Characterization of a functionally expressed dipeptidyl aminopeptidase III from Drosophila melanogaster Claire Mazzocco 1, *, Kayoko M. Fukasawa 2 , Patrick Auguste 3 and Jacques Puiroux 1, † 1 Laboratoire des Re ´ gulations Neuroendocriniennes, Universite ´ Bordeaux I, Talence, France; 2 Department of Oral Biochemistry, Matsumoto Dental College, Shioriji, Nagano, Japan; 3 Laboratoire des Facteurs de Croissance et de la Diffe ´ renciation Cellulaire, INSERM EPI-0113, Universite ´ Bordeaux I, Talence, France A Drosophila melanogaster cDNA clone (GH01916) encoding a putative 723-residue long (82 kDa) protein (CG 7415) and displaying 50% identity with mammalian cytosolic dipeptidyl aminopeptidase (DPP) III was func- tionally expressed in Schneider S 2 cells. Immunocytochemi- cal studies using anti-(rat liver DPP III) Ig indicated the expression of this putative DPP III at the outer cell mem- brane and into the cytosol of transfected cells. Two protein bands (82 and 86 kDa) were immunologically detected after PAGE and Western blot of cytosol or membrane prepared from transfected cells. Western blot analysis of partially purified D. melanogaster DPP III confirmed the over- expression of these two protein bands into the cytosol and on the membranes of transfected cells. Despite the identification of six potential glycosylation sites, PAGE showed that these protein bands were not shifted after deglycosylation experi- ments. The partially purified enzyme hydrolysed the insect myotropic neuropeptide proctolin (Arg-Tyr-Leu-Pro-Thr) at the Tyr-Leu bond (K m  4 l M ). In addition, low con- centration of the specific DPP III inhibitor tynorphin pre- vented proctolin degradation (IC 50 ¼ 0.62 ± 0.15 l M ). These results constitute the first characterization of an evo- lutionarily conserved insect DPP III that is expressed as a cytosolic and a membrane peptidase involved in proctolin degradation. Keywords: enkephalinase; genome sequencing; insects; neuropeptides; proctolin. Mammalian DPP III was first discovered in the bovine anterior pituitary gland [1] and it has been recently cloned from rat liver as a 738-residue (82 kDa) cytosolic protein [2,3]. This enzyme (EC 3.4.14.4) is a zinc metallopeptidase containing a specific domain HELLGH-18X-E where a zinc molecule is bound to both histidines [4]. DPP III is mainly identified as a cytosolic peptidase, but DPP III was also detected on membranes prepared from the brain of guinea- pig [5] and rat [6]. Angiotensins and enkephalins constitute the preferred substrates of the rat brain cytosolic DPP III [7]. The routes of degradation of the insect myotropic neuropeptide proctolin (Arg-Tyr-Leu-Pro-Thr) have been compared to those of enkephalins. Indeed, a dipeptidyl aminopeptidase activity appears as one major proctolin- degrading peptidase, liberating the N-terminal Arg-Tyr dipeptide [8–11]. This dipeptidyl aminopeptidase activity was compared to the vertebrate DPP III [11] and is mainly recovered as a cytosolic enzyme [8]. Interestingly, a proct- olin-degrading DPP activity is also measured on membranes [8,9] especially those obtained from insect proctolin-rich tissues such as hindgut [10]. None of the presumed proctolinases has been fully characterized yet. We recently purified [12] from hindgut membranes of the cockroach, Blaberus craniifer, a proctolin-degrading protein (76 and 80 kDa) that removes the N-terminal dipeptide from proctolin (K m ¼ 3.8 ± 1.1 l M ) and enkephalins (K m ¼ 4.2 ± 0.8 l M ). The partial sequencing of this puri- fied protein revealed a significant homology with the rat liver cytosolic DPP III that was confirmed by the specific detection of this purified insect protein with anti-(rat liver DPP III) Ig in Western blot analysis. In addition, this sequencing allowed the identification in Drosophila melano- gaster of a set of homologous cDNA sequences and a related genomic sequence (available at the Berkeley Droso- phila Genome Project) that encode a deduced 723-residue long protein (82 kDa) sharing 50% identity with mamma- lian DPP III. This D. melanogaster DPP III is annotated CG7415 [13]. Western blot analysis of crude membrane and soluble material prepared from D. melanogaster showed the presence of two protein bands (82 and 86 kDa) immuno- logically related to vertebrate DPP III [12]. From these results, it could be speculated that an evolutionarily conserved DPP III is present in insects. In the present study, the functional expression of the put- ative D. melanogaster DPP III was attempted in Schneider Correspondence to J. Puiroux, Laboratoire de Neurobiologie des Re ´ seaux, CNRS-UMR 5816, Universite ´ Bordeaux 1, Avenue des Faculte ´ s, 33405 Talence Cedex, France. Fax: + 33 557 962561, Tel.: + 33 557 962569, E-mail: j.puiroux@lnr.u-bordeaux.fr Abbreviations: BDGP, Berkeley Drosophila genome project; DAB, diaminobenzidine; DPP, dipeptidyl aminopeptidase. Enzyme: rat liver DPP III (EC 3.4.14.4). *Present address: Groupe de Recherche pour l’Etude du Foie, INSERM E 9917, Universite ´ Victor Segalen Bordeaux II, 146 rue Le ´ o Saignat, 33076 Bordeaux, France. Present address: J. Puiroux, Laboratoire de Neurobiologie des Re ´ seaux, CNRS-UMR 5816, Universite ´ Bordeaux 1, Avenue des Faculte ´ s, 33405 Talence Cedex, France. (Received 20 March 2003, revised 19 May 2003, accepted 27 May 2003) Eur. J. Biochem. 