Characterization of an endoprotease from rat small intestinal mucosal ...

3 downloads 0 Views 741KB Size Report
and the Association pour la Recherche sur le Cancer. The costs of publication ...... moms, Hormones and the Fragments (Martinez, J., ed) Ellis. Horwood Ltd., in ...
THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1989 by The American Society for Biochemistry and Molecular Biology, Inc

Vol. 264, No. 8, IBsue of March 15, pp. 4460-4465,1989 Printed in U.S.A.

Characterization of an Endoprotease from Rat Small Intestinal Mucosal Secretory Granules Which GeneratesSomatostatin-28 from Prosomatostatin by Cleavage after a Single Arginine Residue* (Received for publication, November 2, 1988)

Margery C. BeinfeldS, Julie Bourdais, Paul Kuks, Alain Morel, and PaulCohen From the Groupe de Neurobiochimie Celluluireet Moliculuire, Universiti Pierre et Marie Curie, Unit6 de Recherches Associi 003 au Centre National de la Recherche Scientifique, 96 Bld Raspail, 75006 Paris, France

We have extracted, characterized, and partially pu- ages were completely blocked by preincubation with rified an enzyme from secretory granules from rat aprotinin. small intestinal mucosa which cleaves a synthetic proAlthough further work is required to clarify the somatostatin substrate on the carboxyl sideof a single physiological role of this enzyme, it appears, in view arginine residue. This substrateLeu-Gln-Arg-Ser-Ala- of its catalytic properties, this endoprotease could be Asn-Ser-NH2 containsthe monobasic site at which involved in the conversion of prosomatostatin to somammalian prosomatostatin is cleaved in vivo to gen- matostatin-28 in intestine mucosal secretory cells. erate somatostatin-28. This activity was released from the granules by osmotic shock followed by extraction with 500 mM KCl. The enzyme had amolecular weight Numerous studies indicate that biologically active peptides of about 55,000, a pH optimum of about 7.5, and a K , for the synthetic substrate of 20 PM. It was partially are synthesized initially as a prohormone containing single or inhibited by diisopropyl fluorophosphate, phenylmeth- multiple copies of the final active peptides. These active anesulfonyl fluoride, iodoacetate, soybean trypsin in- sequences are frequently (although not exclusively) flanked hibitor, and EDTA. It was also very sensitiveto apro- by pairs of basic amino acids which are thought to be cleaved tinin (complete inhibition at 25 pg/ml) but was not first by endopeptidases then furtherprocessed by amino- and inhibited by bestatin, pepstatin, or p-chloromercuri- carboxylpeptidases and, in some cases, by special amidating benzoate. enzymes to yield the final product (for reviews see Refs. 1-4). This endoprotease was unable to cleave three small These cleavages arethought to occur either in the Golgi trypsin and kallikrein substrates (Nu-benzoyl-L-argi- apparatus and/or in secretory vesicles (1, 2, 5). Numerous nine ethyl ester, N"-benzoyl-DL-argininep-nitroani- studies have focused on the identification and purification of lide, and Nu-benzoyl-L-arginine 7-amido-4-methyl- these dibasic cleavage enzymes (6-12), onthe associated coumarin). amino- and carboxypeptidases (9-15), and on the amidating It was unable to cleave either the Arg-Asp bond in enzyme (16, 17). CCK 12 or the Arg-Glu and Arg-Met bonds of synthetic There are a number of important prohormone cleavages peptides corresponding to sequences of anglerfish prowhich occur at single basic residues (mainly arginine) in the upstream from the sosomatostatin I1 situated processing of such peptides as cholecystokinin, atrial natrimatostatin-28 domain. These observationstogether suggest that adjacent amino acids play a role in deter- uretic factor, pancreatic polypeptide, dynorphin A and B, mining the conformational specificity of the monobasic growth hormone releasing hormone, peptide histidine methionine amide, andsomatostatin (reviewed in Ref. 18). An cleavage. This soluble enzyme was also able to cleave three enzymatic activity has been described which willperform the synthetic substrates containingdibasic residues (Arg- monobasic cleavage of cholecystokinin (19), although this Lys or Lys-Arg) on the carboxyl sideof the arginine, activity was not characterized in detail. An enzymatic activity although it did so less rapidly than at the monobasic in atrial secretory granules has been reported which cleaves the Arg-Ser bond in pro-atrial natriuretic factor (20, 21). cleavage sites. When incubated with partially purified prosomatoProsomatostatin offers a particularly suitable model since statin from anglerfish pancreas, significant quantities the 12.5-kDa precursor in mammals, which is encoded by a of somatostatin-28 I1 were produced. All these cleav- single gene (reviewed in Ref. 22) contains both dibasic and monobasic cleavage loci. The first is believed to operate in * This work wassupported in part by National Institutes of Health Grant NS18667 (to M. C. B.) and by funds from the Universitk P. et the production of S-14,' whilethe latteris necessarily involved M. Curie, the Centre National de la Recherche Scientifique (U.R.A. in the production of S-28 (22). Although a candidate for the 003), the Institut National de la Santk et de la Recherche Medicale recognition of the dibasic cleavage site was recently proposed (CRE 87-4002),the Fondation pour la Recherche Medicale Franfaise, and the Association pour la Recherche sur le Cancer. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18U.S.C.Section 1734 solelyto indicate this fact. $ Recipient of a Senior International Fellowship (F06 TW01345) from the Fogarty International Center. This work was performed while M. C. B. was on sabbatical leave from the Dept. of Pharmacology, St. Louis, University Medical Center, 1402 South Grant Blvd., St. Louis, MO 63104.

