Aminopeptidase-B in the rat testes: isolation, functional properties and cellular localization in the seminiferous tubules

ELSEVIER

Molecular

and Cellular Endocrinology

110 (1995) 149-160

Aminopeptidase-B in the rat testes: isolation, functional properties and cellular localization in the seminiferous tubules Sandrine Cadela, Adrian R. Pierotti a,l, Thierry Foulona, Christophe Cr&ninonb, Nicole Barr&, Dominique Segr&ainc, Paul Cohen&* ?kborafoire

de Biochimie des Signaux Rt!gulateurs Cellulaires et Mol&daires, Universite’ Pierre et Marie Curie, Unite’ de Recherches AssociPe au Centre National de la Recherche Scientijique 1682, 96 Boulevard Raspail, 75006 Paris, France

bCEA - Service de Pharmacologic et d’lmmunologie, Dtfpartement de Recherches en Imagerie, Pharmacologic et Physiologie du Centre d’Etudes de Saclay, 91191 Gif-sur-Yvette Cedex, France ‘Laboratoire d’Embryologie et de Biologie de la Reproduction, Fact& de Mtfdecine, 45 rue des Saints P&es, 75006 Paris and Groupe d’Etude de la Reproduction chez le Mrile, Institut National de la Recherche Midicale, Campus de Beaulieu, Universite’ de Rennes I, 35042 Rennes cedex, France Received 27 January

1995; accepted 3 March 1995

Abstract

An aminopeptidase of the B-type, with an apparent Mr 72 000 and pI = 4.9, was isolated from rat testes and characterized. The enzyme was able to remove only Arg and/or Lys residues from L-amino acid &rtaphthylamide derivatives and from the N-terminus of several peptides. No cleavage occurred in the case of Arg-Pro bonds as found in bradykinin and substance P.-The enzyme was sensitive to cysteinyl reagents and to aminopeptidase inhibitors, such as bestatin, amastatin and arphamenines A and B. The aminopeptidase activity, tested with L-Arg b-naphthylamide and with Argo-Met-enkephalin as substrates, was inhibited by o-phenanthroline, and restored by 2n2+ suggesting its metallopeptidase character. The partial characterization of an aminopeptidase-B activity in rat brain cortex identified a protein which is biochemically and immunologically related to the testis enzyme. By immunohistochemistry, the aminopeptidase-B was found to be particularly abundant in the seminiferous tubules at late stages of spermatogenesis and was clearly detected in a restricted area of elongated spermatids. Remarkably, the enzyme was observed to concentrate massively in the residual bodies. Since this aminopeptidase-B was able in vitro to trim out N-terminal Arg and/or Lys residues from peptides mimicking processing intermediates, it is proposed that this enzyme may be involved in propeptide and proprotein processing mechanisms in the course of spermatid differentiation. Keywords: Exopeptidase;

Aminopeptidase-B;

Processing;

Testis; Spermatogenesis;

Residual bodies

1. Introduction

Abbreviations: Ap-B, aminopeptidase-B; L-Arg B-NA, t_-arginine /3naphthylamide; CFP-AAF-pAB, N-[I-(R,S)-carboxy-3-phenyl propyl]

Ala-Ala-Phe-para-aminobenzoate; DTT, dithiothreitol; FPLC, fastperformance liquid chromatography; His-HCI, histidine hydrochloride; MES, 2-[N-morpholino] ethane sulfonic acid; NEM, N-ethyl maleimide; PBE, polybuffer exchanger; PCMB, p-(chloromercuri)benzoate; PCMPS, p-(chloromercuri)-benzenesulfonic acid; phosphodiepryl 03. N-(phenylethylphosphonyl)-GIy-L-Pro-L-~inohex~oic acid; PMSF, phenylmethylsulfonylfluoride; TPCK, L-1-chloro-3(4tosylamido)-4-phenyl-2-butanon. * Corresponding author, Tel.: +33 1 42 22 81 8.5; Fax: +33 1 42 22 13 98. ’ Present address: Glasgow Caledonian University, Department of Biological Sciences, Cowcaddens Road, Glasgow G4 OBA, UK.

0 199.5 Elsevier Science Ireland Ltd. All rights reserved 0303-7207/95/$09.50 SSDI 0303-7207(95)03529-G

A number of peptide and protein messengers involved in many aspects of cellular regulation are produced in secretory cells by selective proteolytic activation of macromolecular

biosynthetic

tide bond hydrolysis basic

amino

Gluschankof

acids,

precursors.

In most

cases

pep-

occurs at the level of Arg and Lys arranged

and Cohen,

as doublets

1987; Darby

(Cohen,

1987;

and Smyth,

1990)

(Bourdais and Cohen, 1991; Devi, 1991) although other arrangements are found. The recent discovery of a family of endoproteases related to the subtilisinlike enzyme product of the yeast KEX2 gene supports the original idea that, during processing, cleavage occurs on the C-terminus of basic residues (Docherty and Steiner, or singlets

150

S. Cadel et al. I Molecular and Cellular Endocrinology 110 (1995) 149-160

1982). Then, the activity of a carboxypeptidase of the Btype is required to trim out unwanted basic extensions from the C-terminus of the generated fragments (Fricker, 1988). This scheme is likely to apply to a large number of pro-peptides and proteins. However, in some cases, processing may also involve endoproteases hydrolysing peptide bonds on the N-terminus of basic residues (Peng Loh et al., 1985; Gomez et al., 1988; Azaryan and Hook, 1994; Chesneau et al., 1994). Indeed, the in vivo metabolic pathways of a number of precursors indicate the generation of intermediary peptide fragments bearing an Arg or Lys extension on their N-terminus (Peng Loh et al., 1985; Inagami, 1989). These observations raised the hypothesis that an aminopeptidase-B (Ap-B) like activity is required to remove the extra basic residues and to produce active fragments. Despite the description of some Ap-B activities in various tissue extracts (Kawata et al., 1980; Saderling, 1983; Gainer et al., 1984; Mantle et al., 1985; McDermott et al., 1988; Castro et al., 1989; Flores et al., 1993), the possible involvement of these enzymes in processing mechanisms has received little attention and only a few of these exopeptidases have been studied in this context (Gainer et al., 1984; Gluschankof et al., 1987; Castro et al., 1989). Previous work from this laboratory has identified in rat brain cortex extracts an enzyme complex originally called somatostatin-28 convertase for its ability to generate, in vitro, both somatostatin- 14 and somatostatin-28-( l-l 2) from the octacosapeptide somatostatin-28 substrate (Gluschankof et al., 1987; Gomez et al., 1988). Although, a preliminary characterization of the brain aminopeptidase-B component of this peptidase tandem was performed, the recent finding that the associated endopeptidase called mD convertase (nardilysin; EC 3.4.24.61; Chesneau et al., 1994) is present massively in the testis has led us to consider this tissue as a source for this exopeptidase-B. Except for the preliminary work from Vanha-Pertulla (1973) no such enzyme has been unequivocally identified and adequately characterized in this tissue. Since the testis is a major organ for hormonal regulations and in view of the number of peptides and precursors which have been reported to be present in this complex autocrine-paracrine gland (Kilpatrick et al., 1987; Yoshikawa and Aizawa, 1988), we have attempted to isolate and to characterize the functional properties of an Ap-B activity from rat testes extracts. This enzyme, which belongs to the metallopeptidase family, exhibits a strict selectivity for Arg-Xaa and Lys-Xaa bonds in peptide substrates. This exopeptidase-B is biochemically and immunologically related to the rat brain cortex Ap-B. It appears to be highly present in the germ line at a late stage of spermatogenesis and to concentrate in the residual bodies of the seminiferous tubules. It is proposed that this basic amino acid specific aminopeptidase may par-

