Proc. Natl. Acad. Sci. USA Vol. 81, pp. 6662-6666, November 1984 Biochemistry

Enzymes processing somatostatin precursors: An Arg-Lys esteropeptidase from the rat brain cortex converting somatostatin-28 into somatostatin-14 (endopeptidase/somatostatin radioimmunoassay/neocortex/HPLC/protease assay)

PABLO GLUSCHANKOF, ALAIN MOREL, SOPHIE GOMEZ, PIERRE NICOLAS, CHRISTINE FAHY, AND PAUL COHEN Groupe de Neurobiochimie Cellulaire et Moldculaire, Universitd Pierre et Marie Curie, 96 Boulevard Raspail, 75006 Paris, France

Communicated by Gregorio Weber, July 12, 1984

ABSTRACT The post-translational proteolytic conversion of somatostatin-14 precursors was studied to characterize the enzyme system responsible for the production of the tetradecapeptide either from its 15-kDa precursor protein or from its COOH-terminal fragment, somatostatin-28. A synthetic undecapeptide Pro-Arg-Glu-Arg-Lys-Ala-Gly-Ala-Lys-AsnTyr(NH2), homologous to the amino acid sequence of the octacosapeptide at the putative Arg-Lys cleavage locus, was used as substrate, after 125I labeling on the COOH-terminal tyrosine residue. A 90-kDa proteolytic activity was detected in rat brain cortex extracts after molecular sieve fractionation followed by ion exchange chromatography. The protease released the peptide 1251-Ala-Gly-Ala-Lys-Asn-Tyr(NH2) from the synthetic undecapeptide substrate and converted somatostatin-28 into somatostatin-14 under similar conditions (pH 7.0). Under these experimental conditions, the product tetradecapeptide was not further degraded by the enzyme. In contrast, the purified 15-kDa hypothalamic precursor remained unaffected when exposed to the proteolytic enzyme under identical conditions. It is concluded that this Arg-Lys esteropeptidase from the brain cortex may be involved in the in vivo processing of the somatostatin-28 fragment of prosomatostatin into somatostatin-14, the former species being an obligatory intermediate in a two-step proteolytic mechanism leading to somatostatin14.

of somatostatin-28. Therefore, prosomatostatin offers a model particularly well suited for the study of the proteolytic systems involved in the processing of proneuropeptides and exhibiting a specificity either for basic amino acid doublets and/or for a lone arginine residue. In the hypothalamus, the presence of a 15-kDa immunoreactive form of somatostatin was established by immunochemical analysis of either mouse organ extracts (12, 13) or mRNA translation products (14). Its conversion to both somatostatin-14 and somatostatin-28 by crude hypothalamic extracts was achieved in vitro, demonstrating that it is a potential common precursor to both tetradecapeptide and octacosapeptide (12, 13, 15). To characterize the type of endopeptidase that exhibits a selectivity for stretches of basic amino acids and that is possibly involved in the in vivo production of somatostatin-14 from either its 15-kDa or somatostatin-28 precursors, we have synthesized a peptide substrate. This undecapeptide mimics the amino acid sequence situated on both sides of the Arg-Lys doublet shared both by the 15-kDa precursor and by somatostatin-28. We report here on the detection in brain cortex extracts of an Arg-Lys esteropeptidase. This protease is able to cleave either the synthetic undecapeptide or somatostatin-28 by hydrolysis of the peptide bond situated on the -C- side of the lysine

1

0

residue. This observation is discussed in connection with the possible role of such a selective enzyme in the biosynthetic pathways leading to somatostatin-14.

A number of neural, or hormonal, peptides have been shown to derive biosynthetically from larger molecular weight precursors by post-translational proteolytic cleavage. Elucidation of the amino acid sequences of many of these proforms has clearly underlined the importance of basic amino acid doublets as potential recognition sites for selective proteases. In addition, putative cleavage points corresponding either to a lone lysine or arginine residue, as well as double pairs of basic amino acids, were deduced from the established primary structures of some precursors (1-7). Although structural information on these potential cleavage points is available, little is known at this time about the enzymes involved in recognition of these sites. The predicted sequences of the prosomatostatins from pancreatic tissue (2) have indicated that the tetradecapeptide somatostatin-14 occupies the COOH-terminal end of the precursor and that an Arg-Lys doublet can be found at position -2, -1 from the NH2 terminus of the somatostatin-14 sequence. Furthermore, the finding in both hypothalamic and pancreatic tissues of octacosapeptides called somatostatin28 (8-11) has suggested that the lone arginine residue observed in the precursor (at position -15 from the NH2-terminal alanine of somatostatin-14) may constitute the recognition signal for a specific protease responsible for the release

