Can. J. Chem. 49, 1071-1084 (1971)-Methyl 4-Bromobutyrate

A Synthesis of d,l-Prostaglandin E, Methyl Ester and Related Compounds1

E. S. FERDINANDI' AND G. JUST Department of Chemistry, McGill University, Montreal, Quebec

Received October 21, 1970

A synthesis of d,l-prostaglandin Ez (PGE2) methyl ester is described. Two new acetylenic prostaglandins, methyl lla,15-dihydroxy-9-oxo-prost-5-yn-13-enoate and its Clb epimer, were obtained as intermediates in this synthesis. A solvolytic rearrangement of a bicyclo[2.1.O]hexane to a cyclopentane vinyl system is discussed.

Une synthese de l'ester mkthylique de la d,l-prostaglandine EZ (PGE') est dkrite. Deux nouvelles prostaglandines acktyleniques, le dihydroxy-1 la,15 0x0-9 prostyn-5 en-13-oate demkthyle et son kpimere en CIS, furent obtenus comme intermkdiaires dans cette synthese. Un rkarrangement solvolytique d'un bicyclo[2.1.0] hexane en un vinylcyclopentane est discutk. Canadian Journal of Chemistry, 49, 1070 (1971)

Introduction

There have been a number of syntheses reported of biologically potent prostaglandins (1) and their analogs (2) in the recent literature. We wish to report a synthesis d,l-5-dehydro PGE, and d,l-PGE, methyl esters which follow the general principle published (la) using the modification introduced by Schneider et al. (lb, c). A similar synthesis was described in a recent communication by Schneider (lg). In this paper we wish to describe (Scheme 1): (I) the synthesis of the iodo ester 6b which constitutes the C,-C7 side chain of PGE,; (2) the alkylation of the bicyclic ketone 7 with the iodo ester 6b which gave the keto ester 9; (3) a study of the solvolytic transformation of the bicyclo [3.l .O]hexanone 9 to the acetylenic prostaglandin analog 11 which showed high biological activity; and (4) the final reduction and isomerization of 11 to d,l-PGE, methyl ester 17.

The best synthesis of the seven carbon side- chain 6b was achieved by the condensation of propargyl alcohol with a C, unit, 4-tetrahydropyranyloxybutyl bromide (3b) (Scheme 2).

The bromide 3b was synthesized in four steps starting with y-butyrolactone (1). Treatment of the lactone with saturated aqueous hydrogen bromide at reflux for 15 h gave 4-bromobutyric acid (2a) (4). This was esterified with diazomethane (5), and the resulting ester 2b was reduced with excess lithium aluminum hydride in diethyl ether at -50" to give 4-bromobutanol (3a). Finally, treatment of the bromo alcohol 3a with dihydropyran and a trace of phosphorus oxy-

'Abstracted from the Ph.D. dissertation of E.S.F., McGill University, Montreal, Quebec, 1969.

'Holder of a Canadian Industries Ltd., Fellowship, 1966-1967.

chloride (6) gave 3b. Yields of better than 90% were realized for each step in this sequence.

Reaction of 3b with propargyl alcohol by the method of D'Engenieres et al. (3) gave 7-tetra- hydropyranyloxyhept-2-yn-1-01 (4a) in 5&60% yield. Only moderate yields were realized because of the formation of the two by-products 4b (2.5%) and 4c (25%) which resulted from 0-alkylation of propargyl alcohol and 4b, respectively.

The iodo ester 6b was obtained from 4a by the following sequence. The hydroxy group of 4a was mesylated by first forming the lithium alco- holate with butyl lithium followed by slow addition of mesyl chloride in diethyl ether at 0". The mesylate 5a was then treated with sodium iodide in dry acetone for 1.5 h (7) to give the corresponding iodide 5c. The bromide 5b was similarly prepared by refluxing the mesylate 5a with anhydrous lithium bromide in acetone for 24 h (8). Cleavage of the tetrahydropyranyl ether with dilute methanolic hydrochloric acid gave 7- iodohept-5-yn-1-ol(5d) which was oxidized with excess Jones reagent (17) at 0" to the carboxylic acid 6a in 20% overall yield starting from 4a. Methylation with diazomethane gave the acetylenic iodo ester 6b: i.r. (film) 2240, 1740, 1163 cm-l; 6 3.64 (3H, OCH,) and 3.70 (3H, J = 2 HZ, - CH21); M + 266, M + -I 139.

