Migrate, Differentiate, Proliferate, or Die: Pleiotropic Functions of an Apical "Apoptotic Caspase"

See allHide authors and affiliations

Science's STKE  12 Oct 2004:
Vol. 2004, Issue 254, pp. pe49
DOI: 10.1126/stke.2542004pe49


Caspases, the cysteine proteases that cleave their substrates following an aspartate residue, primarily carry out two distinct functions: (i) activation of proinflammatory cytokines and (ii) execution of apoptosis. These two functions are considered to be unique to individual caspases; thus, some caspases act in apoptosis, whereas others have a role in inflammation. However, this dogma is now being challenged as nonapoptotic functions are ascribed to caspases that, until recently, were only known to function in cell death pathways. Recent work suggests that DRONC, the only initiator cell death caspase in Drosophila, may play a direct or indirect role in cell migration, sperm differentiation, and cell proliferation in addition to its function in cell death.

Programmed cell death (PCD) by apoptosis is a naturally occurring process of cell suicide that plays a crucial role in the development, homeostasis, and defense of metazoans by eliminating unwanted cells (1, 2). Cysteine proteases of the caspase family are main effectors of apoptosis ("apoptotic caspases"), although some caspases, related to caspase-1, primarily function to activate proinflammatory cytokines (3, 4). Recent studies of the initiator cell death caspase in Drosophila, DRONC (Drosophila Nedd2-like caspase), suggest that this caspase, while being a central component of the cell death machinery, also functions in many apparently unrelated nonapoptotic signaling pathways. Two research groups led by Hay and Steller, respectively, have found that active caspases, including DRONC, are required for sperm differentiation (5, 6). Hay’s laboratory has also reported that DRONC is required for compensatory cell proliferation in the Drosophila wing disc in response to forced cell loss (7). Geisbrecht and Montell have shown that DIAP1 (Drosophila inhibitor of apoptosis protein 1)–mediated inhibition of DRONC is required for border cell migration in the Drosophila ovary (8). These unexpected nonapoptotic roles of DRONC challenge the prevailing view of caspase biology and suggest additional functions of the initiator death proteases in metazoans.

The Apoptotic Caspases

It is now well established that all caspases are produced as zymogens that, in most cases, undergo proteolytic processing into two subunits to generate the active enzyme (3, 4, 911). In mammals where cell death execution is complex, the main apoptotic caspases are the so-called "initiator" (also referred to as "apical") caspases, including caspases 2, 8, 9, and 10, and the "effector" caspases 3, 6, and 7 (3, 4). As the name suggests, the initiator caspases are activated first. These caspases are characterized by homotypic protein-protein interaction motifs present in their N-terminal prodomains (3, 4, 912). Caspases 2 and 9, which are the mammalian counterparts of the death protease CED-3 in Caenorhabditis elegans, contain an evolutionarily conserved caspase recruitment domain (CARD) (912). Caspases 8 and 10 contain a pair of death effector domains (DEDs) (912). These domains facilitate the recruitment of initiator caspases by their adaptors, such as Apaf-1 in the case of caspase-9, RAIDD (RIP-associated ICH-1/CED-3 homologous protein with a death domain) for caspase-2, and FADD (Fas-associated death domain protein) for caspases 8 and 10 (912). This interaction between initiator caspases and their adaptors mediates caspase activation by a mechanism that is still not fully understood, but may involve proximity-induced dimerization or oligomerization of procaspase molecules (911). Effector caspases, such as caspases 3, 6, and 7, which lack homotypic interaction motifs, are proteolytically processed and activated by initiator caspases (3, 4, 911). These caspases specifically target a number of cellular proteins for cleavage to irreversibly commit a cell to undergo apoptosis (13).

Caspase-8 plays a key nonredundant function in apoptosis mediated by the death receptors of the tumor necrosis factor (TNF) family, whereas caspase-9 is the main caspase activated in response to cell stress, including cytotoxic insults and growth factor deprivation (3, 4, 14). The function of caspase-2 in either pathway is controversial, but it is believed to be required for apoptosis in some cell types (1519). Because some cells deficient in both caspase-2 and caspase-9 can still undergo cell death in response to stress signals, it has been proposed that these caspases may serve redundant functions as amplifiers of the proteolytic caspase cascade (2022).

