932
COMMUNICATIONS TO THE EDITOR
VOl. 86
forded N-carbobenzoxy-L-alanyl-L-leucyl-L-tyrosyl-L-a partially purified preparation of synthetic A-chain. leucyl - L - valyl- S - benzyl - L-cysteinylglycyl-7-t-butyl-LThe quantitative results of the recombination experiglutamyl - N w- tosyl- L - arginylglycyl-L - phenylalanyl - Lments and the biological assays will be reported in a phenylalanyl-L-tyrosyl-L-threonyl -L - prolyl- Nf-tosyl-Llater communication. 22 lysyl-L-alanine methyl ester (XIV), m.p. 256-259' ; The data on the synthesis of the B-chain presented [ a I z 7 D -33.6' (c 1, dimethylformamide) ( A 4 ~ u lCalcd. . in this report in conjunction with our previous comfor C125H169K21029S.HzO: C ,59.0; H , 6.77; N , 11.5. munication regarding the synthesis of the A-chain* Found: C,58.7; H , 6 . 9 0 ; N , 11.5.); aftertreatment strongly suggest that the structure proposed by Sanger with HBr in acetic acid: Rf60.93, fir9 6.58 X his; amino for the insulin chains is correct. This constitutes a acid ratios in acid hydrolysate. lysl,13argl&benzylunique case where a proposed primary structure for a protein has been confirmed by chemical synthesis. cysteine0.~ithro.~~g1Uo.9781y2.osalal.~~va~~.l~~eu~.19pro~. 97tyr1.97phe1.84; average amino acid recovery, 91%. Du, et ~ l . , have ~ * reported the isolation of crystalline Exposure of X I V to HBr in acetic acid and reaction of insulin, identical with the natural hormone, by rethe resulting product with XI1 yielded the protected combination of natural and B-chains. On this heneicosapeptide ; this, in turn, on exposure to alkali basis, work is now in progress in our laboratory on and then to HBr in acetic acid afforded the C-terminal the synthesis of amounts of the two chains adequate segment of the B-chain Irn-benzyl-L-histidyl-L-leucyl- to permit the isolation of synthetic insulin by combiL - v a l y l - ~ - g l u t a m y l - ~ - a l a n y l - ~ - l e u c y l - ~ - t y r ~ ~ s y l - ~ - l nation e u c y l - of synthetic A- and B-chains. I t is expected ~-valyl-S-benzyl-~-cysteinylglycyl-~-glutamyl-N~tosylt h a t the material thus obtained also will be identical L-arginylglycyl-L-phenylalanyl - L - phenylalanyl - L - tyrowith the natural protein. s y l - ~ - t h r e o n y l - ~ - p r o l y l - ~ f - t o s y l - ~ - l y s y l - ~ - a l ahynine (24) See reference in footnote 1 1 drobromide (XV), [ a l z 7-~24.4' (c 0.9, dimethylform(25) This work was supported by a Research Career Uevelupment Award (GM-K3-15151) from the Public Health Service and a grant (A-3067) from amide) (in none of the usual paper chromatographic the S a t i o n a l Institute of Arthritis and Metabolic Iliseases. Public Health systems employed in these studies did this peptide Service, for which we wish to express our appreciation move from the origin, hence, no paper chromatographic (26) T h e authots wish t o express their appreciation t o Mrs. Jemele Hudson criteria could be obtained) ; amino acid ratios in acid for the enzymatic analyses and amino acid analyses and Mrs. Gudrun Hjorth for technical assistance hydrolysate, lyso.saarg0.96thr1.1~glul.95pr01.1~gly*.~oalal.~5vall.~~leus.