- Email: [email protected]
Programmed cell death in Drosophila
Neuron,Vol. 13, 1269-1274,December,1994,Copyright© 1994by Cell Press
Programmed Cell Death in Drosophila Hermann Steller and Megan E. Grether Howard Hughes Medical Institute Department of Brain and Cognitive Sciences Department of Biology Massachusetts Institute of Technology Cambridge, Massachusetts 02139
During development of the vertebrate nervous system, a large excess of neurons is generated that is subsequently removed by naturally occurring or programmed cell death (PCD; reviewed in Cowan, 1970; Cowan et al., 1984; Oppenheim, 1991; Raft, 1992). These cell deaths display a characteristic ultrastructural morphology, termed apoptosis, that is distinct from the necrotic cell death caused by external insult (Wyllie et al., 1980; Arends and Wyllie, 1991). The induction of apoptosis may often, though not always, require de novo RNA and protein synthesis, suggesting that this process represents an active, genedirected program (Martin et al., 1988; Oppenheimet al., 1990; Raft, 1992). Genetic studies in the nematode Caenorhabditis elegans have led to the identification of specific genes that are required at distinct steps in the cell death program, and these genes have been ordered into a genetic pathway for cell death in this organism (reviewed in Ellis et al., 1991; Hengartner and Horvitz, 1994a, 1994b). Two of the C. elegans cell death genes that control the initiation of death, ced-3 and ced-9, are structurally and functionally homologous to mammalian genes believed to be important for cell death. The ced-3 gene shows significant similarity to a family of cysteine proteases that includes interleukin-ll3 converting enzyme (ICE), and ced-9 appears to be a member of the bcl-2 family (reviewed in Hengartner and Horvitz, 1994a, 1994b). Significantly, overexpression of the human bcl-2 gene protects against PCD in nematodes and can, at least partially, substitute for the loss of ced-9 function (Vaux et al., 1992; Hengartner and Horvitz, 1994c). This indicates that at least some of the components of the apoptotic program have been conserved throughout animal evolution. The idea that apoptosis occurs by an evolutionarily conserved mechanism has received further support from the observation that expression of the baculovirus p35 gene can protect against PCD in nematodes, insects, and mammalian systems (Rabizadeh et al., 1993; Sugimoto et al., 1994; Hay et al., 1994). It appears that the basic cell death program is present in essentially all mammalian cells at all times, and that its activation must be continuously suppressed by extracellular survival signals in order for cells to live (Raft, 1992;Jacobson et al., 1994; Raft et al., 1994).Traditionally, the best studied examples of such survival factors have been the neurotrophins of the nerve growth factor (NGF) family (reviewed in Lindsay et al., 1994; Snider, 1994). Neurons that fail to participate in
Review
functional neuronal circuits are eliminated via PCD (reviewed in Purves and Lichtman, 1985; Purves, 1988; Oppenheim, 1991). Despite recent progress in identifying genes involved in PCD, the precise mechanism of apoptosis is still unknown. In addition, the biochemical pathways that regulate the activation of the cell death program remain to be elucidated. Molecular genetic studies in the fruitfly, Drosophila melanogaster, hold considerable promise for advancing our knowledge in these areas. In Drosophila, as in vertebrates, PCD is under epigenetic control. The number of neurons in the Drosophila nervous system is not genetically predetermined and may vary widely depending on environmental circumstances (e.g., Power, 1943; Nordlander and Edwards, 1968; Fischbach and Technau, 1984; Steller et al., 1987). The elimination of surplus cells by PCD is a major mechanism for determining appropriate cell numbers among interacting populations of neurons (reviewed below). Furthermore, in holometabolous insects, such as Drosophila and Manduca sexta, the nervous system undergoes extensive remodeling during metamorphosis to accommodate the physical and behavioral differences between the larval and adult animal. Again, PCD is widely exploited as a developmental mechanism for remodeling the nervous system. In Drosophila, the induction of cell death is influenced by a number of distinct signals, including the steroid hormone ecdysone (Kimura and Truman, 1990), cell-cell interactions (Fischbach and Technau, 1984; Steller et al., 1987; Magrassi and Lawrence, 1988; Wolff and Ready, 1991; Campos et al., 1992), ionizing radiation, and the arrest of differentiation in numerous mutants (e.g., Fristrom, 1969; Bryant, 1988; Magrassi and Lawrence, 1988; Abrams et al., 1993). In this review, we first summarize advances that have been made in understanding how extrinsic signals regulate the onset of PCD in particular cells and then discuss recent work on genes that regulate many, if not all, PCDs in Drosophila. These results indicate that many different death-inducing signals converge to activate a common cell death pathway.
The Steroid Hormone Ecdysone Controls PCD during Metamorphosis Holometabolous insects, such as Drosophila and Manduca, undergo two major phases of development. Initially, embryonic development transforms the fertilized egg into a worm-like larva (or caterpillar). Later on, postembryonic development or metamorphosis leads to the conversion of the larva into the mature adult form. Traditionally, most studies of PCD in insects have focused on metamorphosis because of the prominence of cell death at this stage (reviewed in Truman, 1992; Truman et al., 1992). Complete metamorphosis occurs in two major phases. In the initial
Neuron 1270
transition from larva to pupa, a crude adult form is produced. This is then followed by maturation of the pupa into an adult. Both of these stages involve largescale cell death. During the larval-pupal transition, larval structures that are no longer needed are removed by cell death. For example, neurons and muscles in the abdomen are utilized for larval locomotion but are no longer needed in the adult, which relies on thoracic structures such as legs and wings to move around. Thus, the extensive abdominal musculature and the neurons that control it are eliminated by PCD during pupation (reviewed in Truman, 1992; Truman et al., 1992). Another round of neuronal cell death occurs much later, once the adult has emerged. Here, PCD is used to eliminate any larval cells that were maintained during metamorphosis, as well as those structures required only for adult emergence itself (Truman, 1983; Kimura and Truman, 1990; Robinow et al., 1993). Both early and late metamorphic cell deaths are controlled by changes in the level of the steroid hormone ecdysone, though in opposite ways. The initial wave of cell death appears to be triggered by the surge of ecdysone at pupation (Truman and Schwartz, 1984; Truman et al., 1992). For example, the death of a particular neuron in Manduca can be prevented if a ligature is placed between the thorax (the source of ecdysteroid) and the abdomen. The death of this neuron can be restored with exogenously supplied ecdysteroid (Weeks and Truman, 1985), indicating that ecdysone acts as a death-promoting signal in this case. Interestingly, ecdysteroid levels have the opposite effect on postmetamorphic cell death. Upon eclosion of the adult, ecdysone levels drop, and it is this decline that induces the late wave of cell death (Truman and Schwartz, 1984; Robinow et al., 1993). Injection of 20-hydroxyecdysone can block normally occurring postmetamorphic cell death (Robinow et al., 1993). How do the ecdysteroids act to mediate PCD? Progress in this area has been made upon cloning of the Drosophila ecdysone receptor (EcR) gene (Koelle et al., 1991). The EcR protein is a member of the steroid receptor superfamily (reviewed in Evans, 1988), and binding of ecdysone is dependent on heterodimerization of EcR with another steroid receptor superfamily member encoded bythe ultraspiracle gene (Yao et al., 1992; Thomas et al., 1993). The EcRgene encodes three different isoforms, EcR-A, EcR-B1, and EcR-B2 (Talbot et al., 1993). While the DNA- and hormone-binding domains are common, each of the three receptor isoforms has a distinct N-terminus. Using isoformspecific monoclonal antibodies, Talbot et al. (1993) showed that EcR-A and EcR-B1 are expressed with distinct spatial and temporal profiles during metamorphosis. Interestingly, almost all the cells in the CNS that will undergo postmetamorphic cell death express high levels of the EcR-A protein (Robinow et al., 1993). This selective hyperexpression of the EcR-A isoform in doomed neurons may explain the apparent paradox of why only some neurons die in response to a
decline in hormone titer. On the other hand, elevated levels of EcR-A expression cannot be sufficient for determining which neurons die, since two neurons, n6 and n7, survive despite their high EcR-A expression levels. Furthermore, cell death can be delayed in the absence of ecdysone by enforcing sustained ecdysis behavior (by trapping the insect in a pupal case; Truman, 1983; Kimura and Truman, 1990). Finally, decapitation of adults soon after eclosion will prevent postrnetamorphic death in the CNS, even though ecdysone titers will decline under these conditions. This has led to the postulation of a "head factor" that is required for death. The molecular nature of this factor remains unknown, but it is possible that neuronal activity is required for the death-inducing activity. Truman et al. (1992) have proposed a two step model for how EcR-A acts to effect PCD. In this model, the binding of ecdysteroid to EcR-A induces the transcription of one or several early genes that are transcriptional regulators of the genes needed to carry out a cell death program. The continued presence of an ecdysteroid-EcR-A receptor complex, however, also serves to suppress the expression of the death-related genes regulated by the hypothetical early gene(s). Onlywhen ecdysteroid levels drop is this suppression relieved, and the early gene(s) can activate transcription of death genes. The observation that treatment with either actinomycin D or cycloheximide can reduce the amount of cell death in the abdomen (Fahrbach et al., 1993) is consistent with the model's requirement of de novo RNA and protein synthesis to carry out postmetamorphic cell death. The relevant downstream targets of the ecdysone receptor complex have yet to be identified, but the cell death gene reaper (see below) is an excellent candidate for such a gene.