The Fly Caspases, No Less Complicated

There are seven caspases in Drosophila, two of which, DREDD (death-related CED-3/Nedd2-like protein) and DRONC, qualify as initiator caspases on the basis of the presence of homotypic domains in their N-terminal prodomain regions (2330). DREDD contains a pair of DEDs and thus appears structurally similar to caspase-8. However, the function of DREDD in cell death is less clear. In fact, most available data suggest that DREDD is required in the innate immune response pathway in Drosophila, rather than for PCD (31). There is only one initiator caspase in the fly, DRONC, which resembles mammalian caspases 2 and 9 in domain organization (26). Given that DRONC is the only CARD-containing caspase in the fly, it is believed to be a nonredundant apical component of the Drosophila cell death machinery.

There are a number of studies that support this assertion. Expression of the dronc gene is transcriptionally regulated in response to developmental cell death cues (26, 32). The steroid hormone ecdysone, which controls PCD of the larval salivary glands and midgut, stimulates dronc expression, and dominant-negative (DN) versions of DRONC delay the removal of salivary glands (26, 3337). In cultured cells, dronc ablation by RNA interference (RNAi) inhibits cell death induced by ecdysone (34). In Drosophila embryos, dronc RNAi ablates normal PCD (38). Furthermore, genetic studies show that DRONC is required for cell death mediated by REAPER, HID, and GRIM, the DIAP1 antagonists that are upstream regulators of PCD in the fly (38, 39).

Five Drosophila caspases lack any homotypic motifs in their N-terminal regions and thus are likely to be downstream effector caspases (30). Two of these, DRICE (Drosophila ICE) and DCP-1 (Drosophila caspase-1), are the most abundant and share the most sequence similarity to each other and to mammalian caspase-3. As in mammals, DRICE and DCP-1 are targets for DRONC-mediated processing and activation; however, activation of effector caspases may also be independent of DRONC in certain contexts (6).

DRONC Activation in Fly: DIAP1 Removal Is the Key

DIAP1 is a direct inhibitor of both DRONC and the effector caspases DRICE and DCP-1 and thus acts as a guardian against the toxic effects of caspases (4042). DIAP1 consists of two BIR (baculoviral IAP repeat) domains and a RING type of ubiquitin ligase domain (40). The DIAP1 BIR1 domain interacts with DRICE, whereas the BIR2 domain binds DRONC (42). Embryos in which diap1 has been mutated and cells in which DIAP1 has been ablated show spontaneous caspase activation and PCD (43, 44).

The prevailing model of DRONC activation predicts that in healthy cells DIAP1 is bound to DRONC, preventing its interaction with DARK (Drosophila Apaf-1–related killer), the Apaf-1–like adaptor protein in Drosophila (Fig. 1) (11, 32, 42, 4548). Upon signals that trigger cell death, REAPER, HID (head involution defective), and GRIM bind DIAP1. This results in DIAP1 autoubiquitination and proteasome-dependent degradation, allowing DRONC activation through interaction with DARK. Interestingly, unlike in mammals—where activation and oligomerization of Apaf-1 requires binding to cytochrome c, which is released from mitochondria (49, 50)—in flies, DARK-mediated DRONC activation can occur in the absence of cytochrome c release from mitochondria (51, 52). The trigger for DRONC activation is thus counteraction of the inhibitory function of DIAP1. This can be achieved by increasing the intracellular concentrations of REAPER, HID, or GRIM, or by elevating the concentration of DRONC, DARK, or both. Both of these scenarios are seen during developmental PCD in the fly (32).

Fig. 1.