~otyrl.gophen.40 (S-benzylcysteine and ImBIOCHEMISTRY DEPARTMENT PANAYOTIS G. K A T S O Y A N N I S ~ ~ . ~ ~ USIVERSITYOF PITTSBURGH KOGHEIFUKUDA benzylhistidine not determined) ; average amino acid SCHOOL OF MEDICINE A N D R E WTOMETSKO recovery, SiY,. PITTSBURGH, PENNSYLVANIA KENJISUZUKI Conversion of IX to the corresponding azide and MASOHAR TILAK coupling of the later product with XV yielded the proRECEIVEDJ A N U A R Y 28, 1964 tected sulfhydryl form of the B-chain of insulin. The protecting groups were removed by treatment with sodium in liquid ammonia and the deblocked product Stable Carbonium Ions. VI.' Nuclear Magnetic was converted to the S-sulfonate form and purified by Resonance Investigation of the Diphenylcarbonium, ion-exchange Chromatography. The S-sulfonate of Diphenylmethylcarbonium, and the B-chain obtained exhibited on ion-exchange chromaPhenyldimethylcarbonium Ions tography a pattern identical with that of the S-sulfonate Sir: of natural B-chainZnand on paper chromatography exhibited a single Pauly-positive spot with the same RfZ1 The triphenylcarbonium ion is the best investigated (0.96) as natural B-chain. Amino acid analysis of the carbonium ion. I t s structure has been established by synthetic material after acid hydrolysis gave a compoultraviolet,*infrared,3and n.m.r. investigations. sition in molar ratios which corresponds to that of In contrast, the diphenylcarbonium ion (benzhydryl the natural B-chain, lys~.a~arg~.o~hisl.~asp~,1~thr~.~4ser~.~6cation) is much less known and has not yet been directly g l u ~ . ~ p r o ~ . ~ ~ g l y ~ . ~ 4 a l a ~ , ~ 4 v a3 l ~ , ~(cystine ~ l e u 4 . 3 ~characterized ~ y ~ ~ . ~ ~ p hwith ~ ~ the exception of ultraviolet investiwas not determined). gations. Gold5 and Den0 and his co-workers6 reported Combination experimentsz2 between the synthetic the ultraviolet absorption of diphenylmethanol in conmaterial and the A-chain, natural or synthetic, procentrated sulfuric acid. Hafner7 and Holmes and vided further proof that the synthetic product is inPettits possibly obtained salts of (C6H5)?CH +, but the deed the B-chain of insulin. As judged by biological structure of these was not more closely characterized, assays, using the mouse-diaphragm method, insulin possibly due to their instability in solution. activity was generated when the synthetic B-chain The diphenylmethylcarbonium ion was investigated was combined with natural A-chain. Furthermore by Gold5 through the ultraviolet spectra of 1, l-dithe amount of activity generated was quantitatively phenylethylene and 1,l-diphenylethanol in sulfuric acid identical with the activity produced when an equal solution, as well as b y O'Reilly and Lefting through the amount of natural B-chain was used for the combination n.m.r. proton spectrum of 1,l-diphenylethanol in sulexperiments.2 3 Considerable insulin activity was also furic acid solution. Although the position of the main obtained when synthetic B-chain was combined with aromatic proton peak and the methyl group could be established, due to fast hydrogen exchange in sulfuric (20) S a t u r a l B-chain was prepared in our laboratory from crystalline zinc-insulin b y a modification of the method of Meienhofer and Brinkhof acid solution, no fine structure was observed. At[ J . Meienhofer and 0. Brinkhof, .\-atwe, 199, 1096 (1Y63j1. (21) T h e K f refers t o a descending paper chromatography in the system . Z -X. 2-butanol-acetic acid-8 .VI urea 1 2 , 5 ~ 1 : 1 1 . 5[Y - C D u . Y . ~ SZhang, 1.u. and C -1,. I'soii, S c i Siiiiia (Peking). 10, 84 (lY6l)l. ( 2 2 ) We are most indebted t o I l r . G . H . Ilixon uf the Department of Biochemistry, University u f British Columbia, who carried out the combination experiments, and t o Dr J . K . Davidson of the Best InFtitute. University of Toronto, who assayed the combination reaction mixtures for insulin activity. ( 2 8 ) A crude preparatiun of B-chain which generated only slight insulin activity upon combination with the A-chain was obtained b y the condensation o f the S terminal tridecapeptide with the C-terminal heptadecapeptide This work was presented ( P . G. K.) a t the Brook Lodge Conference on Proteins and Polypeptides on October 7-9, 1963, Kalamazoo, Mich.
