Proc. Nati Acad. Sci. USA Vol. 80, pp. 2839-2843, May 1983

Identification of human : Complete amino acid sequences of MSEL- and VLDV-neurophysins (evolution/gene duplication/polypeptide processing//neurosecretion) M. T. CHAUVET, D. HURPET, J. CHAUVET, AND R. ACHER Laboratory of Biological Chemistry, University of Paris VI, 96, Bd Raspail, 75006, Paris, France Communicated by Choh Hao Li, January 27, 1983 ABSTRACT Twohuman neurophysins have been purified from Preliminary results on the NH2-terminal amino acid se- acetone-desiccated posterior pituitaries by acidic extraction, mo- quences of human or VLDV-neurophysin (5, 6, lecular sieving, and ion-exchange chromatography. The complete 10) and neurophysin II or MSEL-neurophysin (8, 10) have been amino acid sequence of each has been determined by us- published. These results are in agreement with the presence ing a sequencer and characterizing two sets of overlapping en- of only two types of neurophysins in the gland and with the zymic . The two neurophysins belong to two structural hypothesis that the two and [8-arginine]va- families previously defined as MSEL- and VLDV-neurophysins sopressin and the two neurophysins are cleavage products of according to the nature of the residues in positions 2, 3, 6, and 7. common with (MSEL-neurophysins contain methionine-2, serine-3, glutamic acid- precursors (11, 12). The present work deals the 6, and leucine-7; VLDV-neurophysins contain valine-2, leucine-3, isolation of the two human neurophysins and the determination aspartic acid-6, and valine-7.) Human MSEL-neurophysin has only of the complete amino acid sequences of MSEL- and VLDV- 93 residues instead of 95 usually found in MSEL-neurophysins neurophysins. Apart from minor variations, such as the short- from other mammalian species, probably because of a deletion of ening of the MSEL-neurophysin from 95 to 93 residues, the amino acids 91 and 92. Compared with bovine MSEL-neurophy- two human neurophysins can readily be fitted in the two mam- sin, nine variations (seven substitutions and two deletions) are ob- malian neurophysin lines. served. Human VLDV-neurophysin has 93 residues, as do the other mammalian VLDV-neurophysins. There are 11 substitutions when MATERIALS AND METHODS the comparison is made with bovine VLDV-neurophysin. Between the two human neurophysins, there are 26 variations. However, Purification of Human Neurophysins. Pituitary glands, des- the central parts of the (residues 10-70) are nearly iden- iccated in acetone, were collected in various hospitals of Paris tical. Furthermore, in this region identical substitutions are found through France-Hypophyse. Posterior lobes were separated in in positions 29 and 60 of both neurophysins, suggesting either a the laboratory and pulverized. About 450 glands gave 10.5 g of single exon or some relationship between the two corresponding powder titrating at 0.25-0.30 units/mg of pressor activity and genes. 0.30-0.35 units/mg of oxytocic activity. Extraction was usually carried out with fractions of 1 g of Neurophysins are small proteins found associated with neu- material. Each sample was extracted with 0.1 M HCl (50 ml/ rohypophysial hormones in stoichiometric and reversible com- g) for 4 hr at 40C. After centrifugation, the supernatant solution plexes (1). Isolation of neurophysins and neurohypophysial hor- was directly subjected to a molecular sieving by passage through mones from several mammalian species has been carried out a column (2.5 x 170 cm) of Sephadex G-75 equilibrated with often through this complex procedure (2). 0.1 M formic acid. Fractions (3-ml) were collected, and absor- Purification of human neurophysins has been attempted bance at 280 nm was measured. Five peaks were detected; the previously by using either by-products of the gonadotropin pu- fourth (D), containing proteins of Mr 11,000 as estimated by rification (3) or directly from acetone-desiccated posterior pi- sodium dodecylsulfate/polyacrylamide gel electrophoresis (13), tuitary glands (4, 5). Proteins are usually extracted by 0.1 M was used for further purification. HCI at 4°C; after fractionation by molecular sieving, the crude This "crude neurophysin" fraction was freeze-dried, and the neurophysins of Mr 