Trophic Interactions in the Drosophila Nervous System Trophic interactions among neurons have been studied widely in vertebrates (reviewed in Cowan, 1970; Purves, 1988; Raft, 1992), but similar phenomena have also been described in Drosophila. For example, the size of the optic ganglia, the CNS portion of the insect visual system, is always matched perfectly to the variable size of the retina (Power, 1943). There is a significant amount of cell death in the developing optic ganglia of wild-type animals, and the morphology of these deaths resembles mammalian apoptosis (Nordlander and Edwards, 1968; Fischbach, 1983; Fischbach and Technau, 1984). The amount of cell death is increased dramatically in mutants that interfere with retinal development (Fischbach, 1983; Fischbach and Technau, 1984; Steller et al., 1987). Apparently, the continued survival of neurons in the optic ganglia depends on retinal input, and neurons that fail to form functional neuronal circuits are eliminated by PCD. More recently, retrograde trophic effects in the Drosophila visual system have also been described (Campos et al., 1992). Even though compound eyes
Review 1271
can develop autonomously, the continued survival of photoreceptor neurons depends on connections with the optic ganglia. In animals which have compound eyes that are not connected with the optic ganglia, photoreceptor neurons degenerate within a few days after eclosion (Campos et al., 1992). These results suggest that retinal neurons must make contacts with their targets to survive. The factors that mediate trophic interactions in Drosophila have not yet been identified. In particular, no homologs of the vertebrate neurotrophins have been isolated to date. On the other hand, two Drosophila tyrosine kinase genes have been cloned that are related to the trk family of neu rotrophin receptors (Lindsay et al., 1994; Snider, 1994), Dtrk (Pulido et al., 1992) and Dror (Wilson et al., 1993). Both genes are expressed during nervous system development, but their precise developmental function and the nature of their ligands remain to be determined. Role of Cell-Cell Interactions in the Control of Cell Death and Cell Survival In Drosophila, cell-cell communication may act to prevent or promote the death of specific cell types. For example, mutations in the decapentaplegic (dpp) gene, which encodes a TGFI~ homolog, cause ectopic cell deaths, but these deaths can be prevented if mutant and wild-type tissues are cultured together (Bryant, 1988). This indicates that the survival of certain cells depends on dpp protein, and that this protein can be supplied as a diffusible survival factor by wildtype cells. In other instances, cell survival depends on more local interactions that appear to be mediated through direct cell-cell contacts. Magrassi and Lawrence (1988) have reported that cell death in fushi tarazu (ftz) mutants affects ftz-expressing cells as well as their neighbors, suggesting that local cell-cell interactions are necessary for cell survival. On the other hand, cell interactions may also promote the induction of specific cell deaths. Many of the natu rally occurring cell deaths in the developing Drosophila retina require the product of the irregular chiasm C-roughest (irreC-rst) gene. Mutations in this gene lead to the survival of cells that fail to occupy appropriate positions in the developing ommatidia. These cells would normally be eliminated by PCD (Wolff and Ready, 1991). The irreC-rst gene encodes a transmembrane cell surface protein that shares significant homology with the immunoglobulin gene superfamily (Ramos et al., 1993). Interestingly, several mutations in this gene that specifically block cell death lead to truncations of the intracellular C-terminal domain of the protein. This suggests that the irreC-rst protein serves as a suicide receptor for detecting and eliminating cells that fail to occupy appropriate positions in the developing retina. According to this model, the activated receptor would send a signal to induce cell death in mispositioned cells. In this regard, its action as a receptor for a death factor would be comparable to the Apo-l/Fas and tumor necrosis factor receptors
(Trauth et al., 1989; Yonehara et al., 1989; Itoh et al., 1991; Nagata, 1994). The ligand for the irreC-rst protein and its signal transduction mechanism are presently unknown. Signals That Act within Dying Cells In addition to signals from the extracellular environment, cell death in Drosophila is also influenced by intracellular signals that act autonomously within the dying cell. For example, the death of some of the neuroblasts in the insect embryo seems to depend on cell lineage (Bate, 1976; Truman and Bate, 1988; White et al., 1994). Another signal that appears to act within dying cells is the detection of damage caused by ionizing radiation (Abrams et al., 1993; White et al., 1994). Finally, numerous mutations that block cellular differentiation cause the death of those cells that fail to complete their developmental program (e.g., Fristrom, 1969; Dura et al., 1987; Magrassi and Lawrence, 1988; Smouse and Perrimon, 1990; Bonini et al., 1993; Abrams et al., 1993). Apparently, some unknown mechanism inside cells monitors their ability to differentiate terminally and activates the death program if they fail to do so. Another interesting class of mutants that cause death in cells which would normally live is represented bycertain mutations in phototransduction genes (Meyertholen et al., 1987; Johnson et al., 1982; Stark et al., 1983, 1989). These mutations cause light-dependent retinal degeneration, presumably as a result of "stress" that results from executing a defective phototransduction cascade. At this point, nothing is known about the molecular nature of any of the postulated cell-autonomous signaling pathways in Drosophila. Global Regulators of Cell Death Even though PCD in Drosophila can be induced by many different signals, recent evidence suggests that these different signals ultimately converge onto one common cell death pathway. The protein encoded by the baculovirus p35 gene offers protection against cell death in many different cell types and tissues of insects, nematodes, and mammalian cells (Rabizadeh et al., 1993; Sugimoto et al., 1994; Hay et al., 1994). In this regard, its action as a survival factor is reminiscent of the mammalian bcl-2 gene. However, p35 does not display any obvious sequence similarityto bc/-2 family members or any other host-encoded genes, and its mechanism of action remains to be determined. A genetic approach has been taken recently to screen a large fraction of the Drosophila genome for genes that are required for PCD (White et al., 1994). By examining the pattern of apoptosis in embryos homozygous for previously identified chromosomal deletions, it was possible to survey approximately half of the entire Drosophila genome for functions involved in PCD. A single region on the third chromosome (in chromosomal position 75C1,2) was found to be required for all cell deaths that normally occur in the Drosophila embryo (White et al., 1994). Embryos
Neuron 1272
homozygous for Df(3L)H99, the smallest cell death defective deletion available in this interval, contained many extra cells and did not undergo certain morphogenetic movements, but developed a segmented cuticle and began to move. Interestingly, H99 mutants were not only deficient in normal cell death but were also significantly protected against the induction of ectopic PCD by X-irradiation and against ectopic PCD that is otherwise seen in developmental mutants. Such a global blockade of PCD indicates that this deletion removes a function that is required for the induction of apoptosis in response to many different signals. Molecular analysis of the H99 interval has led to the isolation of a gene, reaper, that appears to serve a central control function for the induction of apoptosis (White et al., 1994). Transformation with reaper genomic or cDNA clones is sufficient for inducing apoptosis in H99 mutants. The gene encodes a small peptide of only 65 amino acids that shows no significant homologies with other known proteins and reveals nothing about its precise function in cell death. However, there are good reasons to believe that reaper is a regulator of PCD and not a cell death effector protein, i.e., that it does not bring about death by itself. This prediction is based on the observation that some cell death can be induced in the absence of reaper when H99 embryos are irradiated with very high doses of X-rays. Significantly, the few cells that die under these circumstances have the typical morphology of apoptosis and can be engulfed by macrophages. This indicates that the basic cell death program is intact but cannot be readily activated in H99 deletions. Consistent with the idea that reaper plays a central regulatory function for the activation of PCD, reaper mRNA is specifically expressed in cells that are doomed to die, preceding the onset of death by I-2 hr. In addition, X-irradiation of embryos leads to a rapid induction of reaper mRNA, followed by widespread cell death (Abrams, Lamblin, and Steller, unpublished data). Therefore, multiple signaling pathways for the activation of apoptosis in Drosophila converge onto the reaper gene. Finally, ectopic expression of reaper is sufficient to kill cells that would normally survive (White and Steller, unpublished data). These results indicate that most, if not all, PCDs in Drosophila occur byone common mechanism, and that multiple signaling pathways for the activation of cell death converge to induce reaper gene expression. According to this model, reaper expression would lead to the selective activation of cell death effector proteins that may be present but inactive in most, if not all, cells (Raff, 1992; Raft et al., 1994). Conceivably, such an activation could occur either by stimulating proteins that promote cell death, e.g., proteins of the ced-3/ICE family, or by inhibiting protective functions, such as members of the bcl-2 family. The exact mechanism by which reaper activates cell death remains to be determined, and progress in this area is likely to provide interesting new information about the molecular mechanism underlying apoptosis.
