DECAY, a Novel Drosophila Caspase Related to Mammalian Caspase-3 and Caspase-7*

Caspases are key effectors of programmed cell death in metazoans. In Drosophila, four caspases have been described so far. Here we describe the identification and characterization of the fifth Drosophila caspase, DECAY. DECAY shares a high degree of homology with the members of the mammalian caspase-3 subfamily, particularly caspase-3 and caspase-7. DECAY lacks a long prodomain and thus appears to be a class II effector caspase. Ectopic expression of DECAY in cultured cells induces apoptosis. Recombinant DECAY exhibited substrate specificity similar to the mammalian caspase-3 subfamily. Low levels of decaymRNA are ubiquitously expressed in Drosophila embryos during early stages of development but its expression becomes somewhat spatially restricted in some tissues. During oogenesisdecay mRNA was detected in egg chambers of all stages consistent with a role for DECAY in apoptosis of nurse cells. Relatively high levels of decay mRNA are expressed in larval salivary glands and midgut, two tissues which undergo histolysis during larval/pupal metamorphosis, suggesting that DECAY may play a role in developmentally programmed cell death inDrosophila.

Programmed cell death in metazoans is mediated by caspases, a family of cysteine proteases, which cleave their substrates following an Asp residue (1)(2)(3)(4)(5). A number of caspases have been described in both vertebrates and invertebrates. To date fourteen caspases have been cloned in mammals, some of which play a critical role in apoptosis, whereas others seem mainly involved in the processing and activation of proinflammatory cytokines (1)(2)(3)(4)(5). Although four caspases exist in the nematode Caenorhabditis elegans, only one, CED-3, is essential for all developmentally programmed cell death (6,7). In Drosophila melanogaster four caspases, named DCP-1, DREDD/DCP-2, drICE and DRONC, have been reported so far (8 -12). Among these caspases, DREDD and DRONC contain long prodomains carrying death effector domains and a caspase recruitment domain (CARD), respectively, suggesting that these two caspases may act as upstream (class I) caspases. On the other hand, DCP-1 and drICE lack long prodomains and are thus similar to downstream effector (class II) caspases in mammals. Currently, loss of function mutants are only available for dcp-1. dcp-1 mutation results in larval lethality and melanotic tumors (8). Additionally, DCP-1 is required for Drosophila oogenesis, as dcp-1 mutants show a defect in transfer of nurse cell cytoplasmic contents to developing oocytes (13). The transcript for dredd accumulates in embryonic cells undergoing programmed cell death and in nurse cells in the ovary at a time that coincides with nurse cell death (9). dronc mRNA is widely expressed during development and appears to be up-regulated by ecdysone in larval salivary glands and midgut before histolysis of these tissues (12). The precise roles of drICE and DRONC in programmed cell death in Drosophila have not been established. However, in vitro antibody depletion experiments suggest that drICE is required for apoptotic activity in the S2 Drosophila cell line (14). Accumulation of dronc mRNA in salivary glands and midgut may be required to sensitize these tissues for deletion by apoptosis during metamorphosis. These recent studies suggest that specific caspases may mediate tissue and stage specific programmed cell death during Drosophila development.
To fully understand the role of various caspases in cell physiology, it is important to identify all caspases in a given model organism. In this study, we describe the characterization of DECAY, the fifth Drosophila caspase. DECAY is highly similar to class II executioner caspases such as mammalian caspase-3 and caspase-7. We show that decay gene expression is widespread in developing fly embryos, and DECAY has substrate specificity similar to caspase-3 subfamily of caspases.

EXPERIMENTAL PROCEDURES
Identification and Sequencing of Decay cDNA-DECAY was identified through a homology search with mammalian caspases as a Gen-Bank TM expressed sequence tag (accession number AI259958). The expressed sequence tag clone was obtained from Berkley Drosophila Genome Project in pOT2 vector and was sequenced in full. This clone (clone ID LP3492) contained a 1,797-base pair insert, which was much longer than the 1.1-kilobase predicted size of the transcript (see below). A careful examination of the sequence suggested that ϳ700 base pair of the 3Ј sequence in this clone may be derived from fusion of a heterologous cDNA. The 5Ј 1,101 base pair of the sequence, including a 20 residue poly(A) tail, is likely to represent the authentic full-length decay cDNA containing the entire coding region. This sequence has been deposited in the GenBank TM data base under accession number AF130469.
