Male accessory glands of Drosophila melanogaster make a secreted angiotensin I-converting enzyme (ANCE), suggesting a role for the peptide-processing enzyme in seminal fluid

Peptidyl dipeptidase A (EC 3.4.15.1) is best known for its role in the mammalian renin–angiotensin system (RAS) as the enzyme responsible for the removal of a C-terminal dipeptide (His-Leu) from the decapeptide angiotensin I to generate the vasopressor octopeptide angiotensin II; hence its common name angiotensin I-converting enzyme (ACE)(Corvol et al., 2004). The enzyme also cleaves a dipeptide from the C terminus of the circulating vasodilator bradykinin, which inactivates this peptide hormone(Erdos, 1979). In mammals,there are two isoforms of ACE; a two-domain somatic protein and a smaller single domain germinal isoform that is found exclusively in male germ cells(Corvol et al., 2004). Although somatic ACE is mainly found on the surface of endothelial cells, soluble enzyme is found in body fluids, including blood, cerebrospinal fluid and seminal fluid (Corvol et al.,2004). The precise function of germinal ACE is not known, but mice deficient in germinal ACE are infertile(Hagaman et al., 1998). Somatic ACE and other components of the RAS are also found in male reproductive tissues (Leydig cells, seminiferous tubules, epididymis, testis and prostate gland) with the prostate being a likely source of the angiotensin II that has been found in human seminal fluid(Leung and Sernia, 2003; O'Mahony et al., 2005; O'Mahony et al., 2000). Angiotensin II might affect fertility in the human male because of its ability to stimulate sperm motility, induce the acrosome reaction and increase oocyte penetration (Kohn et al.,1997; Kohn et al.,1998; Vinson et al.,1996; Vinson et al.,1995). By contrast, the pronounced male infertility of mice lacking germinal ACE is independent of the RAS and probably results from failure to cleave a peptide distinct from angiotensin I(Fuchs et al., 2005).

ACE is an evolutionarily ancient dicarboxypeptidase found in a range of invertebrates (e.g. arthropods and annelids) and even in some bacteria(Rawlings et al., 2006). In insects, ACE is a soluble secreted enzyme that is strongly expressed in the male reproductive tissues (Isaac et al.,2007). In the testes of Drosophila melanogaster, germ cells are the major site of ACE (known as ANCE) biosynthesis and male flies homozygous for hypomorphic alleles of Ance are infertile(Tatei et al., 1995). This infertility results from a failure in spermiogenesis, suggesting that one of the functions of germinal ANCE in D. melanogaster is the processing of a regulatory peptide required for spermatid differentiation(Hurst et al., 2003). Another possible role for ANCE is in the processing of peptides in seminal fluid, as has been proposed for human prostate ACE. In D. melanogaster, the male accessory glands (AG) are responsible for the production and secretion of a large number of proteins into the seminal fluid that mix with sperm on ejaculation (Ram and Wolfner,2007). These include several peptide and protein hormones as well as enzymes, stress response proteins and immune defence proteins. The peptide/protein hormones are responsible for a variety of physiological and behavioural responses in the post-mated female, including increased rate of ovulation, loss of receptivity to males, improved sperm storage and increased appetite (Carvalho et al.,2006; Chapman and Davies,2004).

We now show that in addition to expression in the testes, ANCE is also produced in the AG of adult male D. melanogaster and that the peptidase is localised to giant vesicles of the secondary cells. ACE-like peptidase activity of the AGs is reduced in the post-mated male, presumably as a result of transfer to the mated female of the AG products as part of the seminal fluid. We speculate that the AG ANCE functions to cleave the C terminus of peptides with a role in reproduction to alter their biological activity.