Caspase activation in Drosophila and regulation of multiple cellular outcomes. The caspase inhibitor DIAP1 is essential for survival of most fly cells, and its removal is sufficient to induce the activation of the initiator caspase DRONC. Death inducers REAPER, HID, and GRIM (RHG proteins) bind and promote DIAP1 degradation, thus allowing DRONC activation through binding to its adaptor DARK. Active DRONC can mediate a number of nonapoptotic processes such as sperm differentiation, compensatory proliferation (in response to PCD in the Drosophila wing disc), and inhibition of border cell migration in the fly ovary. There is also some evidence suggesting that DRONC processes Notch to regulate neurogenesis (65). The apoptotic function of DRONC is likely to be mediated by the downstream effector caspases DRICE, DCP-1, or both, each of which can be cleaved and activated by DRONC. DIAP1, which prevents activation of DRONC, is also an inhibitor of activated effector caspases. There is growing evidence (6) that in some tissues, DRICE activation can occur independently of DRONC activation and that active DRICE also has context-dependent nonapoptotic roles.

Active Caspases, Not Always Deadly!

Although DRONC is clearly essential for many PCD pathways in the fly, a number of recent studies implicate this caspase in nonapoptotic functions such as sperm differentiation, cell migration, and compensatory cell proliferation (58). The evidence for DRONC’s function in PCD is derived from experiments involving overexpression of the wild-type or the catalytically inactive DN-DRONC mutant proteins, and from RNAi studies in cell lines and in animals (26, 34, 3739, 53). As discussed below, similar experiments have now been used to uncover the nonapoptotic functions of this caspase.

Caspases and Sperm Differentiation

Arama and colleagues found that active caspases play a role during the individualization process of maturing spermatids (5). During sperm differentiation, spermatozoa produced within a germline syncytium undergo a process known as individualization (54, 55), which involves expelling most of the cytoplasm from spermatids (55, 56). Arama et al. found that this process requires caspase activity, because the active caspase DRICE is detected at the onset of individualization and caspase inhibitors block individualization. Using expression of DIAP1, DARK RNAi, and overexpression of DN-DRONC to inhibit caspase activation, Huh et al. showed that DARK- and HID-dependent activation of DRONC occurs at sites of spermatid individualization and that all three proteins are required for the individualization process (6). In addition to DRONC, genetic evidence showed that DREDD and dFADD, an adaptor that is likely to mediate the activation of DREDD (57), are also required for individualization. Interestingly, the effector caspase DRICE is activated throughout the length of individualizing spermatids.

These studies thus show that multiple caspases and caspase regulators, probably acting in a spatiotemporal manner, are required for spermatid individualization. However, it remains a mystery why apoptotic caspase activation machinery and caspases are required for this process. As suggested by Huh et al. (6), caspases may cleave specific proteins to facilitate spermatid individualization. The identities of the caspase substrates that might be required for individualization, and whether they are different from the substrates cleaved by caspases during apoptosis, are unclear.

Caspase Inhibition and Border Cell Migration

A genetic study by Geisbrecht and Montell shows that inhibition of DRONC by DIAP1 plays a role in border cell migration in the Drosophila ovary (8). The Drosophila egg chamber is made of a single oocyte, 15 nurse cells, and a layer of follicular epithelial cells that surround the germline cells. The border cell cluster, which consists of two anterior polar cells (specialized follicular cells) and six to eight neighboring cells, migrates to the border between the nurse cells and the developing oocyte (58, 59). Rac, a member of the Rho family of guanosine triphosphatases (GTPases), regulates this migration and expression of a DN-Rac (RacN17) inhibits this process (60). Geisbrecht and Montell screened for genes that, when overexpressed, suppressed the migration defect caused by DN-Rac and unexpectedly found that the overexpression of diap1 suppressed the migration defect. As stated above, loss of DIAP1 normally results in spontaneous caspase activation and cell death. However, the diap1 loss-of-function mutations, which caused border cell migration defects, did not result in apoptosis of border cells (8). Interestingly, only the DIAP1 BIR domain mutants, not the DIAP1 RING domain mutants, had border cell migration defects. Using pull-down experiments in cultured Drosophila cells, the authors showed that DIAP1, Rac, and profilin (an actin-binding protein that maintains the pool of monomeric actin) physically interacted in the same molecular complex. Overexpression of DIAP1 in S2 cells resulted in cells with serrate or stellate morphology, and this phenotype was further enhanced by coexpression of DN-Rac (8), which suggests that an interaction between DIAP1 and Rac regulates actin dynamics.