Alkylation of Exo-6-(1'-hepteny1)-bicyclo- [3.1 .O]hexan-3-one (7)

The reaction conditions for the alkylation of bicyclic ketone 7 with the iodo ester 6 were found to be extremely critical. Best results were obtained when freshly prepared potassium t-butoxide dissolved in tetrahydrofuran (THF) was added slowly to a THF solution of the ketone and iodo ester at 0-5" (9, lc). It was also necessary to have

FERDINAND1 AND JUST: SYNTHESIS OF PROSTAGLANDIN METHYL ESTER

1 2 3 4 a R = H a R = H a R = H; RI = -0THP b R = C H 3 b R = THP b R = J " - - - - o T H P ; R~ = H

c R = RI = -0THP

4a -+ X,--C=C-,-,-,,OR -+ I , - - C = C v C 0 2 R

5 6 a X = 02SCH3; R = THP a R = H b X = B r ; R = T H P b R = C H 3 c X = I ; R = T H P d X = I ; R = H

1072 CANADIAN JOURNAL OF CHEMISTRY. VOL. 49, 1971

the iodo ester in a large excess relative to the ketone and rigorously dried THF in order to ob- tain maximum yields of the monoalkylated product. Alkylation was only achieved when iodide was the leaving group.

Alkylation of the ketone 7 with a three-fold excess of the iodo ester gave four alkylated products which were separated from the unreacted starting material by column chromatography on silica gel, and then from each other by preparative t.1.c. on silica gel (benzene-ether, 9: 1). The major product was the monoalkylated compound 9, obtained as a mixture of cis and trans (P and a ) isomers in 23% yield: i.r. (film) 1740,1165cm-'; n.m.r. (CDCI,) F 5.6-4.6 (2H, m, HC=CH), 3.62 (3H, s, OCH,); M f 330. The other three products were the di- (M' 468), tri- (M' 606), and tetra- (M' 744) alkylated compounds obtained in 8,11, and < 5% yields respectively.

The g.1.c. of the monoalkylated product (6' 3% SE30 on chromosorb W, 200") indicated that there were two compounds present (R, = 2.90 and 4.75 min) in the ratio 3:7. In the PGE, series, the analogous alkylation products ia and b obtained in the ratio 28 :72 (1 c) had retention times of 3.05 and 4.67 min on the same g.1.c. ~ o l u m n . ~

I

a R = H; Rl =-COICH3

b R = / \ / \ / \ C 0 2 C H 3 : RI = H

The minor component (R, = 3.05 min) had been assigned the trans configuration, since it had been converted to PGF,. (1 c). Based on the close paral- lelism of the two sets of reactions, the minor and major components must be the trans (a) and cis (p) isomers of 9. These isomers were carried through the final sequence of reactions without separation, since it has been found (29) that 8-iso- PGE, methyl ester isomerizes to the PGE, methyl ester in 5% methanolic sodium acetate.

The g.1.c. of the dialkylated product (6', 3% SE30 on chromosorb W, 248") showed two peaks (R, = 3.50 and 5.25 min) in the ratio 1 :5. Too little was known to make a structural assignment for the two dialkylated products. However, it has

'Samples of ia and b for g.1.c. analysis were supplied by Dr. J. E. Pike of the Upjohn Company, Kalamazoo, Michigan.

been shown (25) that the second condensation reaction occurs predominantly at the more sub- stituted carbon.

The tri- and tetra-alkylated products were not eluted off the column at temperatures up to 250".

An alternative route to the keto ester 9 in- volved alkylation of the bicyclic ketone 7 with an excess of the iodide 5c which gave the monoalkylated product 8 in 19% yield. The subsequent acid hydrolysis of the tetrahydropyranyl ether, Jones oxidation of the resulting alcohol to the carboxylic acid, and final esterification with diazomethane, gave a complicated mixture of products from which the keto ester 9 could not be isolated.

The Final Steps in the Synthesis of d,l-PGE, Methyl Ester

The crucial step in this prostaglandin synthesis was the solvolytic transformation of the cyclopropyl ring in 9 to the open chain diol or prostaglandin skeleton (11 and 12) (Scheme 3).

The acetylenic keto ester 9 was converted by the method of Swern et al. (26) to the diol10a in 87% yield (i.r. (film) 3435 cm-'; mass spectrum mle 364 (M')). The diol 10a was then treated with an excess of methanesulfonyl chloride in anhydrous pyridine at - 10". The reaction was worked-up in the cold and gave the corresponding dimesylate lob in 78-85% yields (i.r. (film) 1350, 1180 cm-' (SO,)).