Conclusions
PCD in Drosophila can be induced by many different signals, yet all the available evidence indicates that many, if not all, of these different signaling pathways converge to activate a common death program. At this point, only a few of the relevant molecules have been characterized, and a formidable amount of work remains to be done to understand exactly how diverse signaling pathways activate the cell death program. Fortunately, Drosophila is open to a wide diversity of technical approaches in such an endeavor, most notably the use of genetics and molecular biology. In addition, a variety of cell culture, biochemical, and pharmacological techniques can be readily applied for studying cell death in this system. This provides for an extremely powerful combination of techniques to study apoptosis in an organism with a complexity that is intermediate between that of vertebrates and the nematode C. elegans, in which most of the genetic analysis of PCD has been carried out in the past. Because at least part of the apoptotic program has been conserved between worms, insects, and man, it seems likely that the lessons learned from studying PCD in Drosophila may also be applicable to mammalian systems. References
Abrams, J. M., White, K., Fessler, L. I., and Steller, H. (1993). Programmed cell death during Drosophila embryogenesis. Development 117, 29-43. Arends, M. J., and Wyllie, A. H. (1991).Apoptosis: mechanisms and role in pathology. Int. Rev. Exp. Pathol. 32, 223-254. Bate, C. M. (1976). Embryogenesisof an insect nervous system. I. A map of the thoracic and abdominal neuroblasts in Locusta migratoria. J. Embryol. Exp. Morphol. 35, 107-123. Bonini, N. M., Leiserson,W. M., and Benzer,S. (1993).The eyes absent gene: genetic control of cell survival and differentiation in the developing Drosophila eye. Cell 72, 379-395. Bryant, P. J. (1988).Localizedcell death caused by mutations in a Drosophila gene coding for a transforming growth factor-I~ homolog. Dev. Biol. 128, 386-395. Campos, A. R., Fischbach, K.-F., and Steller, H. (1992). Survival of photoreceptor neurons in the compound eye of Drosophila depends on connections with the optic ganglia. Development 114, 355-366. Cowan, W. M. (1970). Anterograde and retrograde transneuronal degeneration in the central and peripheral nervous system. In Contemporary Research Methods in Neuroanatomy, J. H. Naute and S. O. E. Ebbenson, eds. (Berlin: Springer-Verlag), pp. 215-
251. Cowan, W. M., Fawcett,J., O'Leary, D., and Stanfield, B. (1984). Regressiveevents in neurogenesis.Science225, 1258-1265. Dura, J.-M., Randsholt, N. B., Deatrick, J., Erk, I., Santamaria, P., Freeman, J. D., Freeman, S. J., Weddell, D., and Brock, H. W. (1987). A complex genetic locus, polyhomeotic, is required for segmental specification and epidermal development in D. melanogaster. Cell 51, 829-839. Ellis, R. E., Yuan, J., and Horvitz, H. R. (1991).Mechanisms and functions of cell death. Annu. Rev. Cell Biol. 7, 663-698. Evans, R. M., (1988).The steroid and thyroid hormone receptor superfamily. Science240, 889-895. Fahrbach, S. E., Choi, M. K., and Truman, J. W. (1993).Inhibitory effects of actinomycin D and cycloheximide on neuronal death in adult Manduca sexta. I. Neurobiol. 25, 59-69.
Review 1273
Fischbach, K.-F. (1983). Neural cell types surviving congenital sensory deprivation in the optic lobes of Drosophila melanogaster. Dev. Biol. 95, 1-18. Fischbach, K.-F.,and Technau, G. (1984).Cell degeneration in the developing optic lobes of the sine oculis and small-optic-lobes mutants of Drosophila melanogaster. Dev. Biol. 104, 219-239. Fristrom, D. (1969). Cellular degeneration in the production of some mutant phenotypes in Drosophila melanogaster. Mol. Gen. Genet. 103, 363-379. Hay, B. A., Wolff, T., and Rubin, G. M. (1994). Expression of bacuIovirus P35 prevents cell death in Drosophila. Development 120, 2121-2129. Hengartner, M. O., and Horvitz, H. R. (1994a). Programmed cell death in Caenorhabditis elegans. Curr. Opin. Genet. Dev. 4, 581586. Hengartner, M. O., and Horvitz, H. R. (1994b). The ins and outs of programmed cell death during (7..elegans development. Phil. Trans. R. Soc. (Lond.) B 345, 243-246. Hengartner, M. O., and Horvitz, H. R. (1994(:). C. elegans cell survival gene ced-9 encodes a functional homolog of the mammalian proto-oncogene bcl-2. Cell 76, 665-676. Itoh, N., Yonehara, S., Ishii, A., Yonehara, M., Mizushima, S.-I., Sameshima, M., Hase, A., Seto, Y., and Nagata, S. (1991). The polypeptide encoded by the cDNA for human cell surface antigen Fas can mediate apoptosis. Cell 66, 233-243. Jacobson, M. D., Burne, J. F., and Raft, M. C. (1994). Programmed cell death and Bcl-2 protection in the absence of a nucleus. EMBO J. 13, 1899-1910. Johnson, M. A., Frayer, K. L., and Stark, W. S. (1982). Characteristics of rdgA: mutants with retinal degeneration in Drosophila. J. Insect Physiol. 28, 233-242. Kimura, K.-i., and Truman, J. W. (1990). Postmetamorphic cell death in the nervous and muscular systems of Drosophila melanogaster. J. Neurosci. 10, 403-411. Koelle, M. R., Talbot, W. S., Segraves, W. A., Bender, M. T., Cherbas, P., and Hogness, D. S. (1991). The Drosophila EcR gene encodes an ecdysone receptor, a new member of the steroid receptor superfamily. Cell 67, 59-77. Lindsay, R. M., Wiegand, S. J., Altar, C. A., and DiStefano, P. S. (1994). Neu rotrophic factors: from molecule to man. Trends Neurosci. 17, 182-190. Magrassi, L., and Lawrence, P. A. (1988). The pattern of cell death in fushi tarazu, a segmentation gene of Drosophila. Development 104, 447-451. Martin, D. P., Sch midt, R. E., DiStefano, P. S., Lowry, O. H., Carter, J. G., and Johson, E. M. (1988). Inhibitors of protein synthesis and RNA synthesis prevent neuronal death caused by nerve growth factor deprivation. J. Cell Biol. 106, 829-844. Meyertholen, E. P., Stein, P. J., Williams, M. A., and Ostroy, S. E. (1987). Studies of the Drosophila norpA phototransduction mutant. II. Photoreceptor degeneration and rhodopsin maintenance. J. Comp. Physiol. 161, 793-798. Nagata, S. (1994). Apoptosis regulated by a death factor and its receptor: Fas ligand and Fas. Phil. Trans. R. Soc. (Lond.) B 345, 281-287.