MATERIALS AND METHODS Wistar male rats (100-120 g) were sacrificed by rapid decapitation and the cerebral cortex was immediately removed. The freshly dissected tissue was homogenized in a PotterElvehjem homogenizer in 50 mM phosphate buffer, pH 7.4/200 mM KCl (5 ml of buffer per cortex). The extract was then centrifuged for 15 min at 3000 x g in a refrigerated Beckman TJ-6 centrifuge, and the supernatant was filtered on a Sephadex G-150 column in 0.25 M Tris HCl (pH 7). The fractions exhibiting the proteolytic activity were then pooled and submitted to ion exchange chromatography on a Mono Q (cationic) column using a fast protein liquid chromatography system (Pharmacia, Uppsala, Sweden). Peptides I and II (see Results) were synthesized by an improved procedure of the solid-phase method (16) using benzhydrylamine resin (Beckman). Crude peptides were cleaved from the resin by liquid HF at 0°C and passed through Sephadex G-10 in 0.5 M acetic acid. Final purification was effected by chromatography on carboxymethyl cellulose (CM cellulose). The purity of the peptides was assessed both by HPLC on a VYDAC 201 TP column (250 x 4.6 mm) eluted by a 5%-30% propanol-2 gradient in 0.1% trifluoroacetic

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Abbreviation: Tyr(NH2), tyrosine amide.

6662

Proc. Natl. Acad. Sci. USA 81 (1984)

Biochemistry: Gluschankof et aL acid, and by amino acid analysis on a Beckman TSM automatic analyzer. '25I-labeling was done by the lactoperoxidase method (17). 125I-labeled peptides were purified by ion exchange chromatography on carboxymethyl cellulose (CM52) resin. Peptide I eluted between 150 and 300 mM ammonium acetate buffer (pH 6), and peptide II eluted between 50 and 120 mM of the same buffer. Proteolytic activities were tested as follows: the 125I-labeled peptide I (20,000 cpm) dissolved in 10 Al of 250 mM Tris HCl (pH 7.0) was incubated at 37°C with an aliquot of either crude rat brain cortex extract or fractions from the gel filtration or from the ion exchange chromatography [final vol, 300 ,4 in 0.25 M Tris HCl (pH 7.0)]. At various reaction times, 50 ,4 of the mixture was removed and the reaction was stopped by addition of 600 ,4 of CM52 resin equilibrated in 30 mM ammonium acetate (pH 6.0) (1:10, vol/vol). The mixture was then centrifuged at 10,000 x g on a Beckman Microfuge for 30 sec. The resulting pellet was washed once with the same buffer and then twice with 100 mM ammonium acetate buffer (pH 6.0). The iodinated products retained on the resin were counted after the second wash of each different salt concentration elution. Results were calculated as follows:

specific activity

= (X1

degrading activity

-

X2)

-

(Y1

-

Y2)

x

Y2 =

(t

x1)

(t

y)