Unless otherwise stated chemicals and immunological reagents were purchased from Sigma-Aldrich Co. Poole, Dorset, UK. The Vectastain ABC Kit and Vectashield were obtained from Vector Laboratories Ltd., Peterborough, UK. The Enhanced Chemiluminescence (ECL) Detection Kit was from Amersham Pharmacia Biotech Ltd., UK. The Gemini 5 μ C18 (150 mm×4.6 mm) high-performance liquid chromatography (HPLC) column was purchased from Phenomenex(Macclesfield, Cheshire, UK).

Wild-type D. melanogaster Meigen (Oregon R) were maintained on oatmeal–molasses–agar medium at 25°C(Ashburner and Thompson,1978).

Antibodies were raised against recombinant ANCE expressed in Pichia pastoris as described elsewhere(Houard et al., 1998). For immunohistochemistry, AGs were dissected in 10 mmol l^–1phosphate-buffered saline (PBS;Na₂PO₄–NaHPO₄ buffer, pH 7.4, 150 mmol l^–1 NaCl) and treated with 2% (v/v) hydrogen peroxide in methanol for 5 min, then washed three times in PBS prior to fixation for 20 min in 4% (w/v) paraformaldehyde in PBS. Tissues were blocked in 1% (w/v)bovine serum albumin, 10% (v/v) normal goat serum in PBST [PBS, 0.03% (v/v)Triton X-100] for 1 h at 25°C before incubation in a 1:2000 dilution of primary antibody (either immune or pre-immune serum) in PBST overnight at 4°C. AGs were washed three times in PBST before treatment with the Vectastain ABC Kit according to the manufacturer's instructions. For immunofluorescence assays, AGs were dissected, fixed and incubated with primary antibody as described above. AGs were then incubated in a 1:2000 dilution of FITC-conjugated goat anti-rat IgG in PBST for 1 h at 25°C. After three washes in PBST, tissues were mounted in Vectashield and images were recorded using a Zeiss LSM510 META upright confocal microscope. Samples were excited with the 488 nm laser line of an argon laser running at 5% power output; and emission was collected with a long pass LP505 nm filter. 3D reconstruction and iso-surface rendering were performed in Imaris version 4.0.6 (Bitplane, Zurich, Switzerland) using contours within the Surpass module of the software.

In situ hybridisation experiments were carried out using Ance RNA probes according to the method described previously(Siviter et al., 2000); sense probes were used as a control.

ACE activity was determined by incubating AG tissue or AG secretions that had been collected in 5 μl of PBS with the substrate solution (5 mmol l^–1 Hip-His-Leu in 0.1 mol l^–1Tris–HCl, pH 8.3, 0.3 mol l^–1 NaCl, 10 μmol l^–1 ZnSO₄; total volume 20 μl). After 4 h at 35°C, the enzyme reaction was terminated by reducing the pH to 2.0 with the addition of 8% (v/v) trifluoracetic acid. The final volume was made up to 260 μl with 0.1% (v/v) trifluoracetic acid and the released hippuric acid was quantified by HPLC using a 5 μ C18 (150 mm×4.6 mm) column, as described previously (Lamango et al.,1996).

AG proteins from both Oregon R and tudor (tud) flies were extracted and separated on a 10% sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS–PAGE) gel, transferred to a PVDF membrane, and then incubated with anti-ANCE antibody at a 1:5000 dilution in PBST, 5% (w/v)non-fat dried milk powder as described previously(Houard et al., 1998; Hurst et al., 2003). Bound anti-ANCE antibody was detected by using a horseradish peroxidase-conjugated sheep anti-rabbit Fc antibody and the Enhanced Chemiluminescence Detection Kit(Amersham Pharmacia Biotech Ltd., UK) as described in the manufacturer's instructions. tud mutant flies were used in addition to wild-type Oregon R because the tud males possess much lower levels of testicular ANCE, which might contaminate the protein preparation. Similar levels of ANCE were detected in AG extracts from the mutant and wild-type males, but only blots of the tud AGs are presented.