Nordlander, R. H., and Edwards, J. S. (1968). Morphological cell death in the post-embryonic development of the insect optic lobes. Nature 218, 780-781. Oppenheim, R. W. (1991). Cell death during development of the nervous system. Annu. Rev. Neurosci. 14, 453-501.
M. (1992). Dtrk, a Drosophila gene related to the trk family of neurotrophin receptors, encodes a novel class of neural cell adhesion molecule. EMBO J. 11, 391-404. Purves, D. (1988). Body and Brain: ATrophic Theory of Neural Connections (Cambridge, Massachusetts: Harvard University Press). Puryes, D., and Lichtman, J. W. (1985). Principles of Neural Development. (Sunderland, Massachusetts: Sinauer Associates, Inc.). Rabizadeh, S., La Count, D. J., Friesen, P. D., and Bredesen, D. E. (1993).Expression of the baculovirus p35 gene inhibits mammalian cell death. J. Neurochem. 61, 2318-2321. Raft, M. C. (1992). Social controls on cell survival and cell death. Nature 356, 397-400. Raft, M. C., Barres, B. A., Burne, J. F., Coles, H. S. R., Ishizaki, Y., and Jacobson, M. D. (1994). Programmed cell death and the control of cell survival. Phil. Trans. R. Soc. (Lond.) B 345, 265268. Ramos, R. G. P., Igloi, G. L., Lichte, B., Baumann, U., Maier, D., Schneider, T., Brandst~itter, J. H., Fr6hlich, A., and Fischbach, K.-F. (1993).The irregular chiasm C-roughest locus of Drosophila, which affects axonal projections and programmed cell death, encodes a novel immunoglobulin-like protein. Genes Dev. 7, 2533-2547. Robinow, S., Talbot, W. S., Hogness, D. S., and Truman, J. W. (1993). Programmed cell death in the Drosophila CNS is ecdysone-regulated and coupled with a specific ecdysone receptor isoform. Development 119, 1251-1259. Snider, W. D. (1994). Functions of the neu rotrophins during nervous system development: what the knockouts are teaching us. Cell 77, 627-638. Smouse, D., and Perrimon, N. (1990). Genetic dissection of a complex neurological mutant, polyhomeotic, in Drosophila. Dev. Biol. 139, 169-185. Stark, W. S., Chen, D. M., Johnston, M. A., and Frayer, K. L. (1983). The rdgB gene of Drosophila: retinal degeneration in different alleles and inhibition by norpA. J. Insect Physiol. 29, 123-131. Stark, W. S., Sapp, S. R., and Carlson, S. D. (1989). Photoreceptor maintenance and degeneration in the norpA (no receptor potential-A) mutant of Drosophila melanogaster. J. Neurogenet. 5, 49-51. Steller, H., Fischbach, K.-F., and Rubin, G. M. (1987). disconnected: a locus required for neuronal pathway formation in the visual system of Drosophila. Cell 50, 1139-1153. Sugimoto, A., Friesen, P. D., and Rothman, J. H. (1994). Baculovirus p35 prevents developmentally regulated cell death and rescues a ced-9 mutant in the nematode Caenorhabditis elegans. EMBO J. 13, 2023-2028. Talbot, W. S., Swyryd, E.A., and Hogness, D. S. (1993). Drosophila tissues with different metamorphic responses to ecdysone express different ecdysone receptor isoforms. Cell 73, 1323-1337. Thomas, H. E., Stunnenberg, H. G., and Stewart, A. F. (1993). Heterodimerisation of the Drosophila ecdysone receptor with retinoid X receptor and ultraspiracle. Nature 362, 471-475. Trauth, B. C., Klas, C., Peters, A. M. l., Matzku, S., M611er, P., Falk, W., Debatin, K.-M., and Krammer, P. H. (1989). Monoclonal antibody-mediated tumor regression by induction of apoptosis. Science 245, 301-305. Truman, J. W. (1983). Programmed cell death in the nervous system of an adult insect. J. Comp. Neurol. 216, 445-452. Truman, J. W. (1992). Developmental neuroethology of insect metamorphosis. J, Neurobiol. 23, 1404-1422.
Oppenheim, R. W., Prevette, D., Tytell, M., and Homma, S. (1990). Naturally occurring and induced neuronal death in the chick embryo in vivo requires protein and RNA synthesis: evidence for the role of cell death genes. Dev. Biol. 138, 104-113.
Truman, J. W., and Bate, M. (1988). Spatial and temporal patterns of neurogenesis in the central nervous system of Drosophila melanogaster. Dev. Biol. 125, 145-157.
Power, M. E. (1943).The effect of reduction in numbers of ommatidia upon the brain of Drosophila melanogaster. J. Exp. Zool. 94, 33-72. Pulido, D., Campuzano, S., Koda, T., Modolell, J., and Barbacid,
Truman, J. W., and Schwartz, L. M. (1984). Steroid regulation of neural death in the moth nervous system. J. Neurosci. 4, 274280. Truman, J. W., Thorn, R. S., and Robinow, S. (1992). ProRrammed
Neuron 1274
neuronal death in insect development. J. Neurobiol. ~23, 12951311.
Vaux, D. L., Weissman, I. L., and Kim, S. K. (1992). Prevention of programmed cell death in Caenorhabditis elegans by human bcl-2. Science 258, 1955-1957. Weeks, J. C., and Truman, J. W. (1985). Independent steroid control of the fates of motoneurons and their muscles during insect metamorphosis. J. Neurosci. 5, 2290-2300. White, K., Grether, M., Abrams, J., Young, L., Farrell, K., and Steller, H. (1994). Genetic control of cell death in Drosophila. Science 264, 677-683. Wilson, C., Goberdhan, D. C. I., and Steller, H. (1993). Dror, a potential neurotrophic receptor gene, encodes a Drosophila homolog of the vertebrate Ror family of Trk-related receptor tyrosine kinases. Proc. Natl. Acad. Sci. USA 90, 7109-7113. Wolff, T., and Ready, D. F. (1991). Cell death in normal and rough eye mutants of Drosophila. Development 113, 825-839. Wyllie, A. H., Kerr, J. F. R., and Currie, A. R. (1980). Cell death: the significance of apoptosis. Int. Rev. Cytol. 68, 251-306. Yao, T.-P., Segraves, W. A., Oro, A. E., McKeown, M., and Evans, R. M. (1992).Drosophila ultraspiracle modulates ecdysone receptor function via heterodimer formation. Cell 71, 63-72. Yonehara, S., Ishii, A., and Yonehara, M. (1989). A cell-killing monoclonal antibody (anti-Fas) to a cell surface antigen codownregualted with the receptor for tumor necrosis factor. J. Exp. Med. 169, 1747-1756.
