Title: Eiger, a TNF superfamily ligand that triggers the Drosophila JNK pathway
Abstract: Article17 June 2002free access Eiger, a TNF superfamily ligand that triggers the Drosophila JNK pathway Tatsushi Igaki Tatsushi Igaki Laboratory for Cell Recovery Mechanisms, Brain Science Institute, RIKEN, 2-1 Hirosawa, Wako, Saitama, 351-0198 Japan Department of Cell Biology and Neuroscience, Hachioji, Tokyo, 192-0397 Japan Search for more papers by this author Hiroshi Kanda Hiroshi Kanda Laboratory for Cell Recovery Mechanisms, Brain Science Institute, RIKEN, 2-1 Hirosawa, Wako, Saitama, 351-0198 Japan Division of Molecular Genetics, Osaka University Graduate School of Medicine, 1-3 Yamadaoka, Suita, Osaka, 565-0871 Japan Search for more papers by this author Yuki Yamamoto-Goto Yuki Yamamoto-Goto Laboratory for Cell Recovery Mechanisms, Brain Science Institute, RIKEN, 2-1 Hirosawa, Wako, Saitama, 351-0198 Japan Search for more papers by this author Hirotaka Kanuka Hirotaka Kanuka Laboratory for Cell Recovery Mechanisms, Brain Science Institute, RIKEN, 2-1 Hirosawa, Wako, Saitama, 351-0198 Japan Search for more papers by this author Erina Kuranaga Erina Kuranaga Laboratory for Cell Recovery Mechanisms, Brain Science Institute, RIKEN, 2-1 Hirosawa, Wako, Saitama, 351-0198 Japan Department of Cell Biology and Neuroscience, Hachioji, Tokyo, 192-0397 Japan Search for more papers by this author Toshiro Aigaki Toshiro Aigaki Department of Biological Sciences, Tokyo Metropolitan University, 1-1 Minami-osawa, Hachioji, Tokyo, 192-0397 Japan Search for more papers by this author Masayuki Miura Corresponding Author Masayuki Miura Laboratory for Cell Recovery Mechanisms, Brain Science Institute, RIKEN, 2-1 Hirosawa, Wako, Saitama, 351-0198 Japan Search for more papers by this author Tatsushi Igaki Tatsushi Igaki Laboratory for Cell Recovery Mechanisms, Brain Science Institute, RIKEN, 2-1 Hirosawa, Wako, Saitama, 351-0198 Japan Department of Cell Biology and Neuroscience, Hachioji, Tokyo, 192-0397 Japan Search for more papers by this author Hiroshi Kanda Hiroshi Kanda Laboratory for Cell Recovery Mechanisms, Brain Science Institute, RIKEN, 2-1 Hirosawa, Wako, Saitama, 351-0198 Japan Division of Molecular Genetics, Osaka University Graduate School of Medicine, 1-3 Yamadaoka, Suita, Osaka, 565-0871 Japan Search for more papers by this author Yuki Yamamoto-Goto Yuki Yamamoto-Goto Laboratory for Cell Recovery Mechanisms, Brain Science Institute, RIKEN, 2-1 Hirosawa, Wako, Saitama, 351-0198 Japan Search for more papers by this author Hirotaka Kanuka Hirotaka Kanuka Laboratory for Cell Recovery Mechanisms, Brain Science Institute, RIKEN, 2-1 Hirosawa, Wako, Saitama, 351-0198 Japan Search for more papers by this author Erina Kuranaga Erina Kuranaga Laboratory for Cell Recovery Mechanisms, Brain Science Institute, RIKEN, 2-1 Hirosawa, Wako, Saitama, 351-0198 Japan Department of Cell Biology and Neuroscience, Hachioji, Tokyo, 192-0397 Japan Search for more papers by this author Toshiro Aigaki Toshiro Aigaki Department of Biological Sciences, Tokyo Metropolitan University, 1-1 Minami-osawa, Hachioji, Tokyo, 192-0397 Japan Search for more papers by this author Masayuki Miura Corresponding Author Masayuki Miura Laboratory for Cell Recovery Mechanisms, Brain Science Institute, RIKEN, 2-1 Hirosawa, Wako, Saitama, 351-0198 Japan Search for more papers by this author Author Information Tatsushi Igaki1,2, Hiroshi Kanda1,3, Yuki Yamamoto-Goto1, Hirotaka Kanuka1, Erina Kuranaga1,2, Toshiro Aigaki4 and Masayuki Miura 1 1Laboratory for Cell Recovery Mechanisms, Brain Science Institute, RIKEN, 2-1 Hirosawa, Wako, Saitama, 351-0198 Japan 2Department of Cell Biology and Neuroscience, Hachioji, Tokyo, 192-0397 Japan 3Division of Molecular Genetics, Osaka University Graduate School of Medicine, 1-3 Yamadaoka, Suita, Osaka, 565-0871 Japan 4Department of Biological Sciences, Tokyo Metropolitan University, 1-1 Minami-osawa, Hachioji, Tokyo, 192-0397 Japan *Corresponding author. E-mail: [email protected] The EMBO Journal (2002)21:3009-3018https://doi.org/10.1093/emboj/cdf306 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Drosophila provides a powerful genetic model for studying the in vivo regulation of