1977 2529Pterins.t Part 2. Stereochemistry of Catalytic Reduction of 6-Methyl-and 6,7-Dimethyl-pterin and of 2,4-Diamino-6-methylpteridineBy Wilfred L. F. Armarego and Henning Schou, Medical Chemistry Group, The John Curtin School ofMedical Research, The Australian National University, Canberra, A.C.T. 2600, AustraliaCatalytic addition of two molecules of hydrogen to 7-deuterio-6-trideuteriomethylpterin yields a 0.8 : 1 mixture ofcis - a n d trans- 7 - d eu t e r i o - 6 - t r i d e u t e r i o met h y I - 5,6,7,8 -tetra h yd r o p t er i n . S i m i I a r red u c t i o n of 2,4 - d i a m i n o - 7 -d e uteri o - 6 - (pa r t i a I ) t ri d e u t e r i o methyl p t e r id i n e g ives a 1 : 1 mi xt u re of cis- a n d trans- 2.4 - d ia m i no - 7 - d e uteri o - 6 -(partial) trideuteriomethyl- 5,6,7,8-tetrahydropteridine.Catalytic reduction of 6,7-dimethyl-and 6,7-bis(trideuterio-methy1)pterin. on the other hand, is stereospecificand forms only the cis-5,6,7,8-tetrahydro-derivatives. Reductionof 6,7-dimethylpterin with sodium in ethanol provides a 1 : 1 mixture of cis- and trans-6,7-dimethyl-5,6,7,8-tetra-hydropterin. The stereochemistry of these products was deduced from lH n.m.r. spectroscopy.THE 5,6,7,8-tetrahydro-derivatives of 6-methyl- (1) and6,7-dimethyl-pterin (2) are substrates for the mono-oxygenase enzyme systems which hydroxylate ( S ) -phenylalanine to (S)-tyro~ine,~*~ (S)-tyrosine to (S)-3,4-dihydroxyphenylalanine (dopa) ,294 and (S)-tryptophanto (S)-5-hydro~ytryptophan.~ The first-mentioned sys-tem, which has been studied in great detai1,2.3 consists ofat least two separate enzymes: a hydroxylase whichoxidises phenylalanine, and an NADPH-requiring di-hydropterin reductase which reduces the dihydropterinformed back to its tetrahydro-derivative; and the cycleis then repeated.The tetrahydro-derivatives of thesubstrates (1) and (2) can replace the natural coenzyme5 , 6,7,8-t et rah ydrobiopt erin (3) very effect i ~ e l y . ~ ? Muchevidence has been presented in support of the inter-mediate ' quinonoid ' 6,7-dihydropterin (4), which couldnot be isolated because it rearranges rapidly in the absenceof enzymes to the isomeric biologically inactive 7,S-dihy-d r o ~ t e r i n . ~ * ~ 9 ~ . ~ The available data indicate that thechiral centre at C-6 and the chiral or prochiral centre atC-7 are unaffected in the enzymic cycle.We require,t Pterin is d-aminopteridin-4(3H)-one. For a discussion ofthe term see ref. 2 , p. 2.Part 1, W. L. F. Armarego and B. A. Milloy, Austral. J .Chew., 197'7, in the press.R. L. Blakley, ' The Biochemistry of Folic Acid and RelatedPteridines,' North-Holland, Amsterdam, 1969, p. 293.S. Kaufman, Adv. Enzymol., 1971, 35, 245.however, more direct evidence regarding the relativeconfiguration of the two centres during the cycle. Wealso want to know the relative and absolute stereospeci-ficities at these two centres when the substrate (1) isreduced by folate reductases.In order to solve these problems we needed a methodto identify the stereochemistry of addition of hydrogento C-6 and -7 in 6-methyl- and 6,7-dimethyl-pterin.The lH n.m.r.approach chosen was based on the know-ledge that the vicinal coupling constants for the cis-2-and 3-protons (ca. 2.7 Hz) in substituted 1,2,3,4-tetra-hydroquinoxalines were consistently smaller than thosefor the corresponding protons (ca. 8 Hz) in the trans-isomers.' We describe here the preparation and spectraof 6,7-dimethyl- (5), 6,7-bis(trideuteriomethyl)- (8), 6-methyl- (6), 7-deuterio-g-methyl- (7), 6-trideuterio-methyl- (9), and 7-deuterio-6-trideuteriomethyl- (10)5,6,7,8-tetrahydropterins, and of 2,4-diamino-7-deuterio-6-trideuteriomethyl-5,6,7,8-tetrahydropteridine. Therelative stereochemistry of hydrogen addition to C-6 andH.Ozawa and K. Suzuki, J . Pharm. SOC. Japan, 1971, 91,1250.5 V. Massey and P. Hemmerich, ' The Enzymes,' vol. XII, ed.P. D. Boyer, Academic Press, New York, 1975, p. 240.K. G. Scrimgeour, ' Chemistry and Biology of Pteridines,'ed. W. Pfleiderer, de Gruyter, Berlin, 1975, p. 731.7 R. Aguilera, J-C. Duplan, and C. Nofre, Bull. Soc. chim.France, 1968, 44912530 J.C.S. Perkin I-7 to form these tetrahydro-derivatives can now bededuced from lH n.m.r. data.H H( 1 1 R’=R~=H( 2 ) R’=H, R Z M ~( 3 )H R( 4 ) ( 5 ) R=Me( 6 ) R=H( 7 ) R = DH R(8) R = CD,(9) R = H(10) R - DThe methyl protons of 6,7-dimethylpterin, which isreadily prepared from 2,4,5-triamino-6-hydroxypyrimi-dinium sulphate and biacetyl’g are barely exchanged bydeuterium in deuteriated acid. In