4596
ROLFPAULAND GEORGEIT.ANDERSON
VOl. 82
3000 and 2960 (aliphatic CH); 1620 ( N H ) ; 1600, 1560 and the residue heated t o boiling in 2 N hydrochloric acid and 1545 (C=C, C=K); 1400 (unassigned). (40 rnl.). T h e acidic solution was treated with Norit and Anal. Calcd. for C5HsN5: C, 43.15; H, 6.52; N, 50.33. the filtrate neutralized t o pH 6 with concentrated ammonium hydroxide. T h e solid that deposited was collected by Found: C,42.97; H, 7.01; N, 50.08. Hydrogenolysis of 5-Amino-4-( l-benzylhydrazino)-6- filtration, dissolved in acetone (10 ml.), a small amount of chloropyrimidine. A . Raney Nickel.-Raney nickel (2.0 insoluble material removed by filtration, and the filtrate evaporated to dryness in vacuo. T h e solid was dried in g . wet, washed twice with pH 7 buffer solution) was added to a mixture of 5-amino-4-( l-benzylhydrazino)-6-chloro- vacuo over PZOSa t 56" for 3 hours; yield 150 mg. (48.5%), in mp: fiH 1, pyrimidine (200 mg.) in pH 7 buffer solution (25 ml.) and m.p. 182-183' dec.; spectral data: , , ,A 334 (3.98); pH 7, 226 (11.2); -244 (9.33), 346 (2.78); the resultant mixture refluxed for 2 hours. The mixture was 3265 ( N H ) ; p H 13, 248 (8.46),280 (4.97); v in filtered hot, the filtrate evaporated to dryness in z~ucuo,and 2940 (aliphatic C H ) ; 1655 (C=X); 1600, 1575, 1550 and the residue extracted with chloroform ( 2 X 10 ml.). Evapo1480 (C=C, C=K); 740 and 700 (monosubstituted phenyl). ration of the combined extracts and sublimation of the L4naZ. Calcd. for ClaHjoCISj: C, 55.50; H , 33.85; residue a t 124: (0.1 m m . ) gave a white solid; yield 75 mg., N , 27.00; C1, 13.67. Found: C, 55.06; H, 4.02; S , 111.p. 130-132 . Recrystallization of this material from 27.01; C1, 13.87. benzene gave a white solid, m.p. 136-137'. This material mas identified as 5-amino-4-benzylaminopyrimidineh y ele1,2-Dihydro-l-methylpyrirnido[ 5,4-e]-as-triazine (XXIII, mental analysis and by comparison of its infrared and K = CH3) .--.4 solution of 5-amino-4-( 1-methy1hydrazino)ultraviolet spectra with those of 5-aniino-4-etliylaminopyrimidine (480 mg.) in formic acid (50 ml.) was refluxed pyrimidinelo; spectral data: Amax in mp (E X pH 1, for 0.5 hour and evaporated t o dryness under reduced 290 ( 1 2 . 7 ) ~PH 7, 255 (9.1); 290 (8.3); pH 13, 255 (9.61, pressure. The reddish-brown residue was then heated on a 290 (7.7); v in cm.-l: 3380 and 3200 ( N H ) ; 3060 (aromatic mater-bath under high vacuum until the color of the residue CH); 2920 and 2880 (aliphatic C H ) ; 1680 (ru"); 1600, had changed t o yellow; yield 580 mg. This material 1590, 1570 and 1510 ( C = C , C=N); 1460 (aliphatic C H ) ; decomposes rapidly without melting above 175" and leaves 750 and 700 (monosubstituted phenyl). a residue on combustion. A portion (200 mg.) of this niaterial was extracted with Anal. Calcd. for CjlH12K4: C, 65.98; H , 6.04; N, ether (3 X 100 ml.); the extracts were combined and 37.98. Found: C, 66.09; H , 6.04;. hT, . 28.20. evaporated to dryness in a stream of nitrogen to give a B. Palladium.--A solution of 5-aniino-4-( l-benzylslightly brown solid; yield 100 m g . , m.p. 202' dec. (when hydrazino)-6-~hloropyrimidine(1 .OO 9 , ) in 1 : 1 ethanoltaken rapidly from 175'); spectral data: , , ,A in mp_(e X water (100 ml.) containing magnesium oxide (1.00 g.) was pH 1, 239 (shoulder, 5.1), 331 (4.55); pH l , 239 hydrogenated over a 5y0 palladium-on-charcoal catalyst. 335 (unstable); pH 13, unstable; i j in cm.