270, 3074–3082 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03689.x S 2 cells to investigate its apparent dual cellular location and to examine the characteristics of this enzyme. Immunocyto- chemical studies, Western blot analysis using anti-(rat liver DPP III), partial purification and proctolin degradation studies demonstrate the overexpression of the D. melano- gaster DPP III on membranes and into the cytosol of transfected cells. These results confirm that DPP III can be attached to cell membranes, and they represent the first identification and characterization of an insect DPP III that plays a major role in proctolin degradation. Experimental procedures Materials Diaminobenzidine (DAB), Hepes, hydrogen peroxide, enkephalins, the enkephalin N-terminal dipeptide Tyr-Gly and proctolin were purchased from Sigma (France). All chromatographic materials were from Pharmacia (Uppsala, Sweden). The specific DPP III inhibitor tynorphin (Val- Val-Tyr-Pro-Trp) was a gift from K. Fukasawa (Matsu- moto Dental University, Nagano, Japan). The anti-(rat liver DPP III) was prepared as described by Fukasawa et al.[2]. Goat anti-rabbit Ig with peroxidase labelling was from Boehringer-Mannheim. The proctolin fragments Arg-Tyr and Leu-Pro-Thr were a gift from B. G. Loughton (York University, Ontario, Canada). The expression system in S 2 cells was from Invitrogen. Stable transfection of S 2 cells with a D. melanogaster cDNA encoding a putative DPP III A2.484kbHpaIfragmentoftheD. melanogaster cDNA clone GH01916 (in pOT2 vector) coding a putative DPP III was subcloned into the EcoRV site of pMTA/V5-His B expression vector previously dephosphorylated with calf intestinal alkaline phosphatase 1 . This vector is under the control of the metallothionein promoter. Schneider S 2 cells were cultured at 24 °C in Schneider S 2 cell medium containing L -glutamine (1 m M ), Penicillin-Streptomycin (25 mgÆmL )1 ) and 10% (v/v) heat-inactivated fetal bovine serum. Cells (10 6 cellsÆmL )1 ) were placed in Petri dishes (35 mm) and incubated overnight at 24 °C before transfec- tion. The mixture for cell transfection was prepared by mixing 36 lLof2 M CaCl 2 with 19 lg of expression vector pMTA/DPP III and 1 lg of selection vector pCoHYGRO coding for hygromycin B and adjusted to a final volume of 300 lL with sterilized MilliQ water. The mixture was added dropwise to 300 lL Hepes sterile buffer (50 m M Hepes, 1.5 m M NaH 2 PO 4 ,280m M NaCl, pH 7.1) and incubated for 30 min at room temperature prior to transfection of S 2 cells. A control transfection was carried out using the pMTA expression vector without the HpaIfragmentofthe D. melanogaster cDNA (ÔmockÕ transfection). After over- night incubation at 24 °C, the transfected cells were rinsed twice with 2 mL of complete S 2 medium to remove the precipitates and finally incubated at 24 °Cwith2mLof complete S 2 medium. Stable recombinant cells were selected by adding hygromycin B (300 lgÆmL )1 ) 2 days post-trans- fection. The protein expression was induced by addition of CuSO 4 (final concentration 0.5 m M ) to hygromycin-selected recombinant cells 48 h before measurements. Analysis of transfected S 2 cells The transfection of S 2 cells by the expression vector pMTA/ DPP III was verified by PCR using a Perkin Elmer GeneAmp PCR System 2400 thermocycler. A sense primer designed against the sequence of the pMTA vector (5¢-GGGGATCTAGATCGGGGT-3¢) and an antisense primer specific for the HpaIfragment(5¢-AGCGGAAGT GTGATGCCG-3¢) were used with DNA from transfected, ÔmockÕ transfected or untransfected S 2 cells as template. These primers and a second set of primers (sense primer 5¢-GAATTCGAGGGCTTCGTGGCC-3¢ and antisense primer 5¢-AACGAGTCCTTCGCCTGCCT-3¢) specific for the HpaI fragment were used for RT-PCR. Total RNAs from transfected, ÔmockÕ transfected or untransfected S 2 cells were used after DNAse I treatment to prepare cDNA templates. Immunocytochemistry of stable transfected S 2 cells Immunocytochemical studies were performed on stable recombinant S 2 cells 48 h after induction. Twenty milliliters of cell culture were centrifuged (1000 g, 5 min) and the pelleted cells were rinsed twice with 1 mL NaCl/P i (137 m M NaCl, 2.7 m M KCl, 10 m M Na 2 HPO 4 ,1.8m M KH 2 PO 4 , pH 7.4). Then, the unfixed cells were incubated with anti- (rat liver DPP III) Ig (1/2000) in NaCl/P i with or without 0.1% (w/v) Tween 20 for 1 h at room temperature under slow constant agitation. The cells were centrifuged (1000 g, 5 min) and rinsed with 1 mL NaCl/P i for 15 min under slow agitation. Cells were then incubated for 1 h at room temperature with horseradish peroxidase labeled secondary antibody (1/1000). Cells were rinsed as described above, then incubated in Tris buffer (50 m M Tris/HCl, pH 8) containing DAB (1.39 m M ) for 5 min and finally revealed by addition of 0.01% (v/v) H 2 O 2 . Cells were rinsed twice and mounted on glass slide in 80% (w/v) glycerol and observed with a microscope (Reichert–Jung model Polyvar) and image analysis software (Spot RT, Diagnostic Instru- ments, USA). Partial purification of a putative D. melanogaster DPP III expressed in S 2 cells Typically, 50 mL of cell culture were centrifuged (1000 g, 5 min, room temperature) and the pelleted cells were rinsed twice with 1 mL NaCl/P i (pH 7). The pellet was suspended in 0.5 mL ice-cold Hepes buffer (10 m M Hepes, 5% (w/v) glycerol, 5 m M MgCl 2 , pH 7.2) and homogen- ized with a motor driven Teflon-glass homogenizer. The homogenate was centrifuged (3000 g for 10 min at 4 °Cin a Beckman J2-MC centrifuge). The pellet was discarded and the supernatant was centrifuged (40 000 g for 20 min at 4 °C). The final supernatant (cytosolic sample) was separated from the membrane pellet, filtered (0.45 lm pore size syringe filter) and stored at )80 °C. The membrane pellet was suspended in 1 mL Hepes buffer and recentrifuged (40 000 g for 20 min at 4 °C). The final membrane pellet was suspended in 200 lL Hepes buffer andstoredat)20 °C. A total of 8 L of stable trans- fected S 2 cells induced for 48 h were thus prepared for chromatography and the total cytosolic sample contained Ó FEBS 2003 Characterization of D. melanogaster DPP III (Eur. J. Biochem. 