10,000 are subjected to ion-exchange chro- material (30-40 mg) was chromatographed onto a column (0.4 matography in order to isolate the different types of neuro- x 0.50 cm) of DEAE-Sephadex A-50 by using a discontinuous physins (4, 5). Purification also has been carried out by using ionic strength gradient (0.15 M to 1 M) of pyridine acetate (pH continuous polyacrylamide gel electrophoresis (6-8). Neuro- 5.9). Fractions (1-ml) were collected, and proteins were esti- physins are often termed neurophysin I, II, III, etc. according mated with Folin-Lowry reagent (14). In a typical experiment to their migrations to the anode, but this identification is con- carried out with 33.5 mg of crude neurophysins, about 10.5 mg fused by the presence of truncated forms. Therefore, another of nomenclature, based upon the nature of the amino acids in po- material was not retained on the column; MSEL-neurophy- sitions 2, 3, 6, and 7, has been proposed, and two chemical types sin was recovered between 142-184 ml (2.7 mg); VLDV-neu- of neurophysins, termed MSEL- and VLDV-neurophysins, have rophysin, between 185-300 ml (5 mg); VLDV-neurophysin with been distinguished (9). (MSEL-neurophysins contain methio- contamination, between 300-400 ml (9.9 mg); and MSEL-neu- nine-2, serine-3, glutamic acid-6, and leucine-7; VLDV-neu- rophysin with a four-residue COOH-terminal truncation, be- rophysins contain valine-2, leucine-3, aspartic acid-6, and va- tween 452-568 ml (5.4 mg). Homogeneity was checked by disc line-7.) In each mammalian species, one protein of each type polyacrylamide gel electrophoresis. The overall recovery in has been discovered. proteins was 83%, neurophysins representing 68% of the ma- terial subjected to ion-exchange chromatography and 81.5% of The publication costs of this article were defrayed in part by page charge the recovered proteins. payment. This article must therefore be hereby marked "advertise- Amino Acid Sequences. Both neurophysins were either ox- ment" in accordance with 18 U.S.C. §1734 solely to indicate this fact. idized by performic acid (15) or reduced by dithiothreitol and 2839 Downloaded by guest on September 27, 2021 2840 Biochemistry: Chauvet et al. Proc. Natl. Acad. Sci. USA 80 (1983) alkylated with iodoacetamide (16) in order to cleave the disul- RESULTS fide bridges. They were split either with trypsin (EC 3.4.21.4) Amino.Acid Sequence of Human MSEL-Neurophysin. Amino or with staphylococcal proteinase (EC 3.4.21.19) (17), and re- .'acid compositions oftryptic peptides (T1-T8) and staphylococcal sulting peptides were separated by mapping on What- proteinase peptides (St1-St8) were determined. They account man 3MM paper under conditions as described (18). Peptides for a 93-residue polypeptide chain. Substitution of the usual were numbered as in Chauvet et al (18). Peptides, after acid Pro-60 by an alanine residue allows an additional tryptic cleav- hydrolysis (6 M HCI for48 hr at 105TC in sealed evacuated tubes), age at the level of the bond between Lys-59 and Ala-60 so that were analyzed on a Spinco model 120 B amino acid analyzer two tryptic peptides, T5 and T5', were found instead of the usual fitted with a high-sensitivity cell (19). Amino acid sequences of single T5 identified in MSEL-neurophysins of the other species peptides were determined.by a manual Edman procedure (20) (18). either directly for small peptides or after (i) cleavage by sub- Amino acid sequences were determined either directly or on tilisin, chymotrypsin, elastase, or staphylococcal proteinase, (ii) subfragments. Tryptic and staphylococcal proteinase peptides isolation of subfragments, and (iii) determination of their se- overlap: St1 gives the alignment T1-T2-T3-T4; St4, the align- quences. ment T4-T5; St5, the alignment T5-T5'-T6; St7, the alignment On the other hand, alkylated proteins also were subjected to T6-T7; and St8, the alignment T7-T8. The data permitted the automated degradation (21) in a SOCOSI model P 110 se- determination of the complete sequence (Fig. 1). quencer under conditions as described (18). Phenylthiohydan- Amino Acid Sequence of HumanVLDV-Neurophysin. Amino toin amino acids were identified by TLC (22). acid compositions of tryptic (T1-T6) and staphylococcal pro-