The solvolysis of the dimesylate was carried out in acetone-water (2 : 1) at room temperature (22- 25") for 36 h in a nitrogen atmosphere. The reaction was followed by t.1.c. on silica gel (benzene -ethyl acetate 1 :1) and was found to be more than 95% complete after this time. The solvolysis mixture consisted of five components which could be easily separated by preparative t.1.c. on silica gel using ethyl acetate as the solvent system.

The major products (R, 0.62 and 0.48) were the isomeric hydroxy mesylates 13, obtained in 70- 80% yield (i.r. (film) 3475, 1355, 1180 cm-I). A minor product (R, 0.37) was also a hydroxy- mesylate but was never obtained pure for proper characterization. The other two components, R, 0.29 and 0.19, were obtained in 11-16% yield and consisted of 5-dehydro PGE, methyl ester and its 8-isomer 11, and the corresponding 15-epimers 12 (i.r. (film) 3440, 1740 cm-'; n.m.r. (CDCI,) F 5.9-5.2 (HCLCH); mle 346 (M'-H,O)).

Treatment of a mixture of 11 and 12 with 0.5 M

FERDINAND1 AND JUST: SYNTHESIS OF PROSTAGLANDIN METHYL ESTER 1073

TABLE 1. Results of the bioassay of: A, the synthetic 5-dehydro PGE, methyl esters 11 and 12 for smooth muscle stimulating activity using the rat stomach fundus;* B, the synthetic PGEz methyl esters 16 and 17 for smoothmuscle stimulating activity using

gerbil colont

Weight ( x 10-9g) of material required for

Compounds equivalent contraction Potency

A. Natural PGE, 20 1 .O 8a and p-15-epi-5- Dehydro PGE, methyl ester 12 720 0.028 8a and p-5-Dehydro methyl ester 11 192 0.10

Epimerized 11 160 0.13

Mixture of 11 and 12 256 0.078

B. 8a and j3-15-epi PGE, Methyl ester 16 0.21 8a PGE2 Methyl ester 17 3.38

'Bioassay carried out by Dr. L. Wolfe, Montreal Neurological Institute, Montreal. Quebec.

tBioassay carrled out by Dr. J. R. Weeks, The Upjohn Company, Kalamazoo, Michigan.

sodium hydroxide in methanol at 38-48" for 1.5 h gave 5-dehydro-PGB, methyl ester 14 ( E ~ ~ ~

28 300)(27). Prolonged heating or heating at reflux temperatures for short times caused the chromo- phore to disappear. Presumably 14 decomposed under these more vigorous conditions.

The configuration at C-15 of the products 11 and 12 were assigned on the basis of their bioassay for smooth muscle stimulating activity. This assignment was subsequently confirmed when 11 and 12 were independently converted to the PGE, and 8-iso PGE, methyl esters 15 and 16 respectively, which in turn showed the pre-

dicted activity for smooth muscle stimulation (28). The bioassay on rat stomach fundus for smooth muscle stimulating activity of the prostaglandin-type products obtained from the solvolysis showed that the slower moving compound (R, 0.19) had higher activity than the faster moving one (R, 0.29) (Table 1).

Since it is known4 that PG's with the natural a-configuration at C-15 are more active than the ones having the unnatural P-configuration, the slower moving isomer (R, 0.19) was assigned the

4J. R. Weeks, private communication.

1074 CANADIAN JOURNAL OF CHEMISTRY. VOL. 49. 1971

0:-configuration shown in 11 while the faster moving isomer (R, 0.29) was assigned the P-configuration 12. The comparison of t.1.c. R, values of the PGE, methyl ester, synthesized in the subsequent steps, with that of the natural PGE, methyl ester supported this assignment.

Treatment of 11, a mixture of 80: and P isomers, with 5% methanolic sodium acetate for seven days at room temperature (29) gave a product with increased biological activity (Table 1). This result was consistent with the formation of more of the 80: isomer since PG's having the natural, 80: configuration are known to be more active than those having the 8P ~ o n f i ~ u r a t i o n . ~

Attempts were made to improve the yields of the rearranged product by varying the leaving group, the solvolysis condition, and the type of carbonium ion generated. Because of the difficulty in synthesizing the alkylated bicyclic ketone 9 and since the side chain should not greatly affect the yields of rearranged products, the unalkylated ketone 7 was used as a model in these solvolysis studies. By analogy to the results of wiberg and Ashe (10) where the exo isomer was found to give a cleaner reaction and more ring opened product than the endo isomer, only the exo isomer was considered in this study.

