1424 J.C.S. Perkin IMethods of Peptide Sequencing. Part 11.l Cyclisation of N-Z-Amino-6-nitrophenyl and N-3-Amino-2-pyridyl Derivatives of Amino-acids andPeptidesBy Robert A. W. Johnstone* and T. Jeffrey Povall, The Robert Robinson Laboratories, The University,Liverpool L69 3BXIan D. Entwistle, Shell Research Ltd., Sittingbourne, KentPeptides and arnino-acids were treated with either 1 -fluoro-2.6-dinitrobenzene or 2-fluoro-3-nitropyridine to giveN-nitroaryl derivatives. The latter were reduced by transfer catalysis to N-aminoaryl derivatives, which cyclisedwith, in the peptide case, release of the peptide minus its N-terminal amino-acid residue. The cvclised materialswere readily. identified by mass spectromeiry.amino-acids in a peptide.THE Edman method for sequencing amino-acid residuesin a peptide chain is widely used with various modific-ationse In this method, the free amino-end of apeptide is treated with an isothiocyanate to release theN-terminal amino-acid residue as a thiohydantoin ; theremaining peptide can be subjected repetitively to thesame cycle to yield the sequence of amino-acidresidues.l-Fluoro-2,4-dinitrobenzene and 2-fluoro-3-nitropyridine4 have been used in determining theN-terminal amino-acid in a peptide but a major dis-Part I, R. A. W. Johnstone and T. J. Povall, J.C.S. Pevkin I,See for example, W. A. Schroeder, Methods Enzymol.,3 F. Sanger, Biochem. J., 1946, 39, 607; Bull. SOC. chim. biol.,1075, 1297.eds. C. H. W. Hirs and S. N. Timasheff, 1972, 25, 298.1955, 37, 23; A.Niederwiesser, J . Chromatog., 1971, 54, 216.This cycle of operations was used to obtain a pahal sequence ofadvantage of these reagents is that, in obtaining thederivative of the N-terminal amino-acid, the rest of thepeptide is destroyed.In a variation of this method, 2,4-dinitrophenylderivatives of simple peptides were hydrogenatedcatalytically and treated with acid to form substitutedaminoquinoxalines.6 However the latter are readilyoxidised in air. Other derivatives tried include 4-met hoxycarbonyl-2-nitrophenyl and 5,7-dinitrobenz-4 A. Signor, L. Biondi, A. M. Tamburro, and E. Bordignon,European J . Biochent., 1969, 7, 328; E. D. Bergmann and M.Bentov, J . Oyg. Chem., 1961, m, 1480.5 M. Jutisz and W. Ritschard, Biochinz.Biophys. A d a , 1955,17, 548; V. M. Ingram, ibid., 1956, 20, 577.R. W. Holley and A. D. Holley, J . Amev. Chem. SOC., 1952,74, 54451975 1425imidazol-4-yl; by these methods, the N-terminalamino-acid residue can be removed and the remainingpeptide subjected again to the cycle of operations.The reduction of dinitro-compounds to the corre-sponding monoaminomononitro-compounds can beeffected easily and in high yield through the use of apalladium catalyst and cyclohexene; * this type ofreduction in which hydrogen is transferred to or fromcyclohexene from or to a substrate via a metal catalysthas been used previously in other situation^.