1040
E. E.SMISSMAN, J. T. SUH, hf.
comparison of the infrared spectrum with t h a t of the product obtained from t h e zinc chloride dehydration of spiro[4,5]decdn-l-o1 according t o the procedure of Mayer,14 which has been shon-n t o yield this olefin a s the major product. T h e by-products present were of a pre-
[ C O s T R I B U T I O S FROM THE
OXMAN AND
Vol. 84
R. DANIELS
sumably rearranged acetate (determined by micro saponification), which exhibited a weak absorption band a t 1740 cm.-’, and weak bands appeared in both samples a t 3070 and 743 cm.-’ which possibly is indicative of olefinic inaterial with a double bond exocyclic to one of the ring systems.
DEPARTXESTS O F PHARMACEUTICAL CHEMISTRY O F THE ~ N I V E R S I T VO F KANSAS, LAWRENCE, KANS.,AND THE UXIVERSITY OF WISCOKSIN, MADISON, WIS.]
A Stereospecific Synthesis of D( - )-Shikimic Acid1z2 BY EDWARD E. SXISSXAN, JOHN T. SUH,MICHAEL OXMASA N D RALPHDAKIELS RECEIVED AGGUST16, 1961
A stereospecific synthesis of dl-shikimic acid ( I ) froin the Diels-illder adduct of lians,trarzs-l,i-diaceto~~--1,3-butad~ene (11) and methyl acrylate is reported. T h e optical resolution also was effected.
Naturally occurring shikimic acid was first isolated in 1SS5.3 Fischer and Dangschat reported its structure to be I.4 KO total synthesis of shi-
1
kimic acid had previously been described, although Grewe and co-workers had reported the synthesis of quinic acid5 which Fischer and Dangschat had previously converted to shikimic acid.ti This acid has been shown to be the biological precursor of phenylalanine, tyrosine and tryptophan,I which, in turn, are the parents of the majority of plant alkaloids. I t is also involved in the biosynthesis of lignin,8 one of the major constituents of wood, flavonoids,9 and of other important aromatic compounds. l o The enzymatic synthesis of shikimic acid from ~-erythrose-4-phosphate and phosphoenolpyruvate was reported by Srinivasan, Katagiri and Sprinson.ll Since the acid has been shown to be an important link in aromatic biogenesis, it was desirable to devise a coinplete synthetic scheme whereby carbon-14 could be introduced into specific positions of the iiiolecule. If this could be accomplished, the pathways of forniatioii of various iiaturally occurring aromatic compounds could be followed. (1) This work was supported in part by grants from t h e Sational Institutes of Health, National Science Foundation, and Research Corporation. (1) Thi? work was initially reported in a Communication t o t h c lklitor, J . A m Chem. Soc., 81,2505 (1959), and was later corroborated b y Iangschat, H e l u . Chin?. R c l n , 17, 1200 ( l ! i 3 4 ) ; 18, 1211 (1933); 20, 708 (1937). ( 5 ) R . Grewe, W. Lorenzen and I,. Vining, C h e w . B c i . . , 87, i K (lg.54). (6) H. 0 . I.. Fischer and G. Dangschat, Biochiir~.Biophjr. .4cla, 4,
199 (1520). (7) B. D. Davis, .I. B i d . C h e r i i . , 191, 31: (1551); J . B a c t e i i d , 64, 729, 749 (19.52). 18) F. F. S o r d and W. J Srhubcrt, / . : x p r t e i i t i o , 15, 213 (lO,i!)). (9) E. %’, Underhill, J. E. U’atkins and 4 . C. Keish, ( ‘ a n . J . Hioi h e i r i . oiid P h y s i o l , 36, ‘219 (1957). (IO) G. Ehrensvaard, Cheiii. S o r . S p e c . Piibi , No. 1 2 , 17 (1938). (11) P. R. Srinivasan, X I . Katagiri and D. I3. Sprinson, J . A l r l . Chev7. S O C . , 77, 4543 (1955).
The route to shikiniic acid which proved successful was based on a Diels-Alder reaction of frans,trans - 1,4- diacetoxy - 1,3- butadienelZsl3(11) with methyl acrylate to afford methyl 2P,5P-diacetoxy3-cyclohexene-l a-carboxylate (111) in 93% yield. Although attempted hydroxylation of the double bond in 111 with potassium permanganate led to a mixture of compounds, cis-hydroxylation anti to the two acetoxy groups was successfully achieved by employing osmium tetroxide. The product obtained in this reaction was designated as methyl 23,SB-diacetoxy-3a,4a-dihydroxycyclohexane - 1CYcarboxylate (IV) . The stereochemical assignment of the carbomethoxy function a t 1 differs from that proposed by McCrindle, Overton and Raphael, but is in accord with the results of pyrolysis of the corresponding acetonide V. 0 il
OCC‘H3 1
OCCH, I1
0
1
!