270) 3075 about 50 mg protein. The pooled membrane preparations contained approximately 45 mg proteins and this mem- brane homogenate (1 mg proteinÆmL )1 ) was solubilized by adding a concentrated Chaps solution (10% w/v in cold Hepes buffer) dropwise to a final concentration of 1%. Solubilization was carried out for 1 h at 4 °C under constant agitation and the sample was then ultracentri- fuged (100 000 g for 1 h at 4 °C, Beckman L8-55 centrifuge). The supernatant (solubilized sample) was filtered, added with 90 mL Hepes buffer (to reduce the detergent concentration) and processed for purification. The partial purification of expressed D. melanogaster putative DPP III was performed according to the method already described in [14]. Briefly, cytosolic or solubilized sample was first loaded on a 5-mL HiTrap Q Sepharose cartridge connected to a Pharmacia AKTA FPLC system (Pharmacia, Uppsala, Sweden) delivering 5 mLÆmin )1 . The cartridge was then rinsed with Hepes buffer (con- taining 0.1% w/v Chaps for the solubilized sample) and proteins were eluted by a regular NaCl gradient (from 10 to 500 m M in 25 min) generated with buffer B (Hepes buffer containing 1 M NaCl). Fractions were collected every minute and the presence of DPP III was identified by degradation studies using proctolin as substrate and by Western blot analysis with anti-(rat liver DPP III) Ig. The enzyme fractions were pooled and concentrated by ultrafil- tration (Macrosep Pall Filtron cartridge, molecular mass cut off 2 ¼ 10 000 spun at 4000 g for 90 min at 4 °C). The protein content and the DPP III activity were measured and the concentrated sample was applied to two Superdex HR 200 10/30 columns (connected in line). The isocratic separation of proteins was obtained with Hepes buffer delivered at 0.25 mLÆmin )1 . The enzyme fractions were identified as above. Both separations were monitored at 280 nm. Degrading activity of the functionally expressed putative D. melanogaster DPP III DPP III activity was measured in homogenates of whole transfected cells, in cytosolic or solubilized membrane samples and in partially purified fractions with proctolin (40 l M , 6 nmol) as substrate. The incubation was carried out in presence of bestatin (100 l M ) in a final volume of 150 lLfor15minat25°C under constant stirring. The reaction was stopped by the addition of 5 lLof2 M HCl and centrifugation (Hettich EBA 12R, 16 500 g at 4 °Cfor 8 min). The degradation products were separated on a Pharmacia PepRPC HR 5/5 reversed-phase column con- nected to an FPLC system (pump A: 0.1% v/v trifluoro- acetic acid in MilliQ water; pump B: 0.1% v/v trifluoroacetic acid, 60% v/v CH 3 CN in MilliQ water) delivering 1 mLÆmin )1 under the following gradient condi- tions: from 1 to 15% CH 3 CN for the first 5 min, then from 15 to 30% CH 3 CN for the next 15 min. The detection of neuropeptide fragments was monitored at 206 and 280 nm. The proctolin fragments Arg-Tyr and Leu-Pro-Thr were identified by coelution with standard solutions of di- and tripeptide. The effect of proctolin concentrations (from 1 to 250 l M ) and the effect of tynorphin concentrations (from 100 l M to 0.1 n M ) on the degradation of proctolin (40 l M ) were examined with the putative DPP III partially purified from cytosol or solubilized membrane sample. The effect of the divalent metal ions Zn 2+ and Co 2+ was checked on DPP III activity contained in whole cell homogenates. A curve-fitting computer program ( CRICKET GRAPH ) 3 was used to determine the K m value for proctolin and the IC 50 for tynorphin. Cell surface degradation and internalization of proctolin by stable transfected S 2 cells Transfected S 2 cells were previously induced for 48 h with CuSO 4 (final concentration in culture medium 0.5 m M ). Then, transfected cells were rinsed twice in NaCl/P i and incubated (50 000 cells per tube) in Hepes degradation buffer (100 lL) with proctolin (25 nmol) as substrate. At the end of incubation, the supernatant (incubation medium) was separated from the cells by centrifugation (1000 g for 8 min at room temperature) 4 andaddedwith 5 lLHCl(2 M ). Then, the cell pellet was rinsed twice before cell disruption and centrifugation (13 000 g for 15 min) to isolate the cytosolic sample also added with 5 lLHCl(2 M ) 5 . Proctolin degradation was analyzed by reversed-phase separation of the incubation medium in order to estimate the metabolism of proctolin at the cell surface. The reversed-phase separation of cytosolic samples indicated the rate of neuropeptide degradation after internalization. These results were compared to those obtained with transfected S 2 cells not previously induced. SDS/PAGE and Western blot analysis The samples consisting of transfected, ÔmockÕ transfected or untransfected cells, cytosol or membrane preparations (30 lg per lane) or partially purified expressed enzyme (1 or 2 lg) were prepared in sample buffer (62.5 m M Tris/HCl, pH 6.8, 0.025% bromophenol blue, 10% w/v glycerol and 1% w/v SDS) and electrophoresed [15] on a 7.5% acryl- amide 1 mm-thick mini gel at 20 mA for 150 min. Then the gel was stained with Coomassie Brillant Blue R-250, silver stained (Silver Stain Plus Kit, Bio-Rad), or transferred to a nitrocellulose membrane (Hybond C extra, pore size 0.45 lm, Amersham) for 1 h using a mini-trans-blot apparatus (Bio-Rad) for Western blot analysis. The mem- brane was soaked in Tris buffered saline with Tween-20 (TBS Tween, 20 m M Tris/HCl, 137 m M NaCl, 0.1% v/v Tween-20, pH 7.3) with 5% (w/v) dry low fat milk (used as blocking agent) for 30 min at 37 °C under constant stirring. Then, the membrane was incubated overnight at 4 °Cwith rabbit polyclonal anti-(rat liver DPP III) Ig (1/2000 in TBS Tween under constant slow agitation) and then with goat anti-rabbit Ig (Boehringer-Mannheim) conjugated with horseradish peroxidase (1/2000 in TBS Tween, 30 min at 37 °C). Staining was obtained by incubation with a fresh solution of 1.39 m M DAB (in 50 m M Tris/HCl, pH 7.3 for 5 min at room temperature) added with H 2 O 2 (0.1 lLÆmL )1 Tris buffer). Deglycosylation experiments were carried out on partially purified D. melanogaster DPP III using an enzymatic deglycosylation kit (Bio-Rad, including NANase II, O-glycosidase DS and PNGase F) according to the manufacturer’s instructions. Samples were checked before and after deglycosylation by SDS/PAGE, followed by staining with Coomassie Blue, using fetuin as a glycosylated protein standard. 