^The preparation of a nitroaryl derivative [(I) or (IV)]quinoxalines (111) and could not be isolated as inter-mediates; at the same time, the remaining peptide wasreleased.The deep orange tetrahydroquinoxalines (I1 I)were easily detectable by t.1.c. and afforded simple massspectra by which they could be identified. The normalamino-acids (Tables 1-4) formed cyclic products (111)readily but one or two small variations were noted.Methionine gave the expected product (111; R =CH,*CH,-SMe) but the reaction was much slower thanwith the other amino-acids, probably owing to sulphurpoisoning the catalyst.ONo2 + H,N.CH.CO.NH- oNo2 __LI N NH*FH-CO.NH - R 'N RSCHEMEof a peptide followed by reductive cyclisation to forma tetrahydroquinoxaline (111) or a tetrahydropyrido-pyrazine (VI) with release of the remaining peptideforms the basis of the method described here (seeScheme).In the following discussion, results for dinitro-phenyl derivatives (I) are dealt with and any distinctdifferences from the nitropyridyl derivatives (IV) areconsidered afterwards.With l-fluoro-2,6-dinitrobenzene, amino-acids andpeptides with free amino-groups gave 2,6-dinitrophenylderivatives (I) which, in these small-scale experiments,were purified by t.1.c. The dinitro-compounds (I) were' half-reduced ' to the corresponding monoaminomono-nitro-compounds (11) with a slight excess by weight ofpalladium-charcoal catalyst in refluxing ethanol con-t aining cyclohexene. Under these conditions, the mono-amino-derivatives (11) cyclised to the tetrahydro-K. L. Kirk and L. A. Cohen, J . Oyg. Chem., 1969, 34, 384,390, 396.The nitropyridyl derivatives (IV) behaved similarlybut the reduction step was significantly faster and thecyclisation to pyridopyrazines (VI) much slower, so thatthe intermediates (V) could be isolated.Addition ofacetic acid was necessary to make the cyclisationproceed at a similar rate to that of the dinitrophenylcompounds (I). The nitropyridyl derivative fromdimethyl glutamate on reductive cyclisation gave thepyroglutamyl compound (VII) as well as the expectedproduct (VI; R = CH,-CH,*CO,Me) but the derivativefrom glut amine gave exclusively the p yroglut amylcompound (VII). The derivative from cysteine wasreduced to the cyclic compound corresponding toalanine (VI; R = Me); in MeOD, reductive cyclisationgave a deuterioalanine derivative (VI) containing 15%I.D. Entwistle, R. A. W. Johnstone, and T. J. Povall,E. A. Braude, R. P. Linstead, and I. R. H. Wooldridge,J.C.S. Perkin I, 1976, 1300.J . Chem. SOC., 1964, 35861426 J.C.S. Perkin Iof deuterium (mass spectrometry). The tetrahydro-pyridopyrazines (VI) were colourless but readily detect-able as blue compounds under U.V. light.The reaction of peptides with l-fluoro-2,6-dinitro-benzene or 2-fluoro-3-nitropyridine followed by reductivecyclisation proceeded normally to give, respectively, theHby formation of the methyl ester, and extremelyabundant M - 59 ions (loss of C0,Me) were alwayspresent for compounds (I) and (IV). Other abundantions corresponded either to loss of an amino-acid side-chain or to the side-chain itself.An extremely abundantTABLE 1Mass spectra of 2,6-dinitrophenyl derivatives (I) aGlyb 196,* 209, 255; Val 238, 254, 297; Phe 254,* 286, 291,299, 345; Trp 77, 103, 130,* 196, 254, 325, 384; Ser 196, 226,*254, 285; (pyr)Orn 107, 108, 124, 136,* 150, 164, 176, 236, 252,418,419; Met 61,* 75, 196, 222, 237, 238, 254, 255, 282, 283, 330;met(0,) 222,* 254, 302, 361; Tyr 107,* 196, 254, 255, 302, 361,362; Tyr(Dnp) 167, 254, 273,* 344, 360, 468, 527; His 81,* 167,196, 254, 276, 290, 335, 349; His(Dnp) 247, 276, 335, 442,* 501;Leu-Phe 252,* 399, 412, 427, 443, 458, 459; Gly-Leu-Phe 196,291, 309,* 337, 349, 353, 459, 469, 484, 515, 516a All of type 2,6-(NO2),C,H,.X.OMe