Báo cáo Y học: Characterization of four substrates emphasizes kinetic similarity between insect and human C-domain angiotensin-converting enzyme

doi:10.1046/j.1432-1033.2002.03043.x

Eur. J. Biochem. 269, 3522–3530 (2002) (cid:1) FEBS 2002

Characterization of four substrates emphasizes kinetic similarity between insect and human C-domain angiotensin-converting enzyme

Korneel Hens1, Anick Vandingenen1, Nathalie Macours1, Geert Baggerman1, Adriana Carmona Karaoglanovic2, Liliane Schoofs1, Arnold De Loof1 and Roger Huybrechts1 1Zoological Institute of the Catholic University of Leuven, Laboratory of Developmental Physiology and Molecular Biology, Leuven, Belgium; 2Universidade Federal de Sao Paulo, Escola Paulista de Medicina, Department of Biophysics, Sao Paulo, Brazil

one out of three yolk protein bands of SDS/PAGE-separ- ated fly haemolymph and egg homogenate, thus confirming that these peptides originate from a yolk protein gene product. Kinetic analysis of these peptides and of the peptides Neb-ODAIF and Neb-ODAIF-11)7 with insect ACE and human ACE show both similar and unique properties for insect ACE as compared with human C-domain ACE.

Keywords: ACE kinetics; domain specific substrates; insect physiology; reproduction.

Angiotensin converting enzyme (ACE) was already discov- ered in insects in 1994, but its physiological role is still enigmatic. We have addressed this problem by purifying four new ACE substrates from the ovaries of the grey fleshfly, Neobellieria bullata. Their primary structures were identified as NKLKPSQWISLSD (Neb-ODAIF-11)13), (Neb-ODAIF-11)9), SLKPSNWLTPSE NKLKPSQWI (Neb-ODAIF-2) and LEQIYHL. Database analysis showed significant homology with amino acid sequence stretches as present in the N-terminal part of several fly yolk proteins. An antiserum raised against Neb-ODAIF-11)9 immunostained

ancestral gene. Testicular ACE is transcribed from the same gene as sACE but from another, intragenic promotor [4]. It has a single active domain. In Drosophila melanogaster two isoforms of the enzyme have been found as well, namely AnCE and ACER [5]. This suggests that gene duplication has occurred in both Deuterostomia and Protostomia.

Another difference between the mammal and insect ACE is the presence and absence, respectively, of a membrane anchor at the C-terminal part of the enzyme. As a consequence, mammalian ACE is mainly membrane bound while insect ACE is soluble.

same

Insect ACE was first isolated from head membranes of the housefly Musca domestica in 1994 [1], a long time after the discovery of its mammalian counterpart in horse plasma in 1956. Since this discovery, and after cloning and purification of several insect ACEs it has become clear that insect and mammalian ACE, despite of their evolutionary distance, are structurally remarkably similar. The molecular biological analyses of insect ACEs revealed a high cDNA and amino acid sequence conservation with mammalian ACE, especi- ally around the active site [2]. The enzymatic activity is also well conserved as insect ACE can hydrolyse mammalian ACE substrates such as angiotensin I, substance P, lutein- izing hormone releasing hormone, enkephalins and enkeph- alinamides, hereby displaying the exo- and endopeptidase activities as mammalian ACE [3].

Mammalian ACE occurs in two isoforms. Somatic ACE (sACE) has a wide tissue distribution and has two active domains, probably generated by gene duplication of a smaller

Mammalian sACE is involved in regulating blood pressure and water and electrolyte homeostasis. Indications about the role of insect ACE range from prohormone processing [6] over immunity [7,8], to neurotransmitter inactivation [6]. Several reports indicate a role of ACE in insect reproduction as well. In addition to impaired male fertility following ACE gene knock-out in Drosophila [9], Schoofs et al. found ACE immunoreactivity in the testis of Locusta migratoria, Neobel- lieria bullata and Leptinotarsa decemlineata [6]. These find- ings were complemented by measurements of enzyme activity in Locusta migratoria [10] and Neobellieria bullata [11].

A major problem in resolving a physiological role for insect ACE is the lack of known in vivo substrates. In this respect we observed that fly ovaries are a rich source of ACE-competitive substances [11]. Hence, we describe the purification and characterization of four novel ACE substrates from fleshfly ovaries and discuss their possible physiological functions.

M A T E R I A L S A N D M E T H O D S

Animals, haemolymph and tissue collection

The grey fleshfly Neobellieria bullata was reared as described [12]. Staging of ovarian development was done according to

Correspondence to K. Hens, K. U. Leuven, Laboratory of Develop- mental Physiology and Molecular Biology, Naamsestraat 59, 3000 Leuven, Belgium. Fax: +32 16 32 39 02, Tel.: +32 16 32 42 60, E-mail: korneel.hens@bio.kuleuven.ac.be Abbreviations: Abz, aminobenzoic acid; ACE, angiotensin converting enzyme; ACN, acetonitrile; cACE, C-domain of human angiotensin converting enzyme; Dnp, dinitrophenyl; ESI Q TOF MS, electrospray ionization quadrupole time of flight mass spectrometry; nACE, N-domain of human angiotensin converting enzyme; Neb, Neobellieria bullata; Lom, Locusta migratoria; ODAIF, ovary derived angiotensin converting enzyme interactive factor; PAP, peroxidase antiperoxidase; PVDF, polyvinylidene difluoride; sACE, somatic angiotensin converting enzyme; tACE, testicular angiotensin converting enzyme; TMOF, trypsin modulating oostatic factor. (Received 21 February 2002, revised 5 June 2002, accepted 8 June 2002)

Ovary-derived ACE substrates and insect physiology (Eur. J. Biochem. 269) 3523

(cid:1) FEBS 2002

Pappas and Fraenkel [13]. For collection of haemolymph, a leg was amputated of anaesthetized flies, the haemolymph was drawn from the resulting wound with a capillary and diluted immediately in ice-cold borate buffer [50 mM Borax, 0.3 M NaCl, 0.2 M (NH4)2SO4, pH 7.5].

inhibited by captopril was regarded as ACE activity. To find out if the HPLC-fractions contain an inhibitor for ACE, appropriate amounts of lyophilized fraction material were added to the standard condition setup. Addition of an ACE inhibitor or an ACE substrate results in competition with the tritium-labelled substrate for ACE and appears as a reduction in ACE activity.

