THESYNTHESIS OF OXYTOCIN
June 20, 1954
Anal. Calcd. for CisHaaOsNa: C, 54.4; H , 6.58; N, 13.8. Found: C, 54.3; H, 6.66; N, 13.4. Ethyl Tosyl-L-isoleucylg1ycinate.-A solution of tosyl-Lisoleucyl chloride (prepared from 2.9 g. of tosyl-L-isoleucine and-phosphorus pentachloride) in r5 ml. of anhydrous ether was added slowly t o a suspension of 2.5 g. of ethyl glycinate hydrochloride and 3.5 ml. of triethylamine in 50 ml. of anhydrous ether and the mixture was allowed t o stand a t room temperature for 12 hours. The white precipitate was filtered off and washed with ether. After trituration with water to remove the triethylamine hydrochloride, 3.34 g. of product, m.p. 159-160°, was obtained; yield 89%. Another 0.25 g., m.p. 152-154', was afforded by extraction of the ether filtrate successively with water, dilute HCl, dilute aqueous KHCO, and water, followed by removal of the ether zn vacuo. Recrystallization of the second crop from ethanol raised the m.p. t o 158-160". For analysis, the tosyl dipeptide ester was recrystallized twice from ethanol and then melted a t 160-161". Anal. Calcd. for C1,H28O5NzS: C, 55.1; H , 7.07; N , 7.56. Found: C , 55.1; H , 7.30; N , 7.55. Tosyl-L-isoleucylg1ycie.-The tosyl dipeptide ester (1.6 g.) was dissolved in 9 ml. of ethanol and 5 ml. of 1 N NaOH was added. The mixture was warmed on a water-bath (70') for 1 minute, allowed t o stand a t room temperature for 2 hours, acidified to congo red paper with HCI and then concentrated zn vacuo t o dryness. The residue was dissolved in ethyl acetate plus a few drops of water and this solution was extracted twice with dilute aqueous KHCO3. Acidification of the aqueous phase t o congo red paper precipitated a crystalline product which wasofiltered off and washed T h e m . p . remained withwater; wt. 1 g. (71%), m.p. 185 unchanged after precipitation of the tosyl dipeptide from dilute aqueous KHC03 with HCl; [aIz1D -26.7" ( c 1.1, 0.5 -\' KHCO3). Anal. Calcd. for C I ~ H ~ Z O ~ X C,~ 52.6; S: H , 6.47; N, 8.13. Found: C, 52.3; H, 6.63; N, 8.20.
.
[ COVTRIBUTION FROM
THE
3115
Unchanged ethyl ester could be recovered by concentration o f t h e ethyl acetate phase i n vacuo. Methyl Tosyl-L-isoleucyl-L-1eucinate.-Tosyl-L-isoleucyl chloride (from 4.2 g. of tosyl-L-isoleucine and phosphorus pentadoride) was dissolved in 20 ml. of anhydrous ether and added t o a suspension of 3.6 g. of methyl L-leucinate hydrochloride and 5 ml. of triethylamine in approximately 75 ml. of anhydrous ether. After 12 hours the precipitate was filtered off, washed with $her and triturated with water; wt. 4.6 g., m.p. 145-146 . The ether filtrate was shaken successively with water, dilute HCl, dilute aqueous K H C 0 3 and water. Removal of the ether in vacuo and recrystallization of the residue from ebhanol gave an additional 0.5 g. of product, m.p. 146-147 ; over-all yield 85%. After two recrystallizations from ethanol, the m.p. mas 147148'. Anal. Calcd. for C Z ~ H ~ Z O S NC, Z S58.2; : H , 7.82; N, 6.79. Found: C,58.2; H,7.84; E,6.87. Tosyl-L-isoleucyl-L-leucine.-This tosyl dipeptide was prepared by saponification of 2 g. of methyl tosyl-L-isoleucyl-Lleucinate (dissolved in 20 ml. of ethanol) with 5.5 ml. of 1 N NaOH according t o the procedure described for the preparation of tosyl-L-isoleucylglycine; mt. 1 g., m.p. 152-154'. The ethyl acetate phase containing unreacted ester was concentrated in vacuo and the residual oil was dissolved in 8 ml. of ethanol and saponified in the same manner using 2 ml. f: 1 N NaOH. The second crop (0.3 9.) melted a t 151-153 . For analqsis, a portion was recrystallized from a mixture of ethyl acetate and petroleum ether and dried; m.p. 163165"; [ a I 2 ' D -24.7' ( C 1.1, 0.5 N KHC03). Anal. Calcd. for C I ~ H ~ O O J L SC, : 57.3; H , 7.58. Found: C, 57.2; H , 7.73.