3076 C. Mazzocco et al.(Eur. J. Biochem. 270) Ó FEBS 2003 Sequence comparisons The coding sequences of D. melanogaster and rat liver DPP III were aligned with homologous coding sequences from the worm Caenorhabditis elegans and the protozoan Leishmania major using CLUSTALW 1.8. In addition, the corresponding protein sequences were analyzed for the presence of a signal peptide and transmembrane regions using SIGNALP from the Expasy proteomic tools and TM at the EMBNET [16]. Measurement of protein content Membrane, cytosolic and purified proteins were measured using a commercial protein assay reagent kit (Bio-Rad) with BSA as a standard [17]. Proteins in solubilized samples were measured according to the instructions of a commercial detergent-compatible reagent kit (Bio-Rad), again with BSA as the standard. Results Stable transfection of S 2 cells expressing a putative D. melanogaster DPP III S 2 cells were double transfected with the pMTA/DPP III and pCoHygro vectors and then selected for their resistance to hygromycin. The transfection of hygromycin-resistant S 2 cells with the pMTA/DPP III vector was first verified by PCR, using a sense primer specific for pMTA and an antisense primer designed against the putative D. melano- gaster DPP III HpaI fragment (Fig. 1A) to amplify a 1017 bp fragment (Fig. 1B). The transcription of specific mRNAs was deduced from the amplification of the 1017 bp fragment after RT-PCR with cDNA templates prepared from total RNA of transfected S 2 cells previously induced for 48 h (Fig. 1C, lane 5). When RT-PCR was carried out with these cDNA templates and a second set of primers specific to the core D. melanogaster DPP III sequence (Fig. 1A), a 783 bp fragment was also amplified (Fig. 1C, lane 5). By contrast, the results of RT-PCR experiments with the second set of primers clearly indicated no significant transcription of endogenous DPP III in untrans- fected and ÔmockÕ transfected S 2 cells, and also in transfected S 2 cells that have not been previously induced (Fig. 1C). These results were corroborated by SDS/PAGE (Fig. 2A) and Western blot analysis (Fig. 2B) with anti- (rat liver DPP III) Ig, revealing two protein bands in the expected range of 82 and 86 kDa only in transfected S 2 cells after 48 h induction. By contrast, soluble DPP III related protein could not be detected by Western blot analysis (data not shown) in transfected S 2 cell culture medium after 48 h induction. When immunocytochemical studies were per- formed with anti-(rat liver DPP III) Ig on unfixed trans- fected and induced S 2 cells in the absence of detergent, a significant and presumably surface labeling was observed (Fig. 3A,B). When cells were treated with 0.1% Tween 20, approximately twice as many transfected and induced S 2 cells were stained than in the absence of detergent (data not shown). By comparison, no labeling could be observed after immunocytochemical treatment of untransfected S 2 cells or transfected S 2 cells not previously induced (data not shown). After a 5-min incubation of proctolin (25 nmol) with stably transfected and induced S 2 cells in NaCl/P i containing the aminopeptidase inhibitor bestatine (100 l M ), 6.25 nmol of proctolin and 1.75 nmol Arg-Tyr (and Leu-Pro-Thr, not shown) were recovered into the incubation medium as determined after reversed-phase separation (Fig. 3C). When incubation was carried out for 10 min, only 5 nmol proctolin and 2.75 nmol Arg-Tyr (and Leu-Pro-Thr, not shown) were measured. After 20 min incubation, 2.7 nmol proctolin and 4.55 nmol Arg-Tyr (and Leu-Pro-Thr, not shown) were quantified, indicating a correlated degradation of extracellular proctolin via DPP activity over the 20 min incubation. By comparison, cytosolic contents of induced Fig. 1. Schematic model of the pMTA/DPP III construction vector and verification of the transfection of S 2 cells with this vector. (A) Hatched area indicates the 2.484 kb HpaIfragmentofD. melanogaster cDNA clone GH 01916 coding for a putative DPP III. This fragment was inserted in the pMTA/V5-His B expression vector. Arrows represent the primer pairs (1 and 2; 3 and 4) used in the PCR and RT-PCR experiments, and the approximative sequence area expected to be amplified (1 kb and 0.8 kb, respectively). (B) Transfection of S 2 cells with pMTA/DPP III vector was verified by PCR with DNA from ÔmockÕ transfected and from stable transfected S 2 cells amplified using a set of primers specific for the pMTA/DPP III construction vector (primers 1 and 2). A 1.0 kb fragment (expected size 1.02 kb) was observed after separation on a 0.8% (w/v) agarose gel of PCR prod- ucts obtained with DNA of stable transfected and induced S 2 cells (lane 2). No amplification product could be detected from control PCR with DNA from ÔmockÕ transfected S 2 cells (lane 1). (C) Transcription of putative D. melanogaster DPP III mRNA in transfected S 2 cells was verified by RT-PCR experiments carried out on RNAs extracted from ÔmockÕ transfected and stable transfected S 2 cells using a second set of primers (3 and 4) specific for the HpaI fragment encoding a putative DPP III. Control amplification of a 0.8 kb product (expected size 0.78 kb) was obtained with DNA of the D. melanogaster clone GH 01916 as observed after separation on a 0.8% (w/v) agarose gel (lane 1). No amplification product was observed after RT-PCR using RNA from ÔmockÕ transfected S 2 cells (lane 2). A 0.8 kb fragment was visualized after separation of RT-PCR products performed with RNAs from stably transfected and induced S 2 cells (lane 3). When RT was omitted (control for contaminant DNA extracted from stable transfected S 2 cells) the 0.8 kb PCR product was not detected (lane 4). When using the first set of primers (1 and 2), the 1.0 kb fragment was amplified after RT-PCR using RNA from transfected S 2 cells after induction (lane 5). Ó FEBS 2003 Characterization of D. melanogaster DPP III (Eur. J. Biochem. 