Male AGs were dissected from virgin D. melanogaster males of between 3 and 5 days old and fixed in 2.5% (v/v) gluteraldehyde in 0.1 mol l^–1 Na₂PO₄–NaHPO₄buffer, pH 6.9, for 3 h. Specimens were then washed twice in the buffer before post-fixation for 1 h in 1% (w/v) osmium tetroxide in 0.1 mol l^–1 Na₂PO₄–NaHPO₄, pH 6.9. After washing in two changes of buffer, the specimens were dehydrated by using an ascending ethanol series (five concentrations ranging from 20–100%), each step taking 20 min. After one additional change of 100%ethanol, the specimens were embedded in Araldite(Luft, 1961). Sections(80–90 nm) were cut from the Araldite blocks using an Ultramicrotome(Reichert-Jung Ultracut-E, Leica, Milton Keynes, UK) and were stained with uranyl acetate and Reynolds' lead citrate solution(Reynolds, 1963). Sections were examined using a Jeol 1200EX transmission electron microscope (Jeol UK Ltd, Welwyn Garden City, UK) at an 80 kV accelerating voltage.

AGs were found to hydrolysed the ACE substrate (Hip-His-Leu) to release hippuric acid. The specificity of the enzyme assay was confirmed by the complete inhibition of the activity by captopril, a potent inhibitor of insect ACE (Cornell et al., 1995; Lamango et al., 1996). The distribution of the AG ACE was determined by gently compressing AGs from unmated males with forceps to discharge the luminal contents into PBS. A comparison of the ACE activity present in these secretions with the activity remaining in the compressed AGs showed that around 80% of the active enzyme was in the AG lumen (Table 1). When males were allowed to mate with ten females over a 3-day period, the ACE activity of the AGs was reduced by 70%, consistent with the enzyme being transferred during copulation from the AG to the female in the seminal fluid(Table 1).

To establish the identity of the ACE-like activity found in the male AGs,antibodies that specifically recognise ANCE were employed in western blot analysis of AG proteins (Fig. 1). A single protein band (M_r,72×10³) of the expected size was detected with the ANCE antiserum, establishing ANCE as an AG product and the enzyme most likely to be responsible for the ACE-like activity. A second D. melanogasterACE-like enzyme called ACER (ACE-related) can also cleave Hip-His-Leu in a captopril-sensitive manner and therefore ACER might also contribute towards the dicarboxypeptidase activity of the male AGs. However, ACER was not detected in western blots using antibodies raised to recombinant D. melanogaster ACER. We therefore conclude that ANCE, rather than ACER, is responsible for the ACE-like activity detected in the male AGs.

Immunohistochemistry revealed strong foci of ANCE protein within the secondary cells that are located in the distal third of the AG(Fig. 2). Confocal microscopy of AGs stained with FITC-labelled secondary antibody clearly identified large vesicles of the secondary cells as the subcellular site of the strong ANCE staining (Fig. 3A). The secondary cells generally contained five or six of these vesicles, some of which appeared to be devoid of ANCE. Electronmicroscopy of AG tissue sections showed that these large membrane-bound vesicles contain a dense granule surrounded by a filamentous material (Fig. 3C), as noted elsewhere(Bairati, 1968). Comparison of the micrographs with the fluorescent immunocytochemistry images show clearly that ANCE is localised to the dense large granule of the secondary cell. Three dimensional reconstruction of the secondary cells using confocal microscopy found the volume of the vesicles to be on average 103±14μm³. Around half of this volume (54±15 μm³)was taken up by the dense core granule(Fig. 3B).

In situ hybridisation using digoxigenin-labelled antisense RNA probes for Ance revealed strong expression in the secondary cells of the gland and strong staining in the AG lumen(Fig. 4). In contrast, the level of staining in the main cells did not appear to be significantly stronger than that obtained using control sense probes.