'The abbreviations used are: S-14, somatostatin-14; S-28, SOmatostatin-2b; AF ProSII, anglerfish prosomatostatin II; AP, aprotinin; BAEE, Ne-benzoyl-L-arginine ethyl ester; BAPNA, N"-benzoyl-DL-arginine p-nitroanilide; CCK, cholecystokinin; CRF, ovine corticotropin releasing factor; DABITC, 4-dimethylaminoazobenzene 4'-isothiocyanate; HPLC, high performance liquid chromatography; Np, neurophysin I; OT, oxytocin; pCMB, p-chloromercuribenzoate; PMSF, phenylmethanesulfonyl fluoride; Pro-OT/Np, pro-oxytocinneurophysin I; Pros, mammalian prosomatostatin; STI, soybean trypsin inhibitor; TLC, thin layer chromatography.

4460

Somatostatin-28 Generating Endoprotease (9,10)nothing at the presenttimeisknownaboutthe endoprotease possibly involved in the recognition of the single arginine residue at position -1 on the NH2-terminal side of somatostatin-28 in mammals. Mackin and Noe (23) have recently reported extraction and limited purification of two separate enzymatic activities operating at acidic pH from anglerfish pancreaticsecretory granules which cleave anglerfish prosomatostatin at either monobasic or dibasic sites. In the present study rat intestinal mucosa which contains more somatostatin-28 than 14 (24) was chosenas a tissue source forthe isolation of somatostatin28 generating activity. The strategy utilized was inspired by the methodology previously used in this laboratory in the search for dibasic specific endoproteases (9-12). Therefore a sensitive assay was developed which utilizeda small peptide substrate containingthe mammalian monobasic cleavagesite Arg-Ser and reproducing the prosomatostatinstructure around this locus. The ability of a partially purified, soluble, enzymatic activity to cleave this peptide and a number of other small synthetic substrates with monobasic and dibasic cleavage sites and to cleave anglerfish pancreatic prosomatostatin wasdetermined. MATERIALS ANDMETHODS

Granule Preparation and Enzyme Extraction-The small intestines were removed rapidly following decapitation from Wistar male rats (300 g), rinsed with cold isotonic saline, and the mucosa removed from the muscular tissue by scraping with a glass slide on a cooled glass plate. The primary subcellular fractionation technique follows that used by DeRobertis et al. (25). The mucosa was homogenizedin 10 ml/g tissue weight of0.25 M sucrose containing 10 mM sodium phosphate, pH 7.4, and centrifuged at 1,500 X g for 10 min. The supernatant was retained and the pellet wasrehomogenized in 5 volumes of homogenization buffer and recentrifuged. The pooled supernatants were centrifuged at 7,000 X g for 10 min and the pellet (P2) was saved; then the supernatant was centrifuged at 10,000 X g for 30 min, and the pellet (Pa) and supernatant were saved. The discontinuous sucrose density centrifugation is adapted from that of Noe et al. (26). P2 and P3 were separately homogenized in 20 ml of 1.5 M sucrose in 10 mM sodium phosphate and layered over a cushion of 5 ml of 2 M sucrose in phosphate. A solution of 0.25 M sucrose was layered on top to fill the tube. The tubes were centrifuged a t 100,000 X g for 1 h. After centrifugation the material at the two interfaces was recovered separately by aspiration with a pipette, as was the cloudy solution in between the two interfaces. The interfaces were separately homogenized in 7 ml of 0.25 M sucrose and centrifuged at 100,000 X g for 1 h to remove sucrose and thepellets were recovered. They were lysed bythree cycles of freezing and thawing in phosphate buffer followed by addition ofKC1 at a final concentration of500 mM. The supernatantwas recovered by centrifugation at 100,000 X g for 1 h. A Beckman J21-B centrifuge equipped with a JA-20 rotor and a Beckman L5-65 ultracentrifuge with 50 Ti and 60 Ti rotors were utilized. The biochemical characterization of granule fractions was done in the following way: in the initial granule fraction preparations, samples of all the subcellular fractions were saved and assayed for acid phosphatase activity (as an indication of lysosomal contamination), cholecystokinin immunoreactivity (as anindication of the enrichment of secretory material) and for protein. The phosphatase assay utilized p-nitrophenol as substrate (27), the CCK radioimmunoassay utilized standard methods (28), and protein assay was by the method of Bradford (29). Enzyme Purification and Assays-The soluble enzyme was partially purified by gel filtration on a 1 X 30-cm column of Sephadex G-150 eluted with 10 mM phosphate, pH 7.3, containing5 mM dithioerythritol at a flow rate of 6 ml/h in the cold room. The preliminary test for activity utilized the substrate ProS(6268)NHz (see Table I). Following incubation, the cleavedpeptides were derivatized with DABITC according to themethod of Chang (30) and separated by HPLC. Granule extracts (usually 1 pl) or for more purified enzyme fractions (5-20 pl) were incubated with 1.5 pg (2 nmol) of ProS(62-68)NH2 in 10 pl of water for 1 h at room temperature. The reaction was stopped by addition of 2 volumes of DABITC