Programmed Cell Death in Drosophila Hermann Steller and Megan E. Grether Howard Hughes Medical Institute Department of Brain and Cognitive Sciences Department of Biology Massachusetts Institute of Technology Cambridge, Massachusetts 02139
During development of the vertebrate nervous system, a large excess of neurons is generated that is subsequently removed by naturally occurring or programmed cell death (PCD; reviewed in Cowan, 1970; Cowan et al., 1984; Oppenheim, 1991; Raft, 1992). These cell deaths display a characteristic ultrastructural morphology, termed apoptosis, that is distinct from the necrotic cell death caused by external insult (Wyllie et al., 1980; Arends and Wyllie, 1991). The induction of apoptosis may often, though not always, require de novo RNA and protein synthesis, suggesting that this process represents an active, genedirected program (Martin et al., 1988; Oppenheimet al., 1990; Raft, 1992). Genetic studies in the nematode Caenorhabditis elegans have led to the identification of specific genes that are required at distinct steps in the cell death program, and these genes have been ordered into a genetic pathway for cell death in this organism (reviewed in Ellis et al., 1991; Hengartner and Horvitz, 1994a, 1994b). Two of the C. elegans cell death genes that control the initiation of death, ced-3 and ced-9, are structurally and functionally homologous to mammalian genes believed to be important for cell death. The ced-3 gene shows significant similarity to a family of cysteine proteases that includes interleukin-ll3 converting enzyme (ICE), and ced-9 appears to be a member of the bcl-2 family (reviewed in Hengartner and Horvitz, 1994a, 1994b). Significantly, overexpression of the human bcl-2 gene protects against PCD in nematodes and can, at least partially, substitute for the loss of ced-9 function (Vaux et al., 1992; Hengartner and Horvitz, 1994c). This indicates that at least some of the components of the apoptotic program have been conserved throughout animal evolution. The idea that apoptosis occurs by an evolutionarily conserved mechanism has received further support from the observation that expression of the baculovirus p35 gene can protect against PCD in nematodes, insects, and mammalian systems (Rabizadeh et al., 1993; Sugimoto et al., 1994; Hay et al., 1994). It appears that the basic cell death program is present in essentially all mammalian cells at all times, and that its activation must be continuously suppressed by extracellular survival signals in order for cells to live (Raft, 1992;Jacobson et al., 1994; Raft et al., 1994).Traditionally, the best studied examples of such survival factors have been the neurotrophins of the nerve growth factor (NGF) family (reviewed in Lindsay et al., 1994; Snider, 1994). Neurons that fail to participate in
Review
functional neuronal circuits are eliminated via PCD (reviewed in Purves and Lichtman, 1985; Purves, 1988; Oppenheim, 1991). Despite recent progress in identifying genes involved in PCD, the precise mechanism of apoptosis is still unknown. In addition, the biochemical pathways that regulate the activation of the cell death program remain to be elucidated. Molecular genetic studies in the fruitfly, Drosophila melanogaster, hold considerable promise for advancing our knowledge in these areas. In Drosophila, as in vertebrates, PCD is under epigenetic control. The number of neurons in the Drosophila nervous system is not genetically predetermined and may vary widely depending on environmental circumstances (e.g., Power, 1943; Nordlander and Edwards, 1968; Fischbach and Technau, 1984; Steller et al., 1987). The elimination of surplus cells by PCD is a major mechanism for determining appropriate cell numbers among interacting populations of neurons (reviewed below). Furthermore, in holometabolous insects, such as Drosophila and Manduca sexta, the nervous system undergoes extensive remodeling during metamorphosis to accommodate the physical and behavioral differences between the larval and adult animal. Again, PCD is widely exploited as a developmental mechanism for remodeling the nervous system. In Drosophila, the induction of cell death is influenced by a number of distinct signals, including the steroid hormone ecdysone (Kimura and Truman, 1990), cell-cell interactions (Fischbach and Technau, 1984; Steller et al., 1987; Magrassi and Lawrence, 1988; Wolff and Ready, 1991; Campos et al., 1992), ionizing radiation, and the arrest of differentiation in numerous mutants (e.g., Fristrom, 1969; Bryant, 1988; Magrassi and Lawrence, 1988; Abrams et al., 1993). In this review, we first summarize advances that have been made in understanding how extrinsic signals regulate the onset of PCD in particular cells and then discuss recent work on genes that regulate many, if not all, PCDs in Drosophila. These results indicate that many different death-inducing signals converge to activate a common cell death pathway.
The Steroid Hormone Ecdysone Controls PCD during Metamorphosis Holometabolous insects, such as Drosophila and Manduca, undergo two major phases of development. Initially, embryonic development transforms the fertilized egg into a worm-like larva (or caterpillar). Later on, postembryonic development or metamorphosis leads to the conversion of the larva into the mature adult form. Traditionally, most studies of PCD in insects have focused on metamorphosis because of the prominence of cell death at this stage (reviewed in Truman, 1992; Truman et al., 1992). Complete metamorphosis occurs in two major phases. In the initial
Neuron 1270
transition from larva to pupa, a crude adult form is produced. This is then followed by maturation of the pupa into an adult. Both of these stages involve largescale cell death. During the larval-pupal transition, larval structures that are no longer needed are removed by cell death. For example, neurons and muscles in the abdomen are utilized for larval locomotion but are no longer needed in the adult, which relies on thoracic structures such as legs and wings to move around. Thus, the extensive abdominal musculature and the neurons that control it are eliminated by PCD during pupation (reviewed in Truman, 1992; Truman et al., 1992). Another round of neuronal cell death occurs much later, once the adult has emerged. Here, PCD is used to eliminate any larval cells that were maintained during metamorphosis, as well as those structures required only for adult emergence itself (Truman, 1983; Kimura and Truman, 1990; Robinow et al., 1993). Both early and late metamorphic cell deaths are controlled by changes in the level of the steroid hormone ecdysone, though in opposite ways. The initial wave of cell death appears to be triggered by the surge of ecdysone at pupation (Truman and Schwartz, 1984; Truman et al., 1992). For example, the death of a particular neuron in Manduca can be prevented if a ligature is placed between the thorax (the source of ecdysteroid) and the abdomen. The death of this neuron can be restored with exogenously supplied ecdysteroid (Weeks and Truman, 1985), indicating that ecdysone acts as a death-promoting signal in this case. Interestingly, ecdysteroid levels have the opposite effect on postmetamorphic cell death. Upon eclosion of the adult, ecdysone levels drop, and it is this decline that induces the late wave of cell death (Truman and Schwartz, 1984; Robinow et al., 1993). Injection of 20-hydroxyecdysone can block normally occurring postmetamorphic cell death (Robinow et al., 1993). How do the ecdysteroids act to mediate PCD? Progress in this area has been made upon cloning of the Drosophila ecdysone receptor (EcR) gene (Koelle et al., 1991). The EcR protein is a member of the steroid receptor superfamily (reviewed in Evans, 1988), and binding of ecdysone is dependent on heterodimerization of EcR with another steroid receptor superfamily member encoded bythe ultraspiracle gene (Yao et al., 1992; Thomas et al., 1993). The EcRgene encodes three different isoforms, EcR-A, EcR-B1, and EcR-B2 (Talbot et al., 1993). While the DNA- and hormone-binding domains are common, each of the three receptor isoforms has a distinct N-terminus. Using isoformspecific monoclonal antibodies, Talbot et al. (1993) showed that EcR-A and EcR-B1 are expressed with distinct spatial and temporal profiles during metamorphosis. Interestingly, almost all the cells in the CNS that will undergo postmetamorphic cell death express high levels of the EcR-A protein (Robinow et al., 1993). This selective hyperexpression of the EcR-A isoform in doomed neurons may explain the apparent paradox of why only some neurons die in response to a
decline in hormone titer. On the other hand, elevated levels of EcR-A expression cannot be sufficient for determining which neurons die, since two neurons, n6 and n7, survive despite their high EcR-A expression levels. Furthermore, cell death can be delayed in the absence of ecdysone by enforcing sustained ecdysis behavior (by trapping the insect in a pupal case; Truman, 1983; Kimura and Truman, 1990). Finally, decapitation of adults soon after eclosion will prevent postrnetamorphic death in the CNS, even though ecdysone titers will decline under these conditions. This has led to the postulation of a "head factor" that is required for death. The molecular nature of this factor remains unknown, but it is possible that neuronal activity is required for the death-inducing activity. Truman et al. (1992) have proposed a two step model for how EcR-A acts to effect PCD. In this model, the binding of ecdysteroid to EcR-A induces the transcription of one or several early genes that are transcriptional regulators of the genes needed to carry out a cell death program. The continued presence of an ecdysteroid-EcR-A receptor complex, however, also serves to suppress the expression of the death-related genes regulated by the hypothetical early gene(s). Onlywhen ecdysteroid levels drop is this suppression relieved, and the early gene(s) can activate transcription of death genes. The observation that treatment with either actinomycin D or cycloheximide can reduce the amount of cell death in the abdomen (Fahrbach et al., 1993) is consistent with the model's requirement of de novo RNA and protein synthesis to carry out postmetamorphic cell death. The relevant downstream targets of the ecdysone receptor complex have yet to be identified, but the cell death gene reaper (see below) is an excellent candidate for such a gene.