deuteriated aqueousbase, on the other hand, the exchange is much faster and6,7-bis(trideuteriomethy1)pterin with >98% isotopicpurity is obtained. Catalytic reduction of this productin 3~-hydrochloric acid gave 6,7-bis (trideuteriomethyl) -5,6,7,8- tetrahydropterin hydrochloride (8) in high yieldwithout observable loss of deuterium.The lH n.m.r.spectrum of this compound [see below and Figure l(c)]shows that the configuration is entirely cis. While thiswork was in progress Weber and Viscontini9 demon-strated by a different method that the catalytic reduc-tion of 6,7-dimethylpterin in trifluoroacetic acid gaveexclusively the cis-tetrahydro-derivative (5) (see below).The catalytic reduction of 6-methylpterin to 6-methyl-5,6,7,8-tetrahydropterin provides only one product, al-though the relative stereochemistry of the hydrogen atomsthat add to C-6 and C-7 can be cis or trans.The stereo-chemistry of the addition can, however, be determinedby the reduction of 7-deuterio-6-methylpterin. We haveanalysed the lH n.m.r. spectrum of 6-methyl-5,6,7,8-tetrahydropterin hydrochloride in D20. The axial andequatorial C-7 proton signals are clearly separated and8 H. I. X. Mager, R. Addink, and W. Berends, Rec. Trav.chim., 1967, 86, 833; C. K. Cain, M. F. Mallette, and E. C.Taylor, J . Amer. Chem. SOL, 1946, 68, 1996.R. Weber and M. Viscontini, Helv. Chim. Acta, 1976, 59,2379.their coupling constants with H-6 are observable [seeFigure 2(a) and below].The spectrum of the tetra-hydro-derivative of 7-deuterio-6-trideuteriomethylpterinshould therefore readily show the steric relation of theprotons that have added to C-6 and -7.Our first approach to the synthesis of 7-deuterio-6-methylpterin was by direct deuteriation of 6-methylpter-in. Here, as in 6,7-dimethylpterin, we found that deuteri-ated acid caused very little exchange, other than of thelabile hydrogen atoms on the nitrogen atoms. Deuteri-ated alkali exchanged all the protons on the 6-methyl(b’ RJ I I 14.5 4.08I 7 - 2.0 1.88FIGURE 1 l H N.m.r. spectra of (a) cis-6,7-dimethyl-5,6,7,8-tetrahydropterin hydrochloride in D20 a t 100 MHz; (b) cis-and trans-6,7-dimethyl-5,6,7,8-tetrahydropterin hydrochloridein 0.5~-DCl a t 100 MHz; (c) cis-6,7-bis(trideuteriornethyl)-5,6,7,8-tetrahydropterin hydrochloride in D20 a t 100 MHzgroup, but there was negligible exchange of H-7 underthe variety of conditions tried.We then turned toTaylor’s unambiguous pteridine synthesis; viz. (1 1 ;+ (13 ; R1 = R2 = H) --t (1). Perdeuteriopyruvalde-hyde oxime (deuteriated oximinoacetone) (1 1 ; R1 =R2 = D) of high deuterium content was preparedas described l2 from ethyl aceto-acetate but using10 E, C. Taylor, K. L. Perlman, I. P. Sword, M. Sdquin-Frey,and P. A. Jacobi, J . Amer. Chem. SOG., 1973,95, 6407.11 E. C. Taylor, K. L. Perlman, Y-H. Kim, I. P. Sword, andP. A. Jacobi, J . Amer. Chem. Soc., 1973,95, 6413.l2 L. Vanino, ‘ Handbuch der Praparative Chemie,’ FerdinandEnke Verlag, Stuttgart, 1936, vol.IT., p. 834.R1 = Rz = H) --t (12; R1 = RZ = H, R = C0,Et1977 2531completely deuteriated reagents in deuteriumoxide. The direct deuterium exchange reactions4.2 4.0 3.8FIGURE 2 ‘H N.m.r. spectra of (a) 6-methyl-5,6,7,8-tetra-hydropterin hydrochloride in D20 at 270 MHz; (b) 6-tri-deuteriomethyl-5,6,7,8-tetrahydropterin hydrochloride in D,Oa t 270 MHz ; (c) cis- and truns-7-deuterio-6-trideuteriomethyl-5,6,7,8-tetrahydropterin hydrochloride in D20 at 270 MHzof the oxime were less satisfactory, and in deuteriumoxide containing sodium deuterioxide, the protonsof the methyl group of the oxime were exchanged more(11 100(12)c 0(13)(14)rapidly than the aldehydic proton. Although almostcomplete exchange of the aldehydic proton was possible,it was found that, during exchange of the CD, group togive back a CH, group with aqueous alkali, much self-condensation of the oxime had taken place.This wasrevealed by an increase in the number of C-methylsignals in the lH n.m.r. spectrum. The reaction ofethyl a-aminocyanoacetate toluene-9-sulphonate saltwith the deuteriated oxime proceeded smoothly and gave2-amino-6-deuterio-3-et hoxycarbonyl-5-trideuterio-methylpyrazine 1-oxide (12; R = C02Et, R1 = R2 =D)in which only a small percentage of the deuterium on themethyl group and none of the 6-D had been exchanged.The deuteriated ester reacted with guanidine in the pres-ence of sodium methoxide but provided 6-methylpterin8-oxide (13; R1 = R2 =H) in which almost completeexchange of deuterium at C-7 and in the methyl grouphad occurred.A repetition,of this