with X = amino-acidor peptide residue.In this and other Tables only the moreprominent ions are recorded; the value corresponding to thebase peak is marked by an asterisk and that corresponding tothe molecular ion is italicised.122(XItetrahydroquinoxaline (111) and the tetrahydropyrido-pyrazine (VI) corresponding to the N-terminal amino-acid residue with release of the remaining peptide. Thepeptides H-Leu-Phe-OMe and H-Gly-Leu-Phe-OMe, eachgave the cyclic product corresponding to the AT-terminalamino-acid and a ninhydrin-positive residue, character-ised as the 2,6-dinitrophenyl derivative of H-Phe-OMeor H-Leu-Phe-OMe, respectively. Similarly, the tetra-peptides H-Val-Gly-Leu-Phe-OMe and H-Ala-Tyr-Leu-Phe-OMe gave the cyclic products corresponding tovaline and alanine, respectively.A larger peptide was used to find to what extent thereductive cyclisation step could be repeated along achain.With 20 pmol of the octapeptide H-Glu(0Me)-Glu(0Me)-Ala-Glu (0Me)-Glu (OMe) -Ala-Tyr-Gly-OMeand by using the nitropyridyl derivative, the amino-acidresidues 1-5 were identified by comparison (t.1.c. andmass spectra) of their cyclic products (VI) with authenticsamples. The final glutamic residue was only justdetectable and the next residue (alanine) was not found.During the sequence other unidentified products beganto accumulate, making detection of amino-acid residuesincreasingly difficult. Some of these products may becyclic dipeptides: peptides are known lo to give thesesubstances on refluxing in acetic acid although at highertemperatures and for longer times than those used inthese experiments.The use of l-fluoro-2,6-dinitro-benzene or 2-fluoro-3-nitropyridine is not yet as satis-factory on a small scale as the Edman method. Furtherwork aimed at decreasing the formation of by-productsand extracting impurities at each step is desirable.Mass spectrometric data are shown in the Tables.Molecular ions were produced in low abundance with2,6-dinitrophenyl (I) and 3-nitropyridyl (IV) compoundsand in some cases were accompanied by M + 1 ions ofsimilar abundance. In all the compounds studied, theC-terminus of the amino-acid or peptide was protectedTABLE 2Mass spectra of 3-nitropyridyl derivatives (IV) aPhe 91, 139, 162, 210,* 242, 270, 301; Gly(0Me) 115, 123, 178,206, 210, 238,* 266, 297; Gln 123, 140, 178,* 206, 210, 223, 233,238, 250, 251, 265, 266, 282, 283; Asn 192,* 209, 237, 251, 252,268, 269; Lys 152, 178, 192, 194, 206,* 210, 264, 282, 373, 404,405; Cys 164, 169, 181, 192, 210, 224,* 240, 320, 379, 380;Cys(CH,CO,CH,) 139, 158, 190, 196, 210," 224, 256, 270, 298,329; Leu-Phe 91, 152, 208,* 236, 358, 383, 414, 415All of type 3-nitro-2-pyridyl-X(OMe) with X = amino-acid or peptide residue.TABLE 3Mass spectra of the 3-R-substituted quinoxalines (111) aH(G1y) 146(14), 147(16), 191(24), 192(28), 193(100); Me(A1a)146(22), 192(100), 193(13), 206(41), 207(57) ; Me,CH(Val)146(31), 192(100), 193(29), 235(23) ; Me,CHCH,(Leu) 146(64),192(100), 193(20), 249(16); PhCH,(Phe) 91(89), 146(34), 192(100),283(11); indol-3-ylmethyl(Trp) 77(14), 130(100), 146(6), 192(3),193(4), 322(14); HOCH,(Ser) 146(57), 192(100), 193(48), 223(22);MeSCH,CH,(Met) 146(69), 192(100), 193(18), 221(31), 267(18)a Data of the form R(amino-acid residue) m/e (% relativeabundance).TABLE 4Mass spectra of the 3-R-substituted