270) 3077 S 2 cells contained only traces of proctolin and Arg-Tyr after 5 min incubation. A significant peak of tyrosine (about 15 nmol) was detected and suggested a complete and rapid degradation of the 65–70% internalized neuropeptide despite the utilization of bestatine (data not shown). In addition, no DPP activity could be measured into concen- trated (ultrafiltration) culture medium (data not shown) using proctolin as a substrate. Although no trace of putative D. melanogaster DPP III could be noticed after Western blot analysis of transfected S 2 cells not previously induced, homogenates of uninduced S 2 cells contained proctolin-degrading activities including a weak dipeptidyl aminopeptidase activity of 0.07 ± 0.45 nmol proctolinÆmg protein )1 Æmin )1 . By contrast, a signifi- cant DPP activity was measured in homogenates of trans- fected and Cu 2+ -induced S 2 cells using met-enkephalin (data not shown) or the insect neuropeptide proctolin as substrate (1.66 ± 0.35 nmol proctolinÆmg protein )1 Æ min )1 ). When divalent metal ions (Zn 2+ ) were added to the incubation medium, the rate of met-enkephalin (data not shown) or proctolin degradation was similar (1.59 ± 0.45 nmol proctolinÆmg protein )1 Æmin )1 ). By contrast, the expressed dipeptidyl aminopeptidase activity contained in S 2 cell homogenates was strongly increased by the presence of Co 2+ (13.6 ± 1.4 nmol proctolinÆmg protein )1 Æmin )1 ). Given these results, cytosol and membranes were prepared separately from stably transfected and induced S 2 cells, electrophoresed and analyzed with anti-(rat liver DPP III) Ig after Western blotting. Two protein bands at 82 and 86 kDa, already visualized from whole transfected and induced S 2 cells, were clearly detected in cytosolic samples (Fig. 4A,B). Both protein bands were also observed in membrane samples (data not shown). Although an equivalent amount of cytosolic or membrane protein was initially loaded, the cytosolic 82 and 86 kDa protein bands were approximately twice as intense as those observed from membrane samples, indicating that a proportionally larger amount of over- expressed D. melanogaster DPP III is present into the cytosol than on the membranes of transfected cells. Partial purification and characterization of the putative D. melanogaster DPP III expressed in S 2 cells Cytosolic and membrane DPP activities expressed after 48 h induction in stable transfected S 2 cells were partially purified, first using a 5-mL Hitrap Q cartridge then two size exclusion columns (Table 1). The DPP fractions were identified by degradation studies using proctolin (40 l M ) as the substrate. SDS/PAGE analysis of fractions contain- ing partially purified DPP activitiy revealed two major bands in the range of 82 and 86 kDa in both cytosolic and membrane samples (Fig. 5). Western blot analysis of both partially purified samples confirmed the presence of two protein bands with similar molecular masses (82 and 86 kDa) that were immunologically related to rat liver DPP III (Fig. 5). Both partially purified D. melanogaster DPP III removed Arg-Tyr from the N terminus of proctolin (data not shown) with K m values of 4.1 ± 0.7 l M for the cytosolic enzyme and 3.0 ± 0.6 l M for the membrane enzyme. They both hydrolysed met-enkephalin at the Gly- Gly bond (data not shown). An IC 50 of 0.62 ± 0.15 l M was obtained with the DPP III inhibitor tynorphin (40 l M ) of proctolin degradation induced with D. melanogaster DPP III partially purified from the cytosol of transfected cells (data not shown). The contamination of transfected S 2 cell membranes was investigated by the addition of partially purified cytosolic DPP III to membranes prepared from trans- fected S 2 cells not previously induced. After a routine twice washing of the membrane pellet, Western blot analysis of this membrane sample revealed only traces of DPP III proteins that are not proportionally related to the amount of partially purified DPP III added to this sample. Furthermore, DPP III activity was measured on these ÔcontaminatedÕ membranes when no DPP III acti- vity could be detected on control membranes (transfected cells not previously induced), but this contamination accounted for only a third of the total DPP activity recovered on membranes prepared from transfected S 2 cells after induction. Sequence analysis The comparison of protein sequences identified in animals as putative or fully characterized DPP III showed variable regions at both extremities and consensus regions in the core sequence (Fig. 6). An ancestral molecule of 679 amino acids, found in the protozoan Leishmania major, includes a potential signal peptide with a possible cleavage site between residues 23 and 24 that is not recovered in the studied Fig. 2. Expression of a putative D. melanogaster DPP III protein in stably transfected S 2 cells. (A) Whole cell homogenates (30 lg per lane) prepared from stably transfected cells (cell lines 1a and 1b) after 48 h induction (lanes 1 and 3) or not previously induced (lanes 2 and 4) for the synthesis of D. melanogaster DPP III were separated by SDS/PAGE on a 7.5% acrylamide gel and stained with Coomassie Brillant Blue. The arrow indicates the region where two protein bands at 82–86 kDa are specifically detected in induced cells. (B) Western blot analysis of cell line 1a was performed with anti-(rat liver DPP III) Ig. Transfected S 2 cells were induced for 48 h (lane 2) or not previously induced (lane 1). 