-': (shoulder), After the steady uptake of one equivalent of hydrogen, 3430.and 3220 (NH); 3040 (=CH); 2910 and 2810 (alithere vias a further, but much slower, absorption of hydrogen. T h e catalyst was removed by filtration and the phatic C H ) ; 1660, 1600 and 150G (C=C, C=X). . 4 n d Calcd. for C6H7Nj: C, 48.31; H, 4.73; 5, 46.96. residue was washed with ethanol (25 d.). A 5y0solution Found: C, 47.84; H, 4.90; K, 47.13. of sodium carbonate (50 ml.) was added t o the combined wash and filtrate, and the whole evaporated to dryness. l-Benzyl-l,2-dihydropyrimido[5,4-e]-as-triazine (XXIII, Extraction of the residue with chloroform ( 2 X 50 ml.) K = CHzCeH5).-A solution of 5-amino-4-( 1-benzylhydraand evaporation of the combined extracts gave 550 mg. of zino)-pyrimidine (460 mg.) in 98-10070 formic acid (25 rnl.) material, m.p. 101' with softening from 95". Sublimation was refluxed for 30 minutes, evaporated to d r y n e s , and the of this sample a t 105' (0.1 mm.) gave pure 5-amino-4-(1residue dissolved in 0.5 A' h3-drochloric acid (15 d . 1 . benzy1hydrazino)-pyrimidine(XXI, R = CHyC6Hs);yield The solution was neutralized with 1 N sodium hydroxide, and 510 mg. (59%), m.p. 100-lO1°; spectral data: A,, in mp the oil that deposited extracted with ether (3 X 50 inl.). (E X pH 1, 293 (shoulder) ( 8 . 3 ) ,314 (9.3); pH 7- Evaporation of the ether and trituration of the residue with 265 (7.6), 307 (8.4); pH 13, unstable; ij in cm.?: 3410 a small amount of methanol gave a light yellow solid; and 3300 (XH); 3040 (aromatic C H ) ; 1650 (NH); 1600, yield 220 mg. (46%). 1575, 1550 and 1490 ( C = C , C=N); 1460 (aliphatic C H ) ; -2small sample of this material was recrystallized from 730 and 700 (monosubstituted phenyl). benzene by the addition of Skellysolve C ; spectral data: in xnp ( E X 10-3): p H 1, 335 (4.3); pH 7 , 341 (3.9); Anal. Calcd. for CIlHl~N5:C, 61.37; H , 6.09; N, , , ,X pH 13, unstable; 7 in cm.-l: 3200 ( K H ) ; 2950 (aliphatic 32.54. Found: C,60.93; H , 6.08; N,32.29. l-Benzyl-5-chloro-l,2-dihydropyrimido[5,4-e]-as-triazine C H ) ; 1660 ( C = S ) ; 1600, 1590, 1570 and 1490 (c=c, (XXII).-A solution of 5-amino-4-( l-benzylhydrazino)-6- C=N); 775 and 700 (monosuhstituted phenyl). Anal. Calcd. for C11H1,N5: C, ti:3.98; 11, .1.!!3; N , chloropyrimidine (320 mg.) in formic acid (10 nil.) mas refluxed for 4 hours, evaporated t o a small volume in Z'UCZLO, 31.09. Found: C,68.91; 11,4.91; K,30.96.
N,N'-Carbonyldiimidazole, a New Peptide Forming Reagent' BY ROLFPAULAND GEORGE\V..\NDCRSON RECEIVED FEBRUARY 9, 1960 S,"-Carbotiyldiililiddzole was shown t o be a useful peptide forming reagent Conditions mere worked out for dc\.uidiW racemization in the formation of ethyl carbobenzoxyglycyl-~-phenylalanylglycinate, a sensitive case
I t has been demonstrated by \\Yeland and Schneider2 that peptide derivatives can be synthesized through acylation of the imidazole ring of methyl N-benzoyl-L-histidinate followed by reaction with the appropriate amine. Their method, however, (1) Preliminary communication G W. Anderson a u d K ,Paul, Tms 80, 4423 (19%). ( 2 ) T . Wieland a n d C, Schneider. Ann.. 580, 1.59 ( 1 9 3 ) ; see also 31 Lierginaiin and L. Zervas. %. p l i r s i u l . C h e m . . 175, 142 (lU28).
JOURNAL,
was not suitable for general use because of low yields. It occurred to us that a more direct agent for making acyl-imidazoles might be N,N'-carbonyldiimidazole. This would be a convenient reagent since the by-products, carbon dioxide and imidazole, are innocuous. The carbon dioxide evolution would provide a driving force for the reaction.