Plasmid Constructs-The 0.85-kilobase coding region of decay was amplified from the original pOT2-decay vector by polymerase chain reaction using Pwo polymerase (Roche Biochemicals) and the following oligonucleotides: Primer A, 5Ј-GGCGGATCCGCCGCCATGGCACCAA-GATCCCATACG; and Primer B, 5Ј-CCGGAATTCTCACTTGTCATCG-TCGTCCTTGTAGTCGGTCTTGGGCTTAACACG (nucleotides corresponding to decay sequence are underlined). Primer A contained a consensus Kozak sequence that required alteration of the initiation site from the original sequence and a BamHI cloning site. Primer B con-* This work was supported in part by grants from the Wellcome Trust and Australian Research Council. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM 1. DECAY sequence and its relationship to other known Drosophila caspases. A, deduced amino acid sequence of DECAY consists of 287 amino acid residues. The pentapeptide sequence QACRG, encompassing the catalytic Cys 150 , is underlined. B, an amino acid sequence alignment of the known Drosophila caspases. AC005466 is the GenBank TM accession number of contig of Drosophila genomic sequence that contains the coding region for a putative caspase. The partial sequence for this caspase shown here, which lacks the amino-terminal region, is derived from a single exon. Alignments were obtained using CLUSTAL W program at European Bioinformatics Institute. Residues conserved in at least five caspases are shown in black boxes. Similar residues in at least five caspases or those identical in four caspases are shown in gray boxes. C, phylogenetic relationship between various Drosophila caspases.
tained an EcoRI site and sequence encoding a FLAG tag. Amplified product was purified and cloned directionally into pcDNA3 (Invitrogen). The catalytic Cys 150 residue of DECAY was mutated to a Gly residue by Quickchange mutagenesis (Stratagene) using pOT2-decay as template. Mutant decay cDNA encoding DECAY(C150G) protein was cloned directionally into pcDNA3 as described above. Primer C, 5Ј-GCGAATTC-CATATGCCACCAAGATCCCAT containing a NdeI site and primer D, 5Ј-GGCGGATCCCGGGTCTTGGGCTTAACACGCAG containing a BamHI site (sequence corresponding to decay is underlined in both primers) were used to amplify wild type and catalytic cysteine mutant DECAY for directional cloning into pET32b vector (Novagen).
Transient Transfection-NIH3T3 cells were maintained in Dulbecco's modified Eagle's medium with 10% fetal calf serum. 293T cells were grown in RPMI with 10% fetal calf serum. For cell death assays, 2 ϫ 10 5 cells were plated per 35-mm dish the day before transfection. 1.5 g of pcDNA3-decay, pcDNA3-decay (C150G) or empty vector, were cotransfected with 0.5 g of a ␤-galactosidase expression plasmid (pEF-␤gal) (20). All transfections were carried out using Fugene6 transfection reagent (Roche Biochemicals) according to manufacturer's instructions. Cells were fixed and stained with X-gal at 24 h or 48 h post-transfection, and ␤-galactosidase positive cells were scored for apoptotic morphology as described previously (20 -22). At 24 h post-transfection, subcellular localization of DECAY was determined by fixing cells in 47.5% methanol, 47.5% acetone, 5% formaldehyde, and staining with anti-FLAG monoclonal antibody (Sigma), followed by anti-mouse-fluorescein isothiocyanate (Roche Biochemicals) for 30 min each. Cells were viewed using a fluorescence microscope (Olympus BH2-RFCA).
Northern and in Situ mRNA Analysis-Total RNA from various developmental stages of Drosophila or adult flies was prepared using RNAzol B according to the manufacturer's (Tel-Test Inc.) protocol. Poly(A) ϩ -enriched RNA was prepared using oligo dT magnetic beads (Dynal). Approximately 20 g of total RNA or 1-2 g of poly(A) ϩ RNA was electrophoresed onto a 2.2-M formaldehyde gel and transferred to Biodyne A nylon membrane (Pall Corp.). The blot was hybridized to a 32 P-labeled decay cDNA coding region probe and exposed to Kodak XAR-5 film. For in situ RNA analysis, antisense and sense digoxygeninlabeled riboprobes were prepared using appropriate RNA polymerases from linearized decay cDNA clone. Digoxygenin labeling was performed according to the manufacturer's instructions (Roche Molecular Biochemicals). In situ hybridization to Drosophila embryos and larval tissues was essentially as described (23) with some modifications. Embryos and dissected larval tissues were fixed in 0.1 M HEPES, 50 mM EGTA, 0.01% Nonidet P-40, 4% formaldehyde, pH 6.9 for 20 min. The proteinase step was omitted. Dissected ovaries from 3-day-old adult females were fixed as described for embryos and treated with 50% ethanol and 50% xylene for 30 min, washed in ethanol, then in methanol, and finally in phosphate-buffered saline with 0.01% Triton X-100 (PBS-T). Ovaries were then refixed for 25 min in 4% paraformaldehyde and treated with proteinase K (5 g/ml) for 8 min at room temperature. After hybridization, nonspecifically bound probe was removed by digestion with RNase A (125 g/ml in PBS-T) for 1 h at 37°C. Hybridization was detected using the alkaline phosphatase-coupled secondary antibody detection system according to the manufacturer's instructions (Roche Molecular Biochemicals).