reaction with deuteri-ated reagents and deuteriated solvents would be pro-hibitive in cost. Alternative syntheses were thereforesought. Attempts to prepare the pterin 8-oxide (13;R1 = R2 = H) from 2-amino-3-ethoxycarbonyl-5-methylpyrazine 1-oxide (12; R = C02Et, R1 = R2 = H)by condensation with cyanamide or methyl isothiouron-ium sulphate, or from 2-amino-3-carbamoyl-5-methylpyr-azine 1-oxide with cyanamide or cyanogen bromide, orfrom 3-carbamoyl- or 3-ethoxycarbonyl-2-amino-5-methyl-pyrazine with guanidine and sodium methoxidewere uniformly unsuccessful. These data imply that inthe formation of the pteridine %oxide, guanidine reactsfirst with the nitrile or ester function in (12; R = CN orC0,Et) , and intramolecular cyclization then takes placeonto the 2-amino-group.If this is not the case then thepyrazine 1-oxide (12; R = C02Et, R1 = R2 = H)should be less reactive than 2-amino-3-ethoxycarbonyl-5-methylpyrazine because of the stronger basicity of theamino-group in the latter. Predicting that the nitrilefunction would be more reactive than the ester functionin this system, we condensed 2-amino-3-cyano-5-methyl-pyrazine with guanidine as above and obtained 2,4-di-amino-6-methylpteridine (14; R1 = R2 = H) in goodyields (cf. ref. 11). When the reaction was repeated withZ-amino-3-cyano-6-deu terio-5-me t hylp yrazine [obtainedfrom the oxime (11 ; R1 = R2 = D) by deoxygenation ofthe derived 1-oxide (12; R = CN, R1 = R2 = D)j, itprovided 2,4-diamino-6-methylpteridine (14) in whichC-7 was completely deuteriated and the 6-methyl groupwas ca.65% deuteriated. This result is not surprisingbecause we know that the hydrogen atoms on the methylgroup in 6-methylpterin are exchanged in the presenceof sodium deuterioxide. Catalytic reduction of the di-aminopteridine gave 2,4-diamino-7-deuterio-6-(partial)-trideut eriomet hyl-5,6,7,8-t et rah ydropteridine hydro-chloride which had a lH n.m.r. spectrum consistent witha product from a mixture of cis- and trans-addition ofhydrogen at C-6 and -7 (see below). The integrals in-dicated that the 6-methyl group was 65% deuteriated.2,4-Diamino-6-methylpteridine was relatively stable inwarm hydrochloric acid but, as in the hydrolysis of 2,4-diamino-6-bromomethylpteridine hydrobromide to 6-bromomethylpterin hydrobromide,13 it was converted13 J.A. Montgomery, J. D. Rose, C. Temple, jun., and J. R.Piper, ‘ Chemistry and Biology of Pteridines,’ ed. W. Pfleiderer,de Gruyter, Berlin, 1975, p. 4852532 J.C.S. Perkin Iinto 6-methylpterin hydrobromide on heating with aque-ous 48% hydrobromic acid. A much more satisfactorypreparation of 6-methylpterin involved heating and re-cryst allising the 2,4-diamino-derivative from %-sodiumhydroxide, which yielded the sodium salt of 6-methyl-pterin from which the free base can be isolated in a purestate by acidification. A similar recrystallisation of 2,4-diamino-7-deuterio-6- (partial) t rideuteriome t hylpteri-dine from aqueous 2N-sodium hydroxide gave 7-deuterio-6-met hylpt erin, whereas heating and recryst allising fromof the latter isomer in the crude and purified samples wasconfirmed by the spectrum of 6,7-bis(trideuteriomethy1)-5,6,7,8-tetrahydropterin hydrochloride [Figure 1 (c)]which was prepared under identical conditions.The Jvalue and chemical shifts of the two doublets are con-sistent with the values derived from the multiplet of thenon-deuteriated isomer. Our deductions and those re-ported9 were made only from the knowledge that thecoupling constant observed between H-6 and -7 wassmall. We confirmed this beyond doubt by preparing alH N.m.r.5,6,7,8-Tetrahydropterin hydrochloridecis-6,7-Dimethylcis-6,7-Bis (trideuteriomethyl)trans-6,7-Dime.thy16-Trideuteriomethyl eff(15; R1 = CD3, R2 = Hc)6-Methyl6-Methyldata of 5,6,7,8-tetrahydropteridines at 100 MHz a4.34 b 4.43 b 6-Meb 1.79 (J 6.7)(J 3.1, 6.7) (J 3.1, 6.7) 7-Meb 1.87 (J 6.7) D,O3.94 4.16 6-Mec 1.87 (J 6.8)(J 8.3, 6.8) (J 8.3, 6.8) 7-Med 1.95 (J 6.8) 0.5~-DC1H- 6 H- 7 CH3 Solvent4.34 (J 3.1) 4.43 (J 3.1) O.5N-DC1HA 4.19, HB 4.19, D2O(JAB 3.6, JAC 9.0) (JAB 3.6, JBC -14.2);(JAC 9.0, JBC - 14.2)Hc 3.85,(as above and J 6.4) (as above and J 6.4) 6-Me 1.98 (J 6.4) D,OH A HB 4.43 O.5N-DCl(JAB 3, JAG 10.0) (JAB 3, JBC 13) ;Hc 4.05(JAG 10.0, JBO 13)trans-7-Deuterio-6-trideuteriomethylf H A 4.19 (J 9.0) HB 3.86 (J 9.0) D2O(16)cis~7-Deuterio-6-trideuteriomethyl f H A 4.20 HB 4.20 D2O(17)2,4-Diamino-5,6,7,8-tetrahydropteridine hydrochloride6-Me t hyl 4.19 7-ax 3.84 6-Me 2.03 (J 6.3) D,O(Juic 10.2, Jgem - 14.5)(Juic 3.1, Jgem - 14.5)7-eq 4.20trans- 7-Deuterio- 6- 4.19 (J 10.2) 3.84 (J 10.2) 6-CHD, 2.03br (d) D,Ocis-7-Deuterio- 6- (partial) trideuteriomethyl 4.20 4.20 6-CHD2 2.03br (d) D,Oments may be reversed.Hz.f Data at 270 MHz are almost identical with these although the spectrum appears simpler (see Figure 2).