Amino acids. 8. A novel synthesis of .gamma.-carboxy-L-glutamic acid

Amino acids. 8. A novel synthesis of .gamma.-carboxy-L-glutamic acid from L-5-oxoproline esters. Franz Effenberger, Wolfgang Mueller, Roland Keller, ...
0 downloads 0 Views 1MB Size
J . Org. Chem. 1990,55, 3064-3067

3064

Amino Acids. 8.' A Novel Synthesis of y-Carboxy-L-glutamic Acid from L-5-Oxoproline Esters Franz Effenberger,* Wolfgang Muller,2aRoland Keller,2bWolfgang Wild,2cand Thomas Ziegler Institute for Organic Chemistry, University of Stuttgart, Pfaffenwaldring 55, 0-7000 Stuttgart 80, Federal Republic of Germany Received J a n u a r y 26, 1990

N-Alkyl-~-5-oxoprolineesters L-5react with phosgene to form primarily the cy-chloro enamines 6, which then react further to give the 4-carboxylated compounds 7. Solvolysis of the latter with an alcohol furnishes the alkyl N-alkyl-4-(alkoxycarbonyl)-~-5-oxoprolinates 2L,4DL-8 in good yields. The dibenzyl N-benzhydryl ester 2 ~ , 4 ~ - 8 c can be converted simply to y-carboxyglutamic acid (L-Gla)by way of the N-Boc-protected dibenzyl ester 2L,4DL-10.

Introduction In a previous paper,' we have reported on a simple synthesis of L-proline starting from the L-5-oxoprolinates L-1 (Scheme I). In this process, the first isolable compounds obtained from the reaction of the esters L-1with phosgene are the dichloro derivatives L-2 from which, by cleavage of HCl with a tertiary amine, the products L-3 are accessible.' It thus seemed possible that compounds L-3could react with electrophiles as a-chloro enamines at the P-position with retention of chirality of the molecule. Carboxylation of L-3 would then, for example, open up a simple access to y-carboxy-L-glutamic acid (L-Gla),a compound that was first detected in proteins in 1974.3 Shortly after the discovery of L-Gla, numerous syntheses of racemic DL-Gla were p u b l i ~ h e d . Optically ~ pure L-Gla was obtained for the first time by Schwyzer et aL5through diastereomeric separation of the quinine and ephedrine salts of DL-Gla. In addition to further methods for this resolution, optically induced reactions for the preparation of L-Gla were also developed.6 The preparation of L-Gla starting from derivatives of L - G ~ was u described by four groups: by Danishefsky et ale7from Z-protected benzyl pyroglutamate, by Zee-Cheng and Olson8" and Baldwin et al.8bby practically racemization-free carboxylation of the lithium salt of dibenzyl N-(triphenylmethyl)glutamate,and by Tanaka et aL9 from an L-prolinol derivative. Our attempts to bring about reaction of L-3 in the Pposition by treatment with excess phosgene were unsuc(1) For part 7, see: Drauz, K.; Kleemann, A,; Martens, J.; Scherberich, P.; Effenberger, F. J . Org. Chem. 1986, 51, 3494.

(2) (a) Muller, W. Diplomarbeit, Univ. Stuttgart, 1981. (b) Keller, R. Zulassungsarbeit, Univ. Stuttgart, 1980. (c) Wild, W. Forschungspraktikum, Univ. Stuttgart, 1988. (3) (a) Stenflo, J.; Fernlund, P. W.; Roepstorff, P. Proc. Natl. Acad. Sci. U.S.A. 1974, 71, 2730. (b) Nelsestuen, G. L.; Zytkovicz, T. H.; Howard, J. B. J . Bioi. Chem. 1974,249,6377. (c) Magnusson, S.; Sottrup-Jensen, L.; Ellabac Petersen, T.; Morris, H. R.; Dell, A. FEES Lett. 1974.44. 189. (d) Grav. A. L.: Koke. R. A.: Dearfield. D. W.. 11: Hiskev. R. G: J.'Org. Chem. l"985, 56, 2189.' (e) Kasei, T.; Larsen, 'P. '0. PrG: Chem. Org. Prod. 1980, 39, 173; Chem. Abstr. 1980,93, 72221r. (4) (a) Morris, H. R.; Thompson, M. R.; Dell, A. Biochem. Biophys. Res. Commun. 1975,62,856. (b) Boggs, N. T., 111;Gawley, R. E.; Koehler, K. A.; Hiskey, R. G. J . Org. Chem. 1975,40, 2850. (c) Bajusz, S.; Juhasz, A. Acta Chim.Acad. Sci. Hung. 1976,88, 161. (d) Weinstein, B.; Watrin, K. G; Loie, H. J.; Marrin, J. C. J. Org. Chem. 1976, 41, 3636. (e) Juhasz, A,; Bajusz, S. Int. J. Pap. Protein Res. 1980, 15, 154. (5) Marki, W.; Opplinger, M.; Thanei, P.; Schwyzer, R. Helu. Chim. . Acta 1977, 60, 798. (6) (a) Boggs, N. T., 111; Goldsmith, B.; Gawley, R. E.; Koehler, K. A.; Hiskey, R. G. J . Org. Chem. 1970, 44, 2262. (b) Cerovsky, V.; Jost, K. Collect. Czech. Chem. Commun. 1984. 49. 2562. (i)Danishefsky, S.; Berman, E.: Clizbe, L. A.; Hirama, M. J . Am. Chem. SOC.1979, 101, 4385. (8) (a) Zee-Cheng, R. K. Y.; Olson, R. E. Biochem. Biophys. Res. Commun. 1980, 94, 1128. (b) Baldwin, J. E.; North, M.; Flinn, A.; Moloney, M. G.J . Chem. Soc., Chem. Commun. 1988, 12, 828. (9) Tanaka, K.; Yoshifuji, S.; Nitta, Y. Chem. Pharm. Bull. 1986, 34, 3879 ~~