RESULTS AND DISCUSSION
Identification of DECAY-While searching for new molecules with homology to various mammalian caspases, using TBLASTN program, we identified an expressed sequence tag in the GenBank TM data base which encoded a partial caspase-like molecule. The sequencing of the entire clone revealed that the cDNA has an open reading frame of 287 amino acid residues with a high degree of homology to mammalian caspases, particularly those related to the caspase-3 subfamily (Fig. 1). We named this new molecule DECAY, for Drosophila executioner caspase related to Apopain/Yama. DECAY shares approximately 39% identity (54% similarity) with Spodoptera frugiperda caspase-1, 37% identity (56% similarity) with mammalian caspase-3 and caspase-7, 35% identity (53-55% similarity) with Drosophila caspases DCP-1 and drICE, and 32-33% identity (48 -52% similarity) with caspase-8 and caspase-10.
An alignment of all known Drosophila caspases showed that DECAY is most homologous to an unpublished putative caspase encoded by genomic sequence contained in a data base entry (accession number AC005466), followed by DCP-1 and Cleavage products were electrophoresed on SDS-polyacrylamide gels, transferred to nitrocellulose membranes, and autoradiographed. The truncated PARP cDNA translates into a 38-kDa protein, which when cleaved following a DEVD sequence, generates 24-and 14-kDa fragments. As expected, the catalytically inactive C150G mutant DECAY does not cleave PARP, whereas, DECAY and caspase-3 generate identical sized fragments of 24 and 14 kDa. drICE (Fig. 1, B and C). Similar to drICE and DCP-1, DECAY lacks a long amino-terminal prodomain. DECAY is distantly related to DRONC and DREDD, the two class I Drosophila caspases. Overall, DECAY is more similar to S. frugiperda caspase-1 and mammalian caspase-3 and caspase-7 than all known Drosophila caspases. Interestingly, DECAY is the only Drosophila caspase that carries a QACRG sequence encompassing the putative catalytic Cys 150 residue. In this respect, it is similar to the majority of mammalian caspases, including caspase-3 and caspase-7.
By hybridizing to a filter containing Drosophila genomic P1 clones, we localized the decay gene to the Fas1 contig located within chromosome region 89C6-D4 (data not shown). None of the other published Drososphila caspase genes map to this region.
DECAY Has a Substrate Specificity Similar to the Caspase-3 Subfamily-To confirm that DECAY is indeed a caspase, we expressed full-length wild-type and C150G mutant DECAY fused to 6xHis in E. coli. The majority of the protein expressed in E. coli was insoluble and became inactive upon attempts to purify under both native and denaturing conditions (data not shown). Therefore, we analyzed caspase activity in the soluble fraction of bacterial extracts using fluorogenic peptide substrates. DECAY did not show significant activity on caspase-1 substrate YVAD-amc, however, it efficiently cleaved caspase-3 substrate DEVD-amc ( Fig. 2A). Interestingly, DECAY was substantially more active on the pentapeptide substrate VDVADamc. This peptide is a preferred substrate for caspase-2, which requires a P 5 residue (24). However, in our hands, VDVAD-amc was also a much better substrate for caspase-3, as compared with DEVD-amc ( Fig. 2A). As expected, the C150G mutant DECAY did not exhibit any appreciable caspase activity.
Poly(ADP-ribose) polymerase (PARP) is one of the key cellular substrates of caspase-3 (25, 26). To check whether PARP can serve as a substrate for DECAY in vitro, we incubated a  (18) with recombinant DECAY and measured its cleavage by electrophoresis and autoradiography. As shown in Fig. 2B, this protein was efficiently cleaved by DECAY. The cleavage products generated by DECAY were identical in size to those generated by caspase-3 suggesting that both caspases cleave following the same DEVD sequence in PARP.
Ectopic Expression of DECAY in Cultured Cells-Many caspases, when overexpressed in cultured cells, induce apoptosis to some degree. We therefore analyzed whether DECAY is able to induce apoptosis in transfected cells. In 293T cells, at 24 h following transfections, around 35% of cells transfected with the wild-type decay construct showed apoptotic morphology when compared with cells transfected with the empty vector or an expression construct carrying the C150G mutant DECAY (Fig. 3A). In NIH3T3 cells, by 24 h post-transfection, a small number (ϳ10%) of cells transfected with the wild-type decay were apoptotic (Fig. 3A). By 48 h, decay transfected cells showing apoptotic morphology increased slightly to around 15% (data not shown). This level of cell death induced by DECAY overexpression is similar to that induced by caspase-3 under similar conditions (22,27). We also assessed the effect of DECAY overexpression in MCF-7 cells and Drosophila S2 cells. In both cases levels of apoptosis similar to those seen in NIH3T3 cells were observed (data not shown).