Trophic Interactions in the Drosophila Nervous System Trophic interactions among neurons have been studied widely in vertebrates (reviewed in Cowan, 1970; Purves, 1988; Raft, 1992), but similar phenomena have also been described in Drosophila. For example, the size of the optic ganglia, the CNS portion of the insect visual system, is always matched perfectly to the variable size of the retina (Power, 1943). There is a significant amount of cell death in the developing optic ganglia of wild-type animals, and the morphology of these deaths resembles mammalian apoptosis (Nordlander and Edwards, 1968; Fischbach, 1983; Fischbach and Technau, 1984). The amount of cell death is increased dramatically in mutants that interfere with retinal development (Fischbach, 1983; Fischbach and Technau, 1984; Steller et al., 1987). Apparently, the continued survival of neurons in the optic ganglia depends on retinal input, and neurons that fail to form functional neuronal circuits are eliminated by PCD. More recently, retrograde trophic effects in the Drosophila visual system have also been described (Campos et al., 1992). Even though compound eyes
Review 1271
can develop autonomously, the continued survival of photoreceptor neurons depends on connections with the optic ganglia. In animals which have compound eyes that are not connected with the optic ganglia, photoreceptor neurons degenerate within a few days after eclosion (Campos et al., 1992). These results suggest that retinal neurons must make contacts with their targets to survive. The factors that mediate trophic interactions in Drosophila have not yet been identified. In particular, no homologs of the vertebrate neurotrophins have been isolated to date. On the other hand, two Drosophila tyrosine kinase genes have been cloned that are related to the trk family of neu rotrophin receptors (Lindsay et al., 1994; Snider, 1994), Dtrk (Pulido et al., 1992) and Dror (Wilson et al., 1993). Both genes are expressed during nervous system development, but their precise developmental function and the nature of their ligands remain to be determined. Role of Cell-Cell Interactions in the Control of Cell Death and Cell Survival In Drosophila, cell-cell communication may act to prevent or promote the death of specific cell types. For example, mutations in the decapentaplegic (dpp) gene, which encodes a TGFI~ homolog, cause ectopic cell deaths, but these deaths can be prevented if mutant and wild-type tissues are cultured together (Bryant, 1988). This indicates that the survival of certain cells depends on dpp protein, and that this protein can be supplied as a diffusible survival factor by wildtype cells. In other instances, cell survival depends on more local interactions that appear to be mediated through direct cell-cell contacts. Magrassi and Lawrence (1988) have reported that cell death in fushi tarazu (ftz) mutants affects ftz-expressing cells as well as their neighbors, suggesting that local cell-cell interactions are necessary for cell survival. On the other hand, cell interactions may also promote the induction of specific cell deaths. Many of the natu rally occurring cell deaths in the developing Drosophila retina require the product of the irregular chiasm C-roughest (irreC-rst) gene. Mutations in this gene lead to the survival of cells that fail to occupy appropriate positions in the developing ommatidia. These cells would normally be eliminated by PCD (Wolff and Ready, 1991). The irreC-rst gene encodes a transmembrane cell surface protein that shares significant homology with the immunoglobulin gene superfamily (Ramos et al., 1993). Interestingly, several mutations in this gene that specifically block cell death lead to truncations of the intracellular C-terminal domain of the protein. This suggests that the irreC-rst protein serves as a suicide receptor for detecting and eliminating cells that fail to occupy appropriate positions in the developing retina. According to this model, the activated receptor would send a signal to induce cell death in mispositioned cells. In this regard, its action as a receptor for a death factor would be comparable to the Apo-l/Fas and tumor necrosis factor receptors
(Trauth et al., 1989; Yonehara et al., 1989; Itoh et al., 1991; Nagata, 1994). The ligand for the irreC-rst protein and its signal transduction mechanism are presently unknown. Signals That Act within Dying Cells In addition to signals from the extracellular environment, cell death in Drosophila is also influenced by intracellular signals that act autonomously within the dying cell. For example, the death of some of the neuroblasts in the insect embryo seems to depend on cell lineage (Bate, 1976; Truman and Bate, 1988; White et al., 1994). Another signal that appears to act within dying cells is the detection of damage caused by ionizing radiation (Abrams et al., 1993; White et al., 1994). Finally, numerous mutations that block cellular differentiation cause the death of those cells that fail to complete their developmental program (e.g., Fristrom, 1969; Dura et al., 1987; Magrassi and Lawrence, 1988; Smouse and Perrimon, 1990; Bonini et al., 1993; Abrams et al., 1993). Apparently, some unknown mechanism inside cells monitors their ability to differentiate terminally and activates the death program if they fail to do so. Another interesting class of mutants that cause death in cells which would normally live is represented bycertain mutations in phototransduction genes (Meyertholen et al., 1987; Johnson et al., 1982; Stark et al., 1983, 1989). These mutations cause light-dependent retinal degeneration, presumably as a result of "stress" that results from executing a defective phototransduction cascade. At this point, nothing is known about the molecular nature of any of the postulated cell-autonomous signaling pathways in Drosophila. Global Regulators of Cell Death Even though PCD in Drosophila can be induced by many different signals, recent evidence suggests that these different signals ultimately converge onto one common cell death pathway. The protein encoded by the baculovirus p35 gene offers protection against cell death in many different cell types and tissues of insects, nematodes, and mammalian cells (Rabizadeh et al., 1993; Sugimoto et al., 1994; Hay et al., 1994). In this regard, its action as a survival factor is reminiscent of the mammalian bcl-2 gene. However, p35 does not display any obvious sequence similarityto bc/-2 family members or any other host-encoded genes, and its mechanism of action remains to be determined. A genetic approach has been taken recently to screen a large fraction of the Drosophila genome for genes that are required for PCD (White et al., 1994). By examining the pattern of apoptosis in embryos homozygous for previously identified chromosomal deletions, it was possible to survey approximately half of the entire Drosophila genome for functions involved in PCD. A single region on the third chromosome (in chromosomal position 75C1,2) was found to be required for all cell deaths that normally occur in the Drosophila embryo (White et al., 1994). Embryos
Neuron 1272
homozygous for Df(3L)H99, the smallest cell death defective deletion available in this interval, contained many extra cells and did not undergo certain morphogenetic movements, but developed a segmented cuticle and began to move. Interestingly, H99 mutants were not only deficient in normal cell death but were also significantly protected against the induction of ectopic PCD by X-irradiation and against ectopic PCD that is otherwise seen in developmental mutants. Such a global blockade of PCD indicates that this deletion removes a function that is required for the induction of apoptosis in response to many different signals. Molecular analysis of the H99 interval has led to the isolation of a gene, reaper, that appears to serve a central control function for the induction of apoptosis (White et al., 1994). Transformation with reaper genomic or cDNA clones is sufficient for inducing apoptosis in H99 mutants. The gene encodes a small peptide of only 65 amino acids that shows no significant homologies with other known proteins and reveals nothing about its precise function in cell death. However, there are good reasons to believe that reaper is a regulator of PCD and not a cell death effector protein, i.e., that it does not bring about death by itself. This prediction is based on the observation that some cell death can be induced in the absence of reaper when H99 embryos are irradiated with very high doses of X-rays. Significantly, the few cells that die under these circumstances have the typical morphology of apoptosis and can be engulfed by macrophages. This indicates that the basic cell death program is intact but cannot be readily activated in H99 deletions. Consistent with the idea that reaper plays a central regulatory function for the activation of PCD, reaper mRNA is specifically expressed in cells that are doomed to die, preceding the onset of death by I-2 hr. In addition, X-irradiation of embryos leads to a rapid induction of reaper mRNA, followed by widespread cell death (Abrams, Lamblin, and Steller, unpublished data). Therefore, multiple signaling pathways for the activation of apoptosis in Drosophila converge onto the reaper gene. Finally, ectopic expression of reaper is sufficient to kill cells that would normally survive (White and Steller, unpublished data). These results indicate that most, if not all, PCDs in Drosophila occur byone common mechanism, and that multiple signaling pathways for the activation of cell death converge to induce reaper gene expression. According to this model, reaper expression would lead to the selective activation of cell death effector proteins that may be present but inactive in most, if not all, cells (Raff, 1992; Raft et al., 1994). Conceivably, such an activation could occur either by stimulating proteins that promote cell death, e.g., proteins of the ced-3/ICE family, or by inhibiting protective functions, such as members of the bcl-2 family. The exact mechanism by which reaper activates cell death remains to be determined, and progress in this area is likely to provide interesting new information about the molecular mechanism underlying apoptosis.