ticipate in pro-peptide and pro-protein mechanisms during germ cell differentiation.

processing

2. Materials and methods 2.1. Aminopeptia’ase extraction and purification procedure Ap-B was isolated from 10 male Wistar rats. Testes and cerebral cortices were homogenized (20% w/v) in 100 mM KCl, 50 mM phosphate buffer (pH 7.4) using a Potter-Elvejhem homogeniser. The extract was centrifuged for 15 min at 2000 X g and the proteins of the supernatant were precipitated at pH 4.7 with 1 M ammonium acetate (pH 4.5), 5 mM P-mercaptoethanol, then centrifuged at 7800 X g for 30 min. The recovered supernatant was concentrated to 10 ml. The pooled enzyme solutions were concentrated using an Amicon Ultrafiltration System (model 202 or 402 with an ultrafiltration membrane YM 30). The extract was then subjected to molecular sieve filtration on a Sephadex G-150 column equilibrated and eluted with 250 mM Tris-HCl (pH 7.5); 5 mM pmercaptoethanol. The resulting active fractions were diluted 1:5 with distilled water, applied to a DEAETrisacryl M ion exchange column (100 ml bed volume; Sepracor) pre-equilibrated in 50 mM Tris-HCl (pH 7.5) and eluted with a continuous gradient from 0 to 150 mM KCl. The Ap-B activity was stable for several days at 4°C. The fractions with Ap-B activity were pooled, concentrated and equilibrated with 20 mM sodium phosphate (pH 7.1) by passage through a Sephadex G-25 column. The resulting active fractions were concentrated to 10 ml and purified from serum albumin by FPLC (PharmaciaLKB) using an affinity-blue cartridge (5 ml; Biorad). The effluent proteins were monitored by UV absorbance at 215 nm. Ap-B was eluted with the buffer used to equilibrate the Sephadex G-25 column. The pH of the active fractions resulting from the affinity-blue gel chromatography step was adjusted to pH 5.6 by passage through a Sephadex G-25 column equilibrated in 25 mM His-HCl (pH 5.6) and then applied to a column of polybuffer exchanger (PBE) 94 gel (1 X 35 cm; Pharmacia) equilibrated with the same buffer. Ap-B activity was eluted from the chromatofocusing column using a pH gradient from 5.6 to 4.5 (Polybuffer 74-HCl, pH 4.5, diluted 1:9 with distilled water; Pharmacia-LKB) at a flow rate of 30 ml/h. The collected fractions containing the ApB activity were pooled, adjusted to pH 7.4 and concentrated to 10 ml. When the enzyme was frozen at -20°C in glycerol solution, a complete loss of activity was observed. 2.2. Enzyme assays with L-amino acidp-naphthylamide substrates Ap-B activity was tested with L-Arg fi-NA (Sigma) as substrate. Samples from the post-DEAE fraction (5 or

S. Cadel et al. 1 Molecular

and Cellular

10~1; l~g/~l of proteins) were incubated at 37°C for 30 min in assay buffer (100 mM borate buffer, pH 7.4; 150 mM NaCl) containing 0.2 mM L-Arg /I-NA. The hydrolysis was interrupted by addition of 0.3 ml of freshly prepared colour reagent (Fast Garnet GBC salt; 1 mg/ml in 1 M sodium acetate buffer, pH 4.2; 10% Tween-20). The absorbance was read at 535 nm using a Varian Cary 118 spectrophotometer and the amount of free /Inaphthylamine was calculated using a standard curve. A unit was defined as the amount of enzyme necessary to hydrolyse 1 nmol of substrate per minute under the described conditions. Optimal pH for Ap-B activity was determined in the pH range 5.5-8.25 using 50 mM sodium phosphate (pH 5.5-7.4) and 100 mM sodium borate (pH 7.4-8.25) buffers. The assays were performed, at least three times, with and without the addition of 150 mM NaCl. To determine the effect of inhibitors on enzyme activity, 10~1 of the pooled active fraction obtained after the DEAE purification step were pre-incubated in assay buffer in the presence of selected inhibitors for 30 min at 37°C. Thirty minutes after the addition of the substrate, the reaction was stopped. The percentage of inhibition was measured by comparison with values obtained without inhibitor. The extent of reactivation by various divalent cations after o-phenanthroline inhibition was determined by the following test: lo@ of enzyme preparation were incubated in 1 ml assay buffer containing 5OOpM o-phenanthroline shown to inhibit by approximately 84% the enzymatic activity of Ap-B when measured with the substrate L-Arg /?-NA (200pM final concentration). Then various cations were added (0.1-500 PM range) to restore the enzymatic activity over a 30 min period at 37°C. The percentage of reactivation was calculated by taking as reference the value observed in the presence of cation chelator, the residual 16% activity was arbitrarily considered as 0%. 2.3. HPLC analysis of aminopeptidase activity towards peptide substrates Hydrolysis of synthetic substrates (peptides from Sigma, Neosystem) was carried out in assay buffer at 37°C. The progress of hydrolysis of the Arg and/or Lys amino terminal bond was determined by HPLC. The peptides [Ala17-Tyr20] somatostatin-28-( 13-20), [Ala17-Tyr20] somatostatin-28( 14-20) and Arg-*-Lys”-somatostatin-14 were eluted on a p-Bondapak Cl8 column (3.9 X 300 mm, Waters), using a gradient of acetonitrile (2028.2% over 30 min) in buffer A (0.1% 1-hexanesulfonic acid, 20 mM acetic acid) at a flow rate of 1 mUmin. The other peptides were eluted on a Cl8 ultrabase column (4.6 x 250 mm, SociCtt Francaise Chromato-Colonne) using the following gradients of acetonitrile in buffer A (0.5 ml/l trifluoroacetic-acid) at a flow rate of 0.5 ml/min: 20-80% in 30 min for Argo-Leu-enkephalin, Argo-Met-