The desert locust Locusta migratoria was raised as described [14]. Only yellow coloured, sexually mature males were used for collection of testes.

Identification of the isolated peptides

Synthetic peptides

NKLKPSQWISLSD (Neb-ODAIF-11)13), NKLKPSQ WISL (Neb-ODAIF-11)11), NKLKPSQWI (Neb-ODAIF- 11)9), NKLKPSQ (Neb-ODAIF-11)7), SLKPSNWLTPSE (Neb-ODAIF-2) and LEQIYHL were from Research Gen- etics Inc. (Huntsville, AL, USA). Synthesis and character- istics of the internally quenched fluorescent ACE substrate O-aminobenzoic acid-Phe-Arg-Lys-2,4-dinitrophenyl- Pro[Abz-FRK-(Dnp)P] was as described previously [15].

Purification of mammalian and insect ACE enzymes

Chinese hamster ovary cells expressing recombinant human sACE, cACE and nACE were a kind gift of P. Corvol and A. Michaud (Institut National de la Sante´ at de la Recherche Medicale, Paris). Cell culture and ACE purifi- cation were as described [16].

Purification of Locusta migratoria testicular ACE was as

described [10].

Nanoflow electrospray ionization (ESI) quadrupole (Q) orthogonal acceleration TOF-MS was performed on a Q-TOF system (Micromass, UK). An appropriate volume of the pure active fraction was dried and redissolved in 10 lL of ACN/water/acetic acid (30 : 69.9 : 0.1, v/v/v). One microliter of this sample was loaded in a gold-coated capillary (Protana L/Q nanoflow needle). This sample was sprayed at a typical flow rate of 30 nLÆmin)1 giving extended analysis time in which an MS spectrum as well as several tandem MS (MS/MS) spectra was acquired. During MS/MS fragment ions are generated from a selected precursor ion by collision induced dissociation [18]. Because not all peptide ions fragment with the same efficiency, the collision energy is typically varied between 20 and 35 eV so that the parent ion is fragmented in a satisfying number of different daughter ions. Needle voltage was set at 850 V, cone voltage was 35 V. The fragmentation spectra obtained were combined and transformed into their single charged state by treatment with the MAX-ENT3 software (Masslynx 3.5 software; Micromass, UK).

Tissue extraction and HPLC purification

N-Terminal amino acid sequencing was performed on an Applied Biosystems Procise protein sequencing system (Edman degradation) running in the pulsed liquid mode.

Kinetic studies of the purified peptides

column

Eight thousand mid-vitellogenic (stage 4B) or late-vitello- genic (stage 4C) fly ovaries were dissected, extracted and prepurified as described [11]. Columns and operating conditions of subsequent HPLC (Gilson) fractionations were: (a) Xterra C18 (Waters Xterra RP18, 7.5 · 300 mm, 7 lm), elution with a linear gradient of 0% to 50% ACN in 0.1% trifluoroacetic acid in 120 min and a flow rate of 2 mLÆmin)1. Two-ml fractions were automatically collected; (b) C8 column (Supelco, LC-8DB, 4.6 · 250 mm, 5 lm), elution with a linear gradient of 0% to 50% ACN in 0.1% trifluoroacetic acid in 120 min and a flow rate of 1 mLÆmin)1. Fractions according to absorbance peaks were manually collected; (c) PKB 100 column (Supelco, PKB 100, 4.6 · 250 mm, 5 lm), elution and collection as (b); (d) Hypercarb (Thermoquest, Hypercarb, 3 · 250 mm, 5 lm), elution with a linear gradient of 0% to 81% ACN in 0.01% trifluoroacetic acid in 90 min and a flow rate of 0.3 mLÆmin)1. Fractions were manually collec- ted. Absorbance was measured at 214 nm with a Waters 486 tunable absorbance detector.

ACE inhibition screening

The ACE inihibition assay is based on the ACE activity assay by Ryan et al. [17], modified by Vandingenen et al. [11]. Briefly, ACE-activity in diluted fly haemolymph is measured with a synthetic, tritiated ACE substrate p-[3H]benzoylglycylglycylglycine (Sigma) (¼ standard con- dition). Adding 10 lM final concentration of captopril (Sigma) served as a negative control. Captopril is a strong and specific ACE inhibitor. Only the activity that could be