Conclusions
PCD in Drosophila can be induced by many different signals, yet all the available evidence indicates that many, if not all, of these different signaling pathways converge to activate a common death program. At this point, only a few of the relevant molecules have been characterized, and a formidable amount of work remains to be done to understand exactly how diverse signaling pathways activate the cell death program. Fortunately, Drosophila is open to a wide diversity of technical approaches in such an endeavor, most notably the use of genetics and molecular biology. In addition, a variety of cell culture, biochemical, and pharmacological techniques can be readily applied for studying cell death in this system. This provides for an extremely powerful combination of techniques to study apoptosis in an organism with a complexity that is intermediate between that of vertebrates and the nematode C. elegans, in which most of the genetic analysis of PCD has been carried out in the past. Because at least part of the apoptotic program has been conserved between worms, insects, and man, it seems likely that the lessons learned from studying PCD in Drosophila may also be applicable to mammalian systems. References
Abrams, J. M., White, K., Fessler, L. I., and Steller, H. (1993). Programmed cell death during Drosophila embryogenesis. Development 117, 29-43. Arends, M. J., and Wyllie, A. H. (1991).Apoptosis: mechanisms and role in pathology. Int. Rev. Exp. Pathol. 32, 223-254. Bate, C. M. (1976). Embryogenesisof an insect nervous system. I. A map of the thoracic and abdominal neuroblasts in Locusta migratoria. J. Embryol. Exp. Morphol. 35, 107-123. Bonini, N. M., Leiserson,W. M., and Benzer,S. (1993).The eyes absent gene: genetic control of cell survival and differentiation in the developing Drosophila eye. Cell 72, 379-395. Bryant, P. J. (1988).Localizedcell death caused by mutations in a Drosophila gene coding for a transforming growth factor-I~ homolog. Dev. Biol. 128, 386-395. Campos, A. R., Fischbach, K.-F., and Steller, H. (1992). Survival of photoreceptor neurons in the compound eye of Drosophila depends on connections with the optic ganglia. Development 114, 355-366. Cowan, W. M. (1970). Anterograde and retrograde transneuronal degeneration in the central and peripheral nervous system. In Contemporary Research Methods in Neuroanatomy, J. H. Naute and S. O. E. Ebbenson, eds. (Berlin: Springer-Verlag), pp. 215-
251. Cowan, W. M., Fawcett,J., O'Leary, D., and Stanfield, B. (1984). Regressiveevents in neurogenesis.Science225, 1258-1265. Dura, J.-M., Randsholt, N. B., Deatrick, J., Erk, I., Santamaria, P., Freeman, J. D., Freeman, S. J., Weddell, D., and Brock, H. W. (1987). A complex genetic locus, polyhomeotic, is required for segmental specification and epidermal development in D. melanogaster. Cell 51, 829-839. Ellis, R. E., Yuan, J., and Horvitz, H. R. (1991).Mechanisms and functions of cell death. Annu. Rev. Cell Biol. 7, 663-698. Evans, R. M., (1988).The steroid and thyroid hormone receptor superfamily. Science240, 889-895. Fahrbach, S. E., Choi, M. K., and Truman, J. W. (1993).Inhibitory effects of actinomycin D and cycloheximide on neuronal death in adult Manduca sexta. I. Neurobiol. 25, 59-69.
Review 1273
Fischbach, K.-F. (1983). Neural cell types surviving congenital sensory deprivation in the optic lobes of Drosophila melanogaster. Dev. Biol. 95, 1-18. Fischbach, K.-F.,and Technau, G. (1984).Cell degeneration in the developing optic lobes of the sine oculis and small-optic-lobes mutants of Drosophila melanogaster. Dev. Biol. 104, 219-239. Fristrom, D. (1969). Cellular degeneration in the production of some mutant phenotypes in Drosophila melanogaster. Mol. Gen. Genet. 103, 363-379. Hay, B. A., Wolff, T., and Rubin, G. M. (1994). Expression of bacuIovirus P35 prevents cell death in Drosophila. Development 120, 2121-2129. Hengartner, M. O., and Horvitz, H. R. (1994a). Programmed cell death in Caenorhabditis elegans. Curr. Opin. Genet. Dev. 4, 581586. Hengartner, M. O., and Horvitz, H. R. (1994b). The ins and outs of programmed cell death during (7..elegans development. Phil. Trans. R. Soc. (Lond.) B 345, 243-246. Hengartner, M. O., and Horvitz, H. R. (1994(:). C. elegans cell survival gene ced-9 encodes a functional homolog of the mammalian proto-oncogene bcl-2. Cell 76, 665-676. Itoh, N., Yonehara, S., Ishii, A., Yonehara, M., Mizushima, S.-I., Sameshima, M., Hase, A., Seto, Y., and Nagata, S. (1991). The polypeptide encoded by the cDNA for human cell surface antigen Fas can mediate apoptosis. Cell 66, 233-243. Jacobson, M. D., Burne, J. F., and Raft, M. C. (1994). Programmed cell death and Bcl-2 protection in the absence of a nucleus. EMBO J. 13, 1899-1910. Johnson, M. A., Frayer, K. L., and Stark, W. S. (1982). Characteristics of rdgA: mutants with retinal degeneration in Drosophila. J. Insect Physiol. 28, 233-242. Kimura, K.-i., and Truman, J. W. (1990). Postmetamorphic cell death in the nervous and muscular systems of Drosophila melanogaster. J. Neurosci. 10, 403-411. Koelle, M. R., Talbot, W. S., Segraves, W. A., Bender, M. T., Cherbas, P., and Hogness, D. S. (1991). The Drosophila EcR gene encodes an ecdysone receptor, a new member of the steroid receptor superfamily. Cell 67, 59-77. Lindsay, R. M., Wiegand, S. J., Altar, C. A., and DiStefano, P. S. (1994). Neu rotrophic factors: from molecule to man. Trends Neurosci. 17, 182-190. Magrassi, L., and Lawrence, P. A. (1988). The pattern of cell death in fushi tarazu, a segmentation gene of Drosophila. Development 104, 447-451. Martin, D. P., Sch midt, R. E., DiStefano, P. S., Lowry, O. H., Carter, J. G., and Johson, E. M. (1988). Inhibitors of protein synthesis and RNA synthesis prevent neuronal death caused by nerve growth factor deprivation. J. Cell Biol. 106, 829-844. Meyertholen, E. P., Stein, P. J., Williams, M. A., and Ostroy, S. E. (1987). Studies of the Drosophila norpA phototransduction mutant. II. Photoreceptor degeneration and rhodopsin maintenance. J. Comp. Physiol. 161, 793-798. Nagata, S. (1994). Apoptosis regulated by a death factor and its receptor: Fas ligand and Fas. Phil. Trans. R. Soc. (Lond.) B 345, 281-287.