4461

(2 mg/ml pyridine). After derivatization, the aqueous phase of each assay was dried in a Speed-Vac concentrator and the residue redissolved in 50 pl of 0.5% pyridine acetate buffer at pH 5.3, and onetenth was injected into an HPLC(LDC/Milton Roy) equipped with a variable wavelength detector set at 436 nm. The S5ODS2 column (Prolabo, Paris, France) (4.6 X 250 mm) was eluted at a flow rate of 1ml/min with 0.5% pyridine acetate buffer at pH5.3 (A) containing 30% (v/v) acetonitrile ( B ) .The gradient was from 70% A30% acetonitrile (B) to 50% A50% B in 20 min, remained at 50% A50% B for another 5 min, and then returned to initial conditions in 2 min. Standards for the cleavage reaction were produced with 1ng of bovine pancreatic trypsin (Type XI, Sigma) added to 1.5 figof dried ProS(6268)NHz in 20 pl of0.1 M ammonium acetate buffer at pH 8. The reaction was stopped by addition of 50% pyridine 1 h later. ForCCK 12, we used enterokinase (Sigma) as trypsin cleaves CCK 12 very inefficiently. These cleaved standards were derivatized like the enzyme samples. The identity of all of the cleaved standards andsome of the cleavage products formed with the enzymes was verified by amino-terminal analysis (31) and in some cases amino acid analysis (32). The same methodology was used for the alanine derivative [Ala65]Pr~S(62-68)NH2 and for CCK 12, AF ProSII(l-lO)NH2, and AF ProSII(64-72). and ProFor the dibasic substrates [Ala",Tyr20]S-28(10-20)NH~ OT/Np(l-20) the reaction products were analyzed as described in (9, 11,12) after 2- and 3-h incubation, respectively, at room temperature with 10 pl of the enzyme preparation. After a 2-h incubation with CRF, the reaction products were analyzed by HPLC on a S50DS2 column (4.6 X 250 mm), eluted with 0.8% trifluoroacetic acid (flow rate 1 ml/min). Acetonitrile was raised from 0 to 40% in 20 min, maintained from 40 to 60% (depending on the elution of the substrates) for 5 min, and returned to initial conditions in 2 min. The peptides were detected a t 220 nm. Standard techniques were used, with colorimetric and fluorescent substrates. With BAEE the cleavage product was monitored at 253 nm (33); with BAPNA, at 410 nm (34). For Na-~-arginine-7-amido4-methylcoumarin fluorescence was excited at 380 nm and detected at 460 nm (35). All of these substrateswere readily cleaved by trypsin. To test anglerfish prosomatostatin cleaving activity the purified enzyme preparation was incubated with 20 ng of anglerfish prohormone previously purified on Sephadex G-75 as described in Ref. 36. After 1 h a t room temperature, the reaction was stopped by addition of50% acetic acid, and the reaction products were separated on a Sephadex G-50 column (1.2 X 90 cm) run in 10% acetic acid (v/v) at a flow rate of 6 ml/h. One-ml fractions were collected and assayed for somatostatin-like immunoreactivity with the somatostatin radioimmunoassay (37). The immunoreactive cleavage products were further characterized by HPLC and radioimmunoassay as described in (36). Peptide Substrates and Fragments-All the peptides used in this work were synthesized by a modification of the solid-phase method (38) as described in detail in (39). They were analyzed by TLC, HPLC, amino acid composition, and fast atom bombardment mass spectroscopy according to procedures previously described (9, 12, 39, 40). RESULTS

The sequences of all the substrates used in the study are shown inTable I. Characterization of Secretory Granules-CCK was selected over somatostatin as a marker for the characterization of both CCK (41) and sosecretoryvesicles(eventhough matostatin (24) are present in mucosal secretory cells) because the content of CCK is at least five times higher than somatostatin. The primary subcellular fractions P2 and P3, known respectively as crude mitochondrial and crude microsomal, were both about 3-fold enriched in CCK (pmol/mg total protein) as compared to homogenates. When the upper (between 0.25 and 1.5 M sucrose) and lower (between 1.5 and 2.0 M sucrose) interfacesof the discontinuous sucrose density gradients were assayed for CCK immunoreactivity, the CCK i n the upper interface was undetectable while the lower interface had about a 2-fold enrichment in CCK as compared to either P 2 or P3;6-fold as compared to homogenate. The cloudy solution between the interfaces had little or n o CCK