(1) Part V : J A m Chem. Soc.. in press. ( 2 ) For a review see, L N Ferguson, C h e m K e n . , 43, 385 (1948) (3) D. W. A S h a r p and N. Sheppard, J C h e m Soc., 674 (1957) (4) R B Moodie, T M Connor, and R Stewart, Cnri. J Ciiein , 37. (1959); R . Dehl, W K Waughan, and K S. Berry, J 01'8 C h r m , 2 4 , (1989); R S Berry, K Dehl, and W. K Waughan, J C h r m P h y s , 34, (1961). ( 5 ) V Gold and F. L T y e , J Chem Soc , 2172 (1952) (6) N. C Deno, J . J . Jaruzelski. and A. Schriesheim, J A m Chpm 7 7 , 3 0 4 1 (1955) (7) K . Hafner and H Pelster, A K ~ Chein W , 73, 342 (1961). (8) J Holmes a n d li. Pettit, J O i g Chcm., 28, 1965 (1963) (9) D. E O'Keilly and H P. Leftin, J . Phys. C h e m . . 64, 1335 ( l Y 6 0 )
1402 1616 1460
Soc ,
COMMUNICATIONS TO THE EDITOR
March 5 , 1964
tempts have been madelo to prepare diphenylmethylcarbonium tetrafluoroborate by the reaction between 1,l-diphenylethyl chloride and silver tetrafluoroborate in ether solution. Although, as evidenced b y silver chloride elimination, there appears to be a primary formation of the carbonium ion due to subsequent proton elimination, no stable complex could be isolated. The phenyldimethylcarbonium ion is so far unreported in the literature. I t was possible now to obtain the diphenylcarbonium, diphenylmethylcarbonium, and phenyldimethylcarbonium ions in the form of their solvated hexafluxoantimonate salts in sulfur dioxide and sulfur dioxide-antimony pentafluoride solutions and to record their proton magnetic resonance spectra. In previous work of this series" the isolation of the diphenylacetyl fluoride-antimony pentafluoride complex was reported, which, in contrast to many similar acyl fluoride-antimony pentafluoride complexes of oxocarbonium (acylium) salt nature, was found both in the solid state (based on infrared investigations) and in sulfur dioxide solution (based on n.m.r. investigations) to be only a polarized donor-acceptor complex.
In continuation of this work a more detailed investigation of this complex in sulfur dioxide was carried out. I t was observed t h a t when a sample of the crystalline diphenylacetyl fluoride-antimony pentafluoride complex was dissolved in sulfur dioxide and left to stand for some hours a t --40°, and the n.m.r. proton resonance spectrum then obtained a t the same temperature, besides the previously reported donor-acceptor complex ( I ) , a second, substantially more deshielded species (11) could also be resolved. Data are summarized in Table I. TABLEI NUCLEAR MAGNETIC PROTON RESONANCE' OF THE DIPHENYLACETYL FLUORIDE-ANTIMONY PENTAFLUORIDE SYSTEMIS SOz SOLUTION AT -40' Aromatic ring protons
( CsH5)zCHCOF(neat)
(CsHs)zCHCOF-SbFs (1) Deshielded species (11) 1'
-7
-7.7
60 Mc., p.p.m. from T M S .
-4.97
10
- 7 12 -8.2
CH
-8 5
IPPM
0
1,
Fig. 1.-Spectrum
Substantial deshielding is evident from the position of the ring protons and the aliphatic CH proton a t -9.7 p.p,m. In preparative scale experiments, the evolution of carbon monoxide from the system could be established, thus indicating that the new deshielded species (IT) could be diphenylcarbonium hexafluoroantimonate.
9
8
6Tl
7
0 T I S
( n . m r . ) of diphenylcarbonium ion.
hexafluoroantimonate, in this case isolable as the stable crystalline salt (m.p. 21 1"). Diphenylmethylacetyl fluoride-antimony pentafluoride and phenyldimethylacetyl fluoride-antimony pentafluoride in sulfur dioxide solution similarly yielded besides the acyl fluoride complexes via decarbonylation substantially deshielded species, in all probability the tertiary carbonium ions. The methyl protons were found in these species a t -3,50 and -3.35 p.p.m., respectively; the aromatic ring protons also show deshielding. In order to ascertain that the observed deshielded species in the phenylacetyl fluoride-antimony pentafluoride systems are indeed the decarbonylated phenylcarbonium ions, and to obtain these ions in pure form, the previously described' carbonium ion formation method using antimony pentafluoride as both Lewis acid and solvent was used. Diphenylchloromethane, 1,l-diphenyl-1-chloroethane, and 2-chloro-2-phenylpropane (cu-chlorocumene) gave with excess antimony pentafluoride stable, solvated hexafluoroantimonate complexes of the diphenyl-, diphenylmethyl-, and phenyldimethylcarbonium ions. The nuclear magnetic proton resonance spectra of these ions could be obtained at -30°, using sulfur dioxide containing excess antimony pentafluoride as solvent. Table I1 summarizes the n.m.r. data compared with that of the known triphenylcarbonium ion. The spectrum of the diphenylcarbonium ion is shown in full in Fig. 1 as representative of the resolution obtained using a Varian Model il-60 spectrograph equipped with a low temperature probe.