x

100 100,

Y1

where specific activity represents percent peptide II generated, degrading activity represents percent Asn-125I-labeled tyrosine amide [125I-Tyr(NH2)] generated, t is total cpm in the 50-,u4 reaction mixture, x, is cpm recovered after the 30 mM ammonium acetate wash, x2 is cpm recovered after the 100 mM ammonium acetate wash, and Yi and Y2 are the cpm recovered, respectively, after the 30 mM and 100 mM washes of the resin with peptide I alone. Processing activity was tested on either hypothalamic somatostatin-28 or on 15-kDa precursor incubated with 300 ,4 of the different G-150 Sephadex fractions or the ion exchange chromatography eluate in 0.29 M Tris HCl (pH 7.0) (final vol, 500 ,4) at 37°C. The reaction was stopped with 60 ,u4 of concentrated acetic acid. When somatostatin-28 was substrate, the mixture reaction was directly evaporated in a Speed Vacuum (Savant), and redissolved in 1 M acetic acid for HPLC analysis. The HPLC system consisted of an SP 8000 (Spectra Physics, Santa Clara, CA) apparatus. A reversed-phase ,Bondapak C18 column (300 x 3.0 mm; Waters Associates) eluted at a flow rate of 1 ml min-1 was used. Elution was performed isocratically by a mixture of 25% acetonitrile in aqueous formic acid (1.15%) adjusted to pH 3 with triethylamine phosphate. Radioimmunoassay was done using antiserum 36-38 directed toward the COOH-terminal portion of somatostatin14 as described (12, 13). Counting of 125I was done on a Kontron Analytical MDA 312 y counter (Roche Analytical Kontron, Montigny le Bretonneux, France). NH2-terminal amino acid analyses were carried out by 4N-N-dimethylaminobenzene 4'-isothiocyanate/phenyl isothiocyanate double-coupling method (18). The 15-kDa somatostatin precursor was purified from mouse hypothalamus as reported in ref. 15. Synthetic somatostatin-14 was furnished by Clin-Midy (Montpellier, France), synthetic somatostatin-28 was from CRB (Cambridge, England), and Tyro somatostatin-14 was from Peninsula Laboratories (San Carlos, CA). Anti-somatostatin antiserum 36-38 (12, 13) was provided by the Unite de Radio Immunologie Analytique (F. Dray, Institut Pasteur, Paris).

6663

RESULTS To characterize the brain enzyme involved in the production of somatostatin-14 from its precursors, a routine test was developed using a synthetic radiolabeled peptide as model substrate:

5 10 1 Pro-Arg-Glu-Arg-Lys-Ala-Gly-Ala-Lys-Asn125I-Tyr(NH2) [peptide I].

A partially purified extract from rat brain cortex provided of converting enzyme activity. The rationale for the choice of both substrate structure and enzyme source was as follows. (i) The undecapeptide primary structure mimics the sequence of hypothalamic somatostatin-28 around the ArgLys putative cleavage signal (9-11) a source

10 5 1 Ser-Ala-Asn- Ser -Asn- Pro-Ala-Met-Ala- Pro- Arg-Glu-Arg-Lys25 20 15 Ala-Gly-Cys-Lys -Asn- Phe-Phe-Trp -Lys- Thr- Phe-Thr- Ser-Cys. S

S

(ii) The Tyr1" side chain provides a site for 1251 labeling, allowing a sensitive and rapid detection of the generated reaction product without altering the possible requirement of the protease for an aromatic residue (Phe20 of the somatostatin28 sequence). Preliminary evaluation of the converting activities in three brain areas indicated that the hypothalamus, a major site for somatostatin and other neuropeptide production, contained a high level of substrate degradation activity. In contrast, the neocortex, an area known to possess both intrinsic as well as hypothalamic-derived axon terminals of somatostatinergic neurons (19) appeared to be more adequate for the enzymatic analysis of the selective cleavage (Table 1). Extracts of rat brain cortex were incubated with peptide I at pH 7 and 370C. At the end of the incubation, the reaction products were analyzed by ion exchange. The expected generated peptide

Ala-Gly-Ala-Lys-Asn-125I-Tyr(NH2) [peptide II], and the starting material were eluted from the CM52 column by ammonium acetate concentrations of 80 mM and 200 mM as assessed separately using both reference synthetic peptides I and II (Fig. 1). In contrast, the final product of a contaminant degrading activity, Asn-125I-Tyr(NH2) was not retained by the resin under these conditions. Peptides generated by cleavage either between residues 5 and 6 or 9 and 10 of peptide I carry distinct net charges and behave in a different manner on such an ion exchange. Thus, this procedure allows a rapid and quantitative discrimination between both Table 1. Evaluation of the Arg-Lys specific and degrading activities in three brain areas % of generated products Brain region Asn- 1251-Tyr(NH2) 125I-labeled peptide II 7 90 Hypothalamus Neocortex 18 40 37 20 Hypophysis Extracts were made from three rat brain regions, and they were then fractionated by molecular sieve filtration (as in Fig. 2A). A 50,ul aliquot of peak a activity (obtained as in Fig. 2A) was incubated with peptide I (20,000 cpm) for 2 hr, and the products were analyzed as described in Fig. 1. Results are expressed as % of starting materi-

al.