Endocrinology

110 (1995)

149-160

151

enkephalin, Argo-neurokinin, neurotensin-(8-13), neurotensin-(9-13) and P25-(263-272), 25-75% in 30 min for substance P, 15-65% in 30 min for bradykinin and Argoalpha atria1 natriuretic factor-( l-28). To study the substrate selectivity of Ap-B, 10~1 of the pooled active fraction resulting from the DEAE chromatography step were incubated with 10 pug of substrate in 100~1 of the assay buffer containing phosphodiepryl 03 (1 PM final) for 30 min or 1 h at 37°C. At the appropriate time the reaction was stopped by addition of 10~1 of 1 M HCl. Peptide hydrolysis, visualised on HPLC by separation of products, was quantified by integration of the HPLC peaks. The effect of inhibitors on enzyme activity towards Argo-Met-enkephalin was determined as described above except that 1Opg of the substrate were added and the reaction was stopped after 4 h. The extent of inhibition was expressed as the percentage of the standard run in the absence of the inhibitor. 2.4. Enzyme analysis Protein quantification was determined by the Bradford procedure (Bradford, 1976) using bovine serum albumin as standard. The purified active fractions were pooled, concentrated and analysed by polyacrylamide gel electrophoresis under denaturating (in the presence of 5% /Imercaptoethanol) or non-denaturating conditions on 825% Phast system gradient electrophoresis gels (Pharmacia-LKB). Separation conditions and silver staining of proteins were those recommended by the supplier. Ap-B activity was visualised on non-denaturating by incubating the gel for 1 h at 37°C in assay buffer containing 0.3 mM L-Arg b-naphthylamide. The reaction was stopped and coloured with Fast-Garnet GBC (0.3 mg/ml in 1 M sodium acetate, pH 4.2), and destained in distilled water for 15 min. 2.5. Antibodies Polyclonal antibodies were raised in rabbit by multiple subcutaneous injections of 5Opg of purified aminopeptidase emulsified in Freund’s complete adjuvant. The first booster injection was given 6 weeks later and then monthly thereafter, and rabbits were bled weekly. The titres of the different antisera were tested with an enzyme immunoassay using biotinylated endopeptidase and acetylcholinesterase-labelled avidin (Frobert and Grassi, 1992). 2.6. Western blots Crude extracts from testes and brain were analysed after the precipitation step (see Section 2.1). Two micrograms of protein were loaded under denaturating conditions on 8-25% Phast system gradient electrophoresis gel (Pharmacia-LKB). Immunoblots were performed after electrotransfer on a nitrocellulose membrane (0.45 pm, Schleicher & Schuell) using a Phast transfer apparatus

S. Cadel et al. I Molecular and Cellular Endocrinology 110 (1995) 149-160

152

20

40

60

80

100

Fractions No Fig. I. Ion exchange chromatography of the testes enzyme. Ap-B recovered from the Sephadex G-150 column was submitted to a DEAE Trisacryl column and eluted with a gradient from 0 to 150 rnM KCI in Tris-HCl (pH 7.5). Each fraction (3.5 mi) was tested and analysed for its activity towards L-Arg #f-NA in 0.1 M borate buffer (pH 7.4); 150 mM NaCl (0) and for its protein content (A).Arrows indicate the molarities in KCI.

(Pharmacia-LKB). The polyclonal serum was used at a l/2000 final dilution. Antigen-antibody complex was visualised using an anti-rabbit IgG alkaline phosphatase conjugate @omega Corp.). 2.7. Immunohistochemistry Testes of mature rats (Sprague-Dawley) were fixed either in 4% paraformaldehyde in phosphate buffer saline or in Bouin’s liquid for 12-24 h. The tissues were dehydrated in graded ethanol and embedded in paraffin. Section 7 pm thick were mounted on gelatine-coated slides for immunohistochemistry. Both the pre-immune serum taken as the negative control and the Ap-B specific polyclonal antiserum were used at 11800 final dilution in immunoperoxidase staining and at l/400 final dilution in immunofluorescence. Immunoperoxidase staining was performed using the amplification Vectastain ABC-Elite kit (Vector Laboratories) and diaminobenzidine was used for the detection step (Sibony et al., 1994). Tissues were finally counterstained with Harris haematoxylin. Immunofluorescence (Chesneau et al., 1994) was visualised with a rhodamineconjugated goat anti-rabbit antibody. 3. Results 3.1. Purification of aminopeptidase-B The soluble fraction of a crude rat testes homogenate was subjected to a four step purification procedure (see Section 2). In the second step, Ap-B activity was eluted from the DEAE-Trisacryl M column with 80 mM KC1 (Fig. 1) whereas a second peak of aminopeptidase-like activity was eluted by 140 mM KCl. Although this enzyme also cleaved basic amino acids /I-NA, it was distinct

from Ap-B (data not shown). The affinity blue chromatography (third step) was then used to remove serum albumin. Finally, at the fourth step, Ap-B was eluted from the chromatofocusing column at pH 4.9 indicating a p1 value of approximately 4.9. Starting from the postprecipitation extract, about 1OOO-fold purification of ApB was achieved. This number is probably underestimated since contaminating proteases interfered with the Ap-B assay in the early stages of purification. Subsequent gel electrophoresis under denaturating conditions showed only one band with a molecular mass around 72 kDa (Fig. 2; lanes A, B), a value consistent with the apparent molecular mass deduced from the behaviour of the Ap-B activity on Sephadex G- 150 molecular sieve filtration (not shown). The presence of a unique band suggests that the active enzyme is made of a single polypeptide chain. Likewise, when the gel obtained after electrophoresis under non-denaturating conditions was revealed by incubation with L-Arg /!I-NA, a single, but slightly broad band appeared, indicating its association with enzyme activity (Fig. 2; lane D). However, two neighbouring bands were observed on non-denaturating gel electrophoresis (Fig. 2; lane C), suggesting a microheterogeneity of the enzyme. Polyclonal antibodies were raised against the purified testes enzyme preparation. Western blot analysis of crude extracts from rat testes and brain clearly showed that a protein with the same electrophoretic mobility and immunologically related to the testes Ap-B was present in the brain extract (Fig. 2; lanes E, F, G). Following the chromatofocusing step, which produces an electrophoretically pure enzyme (Fig. 2), the activity of Ap-B towards peptides was found drastically reduced. This effect was less marked when L-Arg /!I-NA was used as substrate. To circumvent this problem, tests were car-

kDa

A6

CD “^

E

FG

k0.a

94 )

4 94

4637*

467 443

23&m *

a43020.1

Fig. 2. Polyacrylamide gel electrophoresis of aminopeptidase-B. Samples were run on 8-25% polyacrylamide gels. Lanes A and G, molecular mass markers (phosphorylase B 94 kDa, bovine serum albumin 67 kDa, ovalbumin 43 kDa, carbonic anhydrase 30 kDa, trypsin inhibitor 20.1 kDa, a-lactalbumin 14.4 kDa). Lane B, an aliquot (300 ng of protein recovered from the chromatofocusing purification step) was run in denaturating conditions, the gel was silver stained to visuahse the protein. Lanes C and D, two ahquots (300 ng each) were run in nondenaturing conditions; the gel was silver stained (C) or incubated with the substrate with L-Arg/?-naphthylamide for 1 h at 37°C as described in Section 2 (D). Lanes E and F. two aliquots (2pg) of crude extracts from testes (E) and brain cortex (F) were run in the presence of SDS/& mercaptcethanol and transferred on a nitrocellulose membrane. Immunoblots were performed as described in Section 2.