Somatostatin-28 Generating Endoprotease

4462

TABLE I Amino acid sequence of peptide substrates and the fragments produced bythe endopeptidase The enzyme wasincubated with 2 nmol of the monobasic substrates for 1h and 2 nmol of the dibasic substrates for 2 or 3 h at room temperature (see Fig. 4 for specific details). With monobasic substrates the reaction was terminated by addition of 2 mg/ml DABITC in pyridine, and the reaction products were analyzed by HPLC following derivatization and extraction. Withdibasic substrates the reaction was stopped by addition of acetic acid to a final concentration of lo%, and the reaction products were analyzed by HPLC optical detection at 220 nm without derivatization. The arrows indicate the peptide bond hydrolyzed by the enzyme. MONOBASIC SUBSTRATES

1 Leu-Gln-EIq-Ala-Ala-Asn-Ser 1

ProS(62-68)NH2: Leu-Gln-&g-Ser-Ala-Asn-Ser [Ala651Pros(62-68)NH2:

NH2 NH2

AF PrOSII(1-10)NH2

: Gln-Leu-Asp-&g-Glu-Gln-Ser-Asp-Asn-Gln

AF ProSII(64-72)

: Ala-Thr-Glu-Gly-&g-Met-Asn-Leu-Glu

CCX 12

: Ile-Ser-Asp-&-g-Asp-Tyr-Met-Gly-Trp-Met-Asp-Phe

DIBASIC S

U

B

S

T

NH2

W

NH2

W

1

[Ala17,Tyr20]S-28 (10-20)NH2 : Pro-Arg-G1u-Ar~-LvS-A1a-Gly-A1a-Lys-Asn-Tyr NH2

[Ala17,Tyr20]s-2~(15-20)NH2 :

Ala-Gly-Ala-Lys-Am-Tyr NH2 m-Ala-Gly-Ala-Lys-Asn-Tyr

[Ala17,TYr201S-28(14-20)NH2 : [Ala17,Tyr20]S-28(13-20)NH2 :

NH2

Arq-Lvs-Ala-Gly-Ala-Lys-Asn-Tyr NH2

1

[TYr321 CRF(31-41)NH2

: Ala-Tyr-Ser-Am-Am-Lvs-Leu-Leu-Asp-Ile-Ala

( T ~ r CRF(31-36) ~ ~ l

: Ala-Tyr-Ser-ASn-ArQ-Lvs

( T ~ r CRF(31-35) ~ ~ l

: Ala-Tyr-Ser-Asn-m

( T ~ r CRF(31-34) ~ ~ l

: Ala-Tyr-Ser-Asn

NH2

1

PrO-OT/Np(l-20) : C~s-Tyr-lle-Gln-ASn-C~s-Pro-Leu-Gly-Gly-~-Ala-Val-Leu-Asp-Leu-A~p-val-Ar~ Pro-OT/Np(l-l2) : CJs-Tyr-Ile-Gln-Am-C;s-Pro-Leu-Gly-Gly-I

Pro-OT/Np(l-11) : Cys-Tyr-Ile-Gln-Asn-C;s-Pro-Leu-Gly-Gly-& Pro-OT/Np(l-10) : Cjrs-Tyr-Ile-Gln-Asn-Cjrs-Pro-Leu-Gly-Gly Pro-OT/Np(13-20):

either. This degree of enrichment of CCK in both P2 and P3 and in the lower interface is similar to that found in purified synaptosomes from rat brain (42). Enzymatic activity recovered from the upper interface cleaved the synthetic substrate but none of the products coeluted with products produced by trypsin treatment of the substrate. In contrast, activity extracted from the lower interface of both P2 and P3 yielded the correct products (along with other unidentified products). Thus the lower interface was used as a source of enzyme. Subsequently when it was found that activity from both P2 and P3 yielded the same fragments, the two extracts were pooled, although they were run on separate discontinuous sucrose density gradients. This final centrifugation eliminated most of the phosphatase activity. The lower interface derived from P2 or P3 contained from 0.5 to 2% of total tissue phosphatase activity. Enzyme Preparation and Characterization-The elution of the soluble enzyme on a column of Sephadex G-150 is shown in Fig. 1. The soluble activity elutes as a rather broad peak after bovine serum albumin with an approximate molecular weight of about 55,000. A second peak which eluted later from the column (fraction 37) probably represented a small amount of authentic pancreatic trypsin recovered with the secretory granules. The size of this peak was variable and appeared to be related to how thoroughly the intestine was rinsed prior to removal of the mucosa. This enzyme wasactive at neutral or basic pH anddisplayed a rather broad pH optimum, as shown in Fig. 2. After Sephadex filtration, the preparation which contained very little protein (20 pl was below the detection limit of the Bradford assay) yielded few products in addition to Leu-Gln-Arg and Ser-Ala-Asn-Ser-NHz expected oncleavage of ProS(6268)NHz after the Arg residue (Fig. 3). Pre-derivatization of

Ala-Val-Leu-Asp-Leu-Asp-Val-Arg

ap eo



Y

>b.0.