MSEL - NEUROPHYSINS

1 2 5 10 15 20 25 Bovine AlcMet-Ser- AspGlu LeuArilnCys -lCys,;lyprGlyGy Lys.Gly-ArgCys.PhenGly.Pro-Ser Ovine Porcine Equine Whole Rat TL._ M. Human

29 30 35 36 40 45 48 50 Bovine lie -Cys-Cys-Gly-AspGluA-eu.Gly.Cys-Phe.ValGly-Thr-Ala-Glu-AkbLeuArgCysGln-Glu-Glu-Asn-Tyr-Leu Ovine Porcine Equine Whale Rat Human -Ab .

55 60 65 70 72 75 Bovine ProSerIPro-Cys-GlnSerGly-Gln Lys-ProICyscGlyiS-Gly.ArCysAc-AbAaGly-Ile-Cys-Cys-Asn Ovine Porcine Equine Whale Rat Human Ain. Dh. ~V^I

8081 85 89 90 91 92 94 95 Bovine AspGlu-SeraCys.Val -ThrGlu-Pro.G;u.CysnArgGlu-Glr 1-Gly-PhlP-oArgArg-Val Ovine *Porcine __Ala-Ser-L Leu Ala Equine Whale Ala-Set Ala Rat __Ala PhePlw( 1 Leu-Thr Human -PH.-Hs I Ala FIG. 1. Comparison ofbovine (23), ovine (23), porcine (24), equine (25), whale (26), rat (27), and human MSEL-neurophysins. Solid lines indicate residues identical with those of bovine protein. Brackets span an assumed deletion. Downloaded by guest on September 27, 2021 Biochemistry: Chauvet et al. Proc. NatL Acad. Sci. USA 80 (1983) 2841

1 2 3 5 7 9 10 15 20 Bovine Ala-Val-Leu-Asp-Leu-Asp-Val-Arg-Thr-Cys-Leu-Pro-Cys-G ly-Pro-Gly-Gly-Lys-Gly-Arg Porcine Lys Equine - Ala Lys Rat - Ala Met - Lys Human - Ala-Pro Lys

20 25 29 30 35 40 Bovine Cys-Phe-Gly-Pro-Ser-I le-Cys-Cys-Gly-Asp-Glu-Leu-Gly-Cys-Phe-Val-Gly-Thr-Ala-Glu Porcine Equine Rat Ala Human Asn Ala-Glu

45 50 55 60 Bovine Ala-Leu-Arg-Cys-Gln-Glu-Glu-Asn-Tyr-Leu-Pro-Ser-Pro-Cys-Gln-Ser-oGly-Gln-Lys-Pro Porcine Equine Rat Human Ala

64 65 69 70 72 75 80 Bovine Cys-Gly-Ser-Gly-Gly-Arg-Cys-Ala-Ala-Ala-Gly-I le-Cys-Cys-Ser-Pro-Asp-Gly-Cys-His Porcine Glu Asn Arg Equine Leu Rat Thr Arg Human Val-Leu - Leu

81 84 85 86 87 89 90 93 Bovine Glu-Asp-Pro-Ala-Cys-Asp-Pro-Glu-Ala-Ala-Phe-Ser-Gln Porcine Phe Thr Equine Ala Ser - His-Asp Rat Thr Ser Human Ala Ala Thr

FIG. 2. Comparison of bovine (28), porcine (28), equine (29), rat (27), and human VLDV-neurophysins. Solid lines indicate residues identical with those of bovine protein.