S. Cadel et al. /Molecular

and Cellular

Table 1 Substrate specificity of testes Ap-B activity Aminoacyl-2-naphthylamide

Relative rate of hydrolysis (%)

L-Arginine L-Lysine L-Histidine Benzoyl-DL-arginine L-Arginyl-arginine L-Alanine L-Leucine L-Valine L-Isoleucine L-Proline L-Phenylalanine DL-Phenylahmine DL-Methionine L-Tryptophan L-Glycine L-Serine L-Tyrosine L-Asparagine L-Glutamine L-Glutamic acid L-Aspartic acid L-Pyroglutamic acid

100 66 0 0 4 2 0 1 2 3 0 5 3 1 1 5 1 0 0 1 0 0

Reactions were carried out in 100 mM sodium borate buffer (pH 7.4), containing 150 mM NaCl. The rate of hydrolysis of L-Arg /?-NA was taken as 100% and all values were calculated relative to this standard.

ried out with the post-DEAE enzyme preparation. However, this preparation appeared to contain a contaminating endoproteolytic activity, which had been tentatively attributed by others to an intrinsic property of Ap-B (Siiderling, 1983; Mantle et al., 1985; McDermott et al., 1988). When a sample of the post-DEAE fraction was subjected to electrophoresis under non-denaturating conditions and stained for activity (see Section 2), this revealed two distinct bands. The faster moving band gave an intense reaction which corresponded to the Ap-B activity recovered after post-chromatofocusing,.whereas the slower migrating one gave a faint reaction. Microsequencing of this slower migrating protein identified unequivocally thimet-oligopeptidase (EC 3.4.24.15), an endoprotease previously isolated from rat brain and testes (Orlowski et al., 1983, 1989). This contaminant could be inhibited by N-[ I-(R,S)-carboxy-3-phenyl propyllAla-Ala-Phe para-aminobenzoate (CFP-AAF-pAb) and N-(phenylethylphosphonyl)- Gly- L-Pro-L-aminohexanoic acid (phosphodiepryl 03), both inhibitors of thimet oligopeptidase (Orlowski et al., 1988; Barelli et al., 1992; gifts of Drs M. Orlowski and F. Checler, respectively). Enzymatic assays were performed in the presence of phosphodiepryl 03 (1 ,uM final) which was shown to have no inhibitory effect on Ap-B activity (see Section 3.5). 3.2. Characterization of aminopeptidase-B activity using L-amino acid t%uxphthylamides as substrates

Assays with the L-amino acid /3-naphthylamides were

Endocrinology

153

110 (199s) 149-160

performed under the conditions described. The activity of Ap-B on a series of L-amino acid /?-naphthylamides is shown in Table 1. Only the arginyl and lysyl derivatives were hydrolysed. 3.3. Effects of halides on enzyme activity The effect of halides was determined by assaying ApB activity from testes in the presence of different concentrations of halides ranging from 0 to 0.5 M (Fig. 3). Addition of 0.2 M NaCl or KC1 resulted in a 3-fold increase in activity compared to the values obtained without halides. The activity was approximately doubled in the presence of 0.2 M LiCl or KBr but was not changed by the addition of NaF, and was completely inhibited with NaI in the same range of salt concentrations (not shown). 3.4. Kinetic constants Kinetic constants for the reaction of Ap-B with t.-Arg #l-NA and L-Lys /3-NA were estimated from Lineweaver Burk plots of l/v against l/[S] and Hanes-Woolf plots of [S]/v against [S] in the substrate concentration range l3OOpM. The following values were obtained: Arg j?-NA, K,,, = 20 PM, V,, = 1 nmol min-l; LYS p-NA, K, = 36 PM, V_ = 0.7 nmol mint. 3.5. Optimal pH and inhibitor profile The optimal pH for Ap-B activity was determined in a pH range of 5.5-8.25 using either sodium phosphate or sodium borate buffers (Fig. 4). Ap-B from rat testes exhibited a broad curve with an optimum around pH 7.2. In the presence of 150 mM NaCl, the pH optimum was shifted to 6.9. Millimolar concentrations of EDTA, EGTA and ophenanthroline significantly inhibited Ap-B activity, strongly suggesting the metallopeptidase character of the

04 0

100

200

3cKl

4lxl

500

Halide cont. mM

Fig. 3. Effect of halides on Ap-B activity. Activity was measured using the standard assay method described in Section 2 with the appropriate concentration of halide ions added to the assay buffer (NaCl (0). KCI (0). LiCl (O), NaF (A), KBr (+)). All activities are expressed relative to the maximum activity of the enzyme observed in the presence of 200 mM NaCl taken as 100%.

154

S. Cudel et al. I Molecular and Cellulur Endocrinology 110 (1995) 149-160

Table 2 Effect of inhibitors on activity of Ap-B from rat testes (see Section 2) Peptidase class

Inhibitors

Final cont.

8 of inhibition Substrate R”-Met-Enk

Metallo

o-Phenanthroline

EDTA

EGTA

hg. 4. pH aependence of Ap-B activity from rat testes. Enzyme activity was observed in the range 5.5-8 with the following buffer system: pH 6-7.5, 50 mM sodium phosphate (0). pH 7.4-8, 0.1 M sodium borate (A). The relative activity is expressed as a percentage of the highest value obtained (taken as 100%). In all cases, the tests were performed with (straight line) and without (dotted line) 150 mM NaCI. Experiments were run three times.