(partial) trideu teriomet hylConcentration 20 mg in 0.5 ml; tetramethylsilane as external standard. b Assignments taken from ref. 9. c*d Tentative, assign-Computer-simulated spectrum from the experimental signal positions (by M. J. Whittaker), error f O . l9 From ref.14.Complex multiplet.%-sodium deuterioxide gave the deuterio-compound (1 ;R1 = R2 = D) in excellent yields. Catalytic reductionof these compounds furnished the t et rahydro-derivat ives(7) and (lo), respectively, without loss of deuterium atc-7.lH N.m.r. Spectra and Stereochemistry.-The spectrumof 6,7-dimet hyl-5,6,7,8- tetrahydropt erin hydrochloridein deuterium oxide [Figure l(a) and Table] has elevenlines (theory: fourteen lines) for H-6 and -7, and twodoublets for Me-6 and -7. The coupling constantsfor the methyl doublets (6.7 Hz) are almost identical,making the vicinal coupling constant for H-6 and -7 (3.1Hz) readily observable. The small coupling constantsuggests a cis-stereochemistry consistent with a half-chairconformation (15; R1 = R2 = Me) or equilibratinghalf-chair conformations (15) (15a), as has previous-ly been postulated for such sy~tems.7*~~~* The spreadof the signals for cis H-6 and -7 (ca.0.5 p.p.m.) couldconceivably be masking signals from the trans-isomerwhich may be a contaminant. However, the absencel4 R. A. Archer and H. S. Mosher, J. Org. Chem., 1967, 32, 1378.l 5 R. Weber and M. Viscontini, Helv. Chim. Acta, 1975, 58,1772.ca. 1 : 1 mixture of cis- and trans-6,7-dimethyl-5,6,7,8-tetrahydropterin hydrochloride by reduction of 6,7-dimethylpterin with a large excess of sodium in ethanol.The lH n.m.r. spectra of the isomers were clearly separa-ted [see Figure l(b)] and the coupling constant betweenH-6 and -7 in the trans-isomer was significantly largerthan that in the cis-isomer (see Table).The spectrum of 6-methyl-5,6,7,8-tetrahydropterinhydrochloride and its 6-deuterio-derivative in 0 .5 ~ -deuterium chloride at 100 MHz had been reportedpreviously l5 and the J values were deduced by compu-tation. The quartet for H-7 was assigned from the spec-trum of the 6-deuteriated derivative. We attemptedunsuccessfully to convert the ABC pattern of signals fromH-6, Ha-7 and H,,-7 into a first-order spectrum by vary-ing the solvents [e.g. CF3*C02H, D,SO,, (CD,),SO,(CD,),N*CDO, and DCl]; however our spectrum of thehydrochloride in deuterium oxide was similar to the onein o.S~-deuterium chloride reported. The complicatingfactor in this spectrum is the further coupling of H-6 withthe 6-methyl group which makes the theoretical spectrumof H-6 consist of sixteen lines.In a further attempt toobtain a first-order spectrum we measured this compoun1977 2533at 270 MHz [Figure 2(a)]. The signal pattern for H-6and -7 [i.e. in (15; R1 = Me, R2 = H)] at this field wasslightly different from the one at 100 MHz, but there ap-peared to be very little dispersion of H-6 and H,,-7.The pattern was, however, simplified in the spectrum of(18) (19)6- t rideu teriomet hyl-5,6,7,8-t et rah ydropt erin hydrochlor-ide (9) [Figure 2(b)] ; the quartet from Hc in (15; R1 =CD,, R2 = Hc) is clearly visible and the downfield signalscontain the quartet from HB (see simulated spectrum inTable) confirming that the equilibrium (15; R1 = CD,,RZ = Hc) w (15a; R1 = CD,, R2 = Hc) is largely infavour of the conformer with the 6-methyl group pseudo-equatorial, 6-HA pseudoaxial, 7-HB pseudoequatorial,and 7-Hc pseudoaxial (see ref.15). In this example alsothe spectra in deuterium oxide at 100 and a t 270 MHzwere similar, with a very small separation in chemicalshifts between HA and HB [in formula (15; R1 = CD,,R2 = Hc)]. The computer-simulated spectra (kindlycomputed by M. J. Whittaker) together with those re-ported for 6-methyl-5,6,7,8-tetrahydropterin hydro-chloride in 0.5~-deuterium chloride are in the Table.The data indicate that these spin systems are more likean AA’B than an ABC system, and the-computed JAR,J A C , and values are very close (within s O . 