3 I-

0 2 0

4 0

15

20

25

30

35

40

FRACTION NUMBER FIG. 1. Sephadex gel filtration of soluble endoprotease activity. Twenty-microliter aliquots of Sephadex G-150 fractions were incubated for 60 min a t room temperature with 2 nmol of ProS(6268)NHz. The reaction was stopped by addition of 40 p1 of 2 mg/ml DABITC in pyridine. The reaction products were analyzed by HPLC preceded by derivatization and extraction.

the substrate with DABITC increased the recovery of LeuGln-Arg. In thepresence of bestatin, even though the enzyme was partially inhibited, the quantity of Leu-Gln-Arg increased. These observations suggest that some residual aminopeptidase activity remains in the preparation. The enzyme preparation actively cleaved ProS(62-68)NHz, such that after 1h at room temperature most, if not all the 2 nmol of substrate had been converted to products. The rate of conversion of substrate to products was linear. Complete

Somatostatin-28 Endoprotease Generating

4463 TABLE I1

Effects of enzyme inhibitors on activityof the endoprotease The test involved preincubation of20 p1 of a pool of the most active fraction from the Sephadex filtration with the inhibitor for 60 min at room temperature prior to addition of ProS(62-68)NH2. After 60 min at room temperature, the reaction was terminated by addition of 40 p1 of DABITC (2 mg/ml) in pyridine, and thereaction products were analyzed by HPLC preceded by derivatization and extraction.

100.

f2

80.

X

a 2 eo.

*

Inhibitor

Y

> k >, I-

0

Activity

% control

40-

20.

a

I

e

I I

8

0

1

0

pH FIG. 2. Dependence of soluble endoprotease activity on pH. To 30 pl of a pool of the most active fractions from Sephadex gel filtration were added 30 plof 100 mM sodium phosphate buffer adjusted to the final pH indicated on the abscissa of the graph. Two nanomoles of ProS(62-68)NH2 were added, and after 60 min at room temperature 80 plof DABITC (2 mg/ml) was added. The reaction products were analyzed by HPLC following derivatization and extraction.

DIPF, 1 mM 27 45 PMSF, 1 mM 5 PMSF, mM 5 100 pCMB, 1 mM AP, 25 pg/ml 0 AP, 100 pg/ml 0 0.2 IAA,” mM 60 41 IAA, 1.0 mM STI, 50 pg/ml 65 pg/mlSTI, 100 33 STI, 200 pg/ml 47 65 EDTA, 2 mM 100 5 Bestatin, p M Bestatin, 50 p M 84 100 Pepstatin, 10 pg/ml IAA, iodoacetamide; DIPF, diisopropylfluorophosphate.

and the reported preference of kallikrein for Arg-Ser, we synthesized thealanine 65 analog [Ala65]ProS(62-68)NH2. The enzyme easily cleaved this substrate with the same K,,, as for ProS(62-68)NHz (about 20 pM). Plasma kallikrein (Sigma) also rapidly cleaved this substrate, but tissue kallikrein (Sigma) cleaved it much less rapidly than ProS(6268)NHz. The selectivity of the monobasic cleavage enzyme was further examined (sequences in Table I). It did not cleave the Arg-Asp bond of CCK 12. Interestingly, this is not afavorable cleavage for trypsineither; however, enterokinase cleaves CCK 12 quite effectively, as first shown by Mutt et al. (43). Two peptides located on the amino-terminal side of anglerfish somatostatin-28 (asdeduced from the results of Hobart et al., as reviewed in Ref. 22) whichcontain single arginine residues (Arg4 and Arga), but whichhave not been reported to be FIG. 3. HPLC separation of reaction products of the endo- cleaved in vivo were also used as substrates. The Arg4-Glu‘ protease with ProS(62-68)NHs. Twenty microliters of a pool of bond of AF ProSII(1-10)NHz and theArga-Met6’ bond of AF the most active fractions from Sephadex gel filtration was incubated ProS(64-72) were unaffected. with 2 nmol of ProS(62-68)NHz for 60 min at room temperature. The selectivity of this endoprotease was also analyzed with The reaction was stopped by addition of 40 pl of DABITC (2 mg/ml), and the products were analyzed by HPLC following derivatization a variety of substrates containing pairs of basic amino acids as in either pro-somatostatin or pro-oxytocin-neurophysin and extraction.Elution of standards: 1 corresponds to ProS(6568)NHz,2 corresponds to ProS(62-64), and 3 corresponds to proS(62(see Table I for amino acid sequence). In addition, a peptide 68)NHz. corresponding to the entire COOH-terminal sequence of CRF was used (Table I). The enzyme cleaved all three substrates, kinetic measurements were not made with this preparation although significantly slower than the monobasic substrates. but the enzyme displayed a K, of about 20 p~ with this In all cases the major cleavage product recovered was that substrate. corresponding to cleavage onthe carboxyl side of the arginine The ability of known enzyme inhibitors to block the activity (Fig. 4). For [Ala’7,Tyr20]S-28( 10-20)NHz this corresponded when tested with ProS(62-68)NHz as substrate is shown in to an Arg-Lys cleavage and produced [Ala’7,Tyr20]S-28(14Table 11. The sensitivity of the enzyme to diisopropyl fluoro- 20)NHz. In addition very small amounts of [Ala’7,TyrZo]Sphosphate (83% inhibition at 1 mM), PMSF (95%inhibition 28(15-20)NHz representing either a cleavage on the carboxyl at 5 mM), ST1 (67% inhibition at 100 pg/ml) indicates that it side of Lys at a Lys-Ala bond or the product of residual is a serineprotease. Since this enzyme is completely inhibited aminopeptidase activity on [Ala’7,Tyr20]S-28( 14-20)NHzwere by aprotinin, it appears to be a kallikrein-like serine protease. detected. No [Ala’7,Tyr20]S-28( 13-20)NHz was detected Therefore the ability of this enzyme to cleave several small (cleavage on the amino-terminal side of arginine at a Glu-Arg colored or fluorescent substrates containing single arginine bond). With Pro-OT/Np(l-20), the major product was Proresidues was examined. It did not cleave BAEE or BAPNA, OT/Np( 1-12), which represents an Arg-Ala cleavage, while but was able to cleave Na-benzoyl-~-arginine-7-amido-4lesser amounts of Pro-OT/Np(l-11) were recovered which methylcoumarin at about 1/30 to 1/60 the rate of ProS(62- represents either acleavage of a Lys-Arg bond or the product 68)NHz. of contaminating carboxypeptidase activity on pro-OT/Np( 1Because of the clear resemblance of the enzyme to kallikrein 12). Some Pro-OT/Np(l3-20) was recovered, which is the