enzyme (Table 2). Ap-B was inhibited by cysteinylproteinase inhibitors such as PCMB, PCMPS, NEM or D’IT suggesting either a direct or an indirect involvement of thiol group(s) in the activity towards L-amino acid pNA. Inhibitors of aminopeptidases such as bestatin and arphamenines A and B, used in the micromolar concentration range, were efficient (Umezawa et al., 1976, 1983), whereas epibestatin, the 2R,3R diastereoisomer of bestatin (Suda et al., 1976), had no effect. Classical inhibitors of either neutral endopeptidase (NEP), angiotensin converting enzyme (ACE), carboxypeptidases or serine-proteases were inefficient. Interestingly, Ap-B activity was inhibited by CFP-AAF-pAb, previously reported to inhibit thimet oligopeptidase (Orlowski et al., 1988). On the other hand, phosphodiepryl 03 which inhibits thimet oligopeptidase and neurolysin (Barelli et al., 1992) had no effect on Ap-B activity. 3.6. Effect of cations on enzyme activity Ap-B was first pre-incubated with o-phenanthroline (500pM) which produced 84% inhibition of activity. The enzyme activity was then measured after the addition of a series of cations ranging from 0.1 to 500,uM (Table 3). While Cu*+ (1OpM) and Mn*+ (1OOpM) had an inhibitory effect, Ca*+, Ni*+ and Mg*+ had almost no effect. Both Zn*+ (1 PM) and Co*+ (100pM) were shown to partially restore the activity. These results suggested that ApB was a zinc metallopeptidase. As may be expected for a zinc metalloprotease, higher Zn*+ concentrations totally inhibited Ap-B activity. 3.7. Determination

of enzyme specificity towards peptides Kinetic measurements of Ap-B activity towards peptide substrates and effect.of inhibitors on the activity towards Argo-Met-enkephalin. From all the peptides tested, the

Cysteine

Dithiothreitol

IOmM I mM I mM I mM I mM

lodoacetamide lodoacetate PCMB PCMPS

Serine

NEP EC 3.4.24. I I

ACE EC 3.4.15.1 Thimet EC 3.4.24.15 24-15/24-I6

500/~M 250yM IOOyM IOmM 5mM I mM IOmM I mM

‘PM I mM

NEM PMSF Aprotinin TPCK Phosphoramidon Thiorphan Kelathorphan Captopril

IFM I mM 500pM I mM 250,uM 5pM SOfcM 50yM 2mM 100pM 25/1M IO/LM

CFP-AAF-pAB

PhosphodieprylO3

‘/‘M 50pM IO,uM ‘PM

Arg-2-NA

88 78 70 89

76 55 27 95 84 34 74 I3

_ _

65 44 _ 56 _ 24 _ _ 0 _ _ _ _ _ _ _ _ 89 6 _ -

loo 77 0 0 100 _ 100 45 0 0 57 0 9 6 50 0 87 72 32 _ 0 0

Leupeptin Puromycin Bacitracin

lOO,uM 50pM IOO,uM

_ _

10 0 0

Carboxy

Pepstatin

100/1M

_

0

Amino

Bestatin

lOO/lM 5OpM

76 64 46

Aspecific

Epibestatin Arphamenine

A

Arphamenine

B

Amastatin

IPM 50pM lOO/lM 50yM IPM IOO/LM 50pM ‘PM 50pM ‘PM

II 4 _ I3 4 65 16

100 68 0 100 _ 100 0

The percentage of inhibition was calculated by taking as reference the amount of substrate conversion by Ap-B in absence of inhibitor (100% activity).

enzyme removed Arg and Lys from the N-terminus, except when a proline residue followed the basic amino

155

S. Cadel et al. /Molecular and Cellular Endocrinology 110 (1995) 149-160

Table 3 Testes Ap-B reactivation by divalent cations Cont.

500 100 10 1 0.1

Cations CoC12 CaCl2

ZnSO4 CuSO4 MgCl2 NiCl2

0 +41 +6.8 0 0

-16 -16 +64 +48 +13

0 0 +5.8 0 0

-16 -16 -16 0 0

0 0 0 0 0

-16 0 0 0 0

MnCl2 -16 -16 0 0 0

The Ap-B activity is expressed as the percentage of recovery, by mference to the value observed in the presence of the cation chelator. Each experiment was run three times. Mean values are given f 5%.

acid, as in substance P, bradykinin and neurotensin-(913). The lowest K, values were obtained with the longest peptides (Table 4). The effects of inhibitors on the activity of Ap-B towards Argo-Met-enkephalin was essentially identical to the effects observed using L-Arg /J-NA as a substrate (Table 2). Ap-B activity was inhibited by divalent cation chelators like o-phenanthroline in the micromolar range whereas a millimolar concentration of EDTA was re-

quired for the same effect. DTT, PCMB and PCMPS were inhibitory at millimolar concentrations confirming that Ap-B activity depends upon one, or multiple, thiol groups. Classical inhibitors of aminopeptidases (bestatin, amastatin) in the micromolar concentration range were highly efficient whereas arphamenines A and B were effective only at much higher concentrations. 3.8. Cellular localization of Ap-B in rat testis Two techniques of immunochemistry were used to determine the cellular localization of Ap-B in the rat testis. To define the step-specific expression of the enzyme during rat spermatogenesis and spermiogenesis, the terminology proposed by Leblond and Clermont (1952) was used (Clermont, 1972; Hess, 1990; Jegou, 1991). According to these authors, a ‘stage’ represents a different combination of germ cell status, including spermatogonia, spermatocytes, and spermatids. In rat testes, 14 stages which are repeated in consecutive order along the length of the seminiferous tubules have been identified. During spermiogenesis, a ‘step’ refers to the maturation and the differentiation of germ cells from round spermatids to spermatozoa. This metamorphosis is a phenomenon which comprises extensive changes in the shape of the

Table 4 Cleavage of peptides by Ap-B from testes enzyme

Vmsx

Vm&L

(pmol/min)

(X 10-7)

111

48

4.3

125

110

8.8

5

4.16

8.3

101

1923

190

173

925

53

14

3.6

2.5

58

25

4.3

36

38

10.5

312

284

Peptides

Argo-Leu-enkephalin Arg&-Tyr-Gly-Gly-Phe-Leu Argo-Met-enkephalin Arg&-Tyr-Gly-Gly-Phe-Met Arg-‘-LysO-somatostatin-14 Arg~-Lys~-Ala-Gly-Cys-Lys-Asn-Phe-Phe-T~-Lys-~-Phe-T~-Ser-Cys [Ala’7-Tyr20j somatostatin-28 (14-20) Ly&Ala-Gly-Ala-Lys-Asn-Tyr-NH2 [Ala17,Ty?oj somatostatin-28 (13-20) Arg&Lysl-Ala-Gly-Ala-Lys-Asn-TyrNH2 Argo-alpha atrial natriuretic factor (l-28) Arg~-Ser-Leu-Arg-~rg-Ser-Ser-Cys-Phe-Gly-Gly-~g-~e-Asp-~-Ile-Gly-Ala-Gln-Ser-Gly-~u-Gly-LysAsn-Ser-Phe-Arg-Tyr Argo-neurokinin A ArgJ-His-Lys-Thr-Asp-Ser-Phe-VaL-Gly-Leu-Met Neurotensin-(8-13) Arg&-Arg-Pro-Tyr-Ile-Leu Neurotensin-(9-13) Arg-Pro-Tyr-Ile-Leu Substance P Arg-Pro-Lys-Pro-Gln-Gln-Phe-Phe-Gly-Leu-Met Bradykinin Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe-Arg P25-(263-272)/HIV peptide Ly&Arg&Trp-Ile-Ile-Leu-Gly-Leu-Asn-Lys