-5.35 -9.7h
* Peak area ratio 9.95: 1
933
TABLEI1 SUCLEAR MAGNETIC PROTON RESONANCE' OF PHESYLCARBONIUM HEXAFLUOROANT~MONATE COMPLEXES ix S02-SbF6 SOLUTION AT -30'
Compound
C +-CHa
CeHsC +(CHa)2 SbFs-3 60 ( CeHj )zC 'CH3 SbFs-3.70 (CsHs)zC'H ShFs(CsHs)rC SbFe- (in neat Son)
-Aromatic orlhob
ring H--. melab para
C+H
Peak area ratio
-7 95
-8 87
-8.56
6:5.04
- 7 53
-7 96
-8.12
3:9 95
-7.92
-8
-7 01
-7.51
49
-8
37 - 9 . 8 9 9 i : l
+
I
I1
X similar decarbonylation of the triphenylacetyl fluoride-antimony pentafluoride systems resulted in the quantitative formation of triphenylcarbonium (10) D worth a n d (111 G and E. B.
W A Snarp. "Advances in Fluorine Chemistry," Vol. I. ButterCo , London, 1960, p 93. A. Olah. W. S. Tolgyesi, S.J. K u h n , M E. Moffatt, I. J Bastien. Baker, J . A m C h e m S a c , 86, 1328 (1963)
a
60 Mc., p.p.m. from T M S .
-7.76
* See ref. 12
The assignment of the deshielding sequence of the ortho and meta ring protons is tentative and arbitrary based only on the assumed analogy with the ring protons in the known triphenylcarbonium ion, where suc-
COMMUNICATIONS TO THE EDITOR
934
cessive deuteration showed the meta protons more deshielded than the ortho protons. l 2
Vol. 86
TABLE I RINGP R O T O N CHEMrCAL SHIFTS FOR
(12) One of t h e referees called attention t o unpublished and independent w o r k of Dr D G F a r n u m o n phenyl and phenylmethyl carbonium ions and suggested simultaneous publication, which was arranged by mutual agreement. Having learned of Dr. Farnum's d a t a , one can say t h a t there is excellent agreement of t h e observed chemical shifts of t h e aliphatic protons and t h e over-all ring proton shifts. However, based on Dr Farnum's d a t a , supported by calculation of theoretical spectra with t h e use of a computer program, t h e deshielding sequence of t h e o v l l i o and mela protons is reversed in t h e diphenyl- and phenylmethylcarbonium ions, as compared with t h a t of the triphenylcarbonium ion.
S O M E PHENYLCARBONIUM
IONS~
para, p pm
mdapara, p p m
0
0
orlho-
Case
Carbonium ion
+
para, orfho, m e f a , r-units r-units r-units
1 C6H5CH2CH2C(OH)za 2 80
2 80 2 80
2 csHsC(oH)~
194
1 73
2 24
-0 21
+0 30
1 85
1 8 8 2 17
+0 03
+ O 32
1 84 1 76
1 5 7 2 22 2 31 2 13
-0 30 +O 55
+O
1 72
1 97
+0 25
+0 40
c
3 (CsHa)2COH
CONTRIBUTION So. 111 GEORGEA . OLAH EXPLORATORY RESEARCH LABORATORY D O I V CHEMICAL O F C A T A D A , LIMITED S A R S I A , OSTARIO, CANAD.4 RECEIVEDNOVEMBER 18, 1963
L
4 C6H,C( 0 H ) C H ; 5 (Ce&)3CC
+O 38 37
T
6 (CsHs)zCCHsC
2 12
7 (C6Hj)nCHd 1 . 6 2 1 54 2 02 - 0 08 +0 40 8 CsHjC( C H S ) ~ ~ 1 . 4 5 1 . 2 0 2 . 0 3 -0 25 +0 58 a Since sulfonation in chlorosulfonic acid was very rapid, t h e n.m.r. spectrum of this substance was determined in concentrated sulfuric acid. * CHs resonance a t T 6.69. CH1 resonance a t T 6.30. > C H resonance a t T 0.19. e CHa resonance a t T 6.43.