Biochemistry: Gluschankof et al.

6664

Proc. NatL Acad Sci. USA 81

0

20

0L %

0

I

05

20

180 100 NH4OAc , mM

260

FIG. 1. Ion exchange separation of peptide I from peptide II. A mixture of peptide I (56,000 cpm) and peptide 11 (30,000 cpm) was applied to a CM52 column (bed vol, 1 ml) previously equilibrated with 20 mM ammonium acetate buffer (pH 6.0). The column was eluted by established steps between 20 mM and 300 mM ammonium acetate buffer (pH 6.0). Peptide I (i) eluted between 180 mM and 260 mM ammonium acetate. Peptide II (o) eluted between 50 mM and 100 mM ammonium acetate.

specific and contaminating degrading endoprotease activities present in rat brain cortex.

The somatostatin converting activity of the brain cortex extract was partially purified by molecular sieve filtration on a Sephadex G-150 column (Fig. 2A). A specific cleaving activity was found associated with fractions eluting as species with a size approximating 90 kDa. The products generated by this particular enzyme fraction were characterized by

(1984)

NH2-terminal determination using the double-coupling method described (18). Results demonstrated the presence of a single alanine residue at the NH2 terminus of the peptide eluted on the ion exchange column by ammonium acetate, ranging from 30 to 100 mM (peptide II). On this fractionation, another peak of proteolytic activity, with an apparent size of 46 kDa, was separated (Fig. 2A). Peptide I was also converted by this fraction into peptide II, but to a much lower extent (Fig. 2A). In addition to this major proteolytic activity, another protease was detected in fractions eluted as a 60-kDa species. This appeared to correspond to a degrading activity generating the dipeptide fragment Asn-125I-Tyr(NH2) as assessed both by ion exchange chromatography and by NH2-terminal analysis. The 90-kDa fraction was therefore submitted to ion exchange chromatography on a cationic resin. Fig. 2B indicates that the activity eluted between 80 and 120 mM NaCl was essentially free of the contaminating degrading activity. This protease fraction was subsequently used as a source of converting activity in all the experiments described below. The enzyme was found to be stable when stored at 40C and retained its full activity for several days under these conditions. The optimum condition for this specific activity was around pH 7 with a rapid decline below pH 6 (not shown). The reaction was linear, with respect to the amount of total protein up to 200 ,ug of protein per assay. Kinetic studies were carried out to determine the effect of various substrate concentrations on the reaction rate. The proteolytic activity behaved in a Michaelis-Menten manner with peptide I as substrate (Fig. 3). Under these conditions, a Km of 20 x 10l M and a Vmax of 2.1 pmol ,ugg per hr were measured. This activity was inhibited both by iodoacetamide at 2.7 mM (100% inhibition) and by diisopropylfluorophosphate at 6 mM (25% inhibition). No effect was detected when the soybean trypsin inhibitor (20 ,uM) was added to the protease assay. Final demonstration for the selectivity of this esteropepti-

0 4-

.

o

Ca 0

co 0

0 .5

CZ

0-

-3

co

A JvlX o.1

100 i

0.2 =

0

co

C

-i

0

0

0) -0 0~

c

z 0.1

a)

0)

0.05 50

.-

z

a1)

