2292 was crystallized from ethyl acetate-ether in 9 7 x yield to afford the NO. 3MM paper, in different solvent systems, at concentrations ~ (c 2.05, dipeptide free acid XXIX, mp 144-145", [ a . I z 5 -29.5" ranging from 30 to 4000 p g ; in this case less than 1 of a tripeptide methanol). tricarboxylic acid could be detected when the chromatograms Anal. Calcd for C Z O H ~ L O ,C, : 58.88; H , 6.86; N, 6.86. were sprayed with bromphenol blue. The solvent systems used Found: C,58.75; H,6.78; N,6.99. were n-butyl alcohol saturated with water (descending), R f0.85 ; N-Carbobenzoxy-a-r-butyl-L-glutamyl-L-alaninePentachloron-butyl alcohol-water-acetic acid (4 :1 : 1) (ascending), RI 0.89; phenyl Ester (XXX). The dipeptide free acid XXIX (4.08 g, 0.01 n-butyl alcohol saturated with water and phenol-water (77: 23) mole) was converted into the pentachlorophenyl active ester as (two-dimensional ascending). Rr 0.89 and 0.72 ; pyridine-water described for the preparation of the a-dipeptide pentachlorophenyl (4:l) (ascending), Rr 0.79; phenyl-citric acid-disodium phosphate ester 111. Dipeptide active ester XXX was isolated in 50% yield buffer, pH 7.6 (100:20) (ascending), Rr 0.79. Similarly N-carboafter crystallization from methanol, mp 132-1 38". Two recrystalbenzoxy-a-t-butyl-~-L-glutamyl-L-alanyl-y-r-butyl-L-glutamic acid lizations from ethyl acetate-ether-petroleum ether raised the (XXXII) (0.030 g) was treated with 90% trifluoroacetic acid (1 ml) melting point to 141-142", [aIz6~ +3.86 (c0.88, chloroform). at room temperature for SO min. After the removal of trifluoroA m l . Calcd for C26H27N207C15: C , 47.50; H, 4.12; N, 4.27. acetic acid under reduced pressure, the residue was chromatoFound: C,47.68; H,4.37; N,4.24. graphed on the same paper and in the solvent systems used for the N-Carbobenzoxy-a-r-butyl-L-glutamyl-L-alanyl-y-r-butyl-L-glu- a-tripeptide, as described above. In most of the solvent systems, tamic Acid Methyl Ester(XXX1). Dipeptide active ester XXX (1.5 g, though a single spot was observed, the R fvalues of the y-tripeptide 0.00228 mole) was coupled with glutamic acid diester hydrochloride tricarboxylic acid were identical with those of the a-tripeptide. IV and the reaction mixture was worked up as described for the The most satisfactory solvent systems were pyridine-water (4 : 1) and a-tripeptide methyl ester V, yielding 7 2 x of the ?-tripeptide methyl phenol-citric acid-disodium phosphate buffer, pH 7.6 (100:20). ester XXXI, after crystallization from ether-petroleum ether, mp The Rr values with these solvent systems for the y-tripeptide tri69-72". Two recrystallizations from the same solvents raised the carboxylic acid were 0.68 and 0.62, respectively. In another -6.8" ~ ( c 2, chloroform). melting point to 72-73", [ a I z 6 experiment 3 mg each of cy- and 7-tripeptides VI and XXXII were Anal. Calcd for C a ~ H ~ s N ~ O C,l 59.2; o: H, 7.3; N, 6.9. Found: treated separately with trifluoroacetic acid (0.5 ml) for 50 min at C, 59.04; H, 7.46; N,7.16. room temperature. After the removal of trifluoroacetic acid, the N-Carbobenzoxy-~-r-butyl-L-glutamyI-L-alanyl-~-r-butyl-L-gluresidues were chromatographed separately and as a mixture. The tamic Acid (XXXII). Tripeptide methyl ester XXXI (0.607 g, conditions of chromatography were identical with those described 0.001 mole) was saponified as described previously and the saponabove; the Rt values of a- and y-tripeptide tricarboxylic acids were ified product, a white solid, was crystallized from ether-petroleum identical with those described above and they separated very well ether to yield 83 % of the tripeptide free acid XXXII, mp 118-119", into two spots. Since only one spot was observed at concentrations [aIz60 + l o " (~0.5,chloroform). ranging from 30 to 4000 pg for both a- and y-tripeptide tricarAtzal. Calcd for C29H43N301~: C , 58.71; H, 7.25; N, 7.09. boxylic acids, it was concluded that they were 99 pure a- and y-pepFound: C, 58.60; H,7.39; N,7.35. tides. Transpeptidation Studies. N-Carbobenzoxy-7-r-butyl-L-glutamAcknowledgment. This work was supported by yl-L-alanyl-?