Sept. 5, 1960
N,N’-CARBONYLDIIMIDAZOLE
Models show t h a t the amide nitrogen of the imidazole can be brought into contact with the carboxylate suggesting a simultaneous bond formation and carbon dioxide elimination. An alternate path is, I + of course, a simple displacement by a second mole/=N + co, cule of imidazole. HE 1 k L It is possible to obtain peptides from acylimidaStaab‘ has shown that S,N’-carbonyldiimidazole zoles by reaction with an amino acid ester hydro(I) is reactive toward amines and alcohols forming chloride instead of the free base and even by using ureas and carbonates. Extending this, we found an aqueous sslution of an amino acid salt. I n the that acids react to give acyl-imidazoles, 2nd tlie re- latter case, the yields are lower. Carbobenzoxyaction can be used in peptide synthesis. Ethyl glycine, for example, was treated with the reagent carbobenzoxyglycyl-L-tyrosinate(II), for instance, in T H F iollowed by a solution of sodium DL-phenylalaninate in water to give a 53% yield of carbowas made in 95% yield by this method. Difficulties were encountered in following the benzoxyglycyl-DL-phenylalanine(IV) . X limitaliterature3 preparation of the reagent. These were tion to the last method is the low solubilities of solved by preparing N,K’-carbonyldiimidazole some amino acid salts in water. Likewise a limitafrom imidazole and phosgene in rigorously dried tion on the use of amino acid ester hydrochlorides is benzene. It was important to use a slight excess their insolubilities in organic solvents. The importance of using precise amounts of asof imidazole to prevent contamination of the product by half-reacted phosgene. Likewise it was sayed N,S’-carbonyldiimidszole was overlooked found important to remove all the imidazole hy- a t first since the examples studied gave reasonablydrochloride from the reagent. Contaminated N,- good yields and little trouble was encountered in purifying them. It was assumed t h a t the desired ?; I-carbonyldiimidazole gave poor yields in peptide synthesis and was deliquescent. The crude rea- reactions went a t a much faster rate than side reacgent was assayed for carbon dioxide on hydrolysis; tions. As more reactions were studied it became the purity ranged from 91 to 100yoand the melting evident, however, that while acylimidazole formation was relatively fast and complete, the reaction of point of the better material was 113-115’. X typical peptide reaction was run by treating a the acylimidazole with an amine to form a peptide solution of 0.010 mole of an acylamino acid in 10 was about as fast as the reaction of amine with unml. of tetrahydrofuran (THF) with 0.010 mole of reacted N,N’-carbonyldiimidazole to form ureas. reagent, adjusted for that quantity from the assay. Ureas being difficult to separate, it was best to avoid The evolution of carbon dioxide was immediately their formation through the use of precise amounts observed. After an hour the desired amino acid or of assayed reagent. On using an excess of reagent peptide ester was added in 0.010 molar quantity. in a preparation of ethyl carbobenzoxyglycyl-LX small amount of heat was evolved. The reaction tyrosinate, the crude product had a poor melting has been worked up after 15 min., but longer stand- point and was difficult to purify. By the time a ing is probably beneficial. The product was isolated reasonable melting point was achieved, the yield by removing the solvent under vacuum followed by had dropped to 64% as compared to 95y0obtained washing the residue with acid, saturated bicar- previously. bonate and finally water. I n another experiment to determine tlie best solIt is critical t o maintain absolute dryness during vent for the reaction little difference was noted the reaction of the acid with the r\’,N’-carbonyldi- among six tried. The Oilly one out of line was imidazole since the reagent decomposes almost in- methylene chloride which gave a higher yield. This stantly on contact with water. The intermediate difference could be accounted for by the fact t h a t acylimidazole is stable toward hydrolysis for short the solvent was boiled off rather than removed a t periods of time. room temperature and the heating made the reacT o get a better picture of the course of the reac- tion go farther. T o test this, t-butyl trifluoroacetioil, cdrbobenzoxyglycine mas treated with the tylglycyl-L-prolinate (IT) was obtained in 53% reagent and the carbobenzoxyglycylimidazole (111) crude yield when prepared in the usual fashion. was isolated and analyzed. It was treated further In another run, after the addition of the t-butyl Lwith ethvl L-tyrosinate to give an 83% yield of prolinate, the reaction mixture was heated oti a ethyl carbobenzoxyglycyl-i,-tyrosinate.A possible steam-bath for 1 hour, giving a higher crude yield, mechanism for the acyl-in-idazole formation would 61yo. However, heating is not recomniended be an attack on the carbonyl carbon of the reagent as a general procedure since it increascs side reacby the oxygen of the acid or a carboxylate ion fol- tions. lowed by elimination of a n imidazole molecule to In several other experiments (see Table I) the give best conditions for the formation of one particular dipeptide were worked out. Mixing the acid and amine followed by the reagent gave no isolatable product. Removal of the reaction solvent at room temperature before work-up gave a slight (2%) improvement, The best yield was obtained in a run where the reaction of reagent with acid was perL, mitted t o go for at least 30 minutes at room temperature before the amine was added.