I1
1
111
VI
1 , I
v
8
OCCHS
1v
(12) a’.Iieppe, G. Schlichtinq, K. Klager and T.’roeycl, A n i i . , 660 1 (1548). (13) H. H. Inhoffen, J. Heimann-Trosietl, H. hIuxfeldt and 11. Kramer, Chem. Be?., 90, 187 (15.57).
March 20, 1962
SYNTHESIS OF
D( -)-SHIKIMIC
1041
ACID
Pyrolysis of IV resulted not in the formation of the desired methyl 3-0-acetylshikimate but in the production of a mixture of aromatic hydroxy esters and acids. Since protection of the free hydroxyl groups was obviously necessary, the acetonide V was prepared. Base-catalyzed elimination of acetic acid from V would be expected to occur readily if the C-1 hydrogen and the 2-acetoxy group were trans and diaxial to each other. If, on the other hand, a cis relationship existed, elimination should take place more readily on pyrolysis. In point of fact, V failed to undergo elimination or epimerization utilizing base under a variety of conditions, including treatment with potassium carbonate in refluxing benzene or toluene and potassium hydroxide or sodium ethoxide in ethanol. However, on pyrolysis V was converted to dl-methyl 3-0acetylshikimate acetonide (VI) in 95y0 yield. Thus, the stereochemistry of V, and hence of IV and 111, is assigned correctly as shown, in contrast to the interpretation of hlccrindle, Overton and Raphael, who proposed the opposite configuration for the C-1 hydrogen (C-1 carboxy group) on the basis that elimination was found to occur (in unspecified yield) by treatment of V with sodium methoxide. These workers also reported, however, that a t 290°, in the presence of magnesium oxide, V was converted to VI in SO%, yield. The conditions of pyrolysis described herein differ only in that an evacuated tube was employed rather than an acetic acid scavenger. On hydrolysis of the acetonide ester VI, dlshikimic acid was obtained in 65% yield. Resolution was accomplished by means of (-)-a-phenylethylaniine, with which D-( -)-shikimic acid (I) readily forms a sparingly soluble salt, which was identical in all respects with the salt of the naturally occurring compound, Experimental
solution was added t o the flask and the mixture was stirred a t -50" for 12 hours. The initial light yellow color of the mixture soon changed to dark brown, indicating the formation of a n osmium complex. The brown osmium complex was precipitated by adding 400 ml. of anhydrous ether, the mixture filtered, and the residue washed with 50 mi. of anhydrous ether. T h e osmate ester was obtained in a yield of 20.4 g. ( 9 3 5 , ) . T h e osmate ester was cleaved by the method of Baran.14 Pyridine (50 m i . ) was added t o the osmate ester. 4 solution of 19.6 g. (0.18 mole) of sodium bisulfite in 330 ml. of water and 300 ml. of pyridine was added to the pyridine solution of the osmate. On shaking the combined solutions, a dark brown oil separated and the solution became reddishbrown in color. -1fter 1 hour, the solution was extracted with methylene chloride. T h e extract was washed with a small amount of water, dried over anhydrous sodium sulfate, and the solvent removed in V U C I L O . -1 white poivder remained, m.p. 166-167", yield 6.13 g. (56%). A n d . Calcd. for CplHlsOs: C , 19.25; H , 6.22. Found: C , 49.04; H, 6.26. Methyl 2p,5p-Diacetoxy-3a,4a-dihydroxycyclohexane-lacarboxylate Acetonide ( V ) .--h.leth>-l2p,5~-diacetoxy-3a,4adihydroxycJ-clohexane-la-carboxylate( I Y ) (6.13 g., 0.021 mole) was dissolved in 150 ml. of anhydrous acetone. *Inhydrous acetone saturated with dry hydrogen chloride was prepared arid 5 ml. of this was added to the solution along with 5 g. of anhydrous calcium chloride. The reaction was allowed to proceed at room temperature for 36 hours with occasional shaking. At the end of this period, 10 g. of anhydrous sodium carbonate was added. The solvent was removed itz t'acuo. The residue was dissolved in a small volume of chloroform and chromatographed oil F l o r i d . The column was eluted with chloroform and the eluate taken to dryness in 'O'UCUO. Crystallization from petroleum ether gave 3.28 g. (48'3) of ccilorless needles, m . p . 