes. Such genes are so highly conserved that cell-death proteins from one species can function in another. In 1992, David Vaux and Stuart Kim, Stanford University, demonstrated that overexpression of the human bcl-2 gene, a human apoptosis inhibitor first discovered as an oncogene in lymphomas, inhibited cell death in worms.¹ When Hengartner and Horvitz cloned the antiapoptotic ced-9 gene, they pulled out the worm's bcl-2 functional homolog.

HID, GRIM, AND REAPER Higher organisms have developed more complex apoptotic strategies than worms. Even though mammals have many versions of some nematode death proteins, it is unclear if all the elements act in the same way. Also, the sheer increase in cell number makes understanding apoptosis in developing mammals or flies that much more difficult. "[It's like] trying to figure out how a real car works by dissecting a toy car," says Hengartner. "You'll get the basics, but anything that is more sophisticated--FM radio, antilock brakes--you just don't have it in the toy car." The worm's fate maps showed that nematode apoptosis can be predicted, but "flies and humans have a highly variable number of cells, so the strategy of development is sloppier," says Rockefeller University's Hermann Steller. "[Flies and humans] overproduce cells [and] ... more signals are used to decide who lives or dies."

Steller and others uncovered new regulatory pathways by studying flies. Particularly interesting are regulators known as inhibitors of apoptosis protein (IAPs), first discovered in baculoviruses (pathogens that attack arthropods). Human IAPs are frequently overexpressed in tumor cells. Steller and others discovered in 1994 that fly proteins, called Reaper, Hid, and Grim, antagonize IAP activity. Steller compares IAPs to car brakes that prevent cells from dying too easily. But the brake must be released when cells need to die; Reaper and Grim do this by binding to IAPs and destroying them by ubiquitination.

Finding mammalian homo-logs of these IAP antagonists had been elusive until groups led by Vaux, now at the Walter and Eliza Hall Institute in Melbourne, and Xiaodong Wang, University of Texas Southwestern Medical Center, independently isolated the same IAP binding protein from mammalian cells. This IAP inhibitor, known as DIABLO (direct IAP binding protein of low Pi) or SMAC (second mitochondrial activator of cell death), antagonizes mammalian IAPs, inducing caspase-dependent apoptosis.^2,3

OF MICE AND HUMANS The link between apoptosis and diseases is attracting the pharmaceutical industry's attention. Tumor cells escape normal suicide programs, while unregulated cell deaths account for the symptoms of neurodegenerative disorders, including Alzheimer and Parkinson diseases. Molecules that block the killer proteins might be used to stop neuronal cell death in degenerative diseases. And Reaper-mimetics could potentially reduce IAP levels and induce apoptosis in tumors.

These efforts will involve taking the concepts learned in flies and worms and testing them in mice by targeted gene inactivation. "Flies will tell us lots of important things about cell death in humans, but to make the transition, we have to look at what pathways are conserved, and see if we can move in the direction of deriving molecules from Reaper to selectively kill cancer cells," says Steller.

Murine knockouts of apoptosis-related genes are providing new clues into the complexity of mammalian cell death pathways and ways in which they might be manipulated. For example, researchers are using these mice to examine the roles of the central core proteins caspase-9 and Apaf-1 in mammalian cell death. At first glance, caspase-9 and Afap-1 knockouts appeared as expected: showing severe abnormalities due to reduced apoptosis in the brain, leading to early postnatal death. However, Josef Penninger's group at the University of Toronto found that postmitotic neurons and lymphocytes die normally in rare surviving Apaf-1 -/- animals. In addition, recent studies by Andreas Strasser and colleagues at the Hall Institute have unveiled a caspase-9-independent apoptotic pathway in lymphocytes.⁴ Strasser's group worked with wild-type mice in which the blood-cell lineage alone was replaced with fetal liver cells from knockout mice. They found many apoptotic hallmarks in these animals' lymphocytes, including cleavage of caspase substrates, suggesting that caspases other than caspase-9 can become the executioner in some tissues.

LONG LIVE THE WORM Rockefeller University's Shai Shaham, taking a classical genetics approach, has found evidence that even though all worm cells make cell death proteins, only a handful succumb. Shaham has uncoupled cell death in worms from differentiation. He studied a set of neurons that is usually sexually dimorphic (fated to die in hermaphrodites but not in males) and found a mutant in which these cells develop into neurons in both genders, but live only in hermaphrodites. "This result argues that the control of cell death isn't a package deal that comes with a master specifier of cell fate," says Shaham. "Rather, there's something that's dedicated to making the decision whether to die or not, separate from the other differentiated features of that cell." How particular cells are chosen for death remains a mystery.

Then, there is the undertaker's role. When a cell dies, it is quickly removed from the animal. Phagocytosis is gaining attention, despite initial skepticism, because the failure to eliminate dead cells could underlie certain autoimmune disorders. Hengartner says that more worm genes affect phagocytosis than cell death itself. "Initially people said, just like [they said about] cell death, 'It's boring, it's garbage-can science'.... But ... there's probably much more there than meets the eye."

Differences between apoptosis in worms and in higher organisms need further study. "The questions to be resolved are whether those differences reflect a lack of understanding from both sides or a different scenario," says Horovitz. Predicting death might be as capricious as ever, but the lessons from apoptosis research have taught the scientific community much about life itself.

Laura DeFrancesco (defrancesco1@earthlink.net) is a freelance writer in Pasadena, Calif.