3078 C. Mazzocco et al.(Eur. J. Biochem. 270) Ó FEBS 2003 metazoan enzymes. In addition, a C-terminal extension of about 20 amino acid residues is found in metazoan DPP III but no specific function could be attributed to this region. Among metazoa, the analysis of D. melanogaster DPP III revealed the presence of six potential glycosylation sites compared with two sites on rat DPP III. However, no significant shift of the partially purified 86 kDa D. melano- gaster DPP III could be observed on SDS/PAGE after deglycosylation experiments (data not shown). In addition, the analysis of D. melanogaster DPP III indicated four strong putative transmembrane fragments, including two adjacent hydrophobic regions, located near the N terminus (Fig. 6), that could be involved in membrane anchorage of D. melanogaster DPP III. By comparison, the analysis of mammalian DPP III resulted in the identifica- tion of two distant transmembrane regions with lower significance (data not shown). Discussion A D. melanogaster cDNA clone coding for a putative DPP III of 82 kDa (CG 7415) was stably transfected in S 2 cells. After 48 h induction, transfected S 2 cells specifically synthesized two protein bands detected at 82 and 86 kDa after SDS/PAGE. Western blot analysis using anti-(rat liver DPP III) Ig confirmed that these protein bands were immunologically related to rat liver DPP III. This putative D. melanogaster DPP III expressed in transfected cells after induction was partially purified by chromatography accord- ing to the measurement of dipeptidyl aminopeptidase activity and the detection of immunologically related DPP III material in the enzyme fractions. The partially purified enzyme is similar to mammalian DPP III in that it hydrolyses small neuropeptides such as met-enkephalin and the insect myotropin proctolin, from which it removes the N-terminal dipeptide, and because it is inhibited by the specific DPP III inhibitor tynorphin [18]. The K m values of the recombinant D. melanogaster DPP III for proctolin (4 l M and 3 l M ) are very similar to the K m value (also 4 l M ) of the presumed DPP III purified in cockroach [12]. The potency of tynorphin against recombinant DPP III (IC 50 ¼ 0.62 l M ) and that purified in cockroach (0.68 l M ) are also very close. In addition, D. melanogaster DPP III is sensitive to divalent metal ions in a similar manner to vertebrate DPP III [19–21]. The deduced amino acid sequence of the putative D. melanogaster DPP III predicted a strong sequence identity ( 50%) with mammalian DPP III [12]. The functional expression of the D. melano- gaster DPP III cDNA in S 2 cells does, indeed, confirm that the translated protein product is a genuine DPP III. Fig. 3. Immunocytochemistry of transfected S 2 cells and proctolin- degrading activity of transfected S 2 cells. (A,B) Anti-(rat liver DPP III) Ig was used to probe transfected S 2 cells (see Experimental proce- dures). After 48 h induction, transfected S 2 cells were stained without a treatment with detergent. Arrows indicate the positively stained cells. (C) Proctolin (25 nmol) was incubated in NaCl/P i with stably trans- fected S 2 cells. After different time intervals, proctolin and its N-ter- minal dipeptide Arg-Tyr were measured in the incubation medium by reversed-phase chromatography. The results are the mean of three experiments ± SEM. Two thirds of this proctolin (approximately 16–17 nmol) were rapidly internalized into S 2 cellsandrecoveredas tyrosine (not shown). Fig. 4. SDS/PAGE and Western blot analysis of transfected S 2 cells. Cytosolic samples were prepared from transfected S 2 cells after induction and electrophoresed. (A) Cytosolic sample (30 lgprotein per lane) was separated on a 7.5% (w/v) acrylamide gel and stained with Coomassie Brillant Blue. Arrows indicate both bands corres- ponding to overexpressed D. melanogaster DPP III proteins visualized at 82 and 86 kDa. (B) Western blot analysis of cytosolic sample probed with anti-(rat liver DPP III) that allowed the detection of both bands at 82 and 86 kDa. Ó FEBS 2003 Characterization of D. melanogaster DPP III (Eur. J. Biochem. 270) 3079 However, a number of features of the D. melanogaster DPP III are different from the mammalian DPP III. For instance, cytosolic rat liver DPP III is expressed in E. coli as a single 82 kDa protein [2,22] that is in agreement with the expected size and coincides with the molecular mass of the DPP III protein deduced after purification from rat [3]. Although the transfection of S 2 cells was also carried out with a single D. melanogaster DPP III cDNA sequence, two major protein bands at 82 and 86 kDa are expressed and detected in transfected S 2 cells. Both bands are also identified after partial purification from transfected S 2 cells. The expected molecular mass of the D. melanogaster DPP III is 81 937, which coincides with the lower protein band revealed at about 82 kDa after Western blot analysis using anti-(rat liver DPP III). The presence of six predicted putative glycosylation sites on the insect DPP III compared with only two sites on rat liver DPP III suggested that the heavier D. melanogaster DPP III protein may result from post-transductional processing of the 82 kDa DPP III to raise the molecular mass of the D. melanogaster DPP III up to 86 kDa. Deglycosylation experiments performed on partially purified D. melanogaster DPP