Endoprotease Generating Somatostatin-28

4464

A

IO

B

n

20 30 40 FRACTION NUMBER

5 - 2 8 II

v

0 N N

n

0

10 20 RETENTION TIME

FIG.5. AF-ProSII processing by the soluble enzyme preparation. A , 20 ng of AF Pros previously purified as described (36)

B +TIME

FIG.4. HPLC separation of reaction products of incubation of the soluble endoprotease with [Ala",Tyra0]S-28(1020)NHz (Panels A and B ) and Pro-OTmP(1-20) (Panels C and D ) . Ten microliters of the endoprotease pool was incubated for 2 h with [Ala'7,Tyr20]S-28(10-20)NH2(Panel A , control; Panel B , preincubation with 25 pg/ml aprotinin) or for 3h with Pro-OT/ NP(1-20) (Panel C, control; Panel D,with aprotinin). The reaction was stopped by addition of acetic acid to a final concentration of 10% and the products analyzed by HPLC with optical detection. The identity of known fragments isindicated, the others were unidentified. amino-terminal fragment produced when Pro-OT/Np(l-12) is formed by hydrolysis of the Arg"-Ala13 bond. With CRF (not shown in Fig. 4)the cleavage waseven slower than with the other dibasic substrates, the major product being [Tyr3'] CRF(31-35) which represents an Arg-Lys cleavage. A small amount of [Tyr3']CRF(31-36) was also observed which represents aLys-Leu cleavage. These resultsindicate clearly that in all cases major cleavage occurred after an arginine residue. The production of both the major and minor products with all of these substrates was completely blockedby 15-min preincubation with 25 Fg/ml aprotinin. Prosomatostatin Processing-The major observation reinforcing the hypothesis that thisactivity might be an endoprotease involved in prosomatostatin processing was the ability to cleave not only synthetic peptides but also the biosynthetic precursor anglerfish prosomatostatin I1 (AF ProSII). Since the cleavage site in S-2811 from anglerfish prosomatostatin is at a single arginine residue and the sequences around this residue exhibit similarity in primary and secondary structure (39) to the mammalian peptide, precursor I1 from anglerfish pancreas was chosen because it is present in large quantities in this fish organ compared with the amount encountered in mammalian brain (1000-fold less). When 20 ng of AF Pros were incubated with the enzyme (10 pl), 70% of the conversion products consisted of a peptide which coeluted with synthetic

were incubated for 1 h at room temperaturein 100 mM sodium phosphate buffer, pH 7.2, with 10pl enzyme preparation. The reaction was stopped with acetic acid at a final concentrationof 50%,and the sample was applied to a Sephadex G-50 column (1.2 X 90 cm) equilibrated and eluted in 10% acetic acid at a flow rate of 6 ml/h. One-tenth of each fraction was dried and assayed for somatostatinlike immunoreactivity. Somatostatin-like immunoreactivity was expressed as pg/ml. B, the fraction identified as S-2811-like (fractions 22-30, a total of 6 ng) from A were pooled and dried under vacuum. The sample was then analyzed by HPLC (36). Each 1-ml fraction was dried and tested for somatostatin-like immunoreactivity. The arrow indicates the elution time of S-2811 (40).