Michaelis–Menten (Km) constants of ACE for the purified peptides were determined using a competition-based assay. ACE activity was measured using the internally quenched fluorogenic ACE-substrate Abz-FRK-(Dnp)P as described [15]. Briefly, ACE activity was monitored in Tris/HCl buffer (0.1 M Tris/HCl, 0.05 M NaCl, 10 lM ZnCl2, pH 7.0) 1 mL final volume with 1 lM final concentration of Abz-FRK- (Dnp)P for sACE, nACE and LomACE or 2.5 lM final concentration for cACE. Cleavage of the fluorogenic substrate at the R–K bond removes the quencher Dnp from the fluorogenic group Abz resulting in the appearance of fluorescence at 420 nm after excitation at 320 nm which was followed continuously for 10 min in a PerkinElmer LS-50B fluorimeter. The initial cleavage rate was deter- mined by nonlinear regression of the data to the function f(x) ¼ y0 + a[1 ) exp(–bx)] describing an exponential increase to a maximum using the software package SPSS SIGMAPLOT. The parameters a and b were used to calculate the slope of the tangent of this function through the origin, corresponding to the initial cleavage rate v0. The same experiment was repeated in the presence of a nonfluorogenic (dark) purified substrate (5 lM to 50 lM final concentra- tion) in a competition experiment. As we use only the initial cleavage rate, the dark substrate acts as a competitive inhibitor as described by Xie et al. [19]. Hence, the initial cleavage rate in presence of a dark substrate is called vi. The initial rate of cleavage of the fluorescent substrate is

3524 K. Hens et al. (Eur. J. Biochem. 269)

(cid:1) FEBS 2002

thyroglobulin as a carrier using the carbodiimide method as described [20]. The resulting conjugate was dissolved in physiological saline and supplemented with an equal volume of complete Freund’s adjuvant (first immunization) or incomplete Freund’s adjuvant (subsequent immunizations). Rabbits were injected subcutaneously with this conjugate at 2-weekly intervals for 20 weeks. Prior to the first immuniza- tion, the rabbits were bled to obtain preimmune serum. The antiserum was characterized in a dot-spot assay. One lL Neb- ODAIF-11)9 in different concentrations (1 lg to 10 pg) was spotted on to a nitrocellulose membrane and immobilized by baking. The spots were incubated with different dilutions of antiserum (1 : 100, 1 : 500 and 1 : 1000). The spots are visualized with the peroxidase antiperoxidase (PAP)-immu- nohistological technique as described [21].

described by the Michaelis–Menten equation: v0 ¼ kcat · E0/(1 + Km/S) with kcat the turnover number of fluorescent substrate cleavage, E0 the concentration of ACE enzyme, Km the Michaelis–Menten constant for cleavage of the fluorescent substrate and S the concentration of fluorescent substrate. As the nonfluorescent substrate is a competitive inhibitor in these experimental conditions, we can describe substrate cleavage as: vi ¼ kcat · E0/ the fluorescent [1 + Km/S · (1 + S¢/K¢m)] with S¢ the concentration of nonfluorescent substrate and K¢m the Michaelis–Menten constant for cleavage of the nonfluorescent substrate. The ratio of vi and v0 can thus be described by: vi/v0 ¼ (1 + Km/ S)/[1 + Km/S · (1 + S¢/K¢m)]. Since the parameters vi and v0 are measured and Km, S and S¢ are known, we can calculate K¢m from this equation.

SDS/PAGE

Peptide interaction with ACE

To determine whether the purified peptide is a competitive inhibitor or a substrate, 10 lM final concentration of peptide Neb-ODAIF-11)13 or LEQIYHL in Tris/HCl buffer (0.1 M Tris/HCl, 0.05 M NaCl, 10 lM ZnCl2, pH 7.0), 100 lL final volume, were incubated with sACE. After 5 h, the reaction was stopped by adding 10 lL captopril 10 lM.

SDS/PAGE using a polyacrylamide gradient (9–12%) in vertical slab gels in combination with a discontinuous buffering system was performed according to Laemmli [22]. The resulting separated proteins were transferred to a polyvinylidene difluoride (PVDF) membrane by electro- blotting. Protein bands were visualized with Coomassie blue staining or with the PAP-immunohistological technique as described [21]. Bands corresponding to the Neobellieria yolk polypeptides were identified as described [23].

R E S U L T S

Purification of ACE-interactive peptides

As a negative control, this experiment was repeated in the presence of 1 lM captopril. Ten microliters of the hydro- lysation products were separated and analysed using capillary liquid chromatography/tandem MS. These experi- ments were conducted using an Ultimate HPLC pump, a column switching device (Switchos) and a Famos autosam- pler (all LC Packings, the Netherlands) coupled to a Q-TOF hybrid quadrupole/TOF mass spectrometer (Micromass, UK). Chromatography was performed using a guard column (l-guard column MGU-30 C18, LC-Packings, the Netherlands) acting as a reverse phase support to trap the peptides. Ten microliters of the sample was loaded on the precolumn with an isocratic flow of MilliQ water with 0.1% formic acid at a flow rate of 10 lLÆmin)1 After 2 min the column switching valve was switched, placing the precol- umn online with the analytical capillary column, a Pepmap C18, 3 lm 75 lm · 150 mm nano column (LC Packings). Separation was conducted using a linear gradient from 95% solvent A, 5% solvent B to 5% A, 95% B in 55 min (solvent A: water/acetonitrile/formic acid, 94.9 : 5 : 0.1, v/v/v; solvent B: water/acetonitrile/formic acid, 19.9/80/0.1, v/v/v). The flow rate was set at 150 nLÆmin)1.

The Ultimate capillary liquid chromatography was connected in series to the electrospray interface of the Q-TOF mass spectrometer. The column eluent was directed through a metal coated fused silica tip (Picotip type FS360- 75-10 D; New Objective, USA). Needle voltage was set at 1400 V, cone voltage at 30 V. Nitrogen was used as nebulizing gas. Tandem MS was carried out in an automa- ted fashion. Peptide masses of interest were automatically selected for fragmentation during the nano-LC tandem MS separation. Argon was used as a collision gas, collision energy was set at 15–35 eV depending on the selected mass.