It was known that the rearrangement of the cyclopropyl carbinyl system gave, under kinetic control, the cyclopropyl carbinol as the major product (12), but that under thermodynamically controlled conditions, the homoallylic system was favored (1 3). In principle, thermodynamic control would be achieved by treating the glycol 20 (Scheme 4) with mineral acid. However, one competing reaction, a pinacol rearrangement (9), and, in the E-series, the known acid instability of the PG which is dehydrated to the PGA compounds (27), made this route unsuitable. The ditrichloroacetate 21 was, therefore, prepared by treating the diol 20 with excess trichloroacetyl- chloride in pyridine and solvolyzed in trichloroacetic acid. Under these conditions, any regener- ated ditrichloroacetate would be resolvolyzed, giving effectively thermodynamic control.

The solvolysis of the ditrichloroacetate was carried out at various temperatures in trichloroacetic acid, trichloroacetic acid - dioxane 1 : 1, and trichloroacetic acid containing some p-tolu- enesulfonic acid. Pyrolysis of the ditrichloroacetate was also attempted. In order to determine if rearrangement had taken place, the reaction mixture was taken up in ether and washed with 5%

sodium bicarbonate and water to remove all the acid solvent. Some of the resulting material was then treated with 0.5% sodium hydroxide in methanol at 50-60" for 10 min. If rearrangement took place, the resulting P-hydroxy ketone 22 would be converted to the conjugated ketone 29 in the same way the PGE, -, were converted to PGB, -, (27). Thiscompound iseasily identified by its U.V. spectrum (h,,,(MeOH) 268 mp, E 26 000). There was no detectable rearrangement in any of the solvolysis reactions carried out in trichloroacetic acid. The t.1.c. of the products on silica gel (benzene-ether, 7:3) showed only recovered ditrichloroacetate. Pyrolysis at 230" gave a small amount of rearrangement along with considerable charring. At 120-140°, no reaction took place.

Solvolysis of the ditrichloroacetate 21 in refluxing aqueous acetone, analogous to the reaction carried out by Lumb and Whitham (14), gave, after 4 days, the starting ditrichloroacetate (65%), the glycol 20 (30%), and the rearranged diol 22 (2%). Because of the long reaction time and poor yield of the rearranged product 22, this reaction was not further studied.

A similar solvolysis of the dibromide 23, prepared by treating the ketone 7 with pyridinium hydrobromide perbromide (15) in anhydrous chloroform,5 was carried out in aqueous acetone. The solvolysis was carried out at room temperature for 112 h. After separation of the components by t.1.c. (benzene-ether 7:3), the reaction mixture was found to consist of the dibromide 23 (63%), the unrearranged bromohydrin 26a (1 5%) (Mf 289, 291 ; mle 210, Mf-79Br, "Br), and a slow moving fraction (5%) which was still a mixture of four compounds. The components of this slow moving fraction had a low R, value which

I I i i i

'Two products were obtained from this reaction. The desired dibromide 21 (55%) was the major product but could only be separated with much difficulty from a minor product. This minor product was not obtained pure. However bands at 1720 and 1630 cm-' in the i.r. spectrum of the mixture suggested that it was an a,b-un- saturated, five-membered ring ketone such as iii. Such a compound would be formed if, during the bromination reaction in a slightly acidic, aprotic solvent, rearrangement of ii to iii occurred as well.

FERDINAND1 AND JUST: SYNTHESlS O F PROSTAGLANDIN METHYL ESTER

NCO OH

suggested that they were diols. This mixture probably contained some rearranged product, but there was not enough material available for proper characterization of its components.

The bromohydrin 26a was independently synthesized by treating the ketone 7 with N-bromo- succinimide (16). Oxidation of 26a with Jones reagent (17) at 0" gave the bromodiketone 27a in good yield (i.r. (CCI,) 1748, 1700 cm-', n.m.r. (CCI,) 6 4.27(t, lH, J = 6.5 Hz, CHBr)).