Not cleaved Not cleaved Not cleaved 11

Peptide substrates (1Opg in borate buffer, pH 7.5, 150 mM NaCl) were incubated in the presence of Ap-B as described in Section 2. The position of the hydrolysed peptide bond, indicated by an arrow, was determined by HPLC separation of the generated product from the intact substrate. Further identification of a product was made by reference to standards or/and by amino acid composition. Absence of arrow indicates that the substrate was recovered intact at the end of the incubation period. K,,, @M) and V,, (pmol min-l) were deduced from either Lineweaver-Burk or/and Hanes-Woolf plots by a linear regression program. The experiments were run two or three times. Mean values are given + 10%.

156

S. Cudel et al. I Moleculur

and Cellular

Endocrinology

110 (1995) 149-160

Fig. 5. Cellular localization of aminopeptidase-B in the rat testes using immunoperoxidase reaction. Shown are 7+m thick section of rat testes immunocytochemically stained with antibodies labelled with peroxidase. Note the arrowheads indicating labelling; (ES) etongated spermatids; (RB) residual bodies; (RS) round spermatids; (SG) spermatogonia; (SP) spermatocytes, pachytene stage; (yS) young spermatids. (A) At low magnification (X 1 IO), the specific labelling is distributed within seminiferous tubules near the lumen. The different stages of spermatogenesis were indicated. (B) Magnification x400; no labelling is observed during stages II-III. (C) x400, immunoreactivity is restricted to the cytoplasm of late spermatids during stage VVI. (D) X400, Ap-B is localized, during stage VII, essentially as dots in the cytoplasm area which gave rise to the residual bodies. (E) x400, at stages VIII-IX the residual bodies were strongly labelled in restricted areas, (F) x400, no immunoreaction could be detected in the elongated spermatids at stages XII-XIII.

S. Cadel et al. I Molecular and Cellular Endocrinology I10 (1995) 149-160

157

Fig. 6. Cellular localization of aminopeptidase-B in the rat testes using immunofluorescence reaction. Shown are 7-pm thick section of rat testes immunocytochemically labelled with rhodamine-conjugated antibodies. Note the arrowheads indicating labelling and (RB) residual bodies. (A) At low magnification (X 110). the Ap-B was located in the adluminal region of the tubules. The different stages of spermatogenesis are indicated. (B) x250, stages VII-VIII, labelling is observed only in residual bodies around the lumen and in the cytoplasm of the Sertoli cells.

spermatid, cell type.

within the cytoplasm

and in the nucleus of this

Immunoperoxidase and immunojluorescence analysis. At low magnification (Fig. 5A), a weak intensity labelling appeared exclusively in the cells near the lumen of the seminiferous tubules in the rat testes, indicating a stage dependence. The immunofluorescence reaction gave similar patterns to those obtained by peroxidase staining. Despite an important background due to the high antibody concentration, Ap-B was found to be localized only in the adluminal region of the tubules (Fig. 6A). No labelling was observed in the Leydig and the Sertoli cells. Ap-B could not be visualised during either stages II and III of spermatogenesis (Fig. 5B). An immunoreactivity restricted to the late spermatids (maturation phase) and which appeared as a small spot located in the cytoplasm of these cells was observed at stages V-VI (Fig. 5C). This signal was not associated with the future head of the spermatozoon, which was already preformed with specific morphological characteristics. Moreover, the immunoperoxidase reaction did not reveal the presence of Ap-B in the flagella of spermatozoa. During stage VII, Ap-B was essentially localized in the caudal cytoplasm area from which was originated the residual bodies in the late spermatids (Fig. 5D; Clermont, 1972). At a later stage (VIII-IX), there was no immunoreactivity detectable in the young spermatids during the beginning of the elongation phase (Fig. 5E), however the residual bodies were now stained in restricted areas. Fusion of the structures intensified the signal probably as a result of Ap-B concentration. Similar results were obtained by immunofluorescence staining. Labelling was observed as a ring of spots around the lumen of the tubules at stages VIII (Fig. 6B). As indicated by immunop-

eroxidase analysis, strongly positive residual bodies deeply inserted in seminiferous epithelium cell cytoplasm could be detected (Fig. 6B). The labelling of the residual bodies allowed us to follow the different phases of their phagocytosis by the Sertoli cells. When residual bodies were deeply inserted in the Sertoli cell cytoplasm, the labelling became slightly diffuse and disappeared during the phagocytosis process. No immunostaining could be detected in the elongated spermatids at stages XII-XIII thus confirming the precise localization of Ap-B during other stages (Fig. 5F). 4. Discussion Endoproteolytic activation of biosynthetic peptide precursors most often occurs at the level of basic amino acid arrangements. The generated intermediates contain unwanted basic residues which must be removed to produce the active peptides. It is now accepted that carboxypeptidases E and B are implicated in processing of Arg and Lys residues generated by endoproteolytic cleavage of precursor molecules on the C-terminus of basic stretches. In contrast, the possible role of aminopeptidase(s) of the B-type in performing a similar processing at the Nterminus of precursor fragments remains unadressed. No clear studies on aminopeptidase-B-like activities on peptide substrates had been performed. To date, Ap-B activity had essentially been studied on naphthylamides and on di- and tri-peptides (Soderling, 1982; Mantle et al, 1985). Gainer et al. (1984) have identified a membrane form of an Ap-B in bovine pituitaries which is able to remove the Arg moieties from Argo-Met-enkephalin. The present report provides experimental evidence in favour of an Ap-B-like enzyme which exhibits a strict selectivity for Arg and Lys residues at the N-terminus of