After prepurification of the crude ovary homogenate on a Megabond Elute C18 cartridge, the activity was restricted to the 60% ACN fraction. This fraction was further separated on a Deltapak C18 while absorbance was followed at 214 nm. Thirty ovary equivalents of each fraction were tested for inhibition activity, revealing inhibition activity in almost every fraction tested. One equivalent is the amount of a sample that would contain the material present in one ovary. The active fraction that eluted after 82 min, corres- ponding to elution at 33% ACN, was selected for further purification because of its high ACE inhibiting capacity (30% inhibition) by applying it on an Xterra C18 semi- preparative column. This time 60 ovary equivalents were tested in the inhibition assay. The activity divided into several different fractions. The further purification scheme is described in Table 1. Absorbance was always monitored at loss during purification and 214 nm. Due to material screening, an increasing number of equivalents of the fractions had to be screened after each purification step. After the first HPLC purification 30 equivalents were tested, after the second 60 equivalents, then 120 equivalents, then 240 and finally 480 equivalents resulting in a pure active fraction after four (Neb-ODAIF-11)13 and LEQIYHL) or five (Neb-ODAIF-11)9 and Neb-ODAIF-2) successive HPLC columns. The final chromatograms are shown in Fig. 1 with the final active fractions indicated.

Preparation of polyclonal antibodies

Identification of the purified peptides

ESI-TOF MS confirmed the purity of the fractions and yielded the mass of the purified peptides summarized in

An antiserum against Neb-ODAIF-11)9 was raised in New Zealand white rabbits. To improve the immunogenicity, 2 mg synthetic Neb-ODAIF1–9 was coupled to 25 mg bovine

Ovary-derived ACE substrates and insect physiology (Eur. J. Biochem. 269) 3525

(cid:1) FEBS 2002

Table 1. Purification procedure of several ACE-competitive peptides. HPLC purification of several ACE-competitive peptides from an extract of 8000 ovary equivalents from the grey flesh fly, Neobellieria bullata. Elution conditions and active fractions are indicated and the sequence of the purified peptide is given.

Step Active fraction Next purification step Resulting active fractions Peptide identified

Megabond Elute Waters Deltapak C18 1 33% ACN 60% ACN fraction 25 · 100 mm, 15 lm Waters Xterra RP18 2 7.8 · 300 mm, 7 lm Supelco Supelcosil LC-8DB 3 26% ACN 27.5% ACN 20% ACN Deltapak C18 33% ACN Waters Xterra 26% ACN Supelcosil LC-8DB 4

5 20% ACN 21% ACN 51.5% ACN NKLKPSQWI

5 31.5% ACN SLKPSNWLTPSE 4.6 · 250 mm, 5 lm Supelco Suplex PKB 100 4.6 · 250 mm, 5 lm Termoquest Hypercarb 3 · 250 mm, 5 lm Termoquest Hypercarb 3 · 250 mm, 5 lm Supelco Supelcosil LC-8DB 3

sequence of the third purified peptide. This peptide resembles Neb-ODAIF but is not completely identical, so it was called Neb-ODAIF-2. The last sequence obtained was LEQIYHL, which shares no sequence similarity with Neb-ODAIF and hence it will not be given an abbreviated name. Neb-ODAIF will be called Neb-ODAIF-11)11 to avoid confusion.

completely

comprises

the dipeptide SD at

For sequence comparison, the sequences were submitted to protein databases using BLAST at NCBI. All entries yielded, among hits with other proteins, stretches of amino acids as present in yolk proteins (yps) of different fly species as the most abundant hits. Neb-ODAIF-11)13 and Neb- ODAIF-11)9 displayed the highest sequence similarity with a yolk protein (yp3) of the bluebottle fly Calliphora vicina, yp3 and yp2 of the housefly Musca domestica and yp1 of

Table 2. Fragmentation of the ion in a subsequent collision induced dissociation experiment resulted in a partial amino acid sequence by a clear series of b and y¢¢ type ions (data not shown). In addition, the amino acid sequence was determined by automated N-terminal sequencing, resolving leucine/isoleucine and lysine/gluta- mine ambiguities with MS/MS sequencing. The first sequence obtained was NKLKPSQWISLSD, a peptide that the previously purified Neb-ODAIF (A. Vandingenen, personal communication) but with the extension of the C-terminus. Hence it was called Neb-ODAIF-11)13. The second peptide, NKLKPSQWI is completely comprised in Neb-ODAIF, but lacks the C-terminal dipeptide SL, so it was called Neb-ODAIF-11)9. SLKPSNWLTPSE is the

20% ACN Suplex PKB 100 20% ACN Suplex PKB 100 21% ACN Waters Xterra 27.5% ACN Supelcosil LC-8DB 4 19% ACN 21% ACN 22% ACN LEQIYHL 19% ACN Supelcosil LC-8DB 4 19% ACN NKLKPSQWISLSD 21% ACN 0–60% ACN in 150 min 0–50% ACN in 120 min 0–50% ACN in 120 min 0–50% ACN in 120 min 0–80% ACN in 90 min 0–80% ACN in 90 min 0–50% ACN in 120 min 0–50% ACN in 120 min 0–50% ACN in 120 min 4.6 · 250 mm, 5 lm Supelco Suplex PKB 100 4.6 · 250 mm, 5 lm Supelco Suplex PKB 100 4.6 · 250 mm, 5 lm

Fig. 1. Final HPLC chromatograms of the purified peptides. Chromatograms of the HPLC runs that resulted in the final purifi- cation of (A) Neb-ODAIF1-13 (B) Neb- ODAIF1–9 (C) Neb-ODAIF-2 and (D) LEQIYHL with absorbance at 214 nm and elution gradient (% ACN) indicated. Active fractions are indicated with an arrow.