cell death. In our large-scale gain-of-function screen, we identified Eiger, the first invertebrate tumor necrosis factor (TNF) superfamily ligand that can induce cell death. Eiger is a type II transmembrane protein with a C-terminal TNF homology domain. It is predominantly expressed in the nervous system. Genetic evidence shows that Eiger induces cell death by activating the Drosophila JNK pathway. Although this cell death process is blocked by Drosophila inhibitor-of-apoptosis protein 1 (DIAP1), it does not require caspase activity. We also show genetically that Eiger is a physiological ligand for the Drosophila JNK pathway. Our findings demonstrate that Eiger can initiate cell death through an IAP-sensitive cell death pathway via JNK signaling. Introduction The intrinsic cell death execution mechanisms, which are regulated by the caspases, Apaf-1 and the Bcl-2 family of proteins, are highly conserved throughout evolution (Meier et al., 2000a; Vernooy et al., 2000; Shi, 2001). In Caenorhabditis elegans, the regulation of cell death is intrinsic and works through a genetic program. In Drosophila and mammals, on the other hand, cell death is largely regulated through extrinsic mechanisms; this is known as social control of cell death (Raff, 1992). Although a number of extrinsic death factors have been identified in vertebrates, until now, none has been reported in invertebrates, despite abundant similarities in the cell death mechanisms between Drosophila and vertebrates. Drosophila is a powerful genetic model for studying the in vivo role of cell death and its physiological regulation. The intrinsic cell death pathway in flies is stimulated by the Drosophila killer proteins, Reaper, Hid and Grim, or a pro-apoptotic Bcl-2 family member, Drob-1/Debcl/dBorg-1/dBok (Vernooy et al., 2000). Ectopic expression of these cell death trigger proteins in the developing Drosophila eye results in a reduced-eye phenotype (Grether et al., 1995; Chen et al., 1996; White et al., 1996; Colussi et al., 2000; Igaki et al., 2000). On the other hand, the pro-apoptotic adapter molecules such as Drosophila Apaf-1 (Dark) or a Drosophila caspase drICE do not cause the reduced eye size (H.Kanuka and M.Miura, unpublished data), suggesting that the small eye phenotype would be generated by an overexpression of the ‘apical’ cell death triggers, the proteins whose activities are regulated at the transcriptional levels or by the levels of their inhibitory molecules. The reduced-eye phenotype should also be generated by the extrinsic death stimuli such as ligand/receptor-mediated signaling. A Drosophila in vivo expression screen has an advantage for identifying the factors that function through such the cell–cell interactions as well as the cell autonomous factors. To identify novel cell death-inducing factors in Drosophila, we screened for genes that had a potential to reduce the eye size when misexpressed in the developing eye. Here, we describe the first invertebrate tumor necrosis factor (TNF) superfamily ligand identified in this screen, Eiger. We provide genetic evidence that Eiger can induce inhibitor-of-apoptosis (IAP)-sensitive cell death through the activation of the Drosophila JNK pathway. Genetic analysis of eiger mutant flies reveals that Eiger is a physiological ligand for the JNK pathway in the eye disc. Our findings also provide a model for a caspase-independent cell death pathway through the JNK signaling that can be blocked by IAP. Results To discover the cell death triggers encoded in the Drosophila genome, we conducted a misexpression screen using the GAL4/UAS system. The GS vector is a P element-based gene search vector with UAS enhancers (Toba et al., 1999). We crossed a collection of 5000 lines harboring the GS vector (GS lines) with an eye-specific GAL4 (GMR-GAL4) strain to screen for genes that generated the reduced-eye phenotype. Six GS lines resulted in greatly reduced eyes in a GAL4-dependent manner, which we called Regg strains (Regg1–6, for reduced-eye generator with a GMR-GAL4 driver). Identification of a Drosophila TNF superfamily ligand, Eiger, as a novel cell death trigger The F1 progeny generated by mating the GMR-GAL4 strain and Regg1 (GS9830) strain (GMR>regg1GS9830) displayed a strong reduced-eye phenotype (Figure 1B) compared with the wild-type eye (Figure 1A). To assess whether the small eye phenotype of GMR>regg1GS9830 flies was generated by the acceleration of cell death, we stained eye discs with acridine orange to detect dying cells (Figure 1C and D). Acridine orange is commonly used as a marker for apoptotic cell death in Drosophila (Abrams et al., 1993). In GMR>regg1GS9830 discs, numerous acridine orange-staining cells were observed, indicating that regg1GS9830 induced massive apoptotic cell death, leading to the reduced-eye phenotype. Using the inverse PCR method, we determined the insertion site of the P element in the Regg1GS9830 strain, and found a predicted gene, CG12919, adjacent to the insertion site. The GS9830 strain was generated by inserting the GS6 vector, which has a green fluorescent protein (GFP) trailer (UAS-GFP-SV40 terminator) near the 3′P end and a UAS enhancer near the 5′P end, so that a misexpression of vector-flanking sequence occurs 5′P side only (T.Aigaki, unpublished data). As expected, CG12919 was the only gene with an elevated expression level in a GAL4-dependent manner (see Supplementary figure S1 available at The EMBO Journal Online). We sequenced a Drosophila EST clone, LP03784, which included the nucleotide sequence of CG12919, and identified the open reading frame (ORF) of a novel gene we named eiger (EDA-like cell death trigger; see below) (Figure 1E). eiger encoded a protein of 409 amino acids with a C-terminal TNF homology domain and a hydrophobic transmembrane domain, indicating that Eiger was the first Drosophila member of the TNF superfamily (Figure 1F). The absence of a signal peptide suggested that Eiger was a type II membrane protein, which is typical of the members of the TNF ligand family. The sequence of the C-terminal TNF domain of Eiger showed highest homology with human EDA-A2 (28% identity), and also showed significant homology with all known TNF superfamily members including RANKL, CD40L, FasL, APRIL, TWEAK, TNF-α and TRAIL (Figure 1G). Figure 1.Identification of Eiger, a Drosophila TNF superfamily protein. (A and B) A misexpression screen identified a GS strain, GS9830 (as Regg1), which resulted in a reduced eye when driven by a GMR-GAL4 driver. Eye phenotypes of wild type (A) and regg1GS9830/GMR-GAL4 (B) are shown. (C and D) Acridine orange staining detected numerous dying cells in third-instar larval eye discs of the regg1GS9830/GMR-GAL4 strain (D) compared with the wild-type strain (C). Many acridine orange-positive cells were observed behind the morphogenetic furrow (arrowhead), corresponding to the expression domain of the GAL4 driver. (E) A novel ORF encoding Eiger was identified from an EST clone (LP03784) containing the nucleotide sequence of CG12919, the expression of which was simulated in a GAL4-dependent manner in the Regg1GS9830 strain. (F) Schematic structures of Eiger, EDA-A2 and human TNF-α. (G) The amino acid sequence of the TNF homology domain of Eiger is aligned with the sequences of human EDA-A2, RANKL, CD40L, FasL, APRIL, TWEAK, TNF-α and TRAIL. Identical and conserved residues are denoted with blocks of black or shading, respectively. The Eiger cDNA sequence has been deposited with DDBJ/EMBL/GenBank (accession No. AB073865). (H–K) Eiger is a type II membrane protein with a cytoplasmic N-terminus. S2 cells were transfected with the expression vectors for HA-Eiger (pUAS-HA-eiger) and GFP (pUAS-GFP), together with a driver plasmid, pWAGAL4 (actin-GAL4). Twenty-four hours after transfection, cells were immunostained with an anti-HA monoclonal antibody and a Cy3-labeled secondary antibody after (H and I) or prior to (J and K) fixation and permeabilization. (L) Schematic structure of the IR expression construct of eiger. Two partial sequences of eiger cDNA were cloned as a head-to-head inverted repeat that was separated by a non-palindromic 180 bp linker into the pUAST vector. (M and N) eiger is the responsible gene for Regg1. The reduced-eye phenotype induced by regg1GS9830 (A) was completely suppressed by eiger-IR (M). Overexpression of exogenous Eiger caused a reduced-eye phenotype (N). Genotypes are regg1GS9830/+; GMR-GAL4/UAS-eiger-IR (M) and UAS-eiger/GMR-GAL4 (N). Download figure Download PowerPoint To examine whether Eiger was indeed a membrane protein, we transfected S2 cells with expression vectors for an N-terminal hemagglutinin (HA)-tagged Eiger and GFP. In permeabilized cells, cell surface staining was observed in GFP-positive cells by anti-HA immunostaining (Figure 1H and I). When immunostained prior to fixation and permeabilization, however, the staining was not detected (Figure 1K), consistent with Eiger being a type II transmembrane protein with an extracellular C-terminus and a cytoplasmic N-terminus. To confirm that eiger was the gene responsible for the reduced-eye phenotype of GMR>regg1GS9830, we generated transgenic flies possessing an inverted repeat (IR) expression construct of the eiger cDNA (Figure 1L) that would specifically inhibit Eiger expression in a GAL4-dependent manner via a mechanism of RNA interference. This construct worked as a specific inhibitor of Eiger expression in a cell culture system (data not shown). We found that the Regg1 phenotype was completely rescued by the eiger-IR transgene (Figure 1M), demonstrating that eiger was indeed the gene responsible for the reduced-eye phenotype of GMR>regg1GS9830. Furthermore, ectopic expression of exogenous Eiger in the eye caused a reduced-eye phenotype similar to that of Regg1GS9830 (Figure 1N). Because regg1GS9830 activated eiger expression in a GAL4-dependent manner without affecting the expression of adjacent genes (see Supplementary figure S1), we used the Regg1GS9830 strain for further genetic analyses. Overexpression of Eiger results in severe developmental defects that are distinct from the defects caused by Reaper Ectopic expression of Eiger in the eye resulted in a reduced-eye phenotype similar to that of Reaper (White et al., 1996), a Drosophila ‘intrinsic’ cell death trigger (see Figure 3A and E). We further analyzed the effects of Eiger overexpression in other tissues using different GAL4 drivers. When overexpressed in the dorsoventral compartments of the wing discs by a vg-GAL4 driver, Eiger entirely blocked wing formation (Figure 2B), whereas Reaper induced only a regional defect (Figure 2D). Ectopic expression of Eiger in precursor cells for the external sensory organs by a sca-GAL4 driver resulted in disorganized macrochaetae in the notum and scutellum (Figure 2F), whereas Reaper induced a complete loss of bristles in these regions (Figure 2G). In the abdomen, on the other hand, sca>regg1GS9830 resulted in a severe developmental defect (Figure 2I), whereas sca>reaper caused a loss of bristles in the tergum (Figure 2J). Thus, Eiger has the potential to induce developmental defects distinct from the defects caused by the cell-autonomous killer protein Reaper. Figure 2.Eiger has the potential to induce severe developmental defects that are distinct from the defects caused by Reaper. Eiger or Reaper was ectopically expressed within the dorsoventral boundaries of the wing disc by a vg-GAL4 driver (B and D). Note that Eiger entirely blocked wing formation (B, arrowheads), whereas Reaper induced only a regional defect (D, arrowhead), compared with wild-type wings (A and B). Ectopic expression of Eiger in precursor cells for the external sensory organs driven by sca-GAL4 resulted in disorganization of the macrochaetae in the notum and scutellum (F, arrowheads), and a severe developmental defect of the abdomen (I), whereas Reaper induced a complete loss of bristles in these regions (G and J), compared with wild-type fly (E and H). sca>GAL4 and UAS-reaper flies were mated at 18°C, because their progeny, bearing both transgenes, died as pupae when generated at 25°C. Download figure Download PowerPoint Figure 3.Eiger activates an IAP-sensitive cell death pathway that does not require caspase activity. (A–H) Genetic interactions of Eiger or Reaper with caspase inhibitory proteins. Genotypes are as follows: regg1GS9830/+; GMR-GAL4/+ (A), regg1GS9830/GMR-GAL4; UAS-p35/+ (B), UAS-diap1/+; regg1GS9830/GMR-GAL4 (C), regg1GS9830/UAS-dronc DN; GMR-GAL4/+ (D), UAS-reaper/GMR-GAL4 (E), UAS-reaper/GMR-GAL4; UAS-p35/+ (F), UAS-diap1/+; UAS-reaper/GMR-GAL4 (G) and UAS-reaper/UAS-dronc DN; GMR-GAL4/+ (H). (I and J) Heterozygosity