In this study we have shown that a soluble ACE-like peptidase is synthesised by the AGs of male D. melanogaster and that the enzyme activity is lost after mating, presumably by transfer to the female in the seminal fluid. A similar transfer of AG ACE occurs in the tomato moth, Lacanobia oleracea, where the enzyme ends up in the female spermatheca and bursa copulatrix (Ekbote et al., 2003). It was not possible to follow the fate of the D. melanogaster AG ANCE, since the enzyme was also expressed in the female reproductive tissues (R. E. Isaac, unpublished). There are six ACE-like genes in the D. melanogaster genome, but only Ance and Acer products are considered to be enzymatically active; the other gene products are predicted to be proteolytically inactive as they are missing key active site residues (Coates et al.,2000). Using antibodies that differentiate between ANCE and ACER,we have shown that the AG dicarboxypeptidase activity can be attributed to ANCE and not ACER. ANCE is expressed in the secondary cells of the AG and becomes incorporated into the dense granular mass that half-fills the giant vesicles of these cells. This is the first identification of a stored component of this unusual organelle. In situ hybridisation using an Ance riboprobe was consistent with relatively strong expression in the secondary cells. It has been suggested previously that the secondary cells of D. melanogaster AGs discharge their contents by a holocrine secretory mechanism (Chen,1984). This mode of secretion might explain our observation of Ance RNA in the AG lumen.

ANCE is just one of several secreted peptidases produced by the AGs of D. melanogaster. In total, eleven peptidases including aminopeptidases, endopeptidases and a γ-glutamyl transpeptidase have already been identified as AG products in D. melanogaster(Mueller et al., 2004; Walker et al., 2006). One of these peptidases, an astacin-like endopeptidase, is involved in the cleavage of the male AG ovulin in the female reproductive tract to produce four products, two of which stimulate ovulation in the first 24 h post-mating(Ravi Ram et al., 2006). ANCE might work in concert with other peptidases to process prohormone polypeptides to biologically active peptides in a similar manner to the local RAS of the mammalian prostate (Leung and Sernia,2003; O'Mahony et al.,2005). The best known biologically active peptide of the male reproductive tissues is the sex peptide, which stimulates egg-laying and reduces receptivity to males in the mated female(Kubli, 2003). However, ANCE is unlikely to be involved in its biosynthesis because the C terminus of the proprotein does not undergo proteolytic processing. In contrast, Dup99B, a peptide made by the male ejaculatory duct, is a potential substrate for ANCE. The mature Dup99B, the C-terminal amino acid sequence of which is almost identical to the sex peptide, elicits the same responses in the post-mated female as the sex peptide when the peptide is injected into virgin female D. melanogaster (Rexhepaj et al.,2003; Saudan et al.,2002). Unlike the sex peptide, the DUP99B proprotein has a basic dipeptide (Arg-Lys) at the C terminus, which must be proteolytically removed to expose the C-terminal cysteine. We have shown previously that peptides with pairs of basic residues at the C terminus make excellent ANCE substrates(Isaac et al., 1998) and it is possible, therefore, that ANCE has a role in cleaving the Arg-Lys dipeptide from the C terminus of a partially processed DUP99B precursor as the AG products mix with the material secreted by the ejaculatory.

We cannot, at present, reach any conclusions regarding the relative importance of the AG and germinal ANCE for male fertility in D. melanogaster. This level of understanding will require tissue-specific knock-down of Ance expression.

List of abbreviations

 
• ANCE

  Drosophila Angiotensin converting enzyme

 
• ACER

  Drosophila Angiotensin-converting enzyme-related

 
• ACE

  Angiotensin 1 converting enzyme

 
• AG

  accessory gland

 
• RAS

  renin-angiotensin system

We thank Adrian Hick, University of Leeds, for assistance with electron microscopy, Pam Gaunt and Carol Sowden for technical support and the laboratory of Pierre Corvol, College de France, for recombinant ANCE and ACER. We also acknowledge the Biotechnology and Biological Research Council for providing studentships to M.W. and C.R. and for supporting the Leeds Bioimaging Facility.