teinase peptides (St1-St5) were determined. Here again a sub- whale (26), rat (27), and man. Rat MSEL-neurophysin is the stitution of Pro-60 by an alanine residue allows an additional only one, with the human MSEL-neurophysin, to have 93 res- tryptic cleavage and the presence of two peptides, T5 and T5', idues, but the COOH-terminal sequence appears more mod- instead of the usual single T5 found in VLDV-neurophysins of ified because a single arginine residue is found instead of the the other species. These peptides account for a 93-residue poly- two consecutive Arg-Arg. Because this basic tandem may be a peptide chain. In some cases, a 92-residue chain with the COOH- signal for some proteolytic processing enzymes, this variation terminal glutamine missing was observed. appears important. Compared with bovine MSEL-neurophy- Amino acid sequences were determined either directly for sin, human MSEL-neurophysin displays nine variations: seven the shorter fragments or with the help of the subfragments for substitutions and two deletions. Four substitutions are located the longer fragments. Staphylococcal proteinase peptides al- in the "invariant" part of the molecule from position 10 to 74 lowed the alignment of tryptic peptides: St1 for the alignment [namely, in positions 29 (Ala/Gly), 60 (Ala/Pro), 70 (Phe/Ala), T1-T2-T3-T4, St3 for T4-T5, and St4 for com- T5-T5'-T6. The and 72 (Val/Ile)] and three substitutions are in the hyper-vari- plete sequence of human VLDV-neurophysin is given in Fig. able from 89 to 95 2. part position [namely, in positions 89 (Phe/ Val), 90 (Phe/Gly), and 95 (Ala/Val)]. The two deletions are supposed to be located in positions 91 and 92. DISCUSSION Human VLDV-Neurophysin. Although two substitutions are Human MSEL-Neurophysin. The amino acid sequence found observed in the four distinctive positions [namely, positions 2 for human MSEL-neurophysin allows clearly the classification (Ala/Val) and 3 (Pro/Leu)], the human VLDV-neurophysin is of this neurophysin in the MSEL-neurophysin family. Not only easily recognized by the positions 6, 7, and 9 and by its COOH- the NH2-terminal sequence is typical of this family but also the terminal sequence 75-93. COOH-preterminal sequence 75-88. The usual hypervariable Compared with bovine VLDV-neurophysin (Fig. 2) there sequence 89-95 is modified here also: there is an apparent are 11 substitutions-2 in the "inter-family variable" sequence deletion of residues 91 and 92 so that the protein has only 93 1-9, 6 in the invariant sequence 10-74 [namely, in positions 25 residues instead of 95. The typical COOH-terminal Arg-Arg- (Asn/Ser), 29 (Ala/Gly), 30 (Glu/Asp), 69 (Val/Ala), 70 (Leu/ Ala/Val sequence is present. Ala), and 72 (Leu/Ile)], and 3 in the variable COOH-terminal Fig. 1 shows the comparison of MSEL-neurophysins from sequence 73-93 [namely, in positions 81 (Ala/Glu), 87 (Ala/Pro), seven species, namely ox (23), sheep (23), pig (24), horse (25), and 90 (Thr/Ala)]. Downloaded by guest on September 27, 2021 2842 Biochemistry: Chauvet et al. Proc. Nad Acad. Sci. USA 80 (1983) 1 2 3 6 7 9 10 15 20 25 MSEL Ak:Met SwAspLouGlbLlrysLro~-y-aC-l-PysP-Gqy.Glly.~Lys-GlyArg-Cys PheGlyroSer V LDV -Ala-Pro Asp-Va - Lys Asn

30 35 40 45 50 MSEL IlI-CysCys-Ala-AspGhosuGly-CysnPh.Ml-Gly.Thr.AoGhu.Ako-&l*ArgCys.Gln-Glu-GlukAsnTyr-Leu VLDV r-- . .

