PCD of Muscles during Metamorphosis
There is extensive loss of the embryonically-derived musculature within the abdomen in Drosophila following adult eclosion (Miller, 1950; Kimura and Truman, 1990). Although some of the persisting muscles continue to function during pupal life and adult emergence, others provide a template for adult myogenesis (Broadie and Bate, 1991; Bate, 1993). Some of these new muscles, including male-specific muscles, depend upon innervation for differentiation and survival (Bate, 1993). Phagocytosis of degenerated (also referred to as histolyzed) muscle is a feature of the larval-pupal transition in Drosophila (Crossley, 1978; Bate, 1993), but not of muscles that die following adult eclosion (Kimura and Truman, 1990; Jones and Schwartz, 2001).
In contrast with Manduca, relatively little is known about the genes involved in the death of Drosophila muscles during metamorphosis. A recent study has demonstrated that the nuclear spindle matrix scaffold proteins East and Chromator play antagonistic roles in the death of the dorsal oblique external muscles during metamorphosis in flies (Wasser et al., 2007). The Chromator protein appears to act late in histolysis, and is required for the complete destruction of dying muscles. EAST, which physically interacts with Chromator, appears to delay the loss of structural integrity in the muscles. The post-eclosion death of abdominal muscles in Drosophila is dependent on the microRNA Let7C (Sokol et al., 2008). Deletion of Let7C ablates muscle death, which can be rescued with ectopic expression. The mRNA targets for Let7C that are essential for muscle death have not yet been determined.
Insights from Other Tissues
Study of PCD during embryogenesis and metamorphosis outside of the neuromuscular system is a growth area in cell death research. Many tissues have been studied, including the fat body and the germ cell lineages. Of particular relevance are two tissues that undergo PCD during the early hours of metamorphosis: the larval salivary glands, and the larval midgut. This is because they have been used effectively to explore the coupling of ecdyster-oid signals and PCD. They also illustrate the co-occurrence of apoptosis and autophagy, and the cell-specificity of PCD pathways.
Larval Salivary Glands
The salivary glands of Drosophila form during embryogen-esis, and constitute the largest secretory organ in the fly body. During larval life the salivary glands secrete digestive enzymes, while at the end of larval life they secrete the "glue proteins" that attach the pupa to the substrate – typically, the wall of the culture vial. In response to the pre-pupal pulse of ecdysone secretion at 10-12 h post-pupariation, the larval salivary glands undergo PCD (Jiang et al., 1997). This hormonal cue serves to upregulate the expression of a number of multiple pro-apoptotic proteins, including Reaper, HID, Dronc, and Dark (Jiang et al., 1997). Some of this transcriptional regulation is direct: for example, EcR/ USP response elements have been identified in both the rpr and dronc promoters (Jiang et al., 2000; Cakorous et al., 2004). Much of this regulation, however, reflects the activation of a cascade of early response genes by 20E, including E74, Broad Complex (BRC), and E93 (Lee and Baehrecke, 2001; Yin and Thummel, 2005). The importance of this indirect regulation is shown by the fact that E93 loss-of-function mutants have severe deficits in salivary gland PCD (Lee et al., 2000). Notably, ectopic Reaper expression does not kill the larval salivary glands before the middle of the third and final larval instar, as prior to this PCD is blocked by high levels of DIAP1 (Yin et al, 2007). Resistance to Reaper is removed during the second half of the third instar by a CREB-binding protein (CBP) mediated downregula-tion of DIAP1 in response to a mid-third instar ecdysteroid pulse (Yin et al., 2007). The role of CBP in steroid-regulated nerve and muscle PCD has not yet been assessed.