S-2811 (40) on Sephadex and HPLC (Fig. 5). The other 30% coeluted with S-1411. Preincubation of the enzyme with aprotinin totally inhibited these cleavages and the precursor was recovered intact after incubation. By contrast, under similar conditions trypsin treatment of AF ProSII degraded the precursor completely and produced no s-28, only some s-14. DISCUSSION

In this report we describe an endoprotease in secretory granules from rat intestinal mucosa. This enzyme cleaves substrates exclusively on the carboxyl side of single arginine residues. It belongs tothe serine protease family and is kallikrein-like based on its sensitivity to aprotinin. However, it did not cleave standard substrates for both plasma and tissue kallikrein and for trypsin. Clearly this enzyme is neither truly trypsin- nor kallikrein-like. Interestingly, the inability of this endoprotease to cleave the Arg-Asp bond in CCK 12 or the Arg4-Glu5and Argm-Met6' bonds of fragments of AF ProSII suggests that theadjacent amino acids influence cleavage perhaps by conferring a preferred conformation (44) around the cleavage site. Thesefeatures, whichobviously distinguish processing enzymes from degrading proteases, have been documented with dibasic-specific endoproteases (12, 45). The enzyme can also cleave dibasic substrates, although at a slower rate. In all cases preferential peptide bond hydrolysis was observed on the carboxyl side of Arg residues

Endoprotease Generating Somatostatin-28 regardless of the position of Arg in the doublet. An important characteristic of the described enzyme is its ability to generate somatostatin-28 from AF ProSII. Since S28 is the major product produced i n vitro compared with S14, it is hypothesized that this Arg-specific endoprotease is responsible for production of the octacosapeptide in vivo in intestinal mucosal cells. This conclusion is reinforced by the fact that theenzyme cleaves neither the Arg-Glu' nor Arg68Met6' bonds of AF ProSII, which arenot reported to be cleaved in vivo (for a review see Ref. 22). The possibility that this enzyme also produces some s-14 cannot beexcluded. Further work with the mammalian prohormone is required to determine if this endoprotease can cleave mammalian prosomatostatin to s-28. The soluble enzyme described here resembles pro-atrial natriuretic processing enzyme in that italso cleaves an ArgSer bond. One of these enzymes (20) has a similar pH optimum and inhibitor sensitivity but issignificantly larger than the enzyme reported here. In addition, our soluble enzyme is unable to cleave BAPNA, BAEE,or Nu-benzoyl-1-argininel-amino-4-methylcoumarin,while the atrial enzyme can cleave benzoyl-Gly-Pro-Arg-2-naphthylamide. The molecular mass of the latterenzyme is 580 kDa (Zl),completely different from our enzyme. Although further studieswill be required to determine whether or not these are related enzymes, it is unlikely that the enzyme reported here has been previously described. Activity is maximal at neutral orbasic pH, although it does have significant activity (40%) at pH 6. It is thought (based on studies on insulin processing) that post-Golgi prohormone processing begins in clathrin-coated secretory granules the interior of which start out atslightly basic or neutral pH but which become increasingly acidic, reaching a pH of 5-6 (3,5, 46,47). It is possible that enzymes like these act early in the acidification process and that itis the acidification itself that inactivates them and limits the extent of their cleavage (see a discussion in Ref. 48). The precise physiological significance of the soluble enzyme remains to be determined, but it appears, at least in termsof its localization in secretory granules and its ability to cleave related small and large substrates, to be a potential candidate for the enzyme which cleaves prosomatostatin in vivo. Acknowledgments-We wish to thank Drs. H. Boussetta and P. Nicolas for the synthesis of numerous peptide substrates and C. Rougeot (Unit6 de Radioimmunologie Analytique, Institut Pasteur, Paris) for the generous supply of somatostatin antiserum 2044.

1.

2. 3. 4. 5. 6. 7. 8.

9. 10. 11.

REFERENCES Docherty, K., and Steiner, D. F. (1982) Annu. Reu. Physiol. 4 4 , 625-638 Loh, Y. P., Brownstein, M. J., and Gainer, H. (1984) Annu. Reu. Neurosci. 7 , 189-222 Cohen, P. (1987) Biochimie 59,87-89 Gluschankof, P., and Cohen, P. (1987) Neurochem. Res. 12,951958 Orci, L., Ravazzala, M., Amherdt, M., Madsen, O., Vassali, J., and Perrelet, A. (1985) Cell 42, 671-681 Mizuno, K., Kojima, M., and Matsuo, M. (1985) Bwchem. Biophys. Res. Commun. 128,884-891 Loh, Y. P., Parish, D. C., and Tuteja, R. (1985) J. Biol. Chem. 260,7194-7205 Nyberg,F., Norstrom, K., and Terenius, L. (1985) Biochem. Biophys. Res. Commun. 131,1069-1074 Gluschankof, P., Gomez, S., Morel, A., and Cohen, P. (1987) J. Biol. Chem. 262,9615-9620 Gomez, S., Gluschankof, P., Lepage, A., and Cohen, P. (1988) Proc. Natl. Acad. Sci. U. S. A. 85, 5468-5472 Clamagirand, C., Camier, M., Fahy, C., Clavreul, C., Creminon,

4465

12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32.