55 60 65 69 70 72 75 MSEL PrSer-ProCys-Gln.Ser-Gly-Gln-Lys.AboCys.Gly-SermGly.Gly-ArgCysAl

76 77 78 081 82 84 8586 87 88 89 90 91 92 93 MSEL AspGlu-Ser.Cys-Val Thr.Glu Pro-Glu.CysArgGlu.Gly-Ph.i-sArgArgAia VLDV ProAspGly His-A-Asp Ab A-Ab-Glu.Ablhr-PheSerGln FIG. 3. Comparison between human MSEL-neurophysin and VLDV-neurophysin. Solid lines in VLDV-neurophysin indicate residues identical with those of MSEL-neurophysin. Although the COOH-terminal Gln-93 can be missing in some termine whether there is a tandem of genes on the same chro- human samples, the five VLDV-neurophysins [namely, ox (28), mosome. pig (28), horse (29), rat (27), and man] characterized to date have Common Macromolecular Precursors. Possible common 93 residues. In contrast to the MSEL-neurophysin family, no precursors for neurohypophysial hormones and neurophysins basic residue(s) is found near the COOH-terminal end, and, have first been searched for by Sachs et al. (31). Direct isolation therefore, the putative processing enzyme that liberates VLDV- of these precursors from rat hypothalamic regions has later been neurophysin from a precursor may have a specificity different carried out by Russel et al. (32). They have found proteins with from that of the processing enzyme of MSEL-neurophysin pre- a Mr of 20,000, precursor being glycosylated and cursors. However, these basic residues may have been re- oxytocin precursor not. Seidah et al. (33) have isolated from hu- moved by a carboxypeptidase B after cleavage of the precursor, man pituitaries a with a Mr of 12,800, which was as observed for proinsulin (30). assumed to be a third moiety of the vasopressin precursor. Re- Comparison Between Human MSEL- and VLDV-Neuro- cent sequence determination by Land et al. (12) of a bovine physins. When the two human neurophysins are compared, 26 cDNA encoding for a 166-residue precursor has shown that this variations are found, and this figure is significantly higher than protein is constituted from NH2-terminal to COOH-terminal the one found for the other mammals (about 19-23). There are ends by a 19-residue signal peptide, [8-arginine]vasopressin, a five substitutions in the distinctive NH2-terminal sequence 1- linking sequence Gly-Lys-Arg, the MSEL-neurophysin pre- 9, five in the invariant sequence 10-74, and 16 in the variable viously identified (23), an arginine residue, and a 39-residue COOH-terminal sequence 75-93, with only Cys-74, Pro-83, glycopeptide previously identified in bovine extract (34). It can and Cys-85 remaining constant (Fig. 3). This latter part of the be assumed that oxytocin and VLDV-neurophysins are frag- molecule is clearly more subjected to variations than are the ments of a similar common precursor. NH2-terminal and central regions. Human neurohypophysial hormones have been isolated by An unexpected finding is the same substitution in both neu- Light and du Vigneaud (35), using the neurophysin complex rophysins in position 60: is replaced by alanine. All the procedure of Acher et al. (36); oxytocin and arginine vaso- other known neurophysins have a proline residue in this po- pressin have been identified by qualitative amino acid com- sition. In the same way, in position 29, is replaced by position determined after a two-dimensional paper chromatog- alanine in both neurophysins. In the rat the same replacement raphy (35). of glycine by alanine in position 29 was also observed in both The human oxytocic and vasopressor peptides now have been neurophysins (27). It