at the diap1 locus (th4) enhances the Eiger-induced eye phenotype. Whereas flies with a half dosage of the diap1 gene show a perfectly normal eye (I), the reduced-eye phenotype caused by Eiger overexpression (A) is strongly enhanced by the heterozygosity of th4 (J), indicating that endogenous DIAP1 negatively regulates the Eiger-stimulated death signal. Genotypes are th4/+ (I) and regg1GS9830/+; GMR-GAL4/th4 (J). (K–M) (DMe)2R staining of the eye discs shows that Eiger weakly but significantly causes caspase activation. Caspase activity was detected with rhodamine-110 fluorescence released from (DMe)2R. Genotypes are as follows: GMR-GAL4/+ (K), UAS-reaper/+; GMR-GAL4/+ (L) and UAS-eiger/+; GMR-GAL4/+ (M). Arrowheads indicate the position of the morphogenetic furrow. Download figure Download PowerPoint Eiger activates an IAP-sensitive cell death pathway that does not require caspase activity In mammals, the binding of death ligands such as TNF-α and FasL to their receptors triggers the activation of caspase-8, leading to the subsequent caspase-dependent cell death cascade (Ashkenazi and Dixit, 1998). To assess whether Eiger stimulates similar death signaling in Drosophila, we co-expressed Eiger and caspase-inhibitory proteins such as baculovirus P35, Drosophila inhibitor-of-apoptosis protein 1 (DIAP1) or a dominant-negative form of DRONC (DRONC-DN), a P35-resistant Drosophila apical caspase (Meier et al., 2000b). P35 and DRONC-DN exhibited only slight suppressive effects on the Eiger-induced eye phenotype (Figure 3B and D), although they strongly suppressed the Reaper-induced eye ablation, as reported previously (White et al., 1996) (Figure 3F and H). These results indicate that Eiger-induced cell death does not require caspase activity differing from the mammalian ‘extrinsic’ cell death system. In contrast, the co-expression of DIAP1 strongly suppressed the Eiger-induced eye phenotype (Figure 3C), suggesting that DIAP1 could block Eiger-activated death signaling through a mechanism that was independent of caspase inhibition. We also assessed genetically whether endogenous DIAP1 negatively regulates the Eiger-induced phenotype. Whereas heterozygous diap1 mutant flies exhibit a normal eye (Figure 3I), GMR>regg1GS9830 flies with a half dosage of the diap1 gene displayed a completely ablated eye phenotype (Figure 3J) compared with the reduced-eye phenotype of GMR>regg1GS9830 flies (Figure 3A), suggesting that DIAP1 is an endogenous inhibitor of Eiger-induced cell death signaling. To assess whether Eiger does cause caspase activation, the eye disc was stained with (DMe)2R, a caspase substrate that contains only an aspartate residue linked to rhodamine-110 (Hug et al., 1999). The eye disc from GMR>reaper flies showed a strong rhodamine-110 fluorescence at the region posterior to the morphogenetic furrow (Figure 3L), compared with the eye disc from GMR-GAL4 flies (Figure 3K). In the GMR>regg1GS9830 eye disc, the fluorescence was detected at lower, but significant, levels in many cells posterior to the furrow (Figure 3M). These data suggest that although caspase activation is not essential for cell death execution, Eiger activates both caspase-dependent and -independent signaling pathways. Eiger triggers the Drosophila JNK pathway Many mammalian TNF superfamily proteins activate both the NF-κB and the JNK pathway, and activation of the latter pathway facilitates cell death (Davis, 2000). To examine whether Eiger activates the JNK pathway, we tested the genetic interactions of Eiger with the components of the Drosophila JNK cascade (Figure 4A–I). The reduced-eye phenotype induced by Eiger (Figure 4B) was strongly suppressed in a heterozygous mutant of Drosophila JNK, basket (bsk) (Figure 4C). In addition, overexpression of a dominant-negative form of Bsk almost completely suppressed the eye phenotype (Figure 4D). Moreover, heterozygosity at the hemipterous (hep) locus, which encodes Drosophila JNKK, suppressed the reduced-eye phenotype much like bsk did (Figure 4E), and its hemizygosity (null background) rescued the phenotype almost completely (Figure 4F). Furthermore, the co-expression of a dominant-negative form of dTAK1 (Drosophila JNKKK) (Mihaly et al., 2001) also