0022-3263/90/1955-3064$02.50/0

Scheme I"

H 1-1

0SC'CI L-2

O+-Cl L- 3

OR: a = Me; b = CH2C6H5= Bz.

Scheme I1

ether

CH

SiMe3 1-4 - .

R1'

'R2

L- 5

L

R R' R2

4a,5a

5b

4c,5c

Me H

Me Ph Ph

Bz

Ph

Ph Ph

cessfu1,l although such a reaction was to be expected on the basis of literature data. For example, N,N-disubstituted carboxamides react with excess phosgene to furnish a-chloro-0-(chlorocarbonyl) enamines via a-chloro enamine intermediates'O while N-(methoxycarbony1)-substituted enamines can be acylated at the P-position even under mild conditions." Apparently, the nucleophilicity of the Pposition in compounds L-3 is weakened by the strong chlorocarbonyl acceptor group at the nitrogen atom to such an extent that carboxylation by phosgene is no longer possible. Hence, we attempted to replace the N-(chlorocarbonyl) substitutent, in L-3 by donor substituents such as, for example, alkyl groups in order to increase the reactivity of the a-chloro enamines so that carboxylation at the /?-position by phosgene would become possible. Discussion Preparation of N-Alkyl-~-5-oxoprolineEsters L-5. The N-substituents of the N-alkyl-~-5-oxoprolinatesL-5 were selected with regard to their stabilities toward acids so that, on the one hand, they would not be cleaved by the hydrogen chloride formed during the phosgenation or ~~~~~~

(10) (a) Buyle, R.; Viehe, H. G. Tetrahedron 1968,24,4217; 1969,25, 3447. (b) Ghosez, L.; Marchand-Brynaert, J. Adu. Org. Chem. 9,421, J. Wiley, N. Y., London, Sydney, Toronto 1976. (11) Shono, R.; Matsumura, Y.; Tsubara, K.; Suginara, Y. Tetrahedron Lett. 1982. 1201.

0 1990 American Chemical Society

J. Org. Chem., Val. 55, No. 10,1990 3065

y-Carboxy-i-glutamic Acid from L-5-OxoprolineEsters

Figure I. Stereoscopic projection of 2~,4n-8b.

elimination and, on the other hand, they could he removed without racemization taking place. We considered the N-benzyl- and N-benzhydryl substituents to be most suitable for this purpose on the basis of their solvolysis rates determined in trifluoroacetic acid by Weygand et d.12 Since previously performed N-alkylations of the alkali metal salts of L-1were accompanied by complete race~nization,'~'~~ we have prepared the N-benzyl- and Nbenzhydrylprolinates 5 by way of the silyl derivatives 4.'" Protodesilylation of L-4 occurred as a side reaction, especially in reactions with benzhydryl halides, during the alkyldesilylation reaction performed with neat substance at 120-130 "C. This side reaction is apparently a result of the well-known hydrogen halide elimination on heating henzhydryl halides. By means of the rapid removal of HBr (passage of dry nitrogen over the reaction mixture or working under vacuum), we were able to suppress the protodesilylation almost completely and thus obtained the optically pure 1-benzyl- (L-5a.b) or l-benzhydryl-3-0~0prolinates ~-5cl'in yields between 60 and 80% from L-4 (Scheme 11). Phosgenation of N-Alkyl-Substituted L-5-Oxoprolinates L-5. In the course of reactions of L-5 with phosgene under varying conditions, it was found that a reaction temperature of about 40 O C , a reaction time of