The authors wish to thank Mr. Joseph Albert for carrying out the microanalyses reported herein. KEW YORK,S. Y .
DEPARTMEXT OF BIOCAEMISTRY, CORXELL UNIVERSITY MEDICAL COLLEGE 1
The Synthesis of Oxytocin1 B Y V I N C E N T DU V I G N E A U D , ? C H A R L O T T E
RESSLGR,JOHN hf. S W A N , 3 G. KATSOYANNIS4
CARLETON
w.
R O R E R T S AND
PANAYOTIS
RECEIVED APRIL28, 1954 A cyclic octapeptide amide ( I ) having the hormonal activity of oxytocin has been synthesized through the condensation of N-carbobenzoxy-S-benzyl-L-cysteinyl-L-tyrosine and the heptapeptide amide L-isoleucyl-L-glutaminyl-L-asparaginyl-Sbenzyl-L-cysteinyl-L-prolyl-L-leucylglycinamide ( IVa) to yield the protected nonapeptide amide V I followed by reduction with sodium in liquid ammonia and oxidation of the resulting sulfhydryl nonapeptide. IT'a was prepared by the condensation of S-benzyl-L-cysteinyl-L-prolyl-L-leucylglycinamide with tosyl-L-isoleucyl-L-glutaminyl-L-asparagine followed by removal of the tosyl group from the condensation product. The biologically active synthetic material thus obtained has been purified by countercurrent distribution and compared with natural oxytocin as t o potency, specific rotation, partition coefficients, amino acid composition, electrophoretic mobility, infrared pattern, molecular weight, enzymatic and acid inactivation and chromatography on the resin IRC-50. The synthetic material and natural oxytocin were also compared with respect to milk ejection and induction of labor in the human as well as rat uterus contraction in vitro. The crystalline flavianates prepared from the synthetic material and from natural oxytocin were found to have the same crystalline form, melting point and mixed melting point. All of these comparisons afforded convincing evidence of the identity of the synthetic product with natural oxytocin. This synthesis thus constitutes the first synthesis of a polypeptide hormone.
Oxytocin, the principal uterine-contracting and milk-ejecting hormone of the posterior pituitary gland,%was obtained from the latter in this Labora(1) A preliminary report of this work has appeared [V. d u Vigneaud, C. Ressler, J. M. Swan, C . W. Roberts, P. G. Katsoyannis a n d S. Gordon, T:HF~:~uB??~A, 75, 4879 (19531 1. ( 2 ) Appreciation is expressed to the Lederle Laboratories Division, American Cyanamid Company, for a research grant which has aided greatly in this study. Acknowledgment is also made t o Parke, Davis a n d Company, Armour a n d Company, a n d Eli Lilly a n d Company for placing a t our disposal posterior pituitary material used as starting material for preparations of the purified oxytocin. (3) Fulbright Scholar on Smith-Mundt grant-in-aid, on leave from Wool Textile Researcn Laboratory, C.S.I.R.O., Australia. (4) Fellow of State Scholarships Foundation of Greece. ( 5 ) T h e uterine-contracting activity of pituitary extracts was reported by H. H. Dale in 1906 [ J . Physiol., 34, 1631, a n d the milk. ejecting activity 11y I . O t t a n d J . C. Scott in 1910 [ P Y O CSOL. . ExO,
tory in highly purified and isolated as a crystalline flavianate.* The purification was effected by application of countercurrent distribution9~'"to posterior pituitary material which B i d . 'Wed., 8 , 481. For a discussion of numerous chemical and hinlogical investigations of t h e posterior pituitary gland including the early work on the subject, reference might be made to the review b y H . Waring and F. W. Landgrebe ["The Hormones," Vol. 2. G . Pincua a n d R. V. Thimann, Ed., Academic Press, Inc., S e w York, N. Y . , 1950, p p , 427-6141. (6) A. H. Livermore a n d V. du Vigneaud, J . B i d . Chem., 180, 303 (1949). (7) J. G. Pierce a n d V. d u Vigneaud, ibid., 182, 359; 186, 7 7 (19.50) (8) I. G. Pierce, S. Gordon a n d V. d u Vigneaud, ibid , 199, R2!1 (1952). (9) I,. C. Craig, i b i d . , 166, 519 (1944). (10) L. C. Craig, Anal. ChPm.,2 2 , 1348 (1950).