ROLFPAULAND GE(IRGE W. ANDERSON
4598
Vol. 82
TABLE I N,N'-Carbonyldiimidazole3(I).--A11 equipment used in this reaction was dried iii an oven a t 115'. A 2.2-1. portion VARIATION OF REACTION CONDITIOKS IN THE FORMATIOX OF of benzene was placed on 5-10 g . of calcium hydride and 2.0 Z.gly-tyr .OEt(L) (11) 1. was distilled into a N i t e r , 3-necked flask containing 112 Reg . (1.65 moles) of imidazole and equipped with a stirrer aiid crystd. drying tube. When the benzene had been distilled over, yield, the condenser was replaced b y a gas inlet tube and the misVariation Solvent 70 A f . p . , O C . ture heated and stirred t o obtain a clear solution. A 38.6Used excess I m C O 1°F 63 125-128 g . (0.40 mole) quantity of phosgene was collected as a liquid Heated before a d d n . of H , t y r . O E t ?'ina Dry Ice condenser arid dibtilled, by warnling by hand, (L) THP 59 127-129.5 into the reaction mixture. -% precipitate formed at this Heated a f t e r addn. of H . t y r . O E t point. After all the phosgene was distilled, the system w a i (L) THF 70 127-128 flushed with dry nitrogen for a few minutes, cooled t o 30" >-one THF 71 137-128 and let stand until the liquid phase was clear (from 1 t o 24 Sone (CHaOCH2)? 67 127-128 lir.), then warmed t o 50" anti filtered. Most of the betiBoiled solvent oR CHzCIz 78 126-127 zene was removed under reduced pressure a t 60". T h e None Pyridine i4 128-127 rcsulting slush was cooled and quickly filtered. Overnight None DMF titi 126 A-128 drJ-ing under reduced pressure gave 50.4 g . of N,X'-carNone (Et0)zPOH 73 136.5-128 bonyldiiinitlazole, a 77y0yield. T h e material, on a capillary Used H B r . H . t y r . O E t ( L ) ; boiled melting point, shrank a t I l l o , started t o liquify a t 1 1 3 O , off solv. CHzClz 80 126.5-127.5 was a cloudy liquid a t 115' and all clear a t 117" (lit.3 Calcd. arnt. of ImzCO used here 1n.p. 115.5-116°). T h e reagent was assayed for Carl, a n d below CN2Cl? YO 12-1-1?5.5 dioxide on hydrolj-sis. A calculated C o n value of 27.2 Removed sol!.. before workup; 2 compared with the 26.7 i 0.5yo COa found b y aualy recrystn. THF 80 121-126.5 showed the final product t o be 98 =k 276 pure. silver K e p t solvent; 2 recrystn, THF 84 125-127 nitrate test 011 a n acidified sample in water solution was R a n a t -13"; 4 recrystn. I1 hl F 79 127-127 n?gative. On several runs yields of 75-8270 and purities Combined Z.gly.OH H.tyr.OEt TIIF Son-crystalfrom 91 t o 99.57, were obtained. (L) then added lrn?CO liaable oil Ethyl Carbobenzoxyglycyl-L-tyrosinate(11).-To a solu30 min. for 1 9 t step, 12 hr. Znd, tion of 2.09 g. (0.010 mole) of carhobenzozyglycine' in 10 vac. off solv., 1 recrystn. THF 95 127-128 ml. of dry T H F was added 1.62 g . (0.010 of s,s'30 min. f o r 1st s t e p , 12 hr. 2 n d , cnrhoriyldiimidazole. One-half hour later, 2.09 g . (0.010 vac. off solv., 1 recrystn. CH2Clz 89 12.;-127 e ~ added. After standing mole) of ethyl ~ - t > - r o s i n a t was overnight, t h e THF was removed by an air stream and 50 Racemization was investigated in the reaction ml. of 1 '1' hl-drochloric acid was added. Cooling of the of carbobenzoxyglycyl-L-phenylalanineand ethyl solution gave a solid. This was washed with water, triglycinate, a very sensitive case.4 I t was necessary turated with 20 ml. of 570 sodium bicarbonate solution, filtered and again washed with water. T h e product was to determine the purity of the phenylalanine deriva- dried in a steam cabinet and yielded 3.93 g . (98%), havirig tive since commercial L-phenylalanine may contain :I melting point of 125.5-127'. Recrystallization from 50D/G a s x u c h as 5.&! DL-material. The sample used for ethanol gave 3.79 g. (95%) of material with a melting poi112 1.0 the racemization study was carefully purified by of 127-128". T h e optical rotation, [ c Y ] * ~ D-k 18.2 ( r ,5, abs. ethanol), and the melting point compare favorably fractional crystallization. It was checked for race- with Lit. vxiues'oof I . 1 2 4 ~ 19.3 i 0.1', 171.p. 125-126.5", mic content and shown to be pure L-isomer. In a n d yield 680jc. running the test on N,N'-carbonyldiimidazolea t I ~ Jnaiting 15 tniii. before t h e addition of thc ctll? 1 Lroom teinperature in T H F , 5% racemic material t i rosinate an 83$7, x-ield of r e u s - s t a l k e d materid, 111 I). was found in the tripeptide thus produced. How- 12C,-12io, !vas obtained.