is tempting to assume that a single deoxy- isolated by two successive molecular sievings on Sephadex G- nucleotide sequence encodes for the central part 10-74 of both 75 and Bio-Gel P-4, followed by paper chromatoelectrophoresis neurophysins and that the differential NH2- and COOH-ter- (20), and their quantitative amino acid compositions are in minal parts are added by different splicings. However, when agreement with those of oxytocin and [8-arginine]vasopressin, compared with the usual sequence 10-74, positions 30 and 69 respectively. are substituted in human VLDV-neurophysin and not in MSEL- If we assume that the general sequence deduced from the neurophysin, and position 70 is substituted by leucine in VLDV- cDNA for the ox vasopressin-MSEL-neurophysin-glycopep- neurophysin and by phenylalanine in MSEL-neurophysin. If tide precursor is valid in the case of man, [8-argininelvaso- we assume that there are two separate structural genes for the pressin may be separated from MSEL-neurophysin by a linking two precursors, some kind of interdependence in the mutations sequence Gly-Lys-Arg but, because of the COOH-terminal or in the expressions seems to exist although no mechanism can shortening of human MSEL-neurophysin (93 residues instead be proposed at the present time. It will be of interest to de- of 95), the junction with the human glycopeptide [which has 39 Downloaded by guest on September 27, 2021 Biochemistry: Chauvet et al. Proc. Natl. Acad. Sci. USA 80 (1983) 2843 residues (33) as does the ox glycopeptide (34)] could be differ- 6. Capra, J. D., Cheng, K. W., Friesen, H. G., North, W. G. & ent. Walter, R. (1974) FEBS Lett. 46, 71-74. The Two Evolutionary Lines. Two evolutionary lines of neu- 7. Robinson, A. G. (1975)J. Clin. Invest. 55, 360-367. 8. Chauvet, M. T., Chauvet, J., Acher, R. & Robinson, A. G. (1979) rohypophysial hormones have been traced in vertebrates, end- FEBS Lett. 101, 391-394. ing in placental mammals by oxytocin and by [8-arginine]va- 9. Chauvet, M. T., Chauvet, J. & Acher, R. (1975) FEBS Lett. 52, sopressin, respectively (2). These lines should correspond to 212-215. two lines of polypeptide precursors, each containing particu- 10. Schlesinger, D. H. & Audhya, T. K. (1981) FEBS Lett. 128, 325- larly a hormonal peptide and a neurophysin (12, 32). A dupli- 328. cation of the ancestral gene, assumed to have likely occurred 11. Brownstein, M. J., Russel, J. T. & Gainer, H. (1980) Science 207, 373-378. between cyclostomes and bony fishes (2), gave two structurally 12. Land, H., Schutz, G., Schmale, H. & Richter, D. (1982) Nature related polypeptides subjected to independent mutations. Ap- (London) 295, 299-303. parently mutations were rare in the hormonal moiety of the 13. Davis, B. J. (1970) Ann. N.Y. Acad. Sci. 121, 403-413. precursors because, for instance, all the bony fishes have iso- 14. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. tocin and , all nonmammalian tetrapods have meso- (1951)J. BioL Chem. 193, 265-275. tocin and vasotocin, and virtually all placental mammals have 15. Hirs, C. H. W. (1956)J. BioL Chem. 219, 611. 