Detailed analyses of the morphology of the dying larval salivary gland cells and the accompanying gene cascades provided a compelling example of the co-existence of cas-pase activation and the formation of autophagosomes. Expression of p35 in salivary glands keeps the cells alive, but does not block the formation of autophagosomes (Lee and Baehrecke, 2001). Confirmation of the simultaneous co-activation of multiple PCD-related programs comes from microarray and SAGE-based studies of gene expression in dying larval salivary glands (Gorski et al., 2003; Lee et al., 2003). Among the several hundred transcripts identified in these studies were known apoptosis-related transcripts, genes associated with autophagy, and non-caspase proteases (Yin and Thummel, 2005).
PCD of larval salivary glands differs from the PCD observed in the embryonic and metamorphosing nervous system, in that the process in the salivary gland destroys an entire organ, whereas in the CNS selected individual cells die while their neighbors survive. PCD in salivary glands is more similar to the phenomenon observed in the ISMs. Whether the distinct selective nature of PCD activation in one case but not the others represents fundamentally or subtly different forms of regulation is unknown.
Larval Midgut
By the time the larval salivary glands begin to degenerate in response to the pre-pupal ecdysone pulse, the death of the larval midgut in response to the late third larval instar ecdysone pulse is already well underway (Lee et al., 2002). However, in contrast to the salivary gland, midgut death occurs in a caspase-independent manner (Denton et al., 2009). Instead, it is completely dependent on the autophagic pathway (Denton et al., 2010). This model may represent the first clear demonstration of an autopha-gic cell death in the absence of a caspase-based component. Functional analyses of previously identified larval salivary gland death genes were in turn identified as acting as 20E-dependent pro-survival or pro-death factors for the midgut (Chittaranjan et al., 2009).
Summary and Conclusions
A Mature Field
The extent of research described in this topic, much of it based on detailed genetic analyses of PCD in insect systems, indicates how thoroughly insect systems are embedded into the broader field of cell death studies. In addition to gaining a better understanding of how PCD matches insect neuromuscular systems to the stage of development, investigators are now linking the molecular machinery of PCD to their mechanistic analyses of the regulation of organ size, aging, and human neurodegenerative disease. Important pathways to watch in the future include the Hippo kinase pathway, which regulates the balance of cell proliferation and apoptosis during organogenesis (Badouel et al., 2009); TOR-dependent autophagy and its relationship to lifespan (Bjedov et al., 2010); and the regulation of cell death related genes by the forkhead box transcription factor FoxO1, which is in turn regulated by phosphorylation by the Parkinson’s Disease-associated leucine-rich repeat kinase 2 (LRRK2) (Gandhi et al., 2009; Kanao et al., 2010).
Broad Phylogenetic Comparisons
Rapid strides in genome sequencing will ultimately result in numerous interesting comparative studies of the molecular machinery of all forms in cell death. There is already a wide range of insect genomes that have been sequenced, including 12 different Drosophila species and several other dipterans, as well as members of the Lepidoptera, Hymenoptera, Coleoptera, Hemiptera, and Phthiraptera. In addition, the genome of Daphnia pulex, a member of the crustacean sister group, has also been sequenced. These data provide a rich resource for comparative analysis of PCD pathways and PCD evolution. Studies of PCD in insects have already raised several pertinent questions, including the basis for distinct molecular palettes and pathways for regulating common cellular outcomes in divergent organisms.
Evolutionary scenarios for the grim-reaper gene complex The small Reaper, Grim, and Sickle proteins are clearly related, and the much larger Hid protein is also likely to share a common origin. Why are four related proteins needed to control DIAP1 inhibition and apoptosis, and how similar are the functions of these proteins? Gene complexes can provide enhanced capabilities for a cell to control specific biological processes by providing gene products that exhibit overlapping yet distinct expression patterns and functions. Presumably,the existence of the reaper, hid, grim, and sickle genes in Drosophila enables cells to control apoptosis and other processes with the exquisite specificity and efficiency that are required for development. It will be fascinating to discover the functional differences that distinguish these proteins. Already, gene expression studies, as well as loss-of-function and gain-of-function mutants, have indicated that these genes exhibit distinct patterns of expression and non-redundant cell killing activities (see, for example, Grether et al., 1995; Robinow et al., 1997; Zhou et al., 1997; Wing et al., 1998, 2002).