1 Hz) tothe values measured directly from the spectrum at 270but not at 100 MHz.The former spectrum is closer to afirst-order spectrum with regard to H,,-7 and H,,-7,but the signals from H-6 are still too broad to be assignedby inspection. The differences in J values between thiswork and that reported l5 may be partly due to the effectof solvent on the equilibrium (15) (15a).It is now possible, with the knowledge of the aboveassignments, to determine the stereochemistry of addi-tion of hydrogen to C-6 and -7 in 7-deuterio-6-trideuterio-methylpterin. The spectrum of the tetrahydro-deriv-ative obtained from catalytic reduction [Figure 2(c)]consists of two doublets (trans J 9.0 Hz) and a singletinside the downfield doublet. The chemical shifts ofthese signals, when compared with those of 6-trideuterio-met hyl-5,6,7,8-t et rah ydropt erin hydrochloride, confirmthat the compound is a mixture consisting of the trans-isomer (a quartet) in the predominant conformation (16)and the cis-isomer (a singlet) in the predominant conform-ation (17), with almost similar chemical shifts for HI andHB in the cis-isomer, although the trans- (16) and cis- (17)conformers predominate.The spectra give time-averaged lines for each isomer due to the conformers inequilibrium. The slightly broadened singlet from thecis-isomer in the spectrum [Figure 2(c)] is not centredbetween the doublet from the trans-isomer and can beexplained if the position of equilibrium of the conformersis slightly difierent in the two isomers.This differencecan be caused by the effect of the deuterium atom on theconformational equilibria, or a deuterium isotope effecton the chemical shift of the geminal proton, and is prob-ably not of steric origin. The cis : trans ratio calculatedfrom the integrals is 0.8 : 1.The spectrum of 2,4-diamino-6-methyl-5,6,7,8-tetra-liydropteridine hydrochloride is quite similar to that ofthe pterin (6), i.e. an AA’B pattern, and the chemical shiftof the C-7 protons can be assigned by inspection from theabove knowledge. 2,4-Diamino-7-deuterio-6- (partial) -t rideut eriomet hylpteridine gave on cat a1 ytic reductionin 3~-hydrochloric acid the corresponding tetrahydro-derivative, which had a spectrum similar to that of thepterin (10). The cis : trans ratio in this case was ca1 : 1.The difference in the stereospecificity between thecat a1 ytic addit ions of hydrogen to 6,7-dimet h ylpt erinand to 6-methylpterin deserves some comment.Un-doubtedly two reduction steps are involved : addition ofone molecule of hydrogen across the 7,8-double bondfollowed by addition across the 5,6-double bond. Wehave checked this point with 6-methylpterin by measur-ing the lH n.m.r. and U.V. spectra of samples withdrawnafter the absorption of 0.4,0.9, 1.5, and 2.0 mol. equiv. ofhydrogen. The spectra showed that the ratios of 6-methylpterin to 6-methyl-7,8-dihydropterin to 6-methyl-5,6,7,8-tetrahydropterinwere0.7 : 1.0 : 0.0; 0.1 : 1.0 : 0.0;0.0 : 1.0 : 0.8; and 0.0 : 0.0 : 1.0, respectively. A samplewithdrawn after absorption of 1 mol.equiv. of hydrogenwas evaporated and the product converted into the di-thionite salt. This proved identical with authentic6-met h yl-7 &dihydropt erin dit hionit e. In 6,7-dimet hyl-pterin addition of hydrogen across the 7,8-double bond isslower than in the above and gave ratios of pterin to di-hydropterin to tetrahydropterin of 0.4 : 1.0 : 0.4; 0.1 :0.8 : 1.0; and 0.0 : 0.0 : 1.0 after absorption of 1.0, 1.5,and 2.0 mol. equiv. of hydrogen, respectively. The inter-mediate was shown to be 6,7-dimethyl-7,8-dihydropteriJ.C.S. Perkin Iby aeration of a sample at the end of the reduction, givinga U.V. spectrum similar to the one obtained from thesample withdrawn after absorption of 1.5 mol. equiv.of hydrogen.The lH n.m.r. spectra are also consistentwith these findings. Aeration of 6,7-dimethyl-5,6,7,8-tetrahydropterin is known to furnish the 7,8-dihydro-derivative.16 The stereospecificity of cis-addition in6,7-dimethylpterin can be explained by adsorption ofthe molecule on the reduced catalyst and the additionof one molecule of hydrogen across the 7,8-double bond.Then either the dihydro-substrate is held on the catalystand a second molecule of hydrogen is added stereospeci-fically, or more likely it is desorbed, and readsorbedstereospecifically because of the encumbrance of the 7-methyl group [Le. structure (19) would be more favouredin the transition state than structure (IS)]. In the caseof 6-methylpterin, after the 7,s-dihydro-compound isformed it must be released into the solution and re-adsorbed on the catalyst, almost randomly because of lackof steric hindrance near C-7, and reduced further acrossN (5) -C (6).EXPERIMENTALElemental analyses were determined by the AustralianNational University Analytical Service Unit ; values forH + D were calculated as before.17 1.r.spectra of solids(KBr) were measured with a Unicam SP 1000 spectrometer,and U.V. spectra with a Unicam SP 1800. lH N.m.r. spectrawere obtained with Varian T6OA and HA100 spectrometers(tetramethylsilane as internal or external lock). 