'D

QCL

(1

50

100

Fractions

Fractions

FIG. 2. Partial purification of the converting Arg-Lys esteropeptidase from rat brain cortex. (A) An extract prepared from rat brain cortex (3 ml) was filtered on a Sephadex G-150 column (1.6 x 70 cm) equilibrated and eluted with 0.25 M Tris HCl (pH 7.0), at a flow rate of 5 ml/hr at 4°C. Converting activity was assayed on a 50-,ul aliquot of each fraction (1 ml) incubated with "25I-labeled peptide 1 (30,000 cpm) at 37°C, pH 7.0, for 1.5 hr. Arrows indicate elution volume of 94 kDa (phosphorylase b), 68 kDa (bovine serum albumin), and 10 kDa (neurophysin). Peak a (ArgLys esteropeptidase activity), peak b, and peak c elute as 90-, 60-, and 46-kDa forms, respectively. (B) Peak a from the Sephadex G-150 fractionation in A was diluted to a final concentration of 50 mM Tris * HCI (pH 7.7) and submitted to an ion exchange chromatography on a Mono Q column, using the fast protein liquid chromatography system. The proteins were eluted by a NaCl gradient. Converting activity was assayed on a 50-,ul aliquot of each fraction (1 ml) incubated with '25I-labeled peptide I (30,000 cpm) at 37°C, pH 7.0, for 3 hr. The Arg-Lys esteropeptidase activity was eluted between 80 and 120 mM NaCl.

Biochemistry: Gluschankof et aL

Proc. Natl. Acad. Sci. USA 81 (1984)

>

0

6665

B

Ln 80

0

EE E 40. C 1/S,

PM

FIG. 3. Determination of Km and Vmax for the Arg-Lys esteropeptidase of cortex extracts. The indicated amount of peptide I was incubated in a mixture containing a constant amount of 125I-labeled peptide 1 (20,000 cpm), 50 Apl of a pool of fractions 60-70 from the ion exchange chromatography (Fig. 2B), and 230 A.1 of 250 mM Tris1HC1 (pH 7.0).

dase was achieved using somatostatin-28 as substrate. The somatostatin-like products generated by the reaction were analyzed by RIA coupled to HPLC, because the procedure allowed a total separation of somatostatin-14 from somatostatin-28 in a single run (Fig. 4A). Reversed-phase HPLC analysis indicated (Fig. 4B) that >55% of the starting somatostatin-28 immunoreactive material was converted into species indistinguishable from synthetic reference somatostatin14 (retention time, 22 min). No degradation of the generated somatostatin-14 was observed under these conditions, because >98% of the starting somatostatin-like immunoreactivity was recovered at the end of the reaction. Under similar conditions, the 15-kDa hypothalamic prosomatostatin form was recovered unmodified when exposed to the protease preparation. On the other hand, when incubation was made using the 60-kDa degrading enzyme fraction (from Fig. 2A), no significant amount of somatostatin-14 was generated (Fig. 4C).

DISCUSSION The 15-kDa and somatostatin-28 immunoreactive forms of somatostatin appear to be associated with somatostatin-14 in both rat brain cortex (20) and cerebral cortical cells (21). A precursor role for these high molecular weight species was inferred from observations made with crude extracts (12, 22) and from attempts to detect the relevant proteases (12, 22). In these reports, the hypothalamus was chosen as a source of enzyme. In most cases the converting activity, which was found associated with crude synaptosomes, led generally to a rapid degradation of the somatostatin-28 into unidentified products with low yields of somatostatin-14 (23). The present enzyme preparation after partial purification shows both a great specificity toward the cleavage signal and generates a single stable product-i.e., somatostatin-14. Only a single alanine residue was found at the NH2 terminus of both the generated somatostatin-14 and peptide II, no lysine or arginine could be detected at this position of the cleavage products. This suggests that the converting activity detected in the cortex acts as an Arg-Lys esteropeptidase with a strict selectivity for the lysine -C- peptide bond. The sensitivity of the peptidase to thiol reagents could be taken as suggestive of a role for SH moieties at the active site. Most of the processing enzyme activities reported up to

0

0

E

c

0

CO 80

40

15

5

25

Time,min FIG. 4. HPLC analysis of reaction products generated by exposure of somatostatin-28 to proteolytic activities. Somatostatin-28 (360 ng) was incubated at 370C, pH 7.0, for 4.5 hr with 300 /11 of: (A) Tris HCl buffer, (B) a pool of fractions 60-70 from the ion exchange chromatography (Fig. 2B), (C) fraction 53 (Fig. 2A). The mixture was then treated and analyzed. (A) More than 98% of the starting somatostatin-28 immunoreactive material was recovered as unchanged somatostatin-28. (B) The converting activity fractions from the fast protein liquid chromatography elution (Fig. 2B) were used as source of enzyme: 55% of somatostatin-14 was produced and 45% of unmodified somatostatin-28 was recovered in the mixture. (C) Peak b from the gel filtration (Fig. 2A) was used as source of enzyme: 50% degradation of the starting material was observed and no somatostatin-14 could be detected. Results are expressed in ng of somatostatin immunoreactivity per fraction (1 ml). S-28, somatostatin-28; S-14, somatostatin-14.