V.
3116
DU
VIGNEAUD.C. RESSLER,J. ?SWAN, *I. C. W. ROBERTS AND P.G. KATSOYANNISVol. 76
had received preliminary purification according to amide is made up of a cystine-containing cyclic the procedure of Kamm and co-workers." Amino pentapeptide amide linked t o a tripeptide amide. acid analysis by the starch column method of Moore One-half of the cystine residue possesses a free and Stein12 showed that hydrolysates of the highly amino group and is joined to the rest of the cyclic purified material contained leucine, isoleucine, portion of the molecule through its carboxyl group, tyrosine, proline, glutamic acid, aspartic acid, while the other half- of the cystine residue is conglycine and cystine in equimolar ratios t o each nected through its amino group to the rest of the other and ammonia in a molar ratio of 3 to any one cyclic portion of the molecule and through its carboxyl group to the tripeptide prolylleucylglycinamino acid.' The active principle appeared to be a polypeptide amide. Synthetic proof of this structure was of molecular weight approximately 1000.' l 1 Evi- obviously desirable. dence was obtained through oxidation with perI n this connection it may be recalled that in formic acidL4and desulfurization with Rancy nickells early experiments2' with less purified oxytocin that the polypeptide was some type of cyclic struc- preparations, it was found t h a t oxytocin could be ture involving the disulfide linkage. Further stud- reduced and reoxidized without appreciable inactivation. Treatment of the reduced active ies including determination of terminal groups, l 3 Ifi-lR degradation with bromine waterl9,l3and determina- material with benzyl chloride resulted in inactivation of sequence of amino acids by Edman deg- tion. Early attempts to regenerate active maradation and by partial hydrolysis with acid,lJ terial from the benzylated oxytocin by debenzylaalong with the assumption that glutamine and tion with sodium in liquid ammonia were not sucasparagine residues were present rather than their cessful. A reinvestigation of this problem reisomers, allowed structure I to be postulated for centlyZ2with highly purified oxytocin preparations led t o the isolation of the biologically inactive S,S'oxytocin. 15,20 dibenzyl derivative of oxytocin, which on treatCsH40H C?H$ ment with sodium in liquid ammonia according to 1 1 NH? 0 CH? 0 CH-CH3 the procedure of Sifferd and du Vigneaud for cleavIt I 'I 1 ing benzyl t h i ~ e t h e r sgave ~ ~ rise t o biologically CH?-&H-C-SH-CH-C-SH-CH active material. A study of the latter led to the IC=O I s eonelrrsion22that t h e ~ h o m o n e ~ h abeen d regenerated I i from its benzylated derivative. o 0 NH S I n terms of structure I, the reduction of oxytocin II II I would represent the opening of the ring a t the diCH?-CII-XH-C-CH-XH-C-CH-( CH, )?