270MHz Spectra were measured with a Brucker HFX-270 spec-trometer by the National NMR Centre (Dr. A. J . Jones). JValues are in Hz. Mass spectra (by Dr. J.K. MacLeod andstaff) were measured with an A.E.I. MS9 instrument.Deuterium oxide ( >99.9yo) was purchased from theAustralian Atomic Energy Commission. All evaporationswere carried out a t <30 "C and 18 mmHg.6,7-Dimethylpterin (7 g) was converted into the sodiumsalt (82%) by recrystallisation from 2~-sodium hydroxide(150 ml). The salt was washed with a little cold waterthen ethanol and dried a t 100 "C, and had m.p. >360 "C(decomp.) (Found: C, 45.3; H, 3.9; N, 32.6; Na, 11.1.C,H,N,ONarequiresC, 45.1; H, 3.8; N, 32.85; Na, 10.8%).The sodium salt of 6,7-bis(trideuteriomethyl)pterin wasprepared in 70% yield by heating 6,7-dimethylpterin (250mg) a t 100 "C in 2~-sodium deuteroxide in deuterium oxide(25 ml) for 24 h in a sealed tube and isolated as above.Ithad no proton n.m.r. signals in 2~-sodium deuterioxide or2~-deuterium chloride.6-MethyZpterin Sodium Salt.-Crude 6-methylpterin, pre-pared from 2,5,6-triamino-4-hydroxypyrimidinium sulphateas before l8 but on a 130 g scale, was shown by lH n.m.r.spectroscopy in ~N-DC~-D,O to contain 36% of 7-methyl-pterin [a 2.77 (7-Me), 2.80 (6-Me), 8.81 (6-H), and 8.93 (7-H)].The pure sodium salt of 6-methylpterin was obtained byrecrystallisation of the mixture from 10 parts of 2~-sodiumhydroxide (50% recovery) and had m.p. >360 "C (decomp.)[Found (after drying a t 150 "C for 12 h): C, 41.4; H, 3.3;l8 J. H. Bieri and M. Viscontini, Helv. China. Acta, 1974, 57,W. L. I;. Armarego, B. A. Milloy, and W. Pendergast, J.C.S.1651.Perkin I , 1976, 2229.N, 34.4; Na, 11.3.C,H6N,Na0,0.25H,O requires C, 41.3;H, 3.2; N, 34.4; Na, 11.3y0]. The free base was preparedby acidifying an aqueous solution of the salt, and collectingand washing (H,O and EtOH) by centrifugation becauseconventional filtration was exceedingly slow. This com-pound and its tetrahydro-derivative were identical withsamples from an unequivocal synthesis.1°6-Trideuteriomethylpterin Sodium Salt.-6-Methylpterin( 1 g) in 2~-sodium deuterioxide (70 ml) was heated in a bomba t 100 "C for 24 h. The sodium salt (600 mg) that crystal-lised on cooling was collected, washed with a little water andethanol, and dried. It had m.p. >360 "C (decomp.)(Found: C, 39.7; H + D,4.7; N, 32.9; Na, 10.7. C,H,D,-N,Na0,0.5H20 requires C, 39.8; H + D, 4.7; N, 33.2;Na, 10.9%).Prolonged heating of the sodium deuterioxidesolution a t 120 "C did not cause H-7 to be displaced bydeuterium.7-Deuterio-6-methylpterin Sodium Salt.-2,4-Diamino-7-deuterio-6-(partial) trideuteriomethylpteridine (200 mg ; seebelow) in 2~-sodium hydroxide (75 ml) was stirred a t 100 "Cfor 9 h. The solution was concentrated until a solid crystal-lised, and was cooled. The yellow sodium salt (120 mg) wascollected as above. A further 60 mg of salt was obtainedfrom the mother liquors. The U.V. spectra and t.1.c. proper-ties were identical with those of authentic non-deuteriated6-methylpterin, and only a sharp 6-methyl signal was presentin the lH n.m.r. spectrum (Found : Na, 11.3. C,H,DN,NaOrequires Na, 1 1.5 yo).7-Deuterio-6-trideuteriomethylpterin Sodium Salt.-Thiswas prepared as above from 2,4-diamino-7-deuterio-6-(par-tial) trideuteriomethylpteridine in 2~-sodium deuterioxide(18 h a t 100 "C) and had m.p. >360 "C (decomp.) (Found:C, 40.8; H + D, 5.2; N, 34.4.C,H,,D4~,N,Na0,0.1H,0requires C, 40.9; H + D, 5.2; N, 34.1%). The U.V. spectraand t.1.c. behaviour were identical with the above, but theproduct had no lH n.m.r. signals in 2~-deuterium chloride.Deuteriated Hydroxyiminoacetone.-Ethyl acetoacetate(23.8 g) was added to a solution of sodium (4.6 g, 1.1 mol.equiv.) in deuterium oxide (300 ml) followed by sodiumnitrite (1 3.2 g) . The mixture was set aside for 24 h, acidifiedto pH 1 with concentrated hydrochloric acid, and extractedthoroughly with ether and dried (Na,S04) .Evaporationgave the deuteriated oxime (14.5 g, 87y0), which was sublimeda t 50 "C and 0.5 mmHg, and had m.p. 65-66 "C (lit.,l* 65 "Cfor non-deuteriated compound prepared by nitrosation ofacetone) ; vmax. 1 660 (CO), 1 445, and 980 cm-l (Found: C,39.7; H + D, 10.0; N, 15.2. C,HD4N0, requires C,39.55; H + D, 9.95; N, 15.4.x) ; m/e 92 (7%, C,D,NO,'+),NO,*+), and 87 (0) ; the CH:NOH signal was absent in the1H n.m.r. spectrum.2-Amino-6-deuterio-3-ethoxycarbonyl-5-(~artiaZ)trideuterio-methylpyrazine l-Oxide.