1.58

S. Cudel et ul. ! Moleculur

und Cellulur

either naphthylamides or peptide substrates of various length. Previous preliminary work from this laboratory on aminopeptidase-B-like activity was performed on rat brain extracts (Glushankof et al., 1987; Gomez et al., 1988). We report here on the purification of an Ap-B of the rat testes and its existence in the rat cortex. A partial characterization of this brain Ap-B showed similar biochemical properties towards Arg B-NA substrates, indicating in addition to the immunological data, that this protein was probably the same one as in the testis (data not shown). The complete purification of Ap-B led to a unique band on denaturating gel electrophoresis and to a doublet under non-denaturating conditions. Both bands seemed to contain enzyme activity. Although it is possible that the two proteins correspond to isoforms of Ap-B, the exact interrelationship between the two bands has not been determined. A limited proteolysis of the larger enzyme form cannot yet be ruled out. The molecular mass of the purified enzyme was determined as 72 kDa, a value in keeping with the report by Mantle et al. (1985) but different from those of other groups (Kawata et al., 1980; Soderling, 1982). A number of authors have described Ap-B as having an intrinsic endoproteolytic activity towards peptides (substance P, bradykinin; Soderling, 1983; Mantle et al.. 1985; McDermott et al., 1988). However, the complete purification of Ap-B described in this study has allowed the unequivocal identification of this contaminating activity as thimet oligopeptidase as shown by sequencing of endolysine-C fragments. This conclusion is also supported by a quantitative and selective inhibition of this endopeptidase activity by phosphodiepryl 03 (Barelli et al., 1992). The enzyme is able to remove one or two basic amino acid extension(s) from a number of peptide substrates ranging in length from 5 to 29 amino acids, except when the basic residue is followed by a proline. This is the case for a number of bioactive peptides (e.g. substance P) and this may be a general mechanism whereby these peptides protect themselves against the action of Ap-B like enzymes. In the case of dibasic substrates, the monobasic intermediary form was hardly detected, suggesting that the catalysis is very rapid. Restoration of enzyme activity by Zn2+ or Co*+ after inhibition by divalent chelators strongly suggests that, like other aminopeptidases, the B type is a metalloenzyme. Noticeably a higher Zn2+ concentration ( IOOpM) had the opposite effect and led to inhibition of the enzyme, a feature already observed in other Zn2+ dependent enzymes such as leukotriene-A4 hydrolase and carboxypeptidase A (Larsen and Auld, 1991; Orning and Fitzpatrick, 1992; Wetterholm et al., 1994). In addition, Ap-B was shown to be inhibited by CFP-AAF-pAB, whose carboxyl group of the carboxy-3-phenylpropyl moieties has been shown to coordinate the catalytic zinc

Endocrmolo~y

II0

(199.5) 149-160

atom of thimet oligopeptidase. Interestingly, TPCK, a known inhibitor of chymotrypsin showed some inhibitory activity at high concentration (50% inhibition at 250pM). This behaviour has already been reported for other Zn*+peptidases, such as mitochondrial intermediate peptidase (Kalousek et al., 1992) and N-Arg dibasic convertase (Chesneau et al., 1994; Pierotti et al., 1994) and could be attributed to some reactivity of the inhibitor with the histidine residues of the active site (Schoellmann and Shaw, 1963). Cloning and sequencing of the cDNA for Ap-B will reveal whether the enzyme has the zinc binding consensus sequence (H E X X H or H X X E H) or if like leucine-aminopeptidase, another Zn2+ metalloprotease, the enzyme does not possess this signature (Jongeneel et al., 1989; Taylor, 1983). One or more SH groups seem to be implicated in the activity, since several sulfhydryl reagents abolished Ap-B activity. Enzyme activity was significantly activated by Clions (NaCl and KCI) when tested towards naphthylamide substrates. In contrast the effect of NaCl was more ambiguous towards peptide substrates. Indeed, depending upon the peptide, either an activation or an inhibition was observed and, in some cases, no effect was detected (data not shown). Similar results have been reported about the differential effects of Cl- ions on angiotensin converting enzyme activity (ACE, EC 3.4.15.1; Shapiro et al., 1983). Ap-B shows a broad range of activity from pH 6 to 7.5 with an optimum around neutrality, it can be anticipated that the enzyme active site is adapted to catalysis in various cell compartments. Interestingly, cellular localization of Ap-B in rat testes appears clearly to be both stage- and step-dependent during spermatogenesis. At late stages of these differentiation processes, the protein seems to be confined to a restricted area of the late spermatid cytoplasm during the maturation phase. To our knowledge, no exclusive and specific antigen labelling of this particular structure has been reported in the literature. Additionally, Ap-B labelling could be defined as being linked to a membranederived structure due to the absence of diffuse labelling at stages V-VI. This hypothesis was further supported by a set of observations. Firstly, Ap-B was identified as a strong spotted signal in residual bodies. Thereafter, there was a decrease of immuno-labelling during the phagocytosis process, in the Sertoli cells, leading to lysosomal degradation of the residual bodies (Morales et al., 1985). Secondly, in addition to the soluble 72 kDa Ap-B, described here, a membrane associated form solubilised by Triton X- 100 and exhibiting a similar cleavage specificity was recovered in the pellet during the purification procedure of the testes enzyme (data not shown). Together, these data support the hypothesis of a membrane associated Ap-B protein. Further studies will determine if the soluble form of Ap-B is related to the membrane form by truncation of an anchoring domain, a maturation process observed previously in several cases (Germain et al.,

S. Cadel et al. I Molecular and Cellular Endocrinology 110 (1995) 149-160

1992) or if it is able to associate with the membrane by interaction of an amphiphilic helix (Fricker et al., 1990; Mitra et al., 1994) or by other mechanisms. Although the nature of this intracellular membrane component could not be easily identified, some hypothesis can be proposed. The immunolocalisation of angiotensin converting enzyme (ACE) (Sibony et al., 1994) and rab-6 (Anthony et al., 1992), both selective markers of the Golgi apparatus did not correspond to the Ap-B pattern, suggesting that the exopeptidase is not significantly represented or concentrated in this compartment. Therefore, the radial body and the annulate lamellae could be proposed as likely candidates (Clermont and Rambourg, 1978). Further studies by electron microscopy might clear up this important issue. The present data support the involvement of such an enzyme at specific steps of spermatid differentiation. The enzyme clearly exhibits a strict selectivity for basic amino termini of peptide substrates indicating that it might intervene at the final steps of propeptide proteolytic maturation. Generation of such basic amino-terminal moieties implies previous endoproteolytic cleavage on the Nterminus of basic residues in singlets or higher order arrangements. Although it has long been thought that endoproteolysis would take place preliminarily on the Cterminus of basic residues, a number of more recent observations indicate that in some cases this could occur on the other side of the basic signals (Peng Loh et al., 1985; Glushankof et al, 1987; Azaryan et Hook, 1994; Chesneau et al., 1994; Pierotti et al., 1994). This is now demonstrated both by the identification of endoproteases performing Xaa-Lys and/or Xaa-Arg cleavages and by the isolation of precursor derived peptide fragments bearing basic amino-terminal residues (Peng Loh et al., 1985; Inagami, 1989). Although the endogeneous physiological substrate(s) of Ap-B in male gonads have not been identified it is tempting to propose that this enzyme intervenes, in tandem with an endoprotease, in the processing steps leading to mature peptide messengers involved in the control of spermatid differentiation.