-COXH? sulfide linkage to give a disulfhydryl open chain I CH2 nonapeptide. Benzylation would then effect the c=o I conversion of the sulfhydryl form to a dibenzyl nonapeptide, with two residues of benzylcysteine taking the place of one of cystine in the original ~H-C-SH-CI-I-C-SH-CH~-CONH~ molecule. Debenzylation would convert the diI benzyl nonapeptide back to the sulfhydryl form CH-CH? CH* from which the ring would be regenerated through I CH(CH3)z 1 the formation of the disulfide by aeration. With the successful regeneration of oxytocin It may be pointed out that this octapeptide from the dibenzyl derivative of natural oxytocin, a (11) 0. Kamm, T. B. Aldrich I. W. Grote, L. W. Rowe a n d E. P. way was opened to an approach to the synthesis Bugbee, THISJOURNAL, 60, 573 (1028). (12) S. Moore and W. H. Stein, J . Biol. Chem., 178, 53 (1949). of oxytocin on the basis that structure I represents (13) C. Ressler, S. Trippett and V. d u Vigneaud. ibid., 204, 80 that of oxytocin. Thus, if the nonapeptide N(1953). carbobenzoxy-S-benzyl-L-cysteinyl-L-tyrosyl - L - iso(14) 1, hf. Sfueller, J. G. Pierce, H. Davoll and V. d u Vignenud, leucyl-L-glutaminyl-L-asparaginyl -S- benzyl-L-cysi h i d . , 191, 309 (1951). (12) R. A . Turner, J . G . Pierce and V. d u Vigneaud, ibid., 193, 369 teinyl-L-prolyl-L-leucylglycinamide(VI) could be ( I 9.51 1, synthesized, the removal of the carbobenzoxy and (10) €I. Davoli, R. A . Turner, J. G. Pierce and V. d u Vigneaud, ibid., benzyl protecting groups with sodium in liquid ani193,303 (1951). (17) H. G. Kunkel, S. P. Taylor, J i . , and V. d u Vigneaud, ibid., rnonia should give the reduced form of oxytocin and 200, 659 (1953). the latter upon oxidation should yield oxytocin. (IS) V. d u Vigneaud, C. Ressler and S. Tiippett, i b i d . , 206, 949 It may be noted that this synthetic approach (17J53). utilizes the type of procedure for the synthesis of (19) J. hl. hlueller, J. G. Pierce and V. d u Vigneaud, ibid.. 204, 857 (1953). cysteine- and cystine-containing peptides intro( 2 0 ) After the proposal of this structure and the degradative studies duced by du Vigneaud and in a synthesis of involved had been submitted t o J . Biol. Chem. (ref. 18) and after the glutathione. I n the latter synthesis N-carbobenzpreliminary report of t h e synthetic proof of t h e structure had been submitted t o THISJ O U R N A L (ref. 1). a n identical proposal for the oxy-7-glutamyl-S-benzylcysteinylglycinewas prestructure of oxytocin by 1%.T u p p y appeared. This proposal was based pared as the key intermediate in which the carboon the d a t a from this Laboratory on composition, molecular weight,
1
I /
terminal groups and on the cyclic structure involving the disulfide linkage, together with his independent studies on the sequence of amino acids involving partial hydrolysis with acid a n d with an enzyme (H. T u p p y , Biochim. et Biophys. A d a , 11, 449 (1953); H. T u p p y a n d 11. Sfichl, Monalsh. Chem., 8 4 , 1011 (1953)l. I t is extremely interesting t h a t the intcrixetation of the d a t a a n d the assumptions made were quite pnrnllel in h o t h laboratories.
(21) R . R.Sealock and V. d u Vigneaud, J . Phortnacol. ErO. Tliernp., 5 4 , 433 (1935). ( 2 2 ) S. Gordon and V. d u Vigneaud, Proc. SOL. E x p . Diol. . l f r , l . , 8 4 , 723 (1953). (23) R. H . Sifferd and V. d u Vigneaud, .I.Rial. Chem., 108, 753 (103.5). (24) V. du Vigneaud and Q. L Miller. i b i d . , 116, 469 (193B).
THESYNTHESIS OF OXYTOCIN
June 20, 1954
3117
SCHzCsHs (CHz)zCONHz
I
+
RNHCHCONHCHCONHCHCOOH
1
I
I I
CHz
NH~CHCON-CHCONHCHCO;i"2COPiHz I / !