pyrido[2 341-pyrazines (VI) aMe(A1a) 93(28), 120(100), 134(15), 148(78), 162(8), 163(75) ;PhCH,(Phe) 91(20), 93(14), 119(5), 120(27), 148(100), 239(10) ;MeO,C*CH,-CH, [Glu(OMe)] 93(34), 119(16), 120(100), 148(96),161(36), 162(13), 175(25), 176(5), 203(14), 204(9), 235(18) ;H,N-CO*CH, (Asn) 120(151), 148(75), 161(100), 162(48), 189(20),206(30) ; 3-amino-2-pyridyl-NHCH2CH% (Lys) 108(25), 109(100)119(9), 120(19), 122(54), 135(45), 136(25), 148(22), 176(5),202(12), 312(50); (Gln) 120(100), 148(47), 161(10), 174(7),175(7), 203(37)a Data of the form P(amino-acid residue) m/e (yo relativeabundance). Material underwent further cyclisation topyroglutamyl derivative; see (VII) .ion at +n/e 264 from the nitropyridyl derivative of lysineis probably due to a fragment of type (VIII).From thenitropyridyl derivative of asparagine, a second spectrumlo J. D. Baty, R. A. W. Johnstone, and T. J. Povall, J.C.S.Chem. Comm., 1973, 3921975 1427was obtained after the compound had been in the massspectrometer for a few minutes; this showed a molecularion at ?n/e 236, probably due to the succinimide (IX)formed by a thermolytic reaction in the spectrometer.The tetrahydroquinoxalines (111) gave simple massspectra having abundant molecular ions and relativelyfew but prominent fragment ions.Two ions at m/e 192and 146 were present in all cases due respectively to lossof the amino-acid side-chain and of both side-chain andnitro-group. Less satisfactory spectra were obtainedfor the disubstituted histidine and pyro-ornithinederivatives. No molecular ions were present but them/e 192 fragment was abundant. In the case of thehistidine derivative, an M + 1 ion was present and inthat of the ornithine derivative, ions due to the side-chain were present (m/e 136, 150, 164, and 176).The tetrahydropyridopyrazines (VI) also gave simplemass spectra with abundant molecular ions ; two otherabundant ions a t m/e 148 and 120 were always foundand correspond to loss of R and R + CO from themolecular ion.The compound (VI; R = CH,-CO*NH,)derived from asparagine gave an abundant ion at m/e161 (loss of CONH, + H), and the glutamine derivativegave a molecular ion at m/e 203 rather than at theexpected value of m/e 220 (loss of NH,), showing that apyroglutamyl compound (VII) had been formed. Inthe case of the disubstituted lysine compound (X), thebase peak was at nzle 109 and other abundant ions weredue to side-chain fragmentation.EXPERIMENTALMass spectra were recorded a t 70 eV on an A.E.I.MS 12 spectrometer by direct insertion. All t.1.c. wascarried out on silica gel. RB Values for 2,6-dinitrophenyl(Dnp) derivatives of amino-acid methyl esters (Dnp-NH*CHR*CO,Me) in benzene-methanol (either 15 : 2 or15 : 1) were respectively: Gly 0.77, 0.70; Val 0.78, 0.72;Phe 0.83, 0.75; Trp 0.68, 0.60; Met 0.73, 0.72; Ser 0.50,0.38; (pyr)Orn 0.66, 0.55; Tyr 0.53, 0.39; tyr(Dnp) 0.70,0.64; His 0.39, 0.17; his(Dnp) 0.72, 0.71.Similarly RFvalues for 3-nitropyridyl (Np) derivatives (Np-NH*CHR*C0,Me) were: Leu 0.80, 0.74; Phe 0.80, 0.72; Glu(0Me)0.73, 0.64; Asn 0.21, 0.11; Gln 0.24, 0.13; cys(Np) 0.77,0.73; Lys(Np) 0.75, 0.69; cys(CH,*CO,Me) 0.72, 0.67;and for tetrahydroquinoxolines (111) were : (derived from)Gly 0.43, 0.29; Ala 0.44, 0.39; Val 0.50, 0.41; Leu 0.52,0.47; Phe 0.53, 0.44; Trp 0.35, 0.24; Tyr 0.24, 0.12;Met 0.49, 0.36; Ser 0.21, 0.08.2 6-Dinitrophenyl Derivatives (I) .