+
*
+
N-( Carbobenzoxyglycy1)-imidazole
ever, ii the reaction was run iii DMF (diinethyl- ~Itljimidaznle (1.91 g . , 0.01 1 mole, forrliaiiiitlc) t i t - 1 O 3 less tli:~ii O.,??;) racemic nia- g. fO.010 mole) of carbnbenznxygl i i i 10 ml. of d r y 'TIIF. After the cffe tcrial was detectctl. solution was diluted with 30 nil. nf The possibility ol using r)tlicr r e a p i t s siiiiilar i o erystallizd and w a s cnllecterl b y filt~itioii. I t was tv:isl:(.d N,S'-carl~r~rivldii~iiidatzole was investigated. X,- w i t h etlier and drird i n a desiccator t o gi\-e I.U2 g.. 111.1'. N ' - C a r t ~ o r i ~ l d i b c i i ~ i r i i i ~ 1W;LS ~ ~ zmade ~ l e s and foulid 1 1 .?--I 18'. 'rlie niother liquors yic.lrletl a11 nt1tliti~ili:d to be less reactii-e t.han the irnidaznle c ~ o n i ~ ~ o u ~ i(c)l,.I 6g . , m . p . 117--118", giving R total !.ieItI o f XO%. 0 1 1 from TIIF-etiier, I .57 g . (60?0), 111.1). Scveral attelripts to make r\.,Tu"-sullotiyldiiniida- rccryst:illization 1 l~l-120°,wasobtairied. zole failed. 'Table I1 gives yields for other pepfor C131H13K:jOI:C , ( i O . Z ; 11, 5.0-5; s , tides niadc with N.K'-carbonyldiimidazole aiid C, 60.41; 11, j.31; N,16.04. coinpares them to literature preparations. An partially decomposed o n standitlg in t l ~ c attempt has been made to quote the higher and opcn :veriiight as determined b y a drop in nielting point t o 80-88 . A vialed sample decomposed slightly over t h e oftinies the highest literature yields. of a month t o 113-118'. Perhaps this was due t o The use of K,:,"-carbonyldiimidazole in the spn- tperiod h r inoisture in the air of t h e vial. thesis of angiotensin is currently under investigaX 0.259-g. (0.001 mole) sample of Z.gly.Im was treated tion. wit!] 0.209 g. (0.001 mole) of ethyl L-tyrosinate in 2 ml., Of Acknowledgments.--We thank R4r. L. Brancone THF. After standing overnight water was added giving olution and a n oil. The oil was crashed wit11 and staff for analyses and Mr. W. Fulmor and acid and mater. Drying in an oven gave 0.362 staff for optical rotations. of material, m . p . 121-126'. Recrystallization 1 g. yield), 111.p. 1?8-186', of ethyl carboExperimental benzosyglycyl-L-tyrosinate.
Melting Points.-All melting points were run on a calibrated Fisher- Johns block unless otherwise indicated. Solvents-All solvents used were dried; T H F u-as distilled from calcium hydride; DMF was dried b y azeotroping with benzene and storing over anhydrous magnesium sulfate.6 (4) G. W. Anderson and 17. 51. Callahan, THISJ O U R K A L , 80, 2002 (1958). ( 5 ) H. A. S t a a b and G. Seel, A n n . , 612, 187 (1968).
( 0 ) A. B. T h o m a s a n d E. G. Rochow, THIS JOURNAL, 7 9 , 1843 (1957). ( 7 ) hI. Bergmann and L. Zervas, B o . , 66, 1192 (1932). ( 8 ) I n some of t h e earlier experiments t h e necessity of having each b a t c h of X.N,-carbonyldiimidazole assayed was n o t recognized. (0) E. Fischer, Bey., 34, 433 (1901). (10) J . R . Vaughan, Jr., a n d R . L. Osato, T m s J O U R N A L . 74, li7C (1952).
N,N'-CARBONYLDIIMIDAZOLE
Sept. 5 , 1960
4599
TABLE I1 PEPTIDES DERIVATIVES PREPARED USINGN,K'-CARBOSYLDIIMIDAZOLE~ Pure yield, Product
NO
1 2 3 4 J
6 7 8 9 10 11 12
Z.gly-tyr.OEt(L)" B .phe-gly.OEt(L)" Z .gly-phe.OEt(DL)' Z .gly-phe.OEt(DL)' Z .gly-phe.OH(DL) Z.gly-phe.OH(L) Phth.phe-gly-gly.OEt(m)' Z .ala-gly.OEt(L) Z .gly-leu .OH( L ) ~ Z.gly-gly.OEt"' TFA.gly-pro .OtBu( L)' Z.phe-tyr.OEt(L,~)* L-HZ
% 95 78 80 82 55 40 56 65 68 60 51 57
I
h4.p..
oc.