CHzCONHz
CH( CHa )CzHs 11, R = Tosyl
tetraethyl pyrophosphite
RNHCHCONHCHCONHCHCONHCHCOI+--CHCONHCHCONHCH~CONH~ I CH(CHa)CzHs
I
I
CHZCONHZ
I V ~R , =H
I
CHz CHz \
+
/
I
CHzCH(CHa)z
CHCONH~HCOOH I
NHR' CHZCBH~OH V, R' = Carbobenzoxy
tetraethyl pyrophosphite
SCHzCsHs I
and the synbenzoxy group was used for protection of the amino thesis of ~-glutaminyl-~-asparagine~~ groupz6and the benzyl group for protection of the thesis of the tosyl tripeptide 1131 are described sulfhydryl group. The protective groups were separately. Condensation of I1 with I11 was effected by the finally removed to yield glutathione by cleavage use of tetraethyl pyrophosphite according to the with sodium in liquid a m m ~ n i a . ~ ~ , ~ ~ The preparation of the desired intermediate general procedure of Anderson, Blodinger and ~ ~ peptide synthesis. The "amide" (VI) for oxytocin was accomplished by coupling N- W e l ~ h e r for carbobenzoxy - S -benzyl - L - cysteinyl - L - tyrosineZ6J7 procedure32 was used with several modifications. (V) with the heptapeptide amide, L-isoleucyl- The protected heptapeptide amide IV was obtained in approximately 40% yield as a white amorphous L-glutaminyl-L-asparaginyl-S-benzyl-L-cysteiny1-Lprolyl-L-leucylglycinamide (IVa), prepared in turn solid.33 Removal of the t 0 s y 1 ~group ~ from IV from the coupling product of tosyl-L-isoleucyl-L- was effected with sodium in liquid ammonia, by glutaminyl-L-asparagine28(11) with S-benzyl-L-cys- which the benzyl group was also removed. The tein yl-L-prolyl-L-leucylglycinamide (111). The syn- resulting reduced compound was then benzylated *dE amirfe II1,ZS t k s p thesis d the ktmppt~ (30) J. M.Swan and V. du Vigneaud, ibid ,76,3110(1954). (25) M Bergmann and L. Zervas, Rcr ,66B,1192 (1932). (26) C. R. Harington and R. V. Pitt Rivers, Biochcm. J., 38, 417 (1944).
(27) C. W. Roberts and V. du Vigneaud. J. B i d . Chem., 304, 871 (1953). (28) Tosyl is used to designate the fi-toluenesulfonyl grouping. (29) C. Ressler and V. du Vigneaud. THIS JOURNAL, 76,3107(1964).
(31) P. G.Katsoyannis and V. du Vigneaud, ibid., 76,3113 (1054). (32) G. W. Anderson. J. Blodinger and A. D. Welcher, i b i d . , 74, 5309 (1952). (33) This protected heptapeptide amide has recently been obtained from glacial acetic acid as micro-needles, m.p. 236.5-237.5". (34) V. du Vigneaud and 0. K. Behrens, J. B i d . Chem., 117, 27 (1937).
with benzyl chloride3j in the same medium to yield compared with that of the purified natural oxythe heptapeptide amide IVa. The latter was con- tocin40 by assaying one against the other with the densed with N-carbobenzoxy-S-benzyl-L-cysteinyl-synthetic material and the natural oxytocin serving L-tyrosine (17) in the presence of tetraethyl pyro- interchangeably as the standard in the assay. phosphite to yield the desired nonapeptide deriva- The detailed procedure of assay as outlined in the tive N-carbobenzoxy-S-benzyl-L-cysteinyl-L-tyro-United States PharmacopeiaYg was rigorously syl - L - isoleucyl - L - glutaminyl - L - asparaginyl - S - followed. The potencies of the synthetic product benzyl - L - cysteinyl - L - prolyl - L -1eucylglycinamide are expressed as percentages of the potency of natural oxytocin with the corresponding standard (VI) in approximately 30% ,yield. VI was then treated with sodium in liquid error. I n four U.S.P. assays in which the synthetic ammonia, by which both the benzyl and carbo- material was used as the standard, it was found to benzoxy groups were removed. The reduced have 9670 (s.e. 7), 947, (s.e. 12), (s.e. 7) and material obtained after evaporation of the ammonia 9S70 (s.e. 3) of the potency of the natural oxytocin. was oxidized by aeration in dilute aqueous solution I n three U.S.P. assays in which the natural oxytocin was used a s t h e standard, t h e synthetic material a t p S 6 :5-and-tested-f.orbiological-activity. The reaction product was assayed by the rat was found t o have 90% (s.e. 9), 83% (s.e. 11) and uterine strip m e t h ~ d and ~ ~ by ~ ~the ' chicken vaso- 89% (s.e. 10) of the potency of the natural oxytocin. depressor method of Coon,38 which utilizes the As an over-all test of our