- t - butyl- L-glutamic acid (VI, 0.030 g) was disgrants from the National Institutes of Health, Public solved in 90% trifluoroacetic acid (1 ml) and left at room temHealth Service (G. M. 06579 and 08795). We wish to perature for 50 min. The trifluoroacetic acid was removed under thank Professor H. Horan for the infrared spectra. reduced pressure and the residue chromatographed on Whatman
A New Synthesis of Oxytocin Using S-Acyl Cysteines as IntermediateslS2 Iphigenia Photaki
Contribution f r o m the Laboratory of Organic Chemistry, University of Athens, Greece. Received December 16, 1965 Abstract: T h e fully S,N-protected derivatives (VI-IX) of the nonapeptide amide, L-cysteinyl-L-tyrosyl-L-isoleucyl~-glutaminyl-~-asparaginyl-~-cysteinyl-~-prolyl-~-leucylglycinamide, have been prepared by a step-by-step synthesis, using new N-protecting groups (the t-butyloxycarbonyl and the o-nitrophenylsulfenyl groups) in addition t o the classical carbobenzoxy a n d new S-protecting groups (the S-benzoyl a n d S-carbobenzoxy groups). T h e selective removal of the S-protecting groups from compounds VI-IX by methanolysis afforded the N-protected oxytoceines (X-XII). Oxidation of the free thiol groups of X a n d XI1 by 1,2-diiodoethane led t o the formation of N-carbobenzoxy- (XIII), a n d N-t-butyloxycarbonyl-oxytocin,(XIV), respectively. The peptide obtained after decarbobenzoxylation of the N-carbobenzoxy-oxytocin (XIII) was purified by countercurrent distribution and partition chrcmatography on Sephadex. T h e isolated material exhibited the chemical, physical, and biological activities of oxytocin.
xytocin was isolated in highly purified form by du Vigneaud, et a1.,3 and i t s structure was postulated4 and proved by ~ y n t h e s i s . ~Since the first,
0
(1) This investigation was supported by the Royal Hellenic Research Foundation, to which the author is greatly indebted. (2) A preliminary communication of part of this work has already appeared: I. Photaki, Experientia, 20, 487 (1964). (3) A. H. Livermore and V. du Vigneaud, J . Biol. Chem., 180, 365 (1949); J. G. Pierce, S. Gordon, and V. du Vigneaud, ibid., 199, 929 ( 1952).
Journal of the American Chemical Society
1 88:lO
classical synthesis of du Vigneaud, several other oxytocin syntheses were reported6- from various labora(4) C . Ressler, S. Trippett, and V. du Vigneaud, J . Biol. Chem., 205, 949 (1953); H. Tuppy, Biochim. Biophys. Acta, 11, 449 (1953); H. Tuppy and H. Michl, Monatsh. Chem., 84, 1011 (1953). (5) V. du Vigneaud, C. Ressler, J. M. Swan, C. W. Roberts, P. G. Katsoyannis, and S . Gordon, J . Am. Chem. SOC.,75, 4879 (1953); V, du Vigneaud, C. Ressler, J. M. Swan, C . W. Roberts, and P. G. I Z-Ile-Glu(NH&--Asp(NH~+Cys-Prc-Leu-Gly-NHz
Z-11-ONP,
2.
+
Z-Glu(NH2)-Asp(NH&-C!ys-Pro-Leu-Gly-NH~
/
\-j
1. HC1-CHsOH
EtsN
V
I
BZL
BZ
BZ
1. HBr-AcOH
VI
Bz Z-&ys-Tyr-Ile-Glu(NH~)-Asp(NHz~~ys-Pr-Leu-Gly-NH~ Z-Cys-ONP
2.
VI1
Z
BZ
BZ
I I NPS~ys-Tyr-Ile-Glu(NHz+Asp(NH,Z-Cys-PrH~
1. HBr-AcOH
VI11 BZ BZ BOC-Cys-Tyr-Ik-Glu(NHz)-Asp(NHd-&p-hc-Leu-Gly-NJ% I
1.
HBr-AcOH
2.
BOC-CYS-ONP
IX
$z
vIII
-*CHsONa
NPS-Cys-Tyr-Ile-Glu(NHz)-Asp(NH~)-Cys-R~Leu~ly-NH2
CHaOH
CHaONa
IX + -
XI
BOC-Cys-Tyr-Ile-Glu(NH2)-AsP(NHZ~ys-PrO-Leu-Gly-NH~
CH30H
x>-
ICHaCHzI
XI1
Z-dys-Tyr-Ile-Glu(NH1)-Asp(NHz&yS-RO-Leu~ly-NHz XI11
XII>-
ICHiCHzI
BOC-~ys-Ty-Ile-Glu(NHz)--Asp(NH~~ys-hO-Leu-Gly-NH~ XIV
HBr-AoOH
XI11
\
\
Cys-Tyr-Ile-Glu(NH2)Asp(NH~)-~ys-Pr~Leu41y-NH2 1.