3526 K. Hens et al. (Eur. J. Biochem. 269)

(cid:1) FEBS 2002

Table 2. Interaction of the purified peptides with different kinds of ACE. Protonated mass as determined by MS and Km values (lM) of the cleavage of the purified peptides with sACE, nACE, cACE and locust testis ACE.

Km (lM)

sACE nACE cACE Lom testes ACE Peptide Protonated mass

459

Peptide interaction with ACE

several Drosophila species. Entering Neb-ODAIF-2 in the search yielded yp3 of Musca domestica and yp1 and yp2 of several Drosophila species. Finally, LEQIYHL was most similar to Drosophila yp1.

The peptides Neb-ODAIF-11)11, Neb-ODAIF-11)9 and Neb-ODAIF-11)7 were already shown to be true substrates by Vandingenen (A. Vandingenen, Zoological Institute of the Catholic University of Leuven, Laboratory of Develop- mental Physiology and Molecular Biology, Belgium, personal communication).

A multiple alignment of the purified peptides with yp1, yp2 and yp3 of Musca and Drosophila and with yp3 of Calliphora is given in Fig. 2. The peptides align N-terminally with the yps. Interestingly, Neb-ODAIF1–13 and Neb-ODAIF-2 align at the same position within the yps suggesting that they are peptides derived from two different yps in Neobellieria. The peptide LEQIYHL aligns a bit further in the yps but still quite close to the N-terminus of the yps.

To determine whether the peptides Neb-ODAIF-11)13 and LEQIYHL are inhibitors or true substrates, these peptides were incubated for 5 h with sACE. The reaction was stopped by addition of the specific ACE-inhibitor

ODAIF1–13 ODAIF1–11 ODAIF1–9 ODAIF1–7 ODAIF*2–13 LEQIYHL 1515.73 1313.74 1114.71 813.9 1359.12 915.64 136.7 17.0 81.5 90.5 311.7 34.7 673.8 150.5 523.7 No competition No competition 159.2 231.4 6.9 62.2 75.3 108.0 15.7 2.7 147.2 112.6 233.8 412.5

Fig. 2. Multiple alignment of the purified pep- tides with several fly yps. Multiple alignment of the purified peptides with several fly yps, namely Drosophila melanogaster (Drome), Musca domestica (Musdo) and Calliphora vicina (Calvi).

Ovary-derived ACE substrates and insect physiology (Eur. J. Biochem. 269) 3527

(cid:1) FEBS 2002

captopril. For both Neb-ODAIF-11)13 and LEQIYHL, no hydrolysation products could be detected using Q-TOF MS after capillary liquid chromatography.

Furthermore, comparison of the absorbance peak cor- responding to the intact peptide showed no significant difference between the control condition and the experi- mental condition (data not shown). These results show that Neb-ODAIF-11)13 and LEQIYHL are either inhibitors or substrates with a very low turnover number.

Km determination of the purified peptides

initial cleavage rates were used to calculate the Michaelis– Menten constants of ACE for the tested peptides. As shown in Table 2, each ACE type was inhibited differently by different peptides, Neb-ODAIF-11)11 being the best overall substrate. All peptides tested were significantly better recognized by the C domain then by the N domain. Neb- ODAIF-11)7 and Neb-ODAIF-2 are almost not recognized by nACE and are not included in Fig. 3B. For Neb- ODAIF-11)7 this might be explained by the fact that only 25 lM was tested. The Km of sACE for Neb-ODAIF-11)11 is half the value of Km for LEQIYHL. The Km of nACE is nearly the same, indicating that Neb-ODAIF-11)11 is more C-domain specific than LEQIYHL. Neb-ODAIF-11)11 proved to be an excellent substrate for locust testis ACE, confirming that this ACE is kinetically more related to the human C-terminal ACE. The peptide LEQIYHL, however, is a good inhibitor for the C domain of human ACE, but not for locust testes ACE, indicating that locust testis ACE shares some but not all of the kinetic properties with C- domain ACE.

Western blotting

Polyclonal antibodies against Neb-ODAIF-11)9 were raised in New Zealand white rabbits as described in Materials and methods. The resulting antibodies were tested using dot spot methods. A 1 : 100 dilution of the antiserum was able to recognize 0.05 pmol of Neb-ODAIF-11)9 (data not shown). As the bulk of the yps is synthesized in the female fat body and transported by the haemolymph to be taken up by the developing oocytes, haemolymph of vitellogenic Neobellie- ria females, egg homogenate and haemolymph of male flies as a negative control were subjected to SDS/PAGE. A

The Km values of ACE for the purified peptides and for Neb-ODAIF-11)11 and Neb-ODAIF-11)7 were determined with recombinant human sACE, cACE and nACE and with the purified Locusta migratoria testis ACE using a compe- tition based assay. The cleavage of a fluorogenic ACE- substrate was followed in the absence and in the presence of the test peptide. For sACE, nACE and locust testis ACE, 1 lM final concentration of fluorogenic substrate was used. For cACE, 2.5 lM final concentration was used as this cACE stock was less active. Different concentrations of peptides were tested to obtain a clear inhibitory effect with 50 lM final concentration being the highest concentration used. For Neb-ODAIF-11)7, 25 lM was used as the highest concentration because of the limited amounts of this peptide available. In Fig. 3 the results of the assays are shown for one representative concentration of peptide with the different types of ACE. Using SPSS SIGMAPLOT, we performed nonlinear regression to the function f(x) ¼ y0 + a[1 ) exp(–bx)] describing an exponential increase to a maximum. The parameters a and b were used to calculate the initial cleavage rate v0 of the fluorogenic peptide. The