-The preceding deuteriated oxime(2.2 g) and ethyl a-aminocyanoacetate toluene-p-sulphonatesalt 2o (7.6 g) in methanol (10 ml) were stirred a t 35 "C for 24 h.The solution was evaporated and diluted with water (40 ml) ;the pH was adjusted to 9 and the mixture extracted withchloroform.Evaporation of the extract gave the deuteri-ated pyrazine oxide (3.3 g), m.p. 132-133.5 "C after sublim-ation a t 140 "C and 0.5 mmHg (lit.,* m.p. 132.5-133.5 "Cfor non-deuteriated pyrazine oxide) ; m/e 185 (28%, C,H,D,-l8 J. Semb, U.S.P. 2,477,426/1949 (Chem. A h . , 1950,44, 1146).l9 W. Kiister, 2. physiol. Chern., 1926, 155, 157; B. Ohta, J .2o D. H. Robinson and G. Shaw, J . C . S . Perkin I , 1972, 1715.91 (100, C3HD4N0,'+), 90 (24, C3H,D,NO,'+), 89 (7, C,H,D,-Pharm. SOG. Japan, 1948, 68, 2261977N302'+), 184 (71, C8H,D3N302'+), 183 (100, C8H9D2N302'+),182 (85, C8HloDN,02"), and 181 (0); vmX. 3 470, 3 335 (NH),1690 (CO), and 1610 cm-l; U.V.data identical with lit.values lo for non-deuteriated oxide; 6 (60 MHz; CDCl,) 1.45(3 H, t, J 7, Me), 2.47 (2 H, s, 5-Me) and 4.51 (2 H, q, J 7,CH,), and 7.3br (2 H, s, NH,), with no signal for H-6(Found: C, 48.3; H + D, 6.7; N, 21.1. C,H,D,N,O, re-quires C, 48.2; H + D, 6.6; N, 21.1%).2-Amino-3-cyano-6-deuterio-5-trideuteriomethyl~yrazine 1-Oxide.-The preceding deuteriated hydroxyiminoacetone(5.8 g) and a-aminomalononitrile toluene-P-sulphonate salt 21(16.9 g) in propan-2-01 (100 ml) were stirred at 25 "C for 4 h.The pyrazine oxide that separated contained some toluenep-sulphonic acid which was removed by dissolving the mix-ture in water, adjusting the pH to 9, and extracting withchloroform. Evaporation gave the deuteriated Pyrazineoxide (74o/b), m.p.188.5-189 "C, after sublimation a t 125 "Cand 0.1 mmHg (1it.,l1 m.p. 187-188 "C for non-deuteriatedoxide) ; U.V. data identical with those of non-deuteriatedoxide; vmaX. 3 405, 3 310 (NH), 2 150 (CN), 1 645, and 1 625cm-1; m/e 156 (2%), 155 (13), 154 (100, C2H2D4N,0'+), 153(39, C,H3D,N40'+), 152 (5, C,H,D,N,O'+), 151 (l), and 150(0) (Found: C, 46.7; H + D, 6.3; N, 36.2. C,H,D,N,Ore-quires C, 46.75; H + D, 6.5; N, 36.3%).2-Amino-3-carbarnoyl-5-methylpyrazine l-Oxide .-2-Amino-3-cyano-5-methylpyrazine l-oxide l1 (5 g) in sul-phuric acid (d 1.8; 25 ml) was heated a t 100 "C for 15 min;the product was poured into water and neutralised withaqueous ammonia, and the yellow solid was collected andsublimed a t 180-190 "C and 0.1 mmHg to give the amide(84y0), m.p.230-230.5 "C (lit.lo 218-219 "C; lit.,22 235-236 "C), as prepared from a-aminocyanoacetamide andhydroxyiminoacetone; vmRX. 3 470, 3 440, and 3 400 (NH),1695, 1670 (amide), 1620 (C=N), and 1 170 cm-l (N-0)(Found: C, 43.1 ; H, 4.8; N, 33.5. Calc. for C,H,N,O,: C,42.8; H, 4.8; N, 33.3%). The amide could not be hydro-lysed further by prolonged heating in sulphuric acid.2- Amino-3-cyano-6-deuterzo-5-( partia1)trideuteriomethyZ-pyrazine.-This was prepared (46% yield) by deoxygenationof the l-oxide with phosphorus trichloride (with one quarterof the proportion used in the reported procedure 11), andafter addition of water it was necessary to concentrate thereaction mixture until turbid, and cool.The pyrazine hadm.p. 174.5-175.5 "C after sublimation (1it.,lo 172-173 "Cfor non-deuteriated pyrazine) ; U.V. spectrum identical withthat of non-deuteriated pyrazine; vmx. 2 140 (CN) and 1 215scm-l (Found: C, 53.1; H + D, 5.6; N, 41.3. C,H,.,D,.,Nrequires C, 53.1; H + D, 5.6; N, 41.3%), H-6 signalabsent in the lH n.m.r. spectrum.2,4- Diamino- 7-deuterio-6-( partial) trideuteriornethylpteri-dine.-The preceding deuteriated pyrazine (5 g) was addedto a niethanolic solution of guanidine [from guanidine hydro-chloride (4.1 g) dissolved in methanol (350 ml) containingsodium (2.6 g)], and the mixture was stirred a t room tem-perature. No solid had separated after 16 h, and the solu-tion was boiled under reflux for 18 h. The yellow solid thatseparated on cooling (4.7 g, 73%) was collected, washed withwater, and dried.The pteridine had m.p. >360 "C (de-camp.), and its t.1.c. properties and U.V. spectra were identi-cal with those of authentic non-deuteriated diaminopteri-dine. The lH n.m.r. spectrum (solvent 2~-deuterium chlor-ide) showed no peak for H-7, and the 6-methyl signal wasweak and broad because of geminal H-D coupling (Found :21 J . P. Ferris, R. A. Sanchez, and R. W. Mancuso, Org. Synth.,Coll. Vol. V, 1973, p. 32.C, 46.7; H + D, 5.6; N, 46.4. C7H,D2N,,0.15H,0 re-quires C, 46.5; H + D, 5.7; N, 46.5%).2,4-Diamino- 6-methyl- 5,6,7,8-tetrahydropteridine Hydro-chloride -2,4-Diamino-6-methylpteridine l1 (2.5 g) wasadded to a pre-reduced suspension of platinum oxide(250 mg) in 3~-hydrochloric acid (250 ml) and shaken withhydrogen at 20 "C and 720 mmHg.After absorption of thetheoretical amount of hydrogen (3 h), the catalyst wasfiltered off, and the filtrate evaporated. The residue wasrecrystallised from ethanol containing a little ethanolichydrogen chloride and gave (quantitative) the tetrahydropter-idine hydrochloride, m.p. >230 "C (decomp.), Amx. (pH 2) 219(log E 4.21) and 274 nm (4.15) [Found (after drying a t 100 "Cfor6h): C,21.3; H,5.0; C1,45.2. C,H,,N6,5HC1,1.75H2Orequires C, 21.3; H, 5.2; C1, 45.0%]. The HC1 content de-creased on further heating but the salt darkened in colour.The lH n.m.r. data are in the Table.cis- and trans-2,4-Diamino-7-deuterio-6-(~artial)trideuterio-methyl-5,6,7,8-tetrahydropteridine Hydrochloride.-This mix-ture was prepared as above from the preceding deuteriateddiaminopteridine and had m.p. 238-240 "C (decomp.) ; itsU.V.spectrum was identical with that of the non-deuteriatedpteridine salt (Found: C, 27.7; H + D, 6.8; C1, 27.1. Calc.for C7HloD,N,,2.3HC1,2H,0: C, 27.8; H + n, 6.8; C1.27.0%).6- Trideuteriomethyl- 5,6,7,8-tetrahydropterin Hydrochlor-ide.-This derivative, m.p. >260 "C (decomp.) (Found: C,31.2; H + D, 7.0; C1, 24.8. C7H,D,N50,1.9HC1,H,0requires C, 31.0; H + D, 6.65; C1, 24.7%) was prepared bycatalytic reduction of 6-trideuteriomethylpterin as above ;the lH n.m.r. spectrum is in Figure 2(b) and in the Table.Similarly cis- and trans-7-deuterio-6-methyl-5,6,7, %tetra-hydropterin hydrochloride, m.p.> 260 "C (decomp.)(Found: C, 30.5; H + D, 6.1; C1, 22.9. Calc. for C,Hlo-DN,0,1.8HC1,1.5H20: C, 30.7; H + D, 6.2; C1, 23.0%)and cis- and trans-7-deuterio-6-trideuteriomethyl-5,6,7,8-tetrahydropterin hydrochloride, m.p. > 260 "C (decomp.)(Found: C, 34.1; H + D, 6.6; C1, 23.6. Calc. for C,H,D4-N50,1.6HC1: C, 34.4; H + D, 6.8; Cl,23.7yo) [lH n.m.r.spectrum in Figure 2(c) and the Table] were prepared bycatalytic reduction as above. The U.V. spectra and t.1.c.behaviour of these salts were identical with those of authen-tic 6-methyl-5,6,7,8-tetrahydropterin hydrochloride. TheHC1 and H,O contents of the crystals varied with thedrying conditions and the nitrogen figures were all consistent-ly ca. 1% too low.cis-6,7-Bis (trideuteriomethyl) -5,6,7,8-tetrahydropterin Hy-drochloride.-This derivative, m.p. > 300 "C (decomp.) .(Found: C, 35.0; H + D, 7.8; C1, 24.6; N, 25.1. C,H7D,-N,0,1.9HC1,0.25H20 requires C, 34.9; H + D, 7.3; C1,24.6; N, 25.4y0), was prepared by catalytic reduction of 6,7-bis(trideuteriomethy1)pterin sodium salt as above and hadthe same U.V. and t.1.c. properties as the authentic non-deuteriated salt. The lH n.m.r. spectrum is in Figure l(c)and the Table.cis- and trans-6,7-Dimethyl-5,6,7,8-tetrahydropterin Hy-drochloride.-To the sodium salt of 6,7-dimethylpterin (426mg) under dry nitrogen in boiling ethanol was added sodiumuntil the U.V. spectrum of a sample a t pH 2 indicated that re-duction was complete. A total of 17 g of sodium was addedduring a reflux period of 48 h. The solution was cooled inan ice-bath and acidified with ethanolic hydrogen chloride22 F. Chillemi and G. Palamidessi, IZ Farmaco, Ed. Sci., 1963,18, 566J.C.S. Perkin I(120 ml) under nitrogen. Sodium chloride was removedby repeated concentration and filtration. The ca. 1 : 1 mix-ture of cis- and trans-6,7-dimethyltetrahydropterin hydro-chloride (282 mg) obtained from the mother liquors wascrystallised from ethanol containing a little ethanolic hydro-gen chloride under nitrogen. The U.V. data of the mixture,m.p. >250 "C (decomp.), were identical with those of thecis-isomer; the lH n.m.r. spectrum is in Figure l(b) and theTable (Found: C, 36.2; H, 5.8; C1, 24.8. Calc. for C,H,,-N,O,1.85HCl,O.lH,O: C, 36.3; H, 5.7; C1, 24.8%). Thetrans-isomer was clearly less stable to aerial oxidation thanthe cis-isomer, with a ti value at 20 "C and pH 2 (h,,lyb. 219nm) of 40 min, to be compared with 5.2 h for the pure cis-isomer. Chromatographic separation of the mixture hasnot yet been achieved.We thank Drs. D. J. Brown and J. H. Lister for discussionsand the Australian National University for a PostgraduateScholarship (to H. S.).[7/701 Received, 26th Apvil, 1977