assay procedure, a series property of oxytocin of lowering the blood pressure of three U.S.P. assays were run in which natural of tire fowl a d has-beerr adopted- by- tlre- W r r i t d oxytocin was-assay& again& iisclf. In-ex& of the States Pharmacopeia as the method of assay for latter three assays a solution of natural oxytocin oxytocin.39 Based on the assumption that the was divided into two parts, one of which was used synthetic material has a potency equivalent to that as the standard and the other as the "unknown." of the purified natural material,* the yield of bio- Values of 917, (s.e. l i ) , 909;) (s.e. i ) and 90(;1, were obtained for the "unknown." It is logically active material, as determined by the two (s.e. I) types of assay, ranged in several runs from PO to apparent from the assay results t h a t the potency 307, of the calculated amount from the protected of the synthetic product is close to that of the nonapeptide VI. The yield approximated closely natural material. The synthetic material was found to be as the yield of oxytocic activity obtainable under similar conditions from the S,S'-dibenzyl derivative effective in the induction of labor in the human as was the purified natural oxytocin.'i1 When the of natural oxytocin.22 For purification, the crude active material was synthetic compound and natural oxytocin were distributed by countercurrent distribution in the tested for milk-ejecting activity in the human, as system 0 . 0 3 ~ oacetic acid-sec-butyl alcohol for 300 has been reported,42they were found indistinguishtransfers. Xnalysis of the distribution indicated able in effectiveness. Spproximately 1 y of either that approximately 9OYo of the biologically active the synthetic or natural material given intramaterial was present in a single peak having a venously to patients induced milk ejection in PO-30 partition coefficient ( K 0.35) in agreement with that seconds. The synthetic material possessed the specific of natural oxytocin. The material in the peak tube .~D i= 1.0" (c 0.53, waterJ was not distinguishable in potency from one of our rotation [ C X ] ~ ~ -26.1 best preparations of natural oxytocin5 when the compared to [a]"D -26.2' (c 0.53, water) for natutwo were assayed repeatedly against one mother. ral oxytocin.g2 I n addition t o having, under the The contents of 1 G tubes representing the bulk of same conditions, a partition coefficient in agreethe material in the peak were combined, concen- ment with that of natural oxytocin in the system trated and lyophilized. Bioassay of the latter used for purification (0.0S70 acetic acid-s-butyl using the Coon method indicated that its potency alcohol) the synthetic substance had the same was close to but somewhat less than that of the partition coefficient ( K = 1.24) as natural oxytocin* in a second solvent system (0.01 >IT ammonia-smaterial in the peak tube. The synthesis of the biologically active material butyl alcohol).6 Determination of the amino acid has been repeated from the beginning on a some- composition gave the expected values. At two what larger scale and the results originally obtained different hydrogen ion concentrations the synthetic were confirmed. At the time of the first synthetic and natzral- material- ex-hibited- the same electrorun the tetrapeptide amide and the protected tri- phoretic mobility on paper.43 No differences were peptide were available only as white amorphous detected in the infrared patterns.4* The synthetic (40) T h e purified natural oxytocin when assayed afi:iiiist tlii. 1.5 1'. solids. In the more recent run it was possible to Standard Powder showed an activity of approxim;itelg i . i O - i 0 0 U.S.l' use these two intermediates in crystalline form and unitslrng. (ref. 8). T h e assignment of an exact unit:ifie < 1 1 course furthermore the countercurrent distribution was extremely difficult due t o the nature of the biuassay and the standard involved. carried through approximately 1500 transfers. (41) T h e uterine-contracting activity in the liumxii II-:IS tested The potency of the final synthetic product was through the kiiidness of I'rofessur R Gordoii I)ou 1,