Fig. 3. Interaction of the purified peptides with different kinds of ACE. Competition assay for (A) sACE (B) nACE (C and D) cACE and (E) locust testis ACE of the fluorogenic substrate Abz-FRK-(Dnp)P and (a) 0 lM-test peptide, (b) 50 lM Neb-ODAIF*2)13, (c) 25 lM Neb-ODAIF1–7, (d) 50 lM Neb-ODAIF1–13, (e) 50 lM Neb-ODAIF1–9, (f) 50 lM LEQIYHL and (g) 50 lM Neb-ODAIF1)11.

3528 K. Hens et al. (Eur. J. Biochem. 269)

(cid:1) FEBS 2002

liberated from yp3. One possibility explaining the yp3 origin of the purified peptides, is that these are the products when the pinocytosed vitellogenins are transformed into vitellins. Perhaps, the N-terminally cleaving off of a small part of yp3 promotes the nearly crystalline packing of vitellins in the yolk platelets. Logically, when this hypothesis is correct, one should not expect immunoreactivity in mature ovarian extracts as vitellins would lack the Neb-ODAIF sequence. However, as follicle cells also synthesize endogenous vitellogenins [24], these follicle cell-derived vitellogenins might be the cause of anti-Neb-ODAIF-11)9 immunoreac- tivity observed in our ovarian extract. Alternatively, these peptides might also be produced by yolk degradation in the prospect of complete hydrolysis during subsequent embry- onic development. Several proteases such as cathepsin and acid phosphatase [25], capable of generating peptide frag- ments, have been identified in insect yolk granules and the identified in Drosophila proteasome complex that is embryos [26] is also thought to break down yps to peptide fragments [27]. Identification of the proteases present in the vitellogenic follicles in combination with the elucidation of the full sequence of the Neobellieria yps will allow unrave- ling the exact digestive pathways of yps. The fact that we purified two substrates from the same location in two different yps (Neb-ODAIF-11)13 and Neb-ODAIF-2), sug- gests that the yps are processed in a controlled manner.

PVDF membrane replica of the separated proteins was immunostained in a PAP-experiment using the anti-Neb- ODAIF-11)9 antiserum in a dilution series [1 : 2500 (A), 1 : 5000 (B)] (Fig. 4). A second PVDF replica was stained with Coomassie blue (C) in order to visualize all protein bands. Yp bands are marked on the Coomassie blue-stained membrane (C) with arrows. The antibody-stained mem- branes showed a single band corresponding to a yp band, and in the lower molecular weight range additional bands were stained. These lower molecular weight bands corres- pond to degradation products of the yp, that are likely to contain the Neb-ODAIF-11)9 sequence. The membranes were photographed, digitalized and enlarged so that the distance between the top of the membrane and the yp bands could be measured accurately. The antibody-stained band corresponds to the lowest yp band (yp3) on the Coomassie blue-stained membrane. Hence it can be concluded that we have developed an antiserum that is specific for Neobellieria yp3. This is a strong indication that the purified Neb- ODAIF-1 sequences are derived from yp3. Neb-ODAIF-2 resembles Neb-ODAIF-1, but is not completely identical. This peptide is probably derived from the same position in one of the two other yps present in Neobellieria bullata.

Fig. 4. Western analysis of fly haemolymph and egg homogenate with anti-Neb-ODAIF1–9 antibodies. Western blot of (1) male fly haemo- lymph (2) female fly haemolymph (3) fly egg homogenate and (4) molecular weight marker stained with (A) 1 : 2500 dilution and (B) 1 : 5000 dilution of anti-Neb-ODAIF1–9 antiserum or (C) Coo- massie brilliant blue. Yp1, yp2 and yp3 as indicated by arrows refer to the three yolk polypeptide bands. The vertical bars refer to the location of yp degradation products.

D I S C U S S I O N

Since no peptide has been proven to be an in vivo substrate for insect ACE to date, assumptions about ACE physiology in insects have to be made very carefully, especially when dealing with an enzyme with such broad substrate specificity as ACE. One potentially endogenous ACE substrate is already known, namely the trypsin modulating oostatic factor Neb-TMOF [28]. This peptide is capable of regulating vitellogenesis and is present in the ovaries. Neb-TMOF is suggested to be released by the ovaries and to be transported through the haemolymph to the midgut. Here, Neb-TMOF terminates the protein meal-induced trypsin biosynthesis. This results in an impaired blood digestion and a lack of amino acids that are needed for yolk synthesis, thus regulating ovarian development. TMOF has been shown to be an in vitro substrate for ACE present in the fly haemolymph [29] and captopril-feeding experiments indi- cate that TMOF is a true endogenous substrate [30]. The purification of several ACE interactive peptides (Neb- ODAIF-11)13, Neb-ODAIF-11)11, Neb-ODAIF-11)9, Neb- ODAIF-2, LEQIYHL) from the fly ovary stresses a putative regulatory role of ACE in vitellogenic or embryogenic events even more. The purified peptides might serve to stop yolk synthesis as the first batch of eggs reach maturity as does Neb-TMOF. If these peptides indeed regulate vitello- genesis, this would represent an autoregulation mechanism based upon the generation of peptides during yp degrada- tion. Neb-ODAIF-11)11 is very well recognized by insect testis ACE (Lom testis ACE: Km ¼ 2 lM) and may be a physiological substrate for insect ACE. If physiological experiments substantiate this hypothesis, these peptides might be used to interfere with insect reproduction and could thus be used in insect pest management.