now show suggestive evidence for this type of active-site residue involvement (24-27). In a number of hormone production systems, processing associated with the cleavage of basic doublets was reported (25-28). Although the detailed mechanism of such enzyme action still remains obscure, the sequential involvement of trypsin-like, followed by carboxypeptidase B-like, enzymes was invoked (29). In the present model, the production of somatostatin-14 results from the hydrolysis of a peptide bond on the lysine -C- side of the Arg-Lys doublet

1

belonging to somatostatin-28. However, release of the NH2terminal fragment of somatostatin-28 (residues 1 -- 12) (30), which implies the removal of the Arg-Lys doublet from the COOH terminus, may involve carboxypeptidase-like enzymes. Therefore, a more detailed structural analysis of the putative 1 -- 14 fragment on the NH2-terminal side of the Arg-Lys doublet is needed. Interestingly the 15-kDa hypothalamic prosomatostatin, which possesses the somatostatin-28 sequence at its COOHterminal end, including the Arg-Lys doublet (positions 13-14 of somatostatin-28), was not processed by the activity con-

6666

Biochemistry: Gluschankof et al.

verting somatostatin-28 into somatostatin-14 under identical experimental conditions. The possibility that this failure may be due to the experimental conditions cannot be excluded. In particular, if another enzyme involved in the direct conversion of 15-kDa prosomatostatin into somatostatin-14 exists and is associated with another cell compartment, extraction conditions may become critical. Conformational restrictions at the Arg-Lys doublet occurring in the 15-kDa precursor can also be invoked. If this were the case in vivo, prior conversion of the 15-kDa precursor into somatostatin-28 could be an obligatory step in the enzymatic cascade leading to somatostatin-14. Relevant to this hypothesis is our recent observation in anglerfish pancreatic islets that the (tyrosine7, glycine-10) derivative of somatostatin-14 is included in a somatostatin-28 II form, which is not processed in vivo into the tetradecapeptide (31). But in vitro generation of this somatostatin analog, called somatostatin-14 IT (2), was obtained by exposure of the somatostatin-28 II form to the brain cortex esteropeptidase preparation (31). Hence, the protease involved in the production of somatostatin-28 from prosomatostatin, which can be hypothesized as being different from the Arg-Lys esteropeptidase herein described, remains to be discovered. Comparison of the amino acid sequences present around the putative cleaving signal in the established sequences of several precursors (1-7) does not reveal striking homologies. Thus, the possibility that the converting activity described here may be involved in the processing of other precursor sequences associated with doublet basic amino acids must be considered. Regional and subcellular distribution of this activity, together with studies of synthetic substrates reproducing the sequence around the cleavage locus of known peptide precursors, may provide a powerful tool to answer these questions. The help of Dr. Yamashiro (University of California at San Francisco) with peptide synthesis is gratefully acknowledged. This work was supported by funds from the Universite Pierre et Marie Curie, the Centre National de la Recherche Scientifique (Equipe de Recherches Associee no. 693 et Programme Interdisciplinaire de Recherches sur les Bases Scientifiques des Medicaments), the Ministere de l'Industrie et de la Recherche (Contract 81-E-0396), the Institut National de la Santd et de la Recherche M6dicale (CRL 814003 and CRE 834006), the Fondation pour la Recherche Medicale Francaise, and the Departement de Biologie du Centre d'Etudes Nucleaires de Saclay for the supply of radiochemicals. S.G. is a recipient of a predoctoral fellowship from the Ministere de l'Industrie et de la Recherche. 1. Nakanishi, S., Inoue, A., Kita, T., Nakamura, M., Chang, A. C. Y., Cohen, S. N. & Numa, S. (1979) Nature (London) 278, 423-427. 2. Hobart, P., Crawford, R., Shen, L. P., Pictet, R. & Rutter, W. J. (1980) Nature (London) 288, 137-141.