III and verified by SDS/PAGE were inconclusive in demonstrating conven- tional glycosylation of the 86 kDa expressed protein. The structure of the 86 kDa expressed DPP III is not yet elucidated but it cannot be inferred from the expression in S 2 cells because similar bands at 82 and 86 kDa have been already detected in soluble samples and membranes pre- pared from fruit flies [12]. In addition, control SDS/PAGE and Western blot experiments were performed with rat liver samples that resulted in the detection of a single protein band at 82 kDa as previously reported [2,3,23]. The expression of the insect DPP III as a membrane protein represents the second major difference with mam- malian DPP III, typically recovered as a cytosolic peptidase. Immunocytochemical studies of transfected S 2 cells with or without detergent indicated that D. melanogaster DPP III is possibly expressed at the cell surface. Even though the contamination of transfected S 2 cell membrane preparations from transfected S 2 cell cytosol containing overexpressed D. melanogaster DPP III was established, this accounted for only 30% of the total membrane DPP III activity recovered from transfected S 2 cell membranes. These results are in line with the detection of native DPP III after Western blot analysis of soluble and membrane samples prepared from D. melanogaster [12] and with the purifica- tion of a presumed DPP III from gut membranes in cockroach [12,14]. Our results also agree with the dual localization of DPP III previously reported in rat [6] and guinea-pig [5] brain cytosol and membranes. The partial purification of D. melanogaster DPP III expressed in transfected S 2 cells was achieved from both cytosol and solubilized membranes after thorough washing, and the characterization of both D. melanogaster DPP III partially purified from cytosol and membrane of transfected cells showed similar K m values to met-enkephalin and proctolin. Fig. 5. SDS/PAGE and Western blot analysis of overexpressed D. melanogaster DPP III. (A) D. melanogaster DPP III partially purified from cytosol of transfected S 2 cells (1 lg) was electrophoresed on a 7.5% (w/v) acrylamide gel and silver stained (lane 1) revealing the presence of two protein bands at 82 and 86 kDa. After Western blot analysis using anti-(rat liver DPP III) Ig, both bands were specifically detected (lane 2). (B) An aliquot from the different steps of separation of solubilized membranes from transfected S 2 cells was electrophored on a 7.5% (w/v) acrylamide gel and silver stained. Solubilized mem- branes of transfected S 2 cells (18 lg, lane 1), positive fractions resulting from anion exchange separation (5 lg, lane 2) and from size exclusion separation (2 lg, lane 3) were electrophoresed to confirm the isolation of two protein bands at 82 and 86 kDa during the process of separ- ation. (C) Western blot analysis of overexpressed D. melanogaster DPP III partially purified from membranes (2 lg, lane 1) probed with anti-(rat liver DPP III) Ig. Both bands at 82 and 86 kDa were thus detected. Table 1. 6 Purification of D. melanogaster cytosolic DPP III overexpressed in S 2 cells. DPP III loaction Purification steps Proteins (mg) Total activity (nmolÆmin )1 ) Specific activity (nmolÆmg )1 min )1 ) Purification factor Cytosol Filtrate 49.32 39.45 0.799 1 Hi-Trap Q 17 207 12.2 15 Superdex 200 2 122.98 61.49 75 Membrane Filtrate 76 380 4.98 1 Hi-Trap Q 7 150 21.44 4.3 Superdex 200 0.25 81.9 312.7 62.79 3080 C. Mazzocco et al.(Eur. J. Biochem. 270) Ó FEBS 2003 The examination of several DPP III sequences indicated that this enzyme was highly conserved during the evolution process from protozoa to mammals and that the extremities of the sequences are mostly variable. No signal peptide could be identified in the D. melanogaster DPP III and this was supported by the absence of detection of recombinant DPP III in transfected S 2 cell culture media as verified by Western blot analysis. Furthermore, no trace of DPP activity could be measured in ultrafiltered culture media of transfected S 2 cells. The D. melanogaster DPP III is not secreted and is not a circulating DPP. The fact that D. melanogaster DPP III contains a putative N-terminal membrane anchor seq- uence with a significantly higher probability than that calculated for mammalian DPP III corroborates our results that argue for a membrane DPP III in insects when it is mainly identified as a cytosolic peptidase in mammals. Furthermore, the analysis of the orientation of the two putative transmembrane helices identified in D. melanogaster DPP III (amino acid residues 44–62 and 81–106) indicates that the resulting loop between 63 and 80 is presumably on the inner side of the membrane, whereas the other protein regions of D. melanogaster DPP III are on the outer side. In these conditions and considering the anchorage of the D. melanogaster DPP III on the plasma membrane, the six potential glycosylation sites and the HELLGH active site of the protein would be exposed on the outer surface of the cell. Thus, one can expect a significant cell surface-located DPP III activity in transfected S 2 cells. This was confirmed by the detection of a substantial amount of the proctolin N-terminal dipeptide Arg-Tyr in the incubation medium (degradation buffer) following incubation of the neuropeptide proctolin with whole transfected and induced S 2 cells (Fig. 3C). These results provide evidence that D. melanogaster DPP III is expressed in S 2 cells as a cytosolic enzyme and also as a membrane peptidase. In terms of proctolin degrading activity, a dipeptidyl aminopeptidase activity was shown to hydrolyse proctolin in insects and was referred as to mammalian DPP III [11] but this was not further investigated. In another study, a dipeptidyl aminopeptidase activity was