127-128 8849.5 90-91 91-92 162.5-163.5 126.5-127.5 163-164.5 98-99 103-101 82.5-83 90-91 158-160
13 2 .asp-gly . O C H Z C ~ H ~ ( L ) ' 35 177-1 79 27 184 14 Zaileu-his OMe(L,L) 87 119.9-120.3 1 5 Z .gly-phe-glyOEt(L)w
Lit. yield,
Recrystn. solvent
[U]r'D
50% ethanol Pet. ether Benzene-pet. ether Benzene-pet. ether 50% ethanol Water Benzene-pet . ether EtOAc-pet. ether EtOAc-pet. ether 50% ethanol Methylcyclohexane 50% ethanol
Water Methanol-water Abs. ethanol
(6)
-I-18.2 f 1 . 0 " (abs. E t O H ) - 4 . 2 =k 1 . 2 ( E t O H )
4-40.7 & 1 . 7 (abs. E t O H ) -21.7 & 0 . 5 (abs. E t O H ) -18.2 i 0 . 5 (1 N XaOH)
-83 ( E t O H ) - 9 . 2 & 0 . 5 (EtOH)
+1 . 8 ( D M F ) -12.2 & 1 . 2 5 ( E t O H )
c/o
Lit. F.P., C.
68h 125-126.8 81d 86-87 86' 9 1-92 86' 91-92 € 1 3 ~ 160-182 12i.j" 162-163 77' 100 62' 104 83 86-87" 69' 89-90 46b 189-160'
50' 50" 96"
181-183 186-189 120-120.5
Z = Carbobenzoxy. * Ref. 10; mixed anhydride method. B = t-butyloxycarbonyl, G. W. Anderson and A. C. McGregor, THISJOURNAL,79, 6183 (1957). Recrystallized t o 35% yield, m.p. 89.5-90", from ethyl acetate-petr. ether; pyrophosphite method. e Free H.phe.OEt(DL) used. f Ref. 12; mixed anhydride method. 0 Using HBr.H.phe.OEt( D L ) ' directly. ~ Ref. 4. a P h t h = phthaloyl; from Phth.phe.OH and H,gly-gly.OEt. j M. Bergmann, L. Zervas, J. S. Fruton, F. Schneider and H. Schleich, J . Biol. Chem., 109, 325 (1935); acid chloride method. The ethyl ester is an oil and was saponified; the yield is the over-all yield. H. S. Goldschmidt and H. Lautenschlager, Ann., 580,68 (19.53); phosphorus trichloride method, followed by saponification of the ethyl ester thus formed. Using HC1,H.gly.OEt directly. Ref. 14a using the pyrophosphite method. 0. Siis, Ann., 572, 104 (1951), reports m.p. 82.5-83" and a 25y0 yield using the phosphorus trichloride method. T F A = trifluoroacetyl. p Ref. 13; pyrophosphite method, crude yield. * Amorphous. I. J. S. Fruton and M. Bergmann, J . Biol. Chenz., 145,262 (1942), using 2.phe.Cl obtain 4670,m.p: 162'. * Amine used as benzenesulfonate; product was recrystallized three times. ' H. K. ,Miller and H. Waelsch, Arch. Bzochem. Biophys., 35, 176 (1952); diethyl chlorophosphite method. * H. Schwarz, F. M. Bumpus and I. H. Page, THISJOURNAL, 79, 5697 (1957); mixed anhydride method. Ref. 4; pyrophosphite method. The general procedure used for the synthesis of Z.gly-tyr,OEt(L) was used in each case.
Ethyl Carbobenzoxyglycyl-DL-phenylaiadnate.-Ethyl DLphenylalaninate was prepared from 2.75 g. (0.010 mole) of ethyl DL-phenylalaninate hydrobromide" on treatment n i t h a n excess of triethylamine in T H F followed by filtration and concentration. A solution of 2.09 g. (0.010 mole) of carbobenzoxyglycine in 10 ml. of d r y T H F was treated with 1.62 g. (0.U10 mole) of K,N'-carbonyldiimidazole. When the effervescence ceased, the previously prepared ethyl DL-phenylalaninate was added. After 15-30 minutes, the mixture was air-dried and 50 ml. of 1 N hydrochloric acid added. On cooling, a solid formed. This was collected, triturated first with water, then with a 5% sodium bicarbonate solution, and again with water. Drying gave 3.20 g. (83y0 yield) of ethyl carbobenzoxyglycyl-DLphenylalaninate with a melting point of 90-91". The material was recrystallized from benzine-petroleum ether, resulting in 3 . 1 0 ~ (80% . yield). product had a melting point of 90-91 , lit. value12 91-92 . This experiment was repeated except t h a t the ethyl DL-phenylalaninate hydrobromide was added directly t o the reaction mixture without first removing the H B r with triethylamine. An 82% yield of a product having a melting point of 91-92' was obtained by this method. Carbobenzoxyglycyl-DL-phenylalanine (IV).-X solution of 2.09 g. (0.010 mole) of carbobenzoxyglycine in 10 ml. of tlricd T H F was treated with 1.62 g. (0.010 mole) of S , S ' carbon>-ldiiinidazole. IVhen the effervescence stopped, a solution of 1.64 g. (0.010 mole) of DL-phenylalanine in 10 nil. ( J f 1 S sodium hydroxide was added. After 15-30 minutes, 30 ml. of 1 A hydrochloric acid was added. The oily liquid thus formed crystallized in a few minutes. The product, carbobenzoxyglycyl-DL-phenylalanine, was collected, washed with water and dried t o give 2.51 g. (707, yield) of material having a melting point of 158-160'.