Proc. NatL Acad Sci. USA 81

(1984)

3. Comb, M., Seeburg, P. H., Adelman, J., Eiden, L. & Herbert, E. (1982) Nature (London) 295, 663-666. 4. Noda, M., Furutani, Y., Takahashi, H., Toyosata, M., Hirose, T., Inayama, S., Nakanishi, S. & Numa, S. (1982) Nature (London) 295, 202-206. 5. Gubler, V., Seeburg, P., Hoffman, B. J., Gage, L. P. & Udenfriend, S. (1982) Nature (London) 295, 206-209. 6. Land, H., Schutz, G., Schmale, H. & Richter, D. (1982) Nature (London) 295, 299-303. 7. Chance, R. E., Ellis, R. H. & Bromer, W. W. (1968) Science 161, 165-167. 8. Pradayrol, L., Jornvall, H., Mutt, V. & Ribet, A. (1980) FEBS Lett. 109, 55-58. 9. Spiess, J., Villarreal, J. & Vale, W. (1981) Biochemistry 20, 1982-1988. 10. Esch, F., Bohlen, P., Ling, N., Benoit, R., Brazeau, P. N. & Guillemin, R. (1980) Proc. Natl. Acad. Sci. USA 77, 68276831. 11. Schally, A. V., Huang, W.-Y., Chang, R. C. C., Arimura, A., Redding, T. W., Millar, R. P., Hunkapiller, M. W. & Hood, L. E. (1980) Proc. Natl. Acad. Sci. USA 77, 4489-4493. 12. Lauber, M., Camier, M. & Cohen, P. (1979) Proc. Nati. Acad. Sci. USA 76, 6004-6008. 13. Morel, A., Lauber, M. & Cohen, P. (1981) FEBS Lett. 136, 316-318. 14. Joseph-Bravo, P., Charli, J. L., Sherman, T., Boyer, H., Bolivar, F. & McKelvy, J. F. (1980) Biochem. Biophys. Res. Commun. 94, 1004-1012. 15. Morel, A., Nicolas, P. & Cohen, P. (1983) J. Biol. Chem. 258, 8273-8276. 16. Merrifield, R. B. (1963) J. Am. Chem. Soc. 85, 2149-2154. 17. Marchalonis, J. J. (1969) Biochem. J. 113, 299-305. 18. Chang, J. Y., Brauer, D. & Wittmann-Liebold, B. (1978) FEBS Lett. 93, 205-214. 19. Epelbaum, J., Willoghby, J. O., Brazeau, P. & Martin, J. P. (1977) Endocrinology 101, 1495-1502. 20. Gomez, S., Morel, A., Nicolas, P. & Cohen, P. (1983) Biochem. Biophys. Res. Commun. 112, 297-305. 21. Robbins, R. J. & Reichlin, S. (1983) Endocrinology 113, 574581. 22. Zingg, H. H. & Patel, Y. C. (1982) Life Sci. 30, 525-533. 23. Zingg, H. H. & Patel, Y. C. (1983) Life Sci. 33, 1241-1247. 24. Fletcher, D. H., Noe, B. D., Bauer, G. E. & Quigley, J. P. (1980) Diabetes 29, 593-599. 25. Docherty, K., Carroll, R. & Steiner, D. F. (1983) Proc. Nati. Acad. Sci. USA 80, 3245-3249. 26. Loh, Y. P. & Gainer, H. (1982) Proc. Natl. Acad. Sci. USA 79, 108-112. 27. Fletcher, D. J., Quieley, J. P., Bauer, G. E. & Noe, B. D. (1981) J. Cell Biol. 90, 312-322. 28. Fricker, L. D. & Snyder, S. H. (1983) J. Biol. Chem. 258, 10950-10955. 29. Docherty, K. & Steiner, D. F. (1982) Annu. Rev. Physiol. 44, 625-638. 30. Benoit, R., Bohlen, P., Ling, N., Briskin, A., Esch, F., Brazeau, P., Ying, S. Y. & Guillemin, R. (1982) Proc. Natl. Acad. Sci. USA 79, 917-921. 31. Morel, A., Gluschankof, P., Gomez, S., Fafeur, V. & Cohen, P. (1984) Proc. Natl. Acad. Sci. USA 81, 7003-7006.