-X
2.
HBr-AcOH
2
3
4
CaHbCO
5
6
7
8
9
oxytocin
oxidation
Z = CeHsCHzOCO
BZ
1
NPS = o-NOzCsHdS BZL = CsHsCH2
BOC
=
ONP
Photaki
(CH,)&OCO, pNOzCeH40,
/ New Synthesis of Oxytocin
2296 carbobenzoxy-L-asparagine (2.65 g, 0.01 mole) and pivalic acid in 30 mi of chloroform was prepared a s described in the literature.31 After allowing the anhydride solution to stand for 3 min a t 5-10", it was added with stirring at -10" to a chloroform (30 ml) suspension of S-benzoyl-L-cysteinyl-L-prolyl-L-leucylglycinamidehydrobromide prepared by decarbobenz~xylation~~ of 0.01 mole of compound I. Then triethylamine (0.01 mole plus a n amount necessary to neutralize the of hydrogen bromide) was added dropwise and the stirring was continued at room temperature until the solution was transformed t o a gelatinous mass. After allowing the mixture to stand overnight the solvent was removed in vflcuo. The residue was triturated twice in a mortar with cold 1 N hydrochloric acid, then with cold water, and twice with a cold 5 % solution of sodium hydrogen carbonate and water, alternately. The residue was dried in a desiccator over phosphorus pentoxide and was recrystallized by dissolving it in 200-250 ml of hot methanol, followed by the addition of an equal volume of water. After allowing the solution t o cool at room temperature and seeding, the product separated out in prisms. The yield was 4.5 g (61 %), mp 214-215", [ ~ ] ' Q D-71.5" (c 2.5, dimethylformamide). A m / . Calcd for CsaHabNiOgS: C , 56.82; H , 6.13; N, 13.25; S, 4.33. Found: C, 56.97; H, 6.33; N, 13.09; S, 4.19. N-o-Nitrophenylsulfenyl-L-asparaginyl-S-benzoyl-L-cysteiny1-Lprolyl-L-leucylglycinamide (Ha) was prepared by coupling N-onitrophenylsulfenyl-L-asparagine with S-benzoyl-L-cysteinyl-L-prolyl-L-leucylglycinamide in the same manner as described for compound 11. After being allowed to stand for 18 hr a t room temperature the mixture was triturated with ethyl acetate. The precipitate was filtered off and washed on the filter with water and with a solution of potassium hydrogen carbonate. The crude product IIa was dissolved in dimethylformamide and precipitated with ether. The yield was 55%, mp 208-210", [ C U ] ' ~ D-126' ( c I , dimethylformamide). AWL Calcd for Cs3H4?N809S2:N, 14.76; S, 8.45. Found: N, 14.95; S, 8.29. Experimental Section N,N '-Bis(carbobenzoxy-L-asparaginyl)-L-cystinyl bis(L-proly1-LFor the coupling reactions anhydrous reactants and dry solvents leucylglycinamide) (IIb). To a suspension of 0.185 g of protected were used; the ether was dry and free of peroxides. Evaporations pentapeptide 11 in 2.5 ml of absolute methanol, 0.51 mi of methawere carried out iii V L I C U ~at 35-40". Capillary melting points were nolic 0.5 N sodium methoxide was added in an atmosphere of determined for all compounds and are not corrected. Prior t o hydrogen with stirring at room temperature.'* Tlie substance analysis the compounds were dried between 20 and 80" under high dissolved completely during the first few minutes of stirring which vacuum over phosphorus pentoxide. was maintained for 20 min. On acidification with 0.4 nil of acetic Thin layer chromatography was performed on silica gel G acid and titration with 0.1 N iodine,'* 2.45 nil of iodine solution (Merck)" using the follo\Jing solvent systems: (a) I-butanol(98% of the theoretical amount) Mas consumed. Tlie mixture was acetic acid-water (100: IO :30),42(b) 1-butanol-acetic acid-waterconcentrated iii w c u o until most of the methanol h a s removed ; pyridine (3O:6:2J:20),l3 (c) I-propanol-ammonia 33 (67 :3?),41 it was diluted with wjater and extracted with chloroform. Preand (d) 1-propanol-water (64:36).4' Identification of the amino cipitation of the cystine peptide was achieved upon neutraliration acids contained in the intermediate protected peptides was perof the aqueous layer with a saturated sol~itionof sodium hydrogen formed by means of "two-dimensional" thin layer chromatography carbonate. The yield was 0.098 g (61 %), mp 191--lC)2' dec (after of tlicir