TF
(11) Made f r o m Dr.-phenylalanine, ethanol a n d d r y hydrogen bromide; m.p. 131-132'. A f d . Calcd. for CIIHlaKO?Br: C, 48.19; H , 5.88; N, 5.11. F o u n d : C . 48.32; H, 0.01; N,5.39. (12) J. R. Vaughan, Jr., a n d R . L. Osatn, THISJ O U R N A L , 73, 5553 (1951).
The product was recrystallized from 35 nil. of 507; ethanul giving 1.97 g. (557, yield) of the dipeptide with a melting point of 162.6-163.5", lit. valuelo 160-162". The above experiment was repeated using L-phenylalanine in place of DL-phenylalanine. X 407, yield of material was obtained which had a melting point of 126.5-127.5' and a n optical rotation [CY]*~D 40.7 f 1.7" (c 2.9, absolute ethanol); the corresponding lit. values4 were m.p. 127.5' and [ c Y ] ~ ~ D38.8 f 0.5" ( c 5, absolute ethanol). &Butyl Trifluoroacetylglycyl-L-prolinate (V).-The solution resulting from the reaction of 1.71 g. (0.01 mole) of trifluoroacetylglycine's in 10 ml. of T H F with 1.74 g. (0.01 mole, 93y0 pure) of iX,K'-carbonyldiimidazole was permitted to stand for a n hour. Then 1.71 g. of t-butyl Lprolinate14 was added. After standing overnight, the solvent rvas removed under vacuum and the residual yellow oil triturated with 35 ml. of N acid. The product crystallized readily. The colorless solid was collected, triturate? with water and dried, giving 1.70 g. (53y0), m.p. 89-90 . One recrystallization from methylcyclohexane gave 1.65 g . (517, yield) of material, m.p. 90-91". I n a second run the reaction was warmed on a steam-bath for 1 hr. after the addition of the amine instead of leaving i t overnight a t room temperature. In this case, the crude yield was 1.97 g. or 61%, m.p. 89-90', [ a I z 5 D - 83" (c 2 , ethanol); lit.14m.p. 89-90", Racemization Studies with N,N'-Carbonyldiimidazo1e.The reaction of carbobenzoxyglycyl-L-phenylalaninewith ethyl glycinate is very sensitive t o ra~emization.~.'6 I t was ~ h o r n nthat ~ ~ ~using ~ tetraethylpyrophosphite as the peptide-forming agent gave no racemization. T o test the purity of the starting dipeptide acid the reaction was run using tetraethylpyrophosphite and no racemic material
+
+
(13) F. Weygand a n d E. Leising, Be?., 87, 248 (1954). (14) G. W. Anderson a n d F. M. Callahan, THISJ O U R N A L , 82, 3369 (19GO). (15) (a) G. W. Anderson a n d R . W. Young, i b i d . , 74, 5307 (1952); (b) G. W. Anderson, J. Blodinger a n d A. D. \Velcher, ibid., 74, 5309 (1952); ( c ) J. It. Vaughan, Jr., ibid.. 74, GI37 (1952).