Using FLAG-tagged DECAY, we further investigated the subcellular localization of DECAY protein in transfected cells by immunofluorescence analysis employing an anti-FLAG antibody and a fluorescein isothiocyanate-coupled secondary antibody. In both NIH3T3 and 293T cells, most of the DECAY-FLAG protein was present in the cytoplasmic compartment (Fig. 3B).
decay mRNA Expression During Drosophila Development-In RNA blots, decay was present as an approximately 1.1-kilobase transcript in most developmental stages, larvae, pupae, and in the adult fly (Fig. 4). Relatively high levels of decay transcript were detected in the adult fly (Fig. 4B). We further analyzed the expression pattern of decay during fly development by in situ hybridization to Drosophila embryos and larval tissues using a digoxigenin-labeled antisense mRNA probe (Fig. 5). decay is expressed at low levels throughout embryogenesis and shows no specific up-regulation at stage 11 (Fig. 5, A-C, and data not shown) when programmed cell death first becomes evident in Drosophila (28). decay mRNA was present in stage 1-4 syncitial embryos (not shown), suggesting that it is maternally deposited into the embryo, because zygotic expression does not begin before stage 5 (29). In stage 6 -7 cellularized embyros, decay mRNA is ubiquitously expressed (Fig. 5B), but in later stages decay mRNA is present at higher levels within the gut (Fig. 5C). We also examined the expression of decay in third instar larval tissues and during oogenesis (Fig. 5, E-L). High levels of decay expression was observed in salivary glands and midgut tissue from third instar larvae (Fig.  5, E and F), preceding the onset of apoptosis in these tissues, which occurs after pupariation (30). Only very low levels of decay expression were observed throughout third instar larval eye imaginal discs and brain lobes (Fig. 5, G and H), which contain apoptotic cells at this stage (31). However, up-regulation of decay expression was not observed in eye disc or brain lobe cells undergoing apoptosis.
During oogenesis decay mRNA is detected in egg chambers of all stages but was present at higher levels in the nurse cells after stage 10a (Fig. 5I and data not shown). In stage 12 egg chambers, decay mRNA was absent from nurse cells that have initiated apoptosis and present at high levels in the developing oocyte (Fig. 5, J and K), consistent with dumping of the nurse cell cytoplasm into the oocyte that occurs at this stage (32). The expression of decay mRNA in egg chambers is consistent with a role for decay in apoptosis of the nurse cells.
We have previously shown that dronc mRNA is up-regulated when isolated salivary glands and midgut from second instar larvae are exposed to ecdysone (12). Because larval salivary glands and midgut show relatively high expression of decay transcript, it was of interest to check whether decay is also regulated by ecdysone. As shown in Fig. 6, no up-regulation of decay transcript was evident in ecysone treated salivary glands and midgut. Under similar conditions, dronc transcript was up-regulated at least 5-fold in response to ecdysone (Fig. 6).
Conclusions-We have described here preliminary characterization of a new Drosophila caspase DECAY. Presence of multiple caspases in Drosophila suggest that cell death pathways in the fly are likely to be complex. DECAY is most similar to caspase-3-like effector caspases and shares a similar substrate specificity. Low levels of decay transcript are widely expressed during Drosophila embryogenesis. Higher expression of decay mRNA in larval salivary glands and midgut suggests a possible role for DECAY in programmed deletion of these obsolete tissues during metamorphosis and tissue remodeling. Additionally, moderate expression of decay mRNA in nurse cells suggests a possible role for DECAY in nurse cell death. Because high levels of decay transcript are also found in the adult animals, DECAY may also be involved in regulating the normal cell turnover in the adult. Generation of loss-offunction decay mutant, or RNA ablation studies would shed further light on the role of DECAY in programmed cell death in Drosophila.
Mammalian caspases have been proposed to belong to two groups. The upstream, initiator, or class I caspases carrying specific protein-protein interaction domains are autoactivated when several molecules are clustered in close proximity follow-ing recruitment via specific adaptors (33). The downstream, effector, or class II caspases require processing by class I caspases, but once activated these caspases (e.g. caspase-3) can also mediate the processing of some class I caspases, which probably serves as a signal amplification mechanism (34). Because DECAY is a class II caspase, we analyzed whether it can be processed by the known Drosophila caspases (data not shown). However, so far we have been unable to identify a protease capable of cleaving DECAY. Further studies are required to delineate the mechanism of DECAY activation.