hydrolysates in the solvent systems c and d. These chrosoftening at 163'), unchanged after recrystallization froin 2 ml of matograms were compared t o that of an equivalent mixture of the ethanol, [ c Y ] ' ~ D- 129' (c 1, dimethylformamide). Thin layer chroexpected amino acids. matography of an hydrolysate showed the presence of all the exN-Carbobenzoxg-S-benzoyl-L-cysteingl-L-prolyl-L-leucyIglycin- pected amino acids, [.e., aspartic acid, cystine, proline, leucine, and amide (I). To a cold solution of 3.2 g (0,011 mole) of L-proIyI-Lglycine. leucylglycinamide in 8 ml of dimethylformamide 5.5 g of p-niAmi. Calcd for Cj6H80N1101&: C, 52.97; H, 6.35; N, 15.44; trophenyl N-carbobenzoxy-S-benzoyl-L-cysteinate was added. S, 5.05. Found: C, 52.76; H, 6.62; N, 15.30; S, 4.86. The solution was allowed to stand at room temperature for 4 days. IV-Carbobenzoxy-L-glutaminyl-L-asparagin?.l-S-benzoyl-L-cysEthyl acetate was added and the solution was washed with cold 1 N teinyl-L-prolyl-L-leucylglycinamide(111). A. A solution of pentahydrochloric acid and with water until the water extracts were peptide hydrobromide prepared by decarbobenzoxylatioi1:'" of neutral to congo red paper. The ethyl acetate layer was dried over 5.18 g (0.007 mole) of compound I1 in 10 ml of dimethylformamide sodium sulfate and evaporated to dryness. The residue was diswas cooled to 0" and 2.96 g of p-nitrophenyl N-carbobenzoxy-Lsolved in 6 ml of dimethylformamide and crystalline compound I separated out upon addition of 400 ml of ether. The yield was 5.8 g (82"j,), mp 160-161 ', unchanged after recrystallization from 70% (44) (a) D. Ben-Ishai and A. Berger, J . Org. Chent., 17, 1564 (1952). D (c 2.5, dimethylformamide). methanol, [ C Y ] ~ ?-73.2" (b) Decarbobenzoxylation of the N-protected peptides was performed A i d . Calcd for CalHJgNjOiS: C, 59.50; H, 6.28; N , 11.19; throughout this work as follows. T o the solution of 0.01 mole of S, 5.12. F o u n d : C, 59.73; H, 6.23; N, 11.12; S, 5.20. carbobenzoxy peptide in anhydrous acetic acid, sufficient hydrogen
hydrogen ion concentrations. Oxytocin prepared as described above possesses an avian depressor activity37 of approximately 380 IU/mg as measured by Dr. W. Chan at the Department of Biochemistry, Cornel1 U n i ~ e r s i t y . ~This ~ value may be increased by ca. 15% if the water and acetic acid content of the lyophilized product are taken into account. The same sample of oxytocin after storage for some months at - 10" was further purified by partition chromatography on Sephadex according to the procedure of Yama~ h i r o . The ~ ~ material so obtained exhibited an avian depressor activity of ca. 380 IU/mg and an oxytocic activity39 of ca. 400 IU/nig as measurd by Dr. I. Krejci at the Research Institute for Natural Drugs, Prague. 40 Oxytocin has also been obtained from N-carbobenzoxy-oxytoceine by another route. The N-carbobenzoxy-oxytoceine was decarbobenzoxylated by hydrogen bromide in acetic acid. The resulting amorphous hydrobromide has been dissolved in water and after adjustment of the pH to ca. 6.8 it was oxidized by aeration. The oxidized solution was concentrated to a volume of approximately 50 ml and subjected to countercurrent distribution as described above. Oxytocin obtained by this route showed the same chemical, physical, and biological properties compared with the product described above.
N-Carbobenzoxy-L-asparaginyl-S-benzoyI-L-cysteinyl-r-proly1-L- bromide in acetic acid was added so that a final volume of 35 ml, and a concentration of 2.5 N hydrogen bromide in acetic acid werc obtaincd. leucylglycinamide (11). A solution of the mixed anhydride of After standing for 30-35 min a t room temperature the mixture \vas
(37) R . A. Munsick, W. H. Sawyer, aiid H. B. Van Dyke, Endocrinology, 66, 860 (1960). (38) D. Yamashiro, Nature, 201, 76 (1964). (39) Oxytocic assays were performed on isolated rat uterus suspended in Munsick's solution: R . H. Munsick, Endocrinologj>,66, 451 (1960). (40) Appreciation is expressed to Dr. I . I