In contrast with mammalian ACE, only single-domain insect ACE has been identified. Because insect ACE was used in the inhibition assay, peptides that are best recog- nized by insect ACE will be preferentially purified. There- fore, comparison of the kinetic parameters of the interaction

Anti-Neb-ODAIF-11)9 antibodies, apart from some yp degradation products, specifically immunostain yp3 and did not recognize yp1 or yp2. As the same antibodies did not reveal any positive protein bands in male haemolymph, it can be concluded that the Neb-ODAIF-11)9 peptide is derived from a yp3 gene product. No data are available at this moment on the mechanisms by which the peptides are

Ovary-derived ACE substrates and insect physiology (Eur. J. Biochem. 269) 3529

(cid:1) FEBS 2002

6. Schoofs, L., Veelaert, D., De Loof, A., Huybrechts, R. & Isaac, R.E. (1998) Immunocytochemical distribution of angiotensin I-converting enzyme-like immunoreactivity in the brain and testis of insects. Brain Res. 785, 215–227.

7. Wijffels, G., Gough, J., Muharsini, S., Donaldson, A. & Eisemann, C. (1997) Expression of angiotensin-converting enzyme-related carboxypeptidases in the larvae of four species of fly. Insect Biochem. Mol. Biol. 27(5), 451–460.

8. Isaac, R.E., Coates, D., Williams, T.A. & Schoofs, L. (1998) Insect angiotensin-converting enzyme: comparative biochemistry and evolution. In Recent Advances in Arthropod Endocrinology (Coast, G.M. & Webster, S.G., eds), pp. 357–378. Society for Experi- mental Biology, Cambridge University Press, UK.

9. Tatei, K., Chai, H., Ip, Y.T. & Levine, M. (1995) RACE: a Drosophila homologue of the angiotensin converting enzyme. Mech. Dev. 51(2–3), 157–168.

10. Vandingenen, A. (2001) Functional analysis of angiotensin con- verting enzyme. In Locusta Migratoria and Neobellieria Bullata. Phd Thesis. K. U. Leuven, Leuven, Belgium.

11. Vandingenen, A., Hens, K., Macours, N., Schoofs, L., De Loof, A. & Huybrechts, R. (2002) Presence of angiotensin converting enzyme (ACE) interactive factors in ovaries of the grey fleshfly Neobellieria bullata. Comp. Biochem. Physiol. 132, 27–35.

between purified peptides and recombinant human sACE, nACE and cACE will provide key information about the enzymatic similarity of insect ACE with the recombinant human ACEs. From the presented data, it is obvious that the purified peptides are all more or less C-domain specific. Indeed, Neb-ODAIF-11)11 and the peptide LEQIYHL are recognized very well by cACE. This may indicate that the circulating form of ACE of Neobellieria that we used to screen the HPLC-fractions for competitive peptides is more kinetically related to cACE. However, although Neb- ODAIF-11)11 is an excellent substrate for Lom testis ACE (Km ¼ 2 lM), the peptide LEQIYHL, also very well recognized by cACE, is almost not recognized by Lom testis ACE. This may indicate that Neobellieria ACE and Locusta ACE have different enzymatic properties. However, it is more plausible that testes of locusts contain a different ACE isoform than the ACE in circulation. The same might be true for the fly as in Drosophila, two isoforms of ACE with different kinetic properties are already known [31]. The fly ovaries are thus a rich source of domain-specific substrates and inhibitors. These might be used as a model to develop domain-specific inhibitors of ACE, which in turn may contribute to better insights into the domain-specific functions of human ACE. We have also purified a peptide that might allow us to distinguish between different isoforms of insect ACE, which again may be useful in the investigation of ACE functionality in insects.

12. Huybrechts, R. & De Loof, A. (1981) Effect of ecdysterone on vitellogenin concentration in hemolymph of male and female Sarcophaga bullata. International J. Invert. Repr. 3(3), 157–168. 13. Pappas, C. & Fraenkel, G. (1978) Hormonal aspects of oogenesis in flies Phormia regina and Sarcophaga bullata. J. Insect Physiol. 24(1), 75–80. 14. Ashby, G.J. (1972) The UFAW Handbook on the Care and Man- agement of Laboratory Animals. Livingstone, London.

A C K N O W L E D G E M E N T S

15. Araujo, M.C., Melo, R.L., Cesari, M.H., Juliano, M.A., Juliano, L. & Carmona, A.K. (2000) Peptidase specificity characterization of C- and N-terminal catalytic sites of angiotensin I-convering enzyme. Biochemistry 39, 8519–8525.

16. Rousseau, A., Michaud, A., Chauvet, M.-T., Lenfant, M. & Corvol, P. (1995) The hemoregulatory peptide N-acetyl-Ser-Asp- Lys-Pro is a natural and specific substrate of the N-terminal active site of human angiotensin-converting enzyme. J. Biol. Chem. 270(8), 3656–3661.

We thank L. Vanden Bosch for technical assistance with HPLC and sequencing, I. Bongaers for assistance with screening and antibody preparation, J. Puttemans and M. Christiaens for figure layout, P. Corvol and A. Michaud (Institut National de la Sante´ et de la Recherche Me´ dicale, Unite´ 36, College de France, Paris, France) for the kind gift of Chinese hamster ovary cells expressing sACE, cACE and nACE. This work was supported by the (cid:2)Instituut voor de Aanmoed- iging van Innovatie door Wetenschap en Technologie in Vlaanderen Vlaanderen(cid:3), by the Research Foundation of the K. U. Leuven (GOA/ 2000/04) and by the FWO (G0356.98 and G0187.00). 17. Ryan, J.W., Chung, A., Ammons, C. & Carlton, M.L. (1977) A simple radioassay for angiotensin-converting enzyme. Biochem. J. 167, 501–504.