identified in locust synaptosomes as a major proctolin-degrading enzyme [8]. In this example, two thirds of DPP activity was measured in the mitochondrial (cytosolic) fraction and one third was recovered in the synaptosome (membrane) fraction. In transfected S 2 cells, D. melano- gaster DPP III is roughly expressed in these proportions in the cytosol and on membranes as an efficient proctolin-degrading enzyme. In conclusion, our results confirm that the putative D. melanogaster DPP III is a genuine DPP III, expressed in the cytosol and on the membranes of transfected S2 cells. The D. melanogaster DPP III herein identified represents the first fully characterized peptidase involved in proctolin degradation. Fig. 6. Alignment of D. melanogaster and rat DPP III with putative homologous proteins identified in the worm C. elegans and the protozoan L. major. Black frames correspond to the identity of residues in at least three sequences, grey frames indicate similarity between residues and dots represent gaps inserted for optimal alignment with CLUSTAL W . Both squares correspond to potential transmembrane segments in D. melanogaster DPP III as determined with TM software at the EMBNET. Ó FEBS 2003 Characterization of D. melanogaster DPP III (Eur. J. Biochem. 270) 3081 References 1. Ellis, S. & Nuenke, J.M. (1967) Dipeptidyl arylamidase III of the pituitary. J. Biol. Chem. 243, 4623–4629. 2. Fukasawa, K., Fukasawa, K.M., Kanai, M., Fujii, S., Hirose, J. & Harada, M. (1998) Dipeptidyl peptidase III is a zinc metallo- exopeptidase. Molecular cloning and expression. Biochem. J. 329, 275–282. 3. Ohkubo,I.,Li,Y.H.,Maeda,T.,Yamamoto,Y.,Yamane,T., Du, P.G. & Nishi, K. (1999) Dipeptidyl peptidase III from rat liver cytosol: purification, molecular cloning and immunohistochemical localization. Biol. Chem. 380, 1421–1430. 4. Fukasawa, K., Fukasawa, K.M., Iwamoto, H., Hirose, J. & Harada, M. (1999) The HELLGH motif of rat liver dipeptidyl peptidase III is involved in zinc coordination and the catalytic activity of the enzyme. Biochemistry 38, 8299–8303. 5. Smyth, M. & O’Cuinn, G. (1994) Dipeptidyl aminopeptidase activities of guinea-pig brain. Int. J. Biochem. 26, 913–921. 6. Alba, F., Arenas, J.C. & Lopez, M.A. (1995) Properties of rat brain dipeptidyl aminopeptidases in the presence of detergents. Peptides 16, 325–329. 7. Lee, C M. & Snyder, S.H. (1982) Dipeptidyl-aminopeptidase III of rat brain selective affinity for enkephalin and angiotensin. J. Biol. Chem. 257, 12043–12050. 8. Isaac, R.E. (1987) Proctolin degradation by membrane peptidases from nervous tissues of the desert locust Schistocerca gregaria. Biochem. J. 245, 365–370. 9. Puiroux, J. & Loughton, B.G. (1992a) Degradation of the neuro- peptide proctolin by membrane bound proteases of the hindgut and ovary of Locusta migratoria and the effects of different inhibitors. Arch. Insect Biochem. Physiol. 19, 193–202. 10. Quistad, G.B., Adams, M.E., Scarborough, R.M., Carney, R.L. & Schooley, D.A. (1984) Metabolism of proctolin, a pentapeptide neurotransmitter in insects. Life Sci. 34, 569–576. 11. Steele, R.W. & Starratt, A.N. (1985) In vitro inactivation of the insect neuropeptide proctolin in haemolymph from Periplaneta americana. Insect Biochem. 15, 511–519. 12. Mazzocco, C., Fukasawa, K.M., Raymond, A.A. & Puiroux, J. (2001) Purification, partial sequencing and characterization of an insect membrane dipeptidyl aminopeptidase that degrades the insect neuropeptide proctolin. Eur J. Biochem. 268, 4940–4949. 13. Adams, M.D., Celniker, S.E., Holt, R.A., Evans, C.A., Gocayne, J.D., Amanatides, P.G., Scherer, S.E., Li, P.W., Hoskins, R.A., Galle, R.F. et al. (2000) The genome sequence of Drosophila melanogaster. Science 287, 2185–2195. 14. Mazzocco, C. & Puiroux, J. (2000) Purification of proctolin- binding protein from the foregut of the insect Blaberus craniifer. Eur. J. Biochem. 267, 2252–2259. 15. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685. 16. Kyte, J. & Doolittle, R.F. (1982) A simple method for displaying the hydropathic character of a protein. J. Mol. Biol. 157, 105–132. 17. Bradford, M.M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of proteins utilizing the principle of dye binding. Anal. Biochem. 72, 248–254. 18. Yamamoto, Y., Hashimoto, J., Shimamura, M., Yamaguchi, T. & Hazato, T. (2000) Characterization of tynorphin, a potent endogenous inhibitor of dipeptidyl peptidase III. Peptides 21, 503–508. 19. Abramic, M., Zubanovic, M. & Vitale, L. (1988) Dipeptidyl peptidase III from human erythrocytes. Biol. Chem. Hoppe Seyler. 369, 29–38. 20. Shimamori, Y., Watanabe, Y. & Fujimoto, Y. (1988) Human placental dipeptidyl aminopeptidase III: hydrolysis of enkephalins and its stimulation by cobaltous ion. Biochem. Med. Metab Biol. 40, 305–310. 21. Vanha-Perttula, T. (1988) Dipeptidyl peptidase III and alanyl aminopeptidase in the human seminal plasma: origin and bio- chemical properties. Clin. Chim Acta. 177, 179–195. 22. Ohkubo,I.,Li,Y.,Maeda,T.,Yamamoto,Y.,Yamane,T.,Du, P.G. & Nishi, K. (2000) Molecular cloning and immuno- histochemical localization of rat dipeptidyl peptidase III. Forensic Sci. Int. 113, 147–151. 23. Abramic, M., Schleuder, D., Dolovcak, L., Schroder, W., Strupat, K., Sagi, D., Peter-Katalini, J. & Vitale, L. (2000) Human and rat dipeptidyl peptidase III: biochemical and mass spectrometric arguments for similarities and differences. Biol. Chem. 381, 1233–1243. 3082 C. Mazzocco et al.(Eur. J. Biochem. 270) Ó FEBS 2003 . Characterization of a functionally expressed dipeptidyl aminopeptidase III from Drosophila melanogaster Claire Mazzocco 1, *, Kayoko M. Fukasawa 2 , Patrick. model Polyvar) and image analysis software (Spot RT, Diagnostic Instru- ments, USA). Partial purification of a putative D. melanogaster DPP III expressed