4600
DEKEKG . S X Y T I I ,
XTSUO NL\G.CvL4TSU i\ND JOSEPH
s. F R U T O N
Vol. 82
was fourid. Procecdiiig from there, a solution of 3.56 g. [having (0.010 mole) of carbobenzoxyglycyl-L-phenylalanine a melting point of 127.5-128.0", and a n optical rotation [a]"D f 38.2" (c 5 , abs. ethanol) compared with values4 38.8 i.0.5" (c 5. abs. ethanol)] of m.p. 127.5' and [ a ] " D in 10 ml. of dried (calcium hydride) dimethj-liormamide was cooled t o - 10' and 1.65 g. (0.010 mole based on 98y0 purity) of N,N'-carbonyldiimidazole was added. When the slow effervescence stopped, 1.03 g. (0.010 mole) of freshlydistilled ethyl glycinate was added. The reaction solution was allowed t o warm t o room temperature and permitted t o stand 15-30 min. Then 50 ml. of 1 N hydrochloric acid was added. JYlien t h e oily liquid thus formed solidified, i t was washed with 20 m1. of a 57, sodium bicarbonate solution and with water. On drying, 4.22 g. (96% yield) with a melting point of l15.~5-l17.0° was obtained. The product was dissolved in 210 ml. of absolute ethanol t o give a 27, solution. After cooling t o O", the solution was seeded with a crystal of ethyl carbobcnzoxyglycyl-DLphenylalarij~lgl~~ciiiate. Fractions were cut as follows:
133.5", the percentage of DL-tripeptide is estimated from the melting points t o be much less than 0.5%. In other reactions, run in the same way, varying oiic or two factors, more racemization was found. Running the reaction in T H F a t room temperature gave 57, DL-tripeptide. Using ethyl glycinate hydrochloride a t room tempcrature in T H F gave 8% DL-tripeptide. Running the rcaction in dimethylformamide a t 0-5' gave 1.3yc raceniization. I n these studies, the melting point of the carbobenzoxyglycyl-L-phenylalanine used was very important since commercial L-phenylalanine may contain several per cent . of the DL-isomer. Ethyl Carbobenzoxyglycyl-DL-phenylalaninateusing N,N'Carbonyldibenzimidazo1e.--A solution of 2.09 g . (0.010 mole) of carbobenzosygl!.cine in 10 1111. of dry THE; was treated with 2.62 g. (0.010 mole) of S,S'-carboiiyldibeiiziniicl~z(~le.b When no effervescence was noted, and the reagent only partially dissolved, another 10 ml. of T H F was added. Again nothing happened and the mixture was heated undcr reflux for 10 min. A solution formed and 1.93 g. (0.010 No. Time, hr. LVt., 6 . SI.p., o c , mole) of ethyl DL-phenylalaninate (freshly distillcd) was 1 3 0,0091 120.0-133.5 added. After heating for 15-30 niin. on a steam-bath, 2 6 ,0097 119.0-128.5 most of the solvent boiled off. X 50-ml. quantity of 1 fi hydrochloric acid was then added. Cooling and scratcli3 10 ,0214 118.5-119.5 ing gave a solid product. This was washed first with wxter, 4 24 2.3201 119.8-120.1 then with 20 nil. of 5 7 , sodiurn bicarbonate solution and Concd. I . 4787 119.9-120.3 finally with water again. The crude product weighed 3.48 Residue 0.3271 g. (777, yield) and had a melting point range of 83.5-58.0". Kecrystallization from 20 ml. of ethyl acetate and 40 ml. of 0.0188 0 .457, DL-Isomer petroleum ether resulted in 2.75 g . of compound with a L-Isomer 3.8202 87% [ a ] * j-12.2 ~ & 1.25' melting range of 85-88', This material was recrystallized (c 2, E t O H ) again, this time from 20 ml. of benzene a n d 40 mi. of peRcsidue 0,3271 troleum ether, giving 2.63 g. (69y0 yield) of ethyl carbobenzoxyglJ-cyl-DL-phenvlalaninate with a melting point of Material 89.5-91.0' as compared t o a previously cited melting point balance 4.1661 of 90-91". Because of the poorer yield and the inure Since the melting point of pure ethyl earhoberizox~-glycj-l- rigorous conditions nccessary for a reaction, N,S'-carbonyltlibenziiiiidazole is considered inferior t o S,N'-carbonyldiDL-phenylalan?-lglyCinatC was reported4 t o be 132-133' and has been found in other reactions by us t o be 133.0- imidazole.
+
[CONTRIBUTIOX FROX THE
DEPARTMEXT O F BIOCHEMISTRY, YALE
UNIVERSITY]
Some Reactions of N-Ethylmaleimide BY DEREKG. SMYTII,? ATSUONAGAMATSU~ AKD
JOSEPH
S. FRUTON
RECEIVED MARCH7, 1960 T h e reactions of S - e t h y l m a l e h i d e (NEM) with the amino group :if peptides, with imidazole and with cysteine have been investigated. IVith the first two classes of compound, an S-acylation reaction appears t o occur, followed in t h e case of imidazole by a catalytic polymerization of S E M . \\'it11 cysteine, reaction proceeds through addition of the thiol t o the olefinic bond of XEM ; in alkaline sulution. the c p t r i n e adduct undergoes an intramolecular transamidation reaction to form a thiazane derivative.
During the course of studies on the action of cysteine-activated cathepsin C4on glycyl-L-histidinamide a t fiH 7.4, X-ethylmaleimide (KELT) was used to facilitate chromatographic examination of the composition of the incubation mixture. The reaction of NEM with sulfhydryl corn pound^^^^ had been used by Hanes, et al.,; to stabilize glutathione and other sulfhydryl peptides in paper (1) This investigation w a s supported by grants from t h e Kational Science Foundation (G-7451) a n d the United States Public Health Service (RG-6452). (2) Alexander Brown Coxe Fellow of the Yale School of Medicine, 1059-1960. (3) James Hudson Brown Fellow of t h e Yale School of hledicine, 1!158-1059. On leave from t h e Depnrtment $ i f Biochemistry Kyushu
7.niversity. F u k u o k a , Japan. (4) N . Izumiya and J . S. Frruton, .I. B i d Chewi., 218, 59 (1