rescued the Eiger-induced phenotype completely (Figure 4G). Misshapen (Msn) is a MAPKKKK that is genetically upstream of the JNK pathway in Drosophila (Su et al., 1998). We found that a half dosage of the msn gene strongly suppressed the Eiger-induced phenotype (Figure 4H). Heterozygosity of Drosophila jun, a target of the JNK pathway, did not show any genetic interaction with Eiger (Figure 4I), raising the possibility that the Eiger-stimulated death-inducing JNK pathway may not require new transcripts that are controlled by Jun. Figure 4.Eiger induces a reduced-eye phenotype through the activation of the JNK pathway. (A–I) regg1GS9830 genetically interacts with components of the Drosophila JNK signaling pathway. The Eiger-induced phenotype is suppressed by heterozygosity at the bsk, hep or msn locus (C, E and H, respectively). Hemizygosity at the hep locus almost completely suppresses the phenotype (F). Co-expression of a dominant-negative form of Bsk or dTAK1 completely inhibits the Eiger-induced eye phenotype (D and G). Heterozygosity at the Drosophila jun locus does not suppress the eye phenotype (I). Genotypes are as follows: wild type (A), regg1GS9830/GMR-GAL4; TM3,Sb/+ (B), regg1GS9830/bsk2; GMR-GAL4/+ (C), regg1GS9830/GMR-GAL4; UAS-bsk DN/+ (D), hep1/+; regg1GS9830/+; GMR-GAL4/+ (E), hep1/Y; regg1GS9830/+; GMR-GAL4/+ (F), regg1GS9830/UAS-dTAK1 DN; GMR-GAL4/+ (G), regg1GS9830/+; GMR-GAL4/msn172 (H) and regg1GS9830/jun2; GMR-GAL4/+ (I). (J and K) Overexpression of Eiger activates the JNK pathway. The JNK activation was monitored in GMR-GAL4/+ (J) and regg1GS9830/+; GMR-GAL4/+ (K) background eye disc by puc-LacZ expression. X-gal staining (2 h) of the eye disc shows dramatic activation of the JNK pathway in the region posterior to the morphogenetic furrow (K, arrowhead), where Eiger is overexpressed. (L) Eiger stimulates the phosphorylation of Bsk in vivo. hs-GAL4/+ or regg1GS9830/+; hs-GAL4/+ larvae were heat shocked at 37°C for 45 min and cultured at 25°C for a further 4 h, and then subjected to western blot analysis with an anti-JNK or an anti-phospho-JNK antibody. (M) A model for the JNK signaling triggered by Eiger. Download figure Download PowerPoint Puckered (Puc) is a dual-specificity phosphatase, the expression of which is induced by the Drosophila JNK pathway to inactivate Bsk, so that puc expression can be used to monitor the extent of activation of the JNK pathway (Adachi-Yamada et al., 1999). To confirm that the JNK pathway was actually activated by Eiger, puc expression level was assessed in the eye disc of GMR>regg1GS9830 flies using a puc-LacZ enhancer-trap allele. The strong induction of puc-LacZ was observed in the region posterior to the morphogenetic furrow of the eye disc (Figure 4K) compared with the control eye disc (Figure 4J). Furthermore, western blot analysis with an anti-phospho-JNK antibody revealed that Bsk was phosphorylated by Eiger overexpression (Figure 4L). These genetic and biochemical data led to a model in which Eiger activates Msn, thereby triggering the JNK signaling pathway, sequentially stimulating dTAK1, Hep and Bsk (Figure 4M). We also tested, using RT–PCR analysis, whether Eiger could stimulate the NF-κB pathway; however, no obvious upregulation of the antimicrobial peptide genes, the target genes of the Drosophila NF-κB pathway, was detected (data not shown). Eiger is predominantly expressed in the nervous system The spatial patterns of eiger expression during embryogenesis and in larval tissues were analyzed by in situ hybridization (Figure 5). Weak signals were detected in pre-blastoderm embryos (stage 1–3), indicating a low level of maternal contribution (Figure 5A). After stage 10, eiger transcripts were predominantly detected in the nervous system of the embryos (Figure 5B, C and G–J). In the third-instar larva, eiger was strongly expressed in the brain hemispheres (Figure 5K) and at the morphogenetic furrow in the eye disc (Figure 5M); it was also expressed at significant levels in many cells posterior to the furrow. Double staining of the eye disc with the eiger RNA probe and an anti-ELAV antibody, which recognizes terminally differentiated neuronal nuclei, revealed that eiger was strongly expressed