Nordlander, R. H., and Edwards, J. S. (1968). Morphological cell death in the post-embryonic development of the insect optic lobes. Nature 218, 780-781. Oppenheim, R. W. (1991). Cell death during development of the nervous system. Annu. Rev. Neurosci. 14, 453-501.
M. (1992). Dtrk, a Drosophila gene related to the trk family of neurotrophin receptors, encodes a novel class of neural cell adhesion molecule. EMBO J. 11, 391-404. Purves, D. (1988). Body and Brain: ATrophic Theory of Neural Connections (Cambridge, Massachusetts: Harvard University Press). Puryes, D., and Lichtman, J. W. (1985). Principles of Neural Development. (Sunderland, Massachusetts: Sinauer Associates, Inc.). Rabizadeh, S., La Count, D. J., Friesen, P. D., and Bredesen, D. E. (1993).Expression of the baculovirus p35 gene inhibits mammalian cell death. J. Neurochem. 61, 2318-2321. Raft, M. C. (1992). Social controls on cell survival and cell death. Nature 356, 397-400. Raft, M. C., Barres, B. A., Burne, J. F., Coles, H. S. R., Ishizaki, Y., and Jacobson, M. D. (1994). Programmed cell death and the control of cell survival. Phil. Trans. R. Soc. (Lond.) B 345, 265268. Ramos, R. G. P., Igloi, G. L., Lichte, B., Baumann, U., Maier, D., Schneider, T., Brandst~itter, J. H., Fr6hlich, A., and Fischbach, K.-F. (1993).The irregular chiasm C-roughest locus of Drosophila, which affects axonal projections and programmed cell death, encodes a novel immunoglobulin-like protein. Genes Dev. 7, 2533-2547. Robinow, S., Talbot, W. S., Hogness, D. S., and Truman, J. W. (1993). Programmed cell death in the Drosophila CNS is ecdysone-regulated and coupled with a specific ecdysone receptor isoform. Development 119, 1251-1259. Snider, W. D. (1994). Functions of the neu rotrophins during nervous system development: what the knockouts are teaching us. Cell 77, 627-638. Smouse, D., and Perrimon, N. (1990). Genetic dissection of a complex neurological mutant, polyhomeotic, in Drosophila. Dev. Biol. 139, 169-185. Stark, W. S., Chen, D. M., Johnston, M. A., and Frayer, K. L. (1983). The rdgB gene of Drosophila: retinal degeneration in different alleles and inhibition by norpA. J. Insect Physiol. 29, 123-131. Stark, W. S., Sapp, S. R., and Carlson, S. D. (1989). Photoreceptor maintenance and degeneration in the norpA (no receptor potential-A) mutant of Drosophila melanogaster. J. Neurogenet. 5, 49-51. Steller, H., Fischbach, K.-F., and Rubin, G. M. (1987). disconnected: a locus required for neuronal pathway formation in the visual system of Drosophila. Cell 50, 1139-1153. Sugimoto, A., Friesen, P. D., and Rothman, J. H. (1994). Baculovirus p35 prevents developmentally regulated cell death and rescues a ced-9 mutant in the nematode Caenorhabditis elegans. EMBO J. 13, 2023-2028. Talbot, W. S., Swyryd, E.A., and Hogness, D. S. (1993). Drosophila tissues with different metamorphic responses to ecdysone express different ecdysone receptor isoforms. Cell 73, 1323-1337. Thomas, H. E., Stunnenberg, H. G., and Stewart, A. F. (1993). Heterodimerisation of the Drosophila ecdysone receptor with retinoid X receptor and ultraspiracle. Nature 362, 471-475. Trauth, B. C., Klas, C., Peters, A. M. l., Matzku, S., M611er, P., Falk, W., Debatin, K.-M., and Krammer, P. H. (1989). Monoclonal antibody-mediated tumor regression by induction of apoptosis. Science 245, 301-305. Truman, J. W. (1983). Programmed cell death in the nervous system of an adult insect. J. Comp. Neurol. 216, 445-452. Truman, J. W. (1992). Developmental neuroethology of insect metamorphosis. J, Neurobiol. 23, 1404-1422.
Oppenheim, R. W., Prevette, D., Tytell, M., and Homma, S. (1990). Naturally occurring and induced neuronal death in the chick embryo in vivo requires protein and RNA synthesis: evidence for the role of cell death genes. Dev. Biol. 138, 104-113.
Truman, J. W., and Bate, M. (1988). Spatial and temporal patterns of neurogenesis in the central nervous system of Drosophila melanogaster. Dev. Biol. 125, 145-157.
Power, M. E. (1943).The effect of reduction in numbers of ommatidia upon the brain of Drosophila melanogaster. J. Exp. Zool. 94, 33-72. Pulido, D., Campuzano, S., Koda, T., Modolell, J., and Barbacid,
Truman, J. W., and Schwartz, L. M. (1984). Steroid regulation of neural death in the moth nervous system. J. Neurosci. 4, 274280. Truman, J. W., Thorn, R. S., and Robinow, S. (1992). ProRrammed
Neuron 1274
neuronal death in insect development. J. Neurobiol. ~23, 12951311.
Vaux, D. L., Weissman, I. L., and Kim, S. K. (1992). Prevention of programmed cell death in Caenorhabditis elegans by human bcl-2. Science 258, 1955-1957. Weeks, J. C., and Truman, J. W. (1985). Independent steroid control of the fates of motoneurons and their muscles during insect metamorphosis. J. Neurosci. 5, 2290-2300. White, K., Grether, M., Abrams, J., Young, L., Farrell, K., and Steller, H. (1994). Genetic control of cell death in Drosophila. Science 264, 677-683. Wilson, C., Goberdhan, D. C. I., and Steller, H. (1993). Dror, a potential neurotrophic receptor gene, encodes a Drosophila homolog of the vertebrate Ror family of Trk-related receptor tyrosine kinases. Proc. Natl. Acad. Sci. USA 90, 7109-7113. Wolff, T., and Ready, D. F. (1991). Cell death in normal and rough eye mutants of Drosophila. Development 113, 825-839. Wyllie, A. H., Kerr, J. F. R., and Currie, A. R. (1980). Cell death: the significance of apoptosis. Int. Rev. Cytol. 68, 251-306. Yao, T.-P., Segraves, W. A., Oro, A. E., McKeown, M., and Evans, R. M. (1992).Drosophila ultraspiracle modulates ecdysone receptor function via heterodimer formation. Cell 71, 63-72. Yonehara, S., Ishii, A., and Yonehara, M. (1989). A cell-killing monoclonal antibody (anti-Fas) to a cell surface antigen codownregualted with the receptor for tumor necrosis factor. J. Exp. Med. 169, 1747-1756.