143-144'. A n a l . Calcd. for CljH2208:C, 54.53; H , 6.71. Found: C, 54.52; H, 6.65. dl-Methyl 3-0-Acetylshikimate Acetonide (VI).-The acetonide of methyl 2@,5a-diacetoxy-3a,4a-dih\-dros)-c ~ c l o h e s a n e - l a - c a r b [ ~ x ~ l a(45 t e mg., 0.136 mmole) was placed in a PJ-rex tube (8 mm. X 150 mm.) sealed a t one end. This tube was evacuated to a pressure of 0.007 nim. and the tube completelj- sealed. I t was then placed in an oven a t 285" for 20 minutes after which i t was removed and allowed t o cool. The tube was opened and tlic contents dissolved in 3 ml. of chloroform. This solution was chromatographed using a silicic acid-chloroform column. T h e first band eluted was evaporated to dryness and yielded 35 nig. (95yc) of a clear viscous oil. Methyl 2~,5~-Diacetoxy-3-cyclohexene-la-carboxylate Anal. Calcd. for ClaHlsOe: C, 57.80; H , G . G G . Found: (III).-A solution of 3.87 g. (0.028 mole) of trans,trnns1,4-diacetoxy-1,3-butadieneand 2.0 g. (0.023 mole) of C , 58.20; H , 6.86. Hydrolysis of &Methyl 3-0-Acetylshikimate Acetonide methyl acrylate (Eastman Organic Chemicals stabilized milligrams (0.2 with 111-droquinoue) in 5 ml. of anhydrous xylene was (VI) to &-Shikimic Acid ( I ).-Fifty-four ncated slowly t o reflux temperature. Small aliquots v-ere inniole) of dZ-meth~-l3-acctylshikimate acetonide (I-)in 5 removed from the reaction vessel every 5 hours and the opti- inl. of GO',; acetic acid vas heated a t 70" for 3 hours. -4fter cal extinction coefficient calculated a t 250 nik. After 37 having cooled t o room temperature, the solvent was reh i n m the solution was concentrated in cucuo t o give 5.9 g . moved in C ~ C I L O . T h e residue \vas dissolved in 8 nil. of 0.1 N alcoholic sodium h>-droxide and allowed to stand for 1.5 of a brown oil. This oil showed no niaximum in the 250 in* region. T h e oil was chromatographed 011 a silicic hours a t room tenioerature. T h e free acid was recovered acid colunin ultilizing chloroform a s a n eluant. O n con- by the use of a n IR-120 cation exchaiige columns yield 24 centration of the eluate, 5.8 g. of a light brown oily sub- mg. (65(,), m.p. 193-195'. stance was obtained. Distillation afforded 5.5 g. (9354) .l?znl. Cdlcd. for CiHloOj: c, 48.27; H, 5.79. Found: of methyl 2@,5B-diacetoxy-3 -cycloliexene- la -carboxylate C , 18.0%;H , 5 . 5 2 . which was obtained a s a colorless liquid, b.p. 152-155' Resolution of di-Shikimic Acid.-To 0.0333 g. (0.33 ( 3 I I ~ I I I .7)z, Z 5 1.4680, ~ Xi'$" 223 m M ( e 1,000). nimoiej of dl-shikimic acid was added 0.0392 g. (0.32 ininole) . I n a i . Calcd. for ClZH1606: C, 56.37; H, 6.32. Found: of ( - 'i-a-plien~letliylariiinc. This mixture was dissolved C , 56.24; H,6.29. in hot absolute ethanol and alloived to stand a t room ternMethyl 2p,5@-Diacetoxy-3-a,4a-dihydroxycyclohexane-l~- pcrature overnight. The white needles which formed were carboxylate (IV).-Eleven grams (0.043 mole) of methyl recrystallized using 1: 1 ethanol-ethyl acetate until constant 2/3,5p-diacetox~--3-c~-clohexene-la-carboxylate (111) was dis- rotation was obtained. The salt exhibited a specific rotasolved in 50 ml. of anhydrous tetrahydrofuran and placed tion [aI2*J"-125'. The salt of the naturally occurring in a 1-1. flask. T h e flask was cooled t o the temperature of shikimic acid had a specific rotation [ a ] 2 9 . 0-123'. ~ The a Dry Ice-trichloroethylene mixture. Eleven grams of same procedure was utilized using ( )-a-phen!-lethylaxnine anhydrous pyridine and 11 g. (0.043 mole) of osmium and the enantiomorph obtained, tetroxide were dissolved in 275 ml. of anhydrous tetrahydrofuran preb iously cooled to Drb- Ice tcmperaturc. This 111) J. 8.H a r d n , ./. O i g ( h c i i i . , 26, 2.i7 l l l ) B O ) .
+
~
~~~~