33. 34.

35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48.

C., and Cohen, P. (1987) Biochem. Biophys. Res. Commun. 143, 789-796 Clamagirand, C., Creminon, C., Boussetta, H., Fahy, C., Nicolas, P., and Cohen, P. (1987) Biochemistry 26,6018-6023 Gainer, H., Russell, J. T., and Loh, P. (1984) FEBS Lett. 1 7 5 , 135-139 Hook, V. Y. H., Mezey, E., Fricker, L. D., Pruss, R. M., Siegel, R. E., and Brownstein, M. J. (1985) Proc. Natl. Acad. Sei. CJ. S. A. 82,4745-4749 Fricker, L.D., and Snyder, S. H. (1983) J. Biol. Chem. 2 5 8 , 10950-10955 Bradbury, A. F., Finnie, M.D.A., and Smyth, D.G. (1982) Nature 298,686-688 Murthy, A. S. N., Mains, R. E., and Eipper, B. A. (1986) J. Biol. Chem. 261,1815-1822 Schwartz, T. W. (1986) FEBS Lett. 200, 1-10 Straus, E., Malesci, A., and Yalow, R. S. (1978) Proc. Natl. Acad. Sei. U. S. A. 7 5 , 5711-5714 Wypij, D. M., and Harris, R. B. (1988) J. Biol. Chern. 263,70797086 Imada, T., Takayanagi,R., and Inagami, T. (1988) J. Biot. Chem. 263,9515-9519 Cohen, P., Kuks, P., Gomez, S., and Morel, A. (1988) in Prohormoms, Hormones and the Fragments (Martinez, J., ed) Ellis Horwood Ltd., in press Mackin, R. B., and Noe, B. D. (1987) J. Biol. Chem. 2 6 2 , 64536456 Patel, Y. C., Wheatley, T., and Ning, C. (1981) Endocrinology 109,1943-1949 De Robertis, E., Pellegrino de Iraldi, A., Rodrigues de Lorenz Arnaiz, G., and Salganicoff, L. (1962) J. Neurochem. 9,23-35 Noe, B. D., Baste, C. A., and Bauer, G . E. (1977) J. Cell Biol. 74, 578-588 Lindhardt, K., and Walter, K. (1963) in Methods in Enzymatic Analysis (Bergmeyer, H. U., ed) pp. 783-785, Academic Press, New York Beinfeld, M. C., Meyer, D. K., Eskay, R. L., Jensen, R. T., and Brownstein, M. J. (1981) Brain Res. 212, 51-57 Bradford, M. M. (1976) Anal. Biochem. 72,248-254 Norman, J. A., and Chang, J-Y. (1985) J. Biol. Chem. 260,26532656 Chang, J-Y. (1983) Methods Enzymol. 91,455-466 Chang, J-Y., Knecht, R., and Braun, D. G. (1983) Methods Enzymol. 91,41-48 Labouesse, J., and Gervais, M. (1967) Eur. J. Biochem. 2 , 215223 Stepanov, V. M., Strongn, A. Y., Izotova, L. S., Abramov, Z. T., Lyublinskaya, L. A., Ermakova, L.M., Baratova, L.A., and Belyanova, L. P. (1977) Biochem. Biophys. Res. Commun. 77, 298-305 Morita, T., Kato, H., Iwanaga, S., Takada, K., Kimura, T., and Sakakibara, S. (1977) J. Biochem. (Tokyo) 8 2 , 1495-1498 Morel, A., Kuks, P. F. M., Bourdais, J., and Cohen, P. (1988) Biochem. Biophys. Res. Commun. 151,347-354 Morel, A., Nicolas, P., and Cohen, P. (1983) J. Biol. Chem. 258, 8273-8276 Erickson, B. W., and Merrifield, R. B. (1976) in The Proteins (Neurath, H., and Hill, R. L., e&) pp. 255-527, Academic Press, New York Argos, P., Taylor, W. L., Minth, C. D., and Dixon, J. E. (1983) J. Biol. Chem. 258,8788-8793 Nicolas, P., Delfour, A., Morel, M., Rholam, M., and Cohen, P. (1986) Biochem. Biophys. Res. Commun. 140,565-573 Usellini, L., Capella, C., Malesci, A., Rindi, G., and Solcia, E. (1985) Histochemistry 8 3 , 331-336 Allard, L. R., and Beinfeld, M. D. (1985) Neuropeptides 6, 239245 Mutt, V., Tatemoto, K., Carlquist, M., and Light, A. (1981) Biosci. Rep. 1,656-659 Rholam, M., Nicolas, P., and Cohen, P. (1986) FEBS Lett. 207, 1-6 Creminon, C., Rholam, M., Boussetta, H., Marrakchi, N., and Cohen, P. (1988) J . Chromatogr. 440,439-448 Loh, Y. P., Tam, W. W. H., and Russell, J. T. (1984) J . Biol. Chem. 259,8238-8345 Orci, L. (1985) Diubetologia 28,528-546 Marx, J. L. (1987) Science 235, 285-286