16. Crestfield, A. M., Moore, S. & Stein, W. H. (1963)J. Biol Chem. oxytocin and [8-arginine]vasopressin. The neurophysin moiety 238, 622-627. seems to have, in mammals, a nearly invariant central part. The 17. Houmard, J. & Drapeau, G. R. (1972) Proc. NatL Acad. Sci. USA NH2-terminal part (residues 1-9) appears to distinguish the va- 69, 3506-3509. sopressin-associated neurophysin (MSEL-neurophysin) from the 18. Chauvet, M. T., Chauvet, J. & Acher, R. (1976) Eur. 1. Biochem. oxytocin-associated neurophysin (VLDV-neurophysin). If this 69, 475-485. part is virtually invariant in the MSEL-neurophysin family, some 19. Spackman, D. H., Stein, W. H. & Moore, S. (1958) AnaL Chem. 30, 1190-1206. substitutions occur in the VLDV-neurophysin family, particu- 20. Chauvet, J. P. & Acher, R. (1972) Biochemistry 11, 916-926. larly in position 2 that can be valine or alanine. In human VLDV- 21. Edman, P.& Begg, G. (1967) Eur. J. Biochem. 1, 80-91. neurophysin, positions 2 and 3 are occupied by alanine and pro- 22. Edman, P. & Henschen, A. (1975) Protein Sequence Determina- line, respectively, instead of valine and leucine found usually. tion, ed. Needleman, S. B. (Springer, Berlin), pp. 232-279. However, on one hand, residues 6, 7, and 9 are those of the 23. Chauvet, M. T., Chauvet, J. & Acher, R. (1975) FEBS Lett. 58, VLDV-neurophysins; on the other hand, the COOH-terminal 234-237. 24. Chauvet, M. T., Codogno, P., Chauvet, J. & Acher, R. (1976) FEBS sequence (residues 75-93) shows clearly the similarity with the Lett. 71, 291-293. VLDV-neurophysin family (Fig. 2). 25. Chauvet, M. T., Codogno, P., Chauvet, J. & Acher, R. (1977) FEBS The question is whether the two types of neurophysins could Lett. 80, 374-376. be recognized in lower vertebrates. If, for instance, a MSEL- 26. Chauvet, M. T., Codogno, P., Chauvet, J. & Acher, R. (1978) FEBS neurophysin shares the same precursor with vasotocin, and a Lett. 88, 91-93. VLDV-neurophysin shares the same precursor with mesotocin, 27. Chauvet, M. T., Chauvet, J. & Acher, R. (1981) Biochem. Bio- the two phys. Res. Commun. 103, 595-603. putative evolutionary filiations vasotocin-vasopressin 28. Chauvet, M. T., Codogno, P., Chauvet, J. & Acher, R. (1979) FEBS and mesotocin-oxytocin would receive great support. Lett. 98, 37-40. 29. Chauvet, M. T., Chauvet, J. & Acher, R. (1981) Biochem. Bio- We thank Dr. Cl. Gros and Prof. Dray (Unite de Radioimmunologie, phys. Res. Commun. 100, 600-605. Institut Pasteur) for supplying human glands. We 30. Steiner, D. F., Cho, S., Oyer, P. E., Terris, S., Peterson, J. D. also thank Mrs. Danielle Thevenet and Miss Christine Jourdain for & Rubenstein, H. (1971) J. BioL Chem. 246, 1365-1374. technical assistance. This investigation was supported in part by grants 31. Sachs, H., Fawcett, P., Takabatake, Y. & Portanova, R. (1969) Re- from Centre National de la Recherche Scientifique (ERA 070563), cent Prog. Horm. Res. 25, 447-492. Delegation Generale a la Recherche Scientifique et Technique (80-7- 32. Russell, J. T., Brownstein, M. J. & Gainer, H. (1979)J. BioL Chem. 0294), and the Fondation pour la Recherche Medicale. 76, 6086-6090. 33. Seidah, N. G., Rochemont, J., Hamelin, J., Benjannet, C. & 1. Acher, R., Manoussos, G. & Olivry, G. (1955) Biochim. Biophys. Chretien, M. (1981) Biochem. Biophys. Res. Commun. 102, 710- Acta 16, 155-156. 716. 2. Acher, R. (1980) Proc. R. Soc. (London), Sect. B 210, 21-43. 34. Smyth, D. G. & Massey, D. E. (1979) Biochem. Biophys. Res. 3. Cheng, K. W. & Friesen, H. G. (1972)J. Clin. Endocrinol Metab. Commun. 87, 1006-1010. 34, 165-176. 35. Light, A. & du Vigneaud, V. (1958) Proc. Soc. Exp. Biol Med. 98, 4. Watkins, W B. (1971)J. Endocrinol. 51, 595-596. 692-696. 5. Foss, I., Sletten, K. & Trygstad, 0. (1973) FEBS Lett. 30, 151- 36. Acher, R., Light, A. & du Vigneaud, V. (1958)J. BioL Chem. 233, 156. 116-120. Downloaded by guest on September 27, 2021