One key question here is whether the effect of DIAP1 on Rac-dependent cell migration relates to DIAP1’s role as a caspase inhibitor. The DRICE and DCP-1 inhibitor p35 did not rescue the DN-Rac–mediated border cell migration defect; however, DN-DRONC was as effective as overexpression of DIAP1. Furthermore, mutations affecting dark also rescued the effects of DN-Rac on border cell migration. These results together strongly suggest that an inhibition of DRONC activation (by DIAP1) is required for border cell migration.

DRONC in Cell Proliferation

Hay and colleagues provide evidence for a role of DRONC activity in compensatory cell proliferation induced by PCD in the Drosophila wing disc (7). It is well known that in the fly wing disc, loss of a large number of cells (for example, after irradiation) is accompanied by compensatory proliferation resulting in virtually normal wings (61, 62). Huh et al. argued that the signals that drive this compensatory proliferation in the wing disc might come from the dying cells. To test this, they activated PCD signaling by overexpressing HID in some cells of the wing disc, but prevented apoptosis by expressing baculoviral p35. Because p35 inhibits the activity of effector caspases (DRICE and DCP-1) but not DRONC, this elegant experimental design allowed them to activate early parts of the cell death machinery without killing cells. Consistent with their hypothesis, Huh et al. observed increased cell proliferation in wing discs that coexpressed p35 and HID (7). When they expressed DN-DRONC along with p35 and HID, they did not exhibit an increase in cell proliferation; this result suggests that DRONC is required for compensatory proliferation. Furthermore, activation of DRONC in the absence of apoptosis was sufficient to drive cell proliferation. The compensatory proliferation model predicts that dying cells provide proliferative signals to neighboring cells. Huh et al. found DRONC-dependent ectopic expression of Wingless in wing discs coexpressing p35 and HID. Because Wingless is a known secreted mitogen in the wing (63), it may be the mediator of cell proliferation in response to cell death signaling in the wing disc.

One surprise in the Huh et al. study was that a decrease in DIAP1 levels by RNAi in wing disc cells expressing p35 failed to produce an increase in cell proliferation. Loss of DIAP1 should result in DRONC activation; therefore, the results do not support the notion that DRONC activity alone is sufficient for providing a proliferative signal. Although the precise mechanism of DRONC-mediated effects observed by Huh et al. requires more investigation, the studies nonetheless support a role for DRONC in providing proliferative signals from cells that are dying by DRONC-dependent apoptosis.

In Wrapping Up . . .

In some ways, the new findings with DRONC are consistent with studies in mammals that suggest that some apoptotic caspases play a context-dependent role in cell proliferation and differentiation in specific tissues (64). However, as with most new and unexpected observations, the nonapoptotic functions of DRONC described in the recent papers need to be treated with some caution. First, most data in these studies were derived from experimental systems that use DN versions of DRONC or caspase inhibitors that may have unknown nonspecific consequences. Furthermore, in all cases the phenotypes appear to be limited to specific cells or tissues. Second, none of the studies has addressed what prevents active DRONC from killing the cells, and the results do not clearly dissect out the molecular basis of the apoptotic and nonapoptotic functions of DRONC. Nonetheless, these studies do provide strong indirect evidence for divergent functions of DRONC and a basis for additional investigation.

How might we reconcile these apparently unrelated nonapoptotic functions of DRONC? Perhaps the most obvious prediction, as discussed by the authors of these papers themselves, is that DRONC proteolytically activates or inactivates specific proteins, which then mediate signaling leading to the various context-dependent outcomes. Perhaps the low levels of DRONC activity required for these nonapoptotic signaling pathways are insufficient for apoptotic execution. Alternatively, active DRONC may be sequestered in specific subcellular compartments to prevent it from cleaving substrates involved in apoptosis. Although many questions remain, the data presented in these recent papers add to a growing body of evidence that suggests that some caspases do much more than simply dismantle cells during PCD.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
  63. 63.
  64. 64.
  65. 65.
View Abstract

Navigate This Article