-The hydrochloride ofserine methyl ester (3 mg) was shaken for 2 h a t 20 "C in asolution of l-fluoro-2,6-dinitrobenzene (7 ing) in aqueous1 yo sodium hydrogen carbonate (1 ml) and ethanol (1 ml).The solution was acidified (~M-HC~) and the yellow productextracted with ethyl acetate and purified by t.1.c.[benzene-methanol (6 : l)]. Other amino-acid and peptide esterswere treated similarly.3-Nitro-2-fiyridyZ Derivatives (IV) .-The hydrochlorideof lysine methyl ester (5 mg) was shaken for 2 h a t 20 "C inethanol-water (80 : 20; 3 ml; adjusted to pH 9 withEt,N) containing 2-fluoro-3-nitropyridine ( 15 mg) . Theproduct was poured into ethyl acetate-water (1 : 1 ; 20 ml)and the organic phase was dried and evaporated to give theyellow bis-3-nitropyridyl derivative, purified by t .l.c.[benzene-methanol(15 : 2)].Other amino-acid and peptideesters were treated similarly.Reduction and Cyclisation Steps.-The 2,6-dinitrophenylderivative of tryptophan methyl ester (4 mg) was heatedunder reflux in ethanol (4 ml) with cyclohexene (0.05 ml)and 10% palladium-charcoal (5 mg) for 60 min. Theyellow solution became orange and was filtered from thecatalyst to give 3,4-dihydro-3-(indol-3-ylmethyl)-5-nitro-quinoxalin-2( lH)-one (I11 ; R = indol-3-ylmethyl), purifiedby t.1.c. [benzene-methanol (6 : l)]. As a further example,the 3-nitropyridyl derivative of asparagine methyl ester(3 mg) was heated under reflux in ethanol-ethyl acetate(1 : 1; 3 ml) in the presence of cyclohexene (0.05 ml), 10%palladium-charcoal (5 mg), and acetic acid (1 ml) for 45min.The yellow solution became colourless and wasfiltered from catalyst to give 3-carbamoylmethyl-3,4-dihydropyrido[2,3-b]pyrazin-2( lH)-one (VI ; R =CH,*CO*NH,) purified by t.1.c. [benzene-methanol (6 : l)].All other reductions and cyclisations were carried outsimilarly; in the absence of acetic acid the 3-nitropyridylderivatives were quickly reduced but cyclised slowly so thatthe intermediate 3-aminopyridyl compounds (V) could beisolated.Sequencing of an Octafieptide.-The protected octapeptideBoc-Glu (OBut ) -Glu (OBut) -Ala-Glu (OBut) -Glu (OBut) -Ala-Tyr-Gly-OMe (25 mg) * was treated with 90% trifluoro-acetic acid followed by methanolic hydrogen chloride togive the hydrochloride of the peptide H-Glu(0Me)-Glu (0Me)-Ala-Glu (OMe) -Glu (OMe) -Ala-Tyr-Gly-OMe ; thelatter was heated under reflux in 90% ethanol (5 ml) withtriethylamine (0.05 ml) and 2-fluoro-3-nitropyridine (16 mg)for 35 min. The solvents were evaporated off to give ayellow solid which was heated under reflux in ethanol(4.3 ml) and acetic acid (0.7 ml) with cyclohexene (0.05 ml)and 10% palladium-charcoal(10 mg) for 2 h. The catalystwas filtered off and the solution evaporated to leave aresidue (A) which was extracted with dichloromethane.The extract was filtered and evaporated in vacuo to give thepyridopyrazine (VI ; R = CH,*CH,*CO,Me), identified byt.1.c. and mass spectral comparison with an authenticsample. The derivatisation and reduction-cyclisationsteps were repeated with residue A to give successively thepyridopyrazines from Glu(0Me) twice, Ala, Glu (OMe) twice,and Glu(0Me) ; a t the sixth cycle, no pyridopyrazine corre-sponding to alanine was detected.[6/148 Received, 23rd January, 19751* Kindly donated by Professor G. W. Kenner
展开▼