R E F E R E N C E S

18. Morris, H.R., Paxton, T., Dell, A., Langhorne, J., Berg, M., Bordoli, R.S., Hoyes, J. & Bateman, R.H. (1996) High sensitivity collisionally-activated decomposition tandem mass spectrometry on a novel quadrupole/orthogonal-acceleration time-of-flight mass spectrometer. Rapid Commun. Mass Spectrom. 10(8), 889–896. 1. Lamango, N.S. & Isaac, R.E. (1994) Identification and properties of a peptidyl dipeptidase in the housefly, Musca domestica, that resembles mammalian angiotensin-converting enzyme. Biochem. J. 299, 651–657.

19. Xie, D., Suvorov, L., Erickson, J.W. & Gulnik, S. (1999) Real- time measurements of dark substrate catalysis. Protein Sci. 8(11), 2460–2464.

2. Quan, G.X., Mita, K., Okano, K., Shimada, T., Ugajin, N., Xia, Z., Goto, N., Kanke, E. & Kawasaki, H. (2001) Isolation and expression of the ecdysteroid-inducible angiotensin-converting enzyme-related gene in wing discs of Bombyx mori. Insect Bio- chem. Mol. Biol. 31(1), 97–103.

20. Hermanson, G.T. (1996) Zero-length cross-linkers. Bioconjugate Techniques, pp. 169–173. Academic Press Inc., New York. 21. Vandesande, F. (1979) A critical review of immunocytochemical methods for light microscopy. J. Neurosci. Methods 1, 3–23. 22. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685. 3. Lamango, N.S., Sajid, M. & Isaac, R.E. (1996) The endopeptidase activity and the activation by Cl– of angiotensin-converting enzyme is evolutionarily conserved: purification and properties of an angiotensin-converting enzyme from the housefly, Musca domestica. Biochem. J. 314(2), 639–646.

4. Howard, T.E., Shai, S.Y., Langford, K.G., Martin, B.M. & Bernstein, K.E. (1990) Transcription of testicular angiotensin- converting enzyme (ACE) is initiated within the 12th intron of the somatic ACE gene. Mol. Cell Biol. 10(8), 4294–4302. 23. Huybrechts, R. & De Loof, A. (1983) Immunological and elec- trophoretical identification of vitellogenin, the mayor yolk pre- cursor protein of Sarcophaga bullata (diptera). The unnecessity of tedious protein purification to study the physiology of vitellogenin synthesis. Annu. Soc. R. Zool. Belg. 113, 45–54.

24. Huybrechts, R., Cardoen, J. & De Loof, A. (1983) (cid:2)In vitro(cid:3) secretion of yolk polypeptides by fat body and ovaries of 5. Taylor, C.A., Coates, D. & Shirras, A.D. (1996) The Acer gene of Drosophila codes for an angiotensin-converting enzyme homo- logue. Gene 181(1–2), 191–197.

3530 K. Hens et al. (Eur. J. Biochem. 269)

(cid:1) FEBS 2002

Sarcophaga bullata (Diptera, Callipharidae). Ans. Soc. R. Zool. Belg. 113, 309–317.

25. Ribolla, P.E., Bijovsky, A.T. & De Bianchi, A.G. (2001) Pro- cathepsin and acid phosphatase are stored in Musca domestica yolk spheres. J. Insect Physio, 47, 225–232.

26. Udvardy, A. (1993) Purification and characterization of a multi- protein component of the Drosophila 26 S (1500 kDa) proteolytic complex. J. Biol. Chem. 268, 9055–9062.

29. Zhu, W., Vandingenen, A., Huybrechts, R., Baggerman, G., De Loof, A., Poulos, C.P., Velentza, A. & Breuer, M. (2000) In vitro degradation of the Neb-trypsin modulating oostatic factor (Neb-TMOF) in gut luminal content and hemolymph of the grey fleshfly, Neobellieria bullata. Insect Biochem. Mol. Biol. 31, 87–95. 30. Vandingenen, A., Hens, K., Macours, N., Zhu, W., Janssen, I., Breuer, M., De Loof, A. & Huybrechts, R. (2001) Captopril, a specific inhibitor of angiotensin converting enzyme, enhances both trypsin and vitellogenin titers in the grey fleshfly Neobellieria bullata. Arch. Insect Biochem. Physiol. 47, 161–167. 27. Giorgi, F., Bradley, J.T. & Nordin, J.H. (1999) Differential vitellin polypeptide processing in insect embryos. Micron 30(6), 579– 596.

31. Houard, X., Williams, T.A., Michaud, A., Dani, P., Isaac, R.E., Shirras, A.D., Coates, D. & Corvol, P. (1998) The Drosophila melanogaster-related angiotensin-I-converting enzymes Acer and Ance-distinct enzymic characteristics and alternative expression during pupal development. Eur. J. Biochem. 257, 599–606. 28. Bylemans, D., Proost, P., Samijn, B., Borovsky, D., Grauwels, L., Huybrechts, R., Van Damme, J., Van Beeumen, J. & De Loof, A. (1995) Neb-colloostatin, a second folliculostatin of the grey fleshfly, Neobellieria bullata. Eur. J. Biochem. 228, 45–49.