While the grim-reaper genes are essential cell death activators in Drosophila, bona fide structural homologs are not found in C. elegans or vertebrates. Indeed, until recently, RHG genes had only been identified in other dipterans, including several Drosophila species, the blowfly Lucilia cuprina, and the mosquitoes Anopheles gambiae and Aedes aegypti (White et al, 1994; Chen et al, 2004; Zhou et al, 2005). However, a recent bioinformatic anlysis identified an RHG gene from the lepidopteran Bombyx mori (Bryant et al., 2009). The predicted proteins share a highly conserved NH2-terminal RHG domain, and most also possess a GH3 domain/Trp block. These RHG proteins are likely to serve as PCD-promoting IAP antagonists. Interestingly, the mosquitoes and silkworm genomes each appear to possess a single RHG family member, as opposed to the four related, linked genes in Drosophila. This suggests that gene duplication events leading to the four linked RHG genes in Drosophila occurred recently, and that, in comparison with the caspases and Bcl-2 family proteins, there has been an extremely rapid divergence of the coding sequences of IAP-antagonists. It will be of interest to clarify the phylogenetic distribution of RHG genes.
Future Directions
It can be confidently predicted that studies of the diverse mechanisms of PCD in insect neuromuscular systems across the lifespan will lead to greater understanding of the molecular mechanisms of metamorphosis, particularly the ways in which developmental hormone signals are coupled to neuronal and muscle survival. Although studies of developmental PCD in insects now encompass other tissues, studies of PCD in the CNS will remain important because they offer the opportunity to study the life (and death) history of individual identified neurons. Studies of PCD in ISMs will continue to offer the challenge of segment-specific death phenotypes. Studies driven by investigators eager to identify the molecular pathologies involved in human neurodegenerative disease and human myopathies are likely to become an ever-increasing force in the field of insect PCD, particularly as researchers become ever more adept at transitioning seamlessly across mammalian and insect model systems (see, for example, Godin et al., 2010). The focus on translational research, however, should not obscure the many fascinating basic questions that remain unanswered concerning PCD and insect neurons and muscle. A partial list includes the following:
• Why are there so many PCD pathways? Can we develop a model that allows us to predict the PCD program a cell will select under specific developmental and environmental conditions?
• What are the downstream effects of PCD in neuromus-cular systems? That is, do dying cells in turn produce signals that influence the phenotype and/or survival of neighboring cells? For example, it has been shown in fly wing that dying cells produce a caspase-dependent signal (wingless) that induces neighboring cells to proliferate. This phenomenon, referred to as compensatory cell proliferation, is thought to help maintain homeostasis in developing tissues (see, for example, Ryoo et al., 2004). The relevance of this phenomenon to other organs and tissues is unknown.
• Is our view of insect PCD skewed because we rely so heavily on a single insect model? For example, there is growing interest in epigenetic regulation of apoptotic genes (Mazin, 2009; Hajji and Joseph, 2010; Jazirehi, 2010), yet the Drosophila melanogaster genome does not encode any methyltransferases – thus one key component of epigenetic DNA modification cannot be studied in this species, although other insect species such as the honey bee have methyltransferase enzymes comparable to those of mammals (Gabor Miklos and Maleszka, 2010).
• Do vertebrate-like neurotrophins function to regulate PCD in insect embryogenesis? During metamorphosis? The recent report describing Drosophila Neurotrophin 1 (DNT-1) should spark new interest in this possibility (Zhu et al, 2008).
• What is the relationship of cell energy budgets to PCD in insect neurons and muscles?
• What role, if any, do microRNAs play in the regulation of nerve and muscle survival?
• Can genetic tools (RNAi, transfection, etc.) be applied to the study PCD in Manduca to facilitate the analysis of the novel cell death genes isolated from this species?
A persisting question in the field is the nature of the relationship between pathological cell death and PCD that is necessary for normal development.