Acknowledgements This work was supported in part by funds from the Centre National de la Recherche Scientifique (CNRS, URA 1682), the Universite Pierre et Marie Curie and from Rhone-Poulenc-Rorer (Programme Bioavenir). S.C. and A.R.P. were respectively recipients of a fellowship from Association de la Recherche contre le Cancer and the Fond&ion pour la Recherche Medicale. We wish to thank F. Checler (Sophia Antipolis, France), M. Orlowski (New York, USA), for their gifts of inhibitors used in this study, J. D’Alayer (Pasteur, France) for partial sequencing of the enzyme, and Drs B. JCgou, A. Prat and M. Draoui for helpful advice in the preparation of the manuscript.

159

References Anthony, C., Cibert, C., Geraud, G., Santa Maria, A., Mare, B., Mayau, V. and Goud, B. (1992) J. Cell Sci. 103,785-796. Azaryan, A.V. and Hook, V.Y.H. (1994) FEBS Len. 341, 197-202. Bamlli, H, Dive, V., Yiotakis, A., Vincent, J.P. and Checler F. (1992) Biochem. 1.287,621-625. Bourdais, J. and Cohen, P. (1991) Ann. Endocrinol. 52.339-347. Bradford, M.M., (1976) Anal. Biochem. 72.248-254. Castro, M.G., Birch, N.P. and Peng Loh, Y. (1989), J. Neurochem. 52, 1619-1628. Chesneau, V., Pierotti, A.R., Barr& N., Cmminon, C., Tougard, C. and Cohen, P. (1994) J. Biol. Chem. 269,2056-2061. Clermont, Y. (1972) Physiol. Rev. 52, 198-236. Clermont, Y. and Rambourg A.. (1978) Am. J. Anat. 151. 191-212. Cohen, P. (1987) Biochimie 69, 87-89. Darby, N.J. and Smyth., D.J. (1990) Biosci. Rep. 10, 1-13. Devi, L. (1991) FEBS L&t. 2, 189-194. Docherty, K. and Steiner, D.F. (1982) Annu. Rev. Physiol. 44, 625638. Flares, M., Aristoy, MC. and Toldra, F. (1993) Biochimie 75, 861867. Fricker, L.D. (1988) Annu. Rev. Physiol. 50, 309-321. Fricker, L.D., Das, B. and Angeletti, H. (1990) J. Biol. Chem. 265, 2476-2482. Frobert, Y. and Grassi, J. (1992) Methods Mol. Biol. 10,65-78. Gainer, H., Russel, J.T. and Peng Lob, Y. (1984) FEBS Lett 175, 135139. Germain, D., Vemet. T., Boileau, G. and Thomas, D.Y. (1992) Eur. J. B&hem. 204.121-126. Gluschankof, P. and Cohen, P. (1987) Neurochem. Res. 12.951-958. Gluschankof, P., Gomez, S., Morel, A. and Cohen, P. (1987) 1. Biol. Chem. 262.9615-9620. Gomez, S., Gluschankof, P., Lepage, A. and Cohen, P. (1988) Proc. Natl. Acad. Sci. USA 85,5468-5472. Hess, R.A. (1990) Biol. Reprod. 43.5255542. Inagami, T. (1989) J. Biol. Chem. 264,3043-3046. Jegou, B. (1991) Ann. N. Y. Acad. Sci. 637,340-353. Jongeneel, C.V., Bouvier, J. and Bairoch, A. (1989) FEBS I&t. 242, 211-214. Kalousek, F., Tsaya, G. and Rosenberg, L.E. (1992) EMBO J. 11, 2803-2809. Kawata, S., Takayama, S., Ninomiya, K. and Makisumi, S. (1980) J. Biochem. 88, 1601-1605. Kilpatrick, D.L., Borland, K. and Jin, D.F. (1987) Proc. Natl. Acad. Sci. USA 84,5695-5699. Larsen, KS. and Auld D.S. (1991) Biochemistry 30.2613-2618. Leblond, C.P. and Clermont, Y. (1952) Ann. N. Y. Acad. Sci. 55, 548573. Mantle, D., Lauffart, B., McDermott, J. R., Kidd, A.M. and Pennington, R.J.T. (1985) Eur. J. Biochem. 147,307-312. McDermott, J.R., Mantle, D., Lauffart, B., Gibson, A.M. and Biggins, J.A. (1988) J. Neurochem. 50, 176-182. Mitra, A., Song, L. and Fricker, L.D. (1994) J. Biol. Chem. 269, 19876-19881. Morales, C., Clermont, Y. and Hermo, L. (1985) Am. J. Anat. 173, 203-217. Orlowski, M., Michaud, C. and Chu, G.C. (1983) Eur. J. Biochem. 135, 81-88. Grlowski, M, Michaud, C. and Mohneaux, C.J. (1988) Biochemistry 27,597-602. Orlowski, S., Reznik, J., Ayala, J. and Pierotti, A.R. (1989) Biochem. J. 261.951-958. Oming, L. and Fitzpatrick F.A. ( 1992) Biochemistry 3 1.4218-4223. Peng Loh, Y., Parish, C.D. and Tujeta, R. (1985) J. Biol. Chem. 260, 7194-7205. Pierotti. A.R., Prat, A., Chesneau, V.. Gaudoux, F., Leseney, A.M.,

160

S. Code1 et al.

I Molecular

and Cellular Endocrinology

Foulon, T. and Cohen, P. (1994) Proc. Natl. Acad. Sci. USA 91, 60786082. Schoellmann, G. and Shaw, E. (1963) Biochemistry 2,252-255. Shapiro, R., Holmquist, B. and Riordan, J.F. (1983) Biochemistry 22, 3850-3857. Sibony, M., Segretain, D. and Gasc, J.M. (1994) Biol. Reprod. 50, 1015-1026. SBderling, E. (1982). Arch. Biochem. Biophys. 216, 105-I 15. Siiderling, E. (1983) Arch. Biochem. Biophys. 220, l-10. Suda, H., Aoyagi, T., Takeuchi, T. and Umezawa, H. (1976) Arch. Biochem. Biophys. 111, 196-200.

I10 (I 995) 149-I

60

Taylor, A. (1983) Trends Biochem. Sci. 18, 167-172. Umezawa, H., Aoyaki, T., Suds, H, Hamada, M. and Takeuchi, T. (1976) J. Antibiot. 29,97-99. Umezawa, H., Aoyagi, T., Okuchi, S., Okuyama, A., Suds, H., Takita, T., Hamada, M. and Takeuchi, T. (1983) J. Antibiot. 36, 15721575. Vanha-Pert&da, T. (1973) 1. Reprod. Fertil. 32,33-44. Wetterholm, A.. Macchia, L. and Haeggstrom, J.Z. (1994) Arch. Biothem. Biophys. 311.263-271. Yoshikawa, K. and Aisawa, T. (1988) Biochem. Biophys. Res. Commutt. 151.664-671.