in the proliferating cells at the furrow (Figure 5O and P). RT–PCR analysis demonstrated that eiger mRNA was expressed at all stages of Drosophila development (data not shown). Figure 5.eiger is predominantly expressed in the nervous system. (A–J) Whole-mount in situ hybridization of wild-type embryos at various stages of development using eiger antisense (A–C and G–H) or sense (D–F) RNA probes. A pre-blastoderm embryo showed low levels of eiger expression (A). After germ band extension, strong staining was evident in the nervous system (B, C and G–J). (K–N) In situ analysis of eiger expression in third-instar larval brains (K and L) and eye discs (M and N) using eiger antisense (K and M) or sense (L and M) RNA probes. A high level of eiger expression was evident in the brain hemispheres (K). In the eye disc, stronger staining was detected at the region posterior to the morphogenetic furrow (M). Double staining of the eye discs with the eiger RNA probe (O) and the anti-ELAV antibody (P) revealed that eiger was strongly expressed in the proliferating cells at the furrow. Download figure Download PowerPoint Eiger is a physiological ligand for the Drosophila JNK pathway To analyze the physiological role of Eiger, we generated a loss-of-function mutant for eiger by imprecise excision of the P element in the Regg1GS9830 strain. We isolated nine mutants harboring deletions that removed parts of the eiger genomic sequences, but leaving right sequences downstream of the P element intact (Figure 6A and B). egr1 and egr3 mutants were homozygous viable with greatly reduced eiger expressions (Figure 6C and E). In the eye disc of the puc-LacZ enhancer-trap line, endogenous JNK activity could be detected in the region posterior to the morphogenetic furrow by long-time (20 h) X-gal staining (Figure 6F, arrowhead). We found that the puc-LacZ expression was dramatically reduced in the eye disc of the eiger mutant (Figure 6G). Notably, the JNK activities at the marginal region of the eye disc, where endogenous eiger expression was not detected (Figures 5M and 6D), were not affected in the eiger mutant (Figure 6F and G, arrows). These findings indicate that Eiger functions as a physiological ligand for the Drosophila JNK pathway. Figure 6.Eiger is a physiological ligand for the JNK pathway. (A) Diagram of the eiger genomic locus and location of the end points of excision mutations in egr1 and egr3 mutant alleles. Exon sequences are shown by closed boxes. (B) PCR analysis of genomic DNA from wild-type and homozygous eiger mutant adult flies using primers a and b [indicated in (A)]. The end points of genomic deletions in eiger mutants were determined by sequencing of the PCR products. (C) RT–PCR analysis of cDNA from wild-type and homozygous eiger mutant adult flies using primers c and d or e and f [indicated in (A)] to determine the expression levels of eiger and CG1371, respectively. GAPDH expression was determined as an internal control. (D and E) In situ analysis of eiger expression in the eye disc from wild-type (D) and egr1/egr1 (E) larvae. (F and G) The JNK activity was monitored in wild-type (F) and egr1/egr1 (G) background eye discs by puc-LacZ expression. Long-time (20 h) X-gal staining was able to detect the endogenous JNK activity in the region posterior to the morphogenetic furrow (F, arrowhead). The JNK activity was dramatically reduced in the eiger mutant (G). Note that the JNK activities in the disc margin, where eiger is not expressed even in the wild-type disc, are not affected in the eiger mutant (F and G). Download figure Download PowerPoint Discussion In our in vivo expression screen, we identified Eiger as a novel cell death trigger molecule in Drosophila. The structure and function of Eiger suggest that the extrinsic cell death-inducing mechanism might be evolutionarily conserved in Drosophila. Genetic evidence reveals that caspase activation is not essential to execute Eiger-induced cell death. The Drosophila extrinsic cell death system might predominantly utilize the caspase-independent pathway, in contrast to the intrinsic cell death system, which is regulated by Reaper, Hid and Grim, and depends completely on caspase activation. Although caspases do take p