Chemical Shifts and Long Range Shielding Effects in the Nuclear Magnetic Resonance Spectra of Phenyldimethyl, Diphenyl, Triphenyl, and Related Carbonium Ions Sir: Analyses of the n.m.r. spectra of triphenylcarbonium and methyldiphenylcarbonium ions have been reported in the l i t e r a t ~ r e . ' - ~We wish to describe results of an analysis of the n.m.r. spectra of a number of phenyl-
The ortho, meta, and para ring proton chemical shifts for a number of phenylcarbonium ions are given in Table I. Values were determined by comparing n.m.r. spectra of chlorosulfonic acid solutions of the appropri-
I
O P
M
P
O
M
P
M
O
Fig, 1.-Calculated (dashed line) and experimental (solid line) n.1n.r. spectra of some phenylcarbonium ions. Chemical shift differences (6) and coupling constants ( J )for the ortho ( O ) ,weta ( M ) ,and para ( P )protons used for the calculation are given in C.P.S. Changes of 0.5 c p.s. in chemical shifts or the large coupling constants gave noticeably poorer agreement. The small coupling constants are probably within 1 C . P . S . of the correct value. (3) Dipl~en\-lcarboniumion: 6 (0-P) -5, 6 ( 0 - M ) +24; J ( 0 , M ) 8.2, (0, P) 1.2, (0,M') 0.5, (0,0') 1.2, ( M , P ) 7.5, ( M , M') 1.7, ( b ) Methyldiphen~lcarhoniumion: 6 (0-P) +15, 6 (M, P ) +24.5; J ( 0 ,M ) 8.2, (0,P) 1.2, ( 0 ,M') 0.5, (0,0') 1.2, (M, P ) 7.5, ( M , M') 1.7. ( c ) Triphenylcarbonium ion: 6 (0-P) $33.5, 6 (M-P) +22; J ( 0 ,M) 7 . 7 , ( 0 ,P) 1.2, (0,Ll') 0.5, ( 0 , 0') 1.2, ( M , P ) 7.5, (M, M ' ) 1.0.
carbonium ions which (1) suggest that shielding of ortho protons by neighboring phenyl rings in di- and triphenylcarbonium ions is more important than has been recog'lized in papers, and (2) permit a qualitative Correlation Of the extent Of phenyl delocalization of the positive charge with the position of the para proton resonance. (1) R Dehl, W. R Vaughan, and R S . Berry, J O r g . Chem , 24, 1616 (1989); J C h e m P h y s . . 34, 1-160 (1961) ( 2 ) D E O'Reilly and H. P Leftin, J P h y s . C h e w . , 64, 1555 (1960). (3) K B. XIoodie, T & Connor, l. and R Stewart, C a n . J . Chem., 37, 1402 (195Cl).
a t e carbinol or carbonyl compound4 with calculated spectra obtained using the computer program of Bothner-By and N a a r - C ~ l i n . ~ The ~ ' calculated and de(4) Solutions were made up to approximately 10% concentration by dropwise addition of a 1 . 1 solution of t h e compound in thionyl chloride t o vigorously stirred chlorosulfonic acid a t -20" (-40' for phenyldimethylcarbonium ion). T h e n . m I . spectra were determined on a Varian A-60 spectrometer a t room temperature [ -40° for phenyldimethylcarbonium ion) Chemical shifts are reported on t h e r-scale with tetramethylammonium fluoroborate (7 6 87) as a n internal standard 5 ( 5 ) D . G . F a r n u m , 41.A . T . Heybey, and B. Wehster. Tefr.nhedmn L i l l e r s . No. 6, 307 (1963). (6) A . A Bothner-By a n d C. Naar-Colin, J . A m . Chem. Soc., 89, 231