When the bromohydrin 26a was treated with 0.5% methanolic sodium hydroxide at 50°, the conjugated hydroxy ketone 29 was obtained in about 30% yield (crude product E,,, 8500). This

compound was probably formed via the epoxide 28 which is known to rearrange to the conjugated ketone under basic conditions (1 1).

Since the generation of a 'normal' carbonium ion gave products of rearrangement in less than 20% yield, it was decided to try to generate a 'hot' carbonium ion (19) (Scheme 4).

Addition of iodine isocyanate (20)+to the olefin 7 gave the iodo isocyanate 24 in good yield: (i.r. (CCI,) 2245 cm- '). The mode of addition follows from subsequent reactions. The isocyanate was hydrolyzed to the corresponding amine in aqueous acetone. The amine was not isolated but treated, in situ, with formic acid and sodium ni-


trite at -4". After 1 h at this temperature and 3.5 h at 25", the only compound isolated was the unrearranged iodohydrin 266 (Mf 336; mle 209, M + -I). If any rearrangement did occur, it was to an extent of less than lo%, and certainly no higher than in the dimesylate solvolysis. Oxida- tion of the iodohydrin 266 with Jones reagent gave the iodo diketone 276 (i.r. (CCI,) 1745,1695 cm-' ; n.m.r. (CCI,), 100 Hz) 6 4.48(t, lH, J = 4.0 Hz, CHI)).

Considerable work has been done to study the cyclopropylcarbinyl solvolysis and to understand the nature of the carbonium ion which was in- volved (21). Although rate enhancement of the solvolysis due to the participation of the cyclopropyl group adjacent to the leaving group has been widely demonstrated (10, 12a-d)6 the exact nature of the carbonium ion is controversial (21, 12e,f, 13a) and the extent of rearrangement in various cyclopropylcarbinyl systems under similar solvolytic conditions is quite unpredictable (10, 14, 22).

The reason for the limited rearrangement in the bicyclo [3.1 .O]hexanone system discussed above was not clear. A number of factors could influence the amount of rearrangement product. The stereochemical arrangement of the leaving groups to the migrating bond (18,30) and the amount of ring strain in the transition state caused by bond breaking and rearranging (12g,h, 22, 23) have been found to greatly influence the type of products obtained in a solvolysis. Recently it was shown (24) in a modified PGE, synthesis that the endo isomer gives about twice as much rearranged product as the exo isomer under the same reaction conditions. In the case of the ditrichloroacetate 21, where no reaction took place at all, the reaction might be very slow. Indeed, it was found that the dibromide solvolysis was much slower than that of the dimesylate but faster than that of the ditrichloroacetate. This order of reactivity of the leaving groups is the expected one (1 8). No extensive data on the reactivity of trichloroacetates, or the solvolyzing ability of trichloroacetic acid, have been reported. Another factor which may limit the amount of

61t was found that the leaving group adjacent to the cyclopropyl ring was the only one solvolyzed in all the reactions that were carried out. Indeed, in the case of the iodohydrin 26b, the iodide could not be solvolyzed when it was refluxed in aqueous acetone for 24 h. Therefore the cyclopropyl ring does participate in the initial solvolysis reaction.

rearrangement is that the neighboring group par- ticipates in the formation of the cyclopropylcarbinyl carbonium ion, minimizing the effect of the cyclopropyl group, although the latter also par- ticipates in the ion formation!

The final step in the synthesis of PGE, methyl ester was the reduction of the acetylenic bond to a cis double bond. Catalytic hydrogenation of the mixture of 5-dehydro PGE, methyl ester and its 8-isomer 11 using Lindlar catalyst poisoned with 1-3% quinoline (32) gave a mixture of PGE, and 8-iso PGE, methyl esters 15 in 63% yield. The corresponding mixture of 15-epi-5-dehydro PGE, methyl ester and its 8-isomer was hydrogenated in the same way to a mixture of 15-epi-PGE, methyl ester and its 8-isomer 16 (Scheme 5).

The mixture 8a- and P-PGE, methyl ester 15 was treated with 3% ethanolic potassium acetate solution for 94 h at 22" (28). The resulting reaction mixture consisted of one major and two minor components. The major product, isolated in 80% yield from the other two by preparative t.1.c. on silica gel (using ethyl acetate as the solvent system), was characterized as 8a-PGE, methyl ester 17 (i.r. (film) 3440, 1740 cm-'; n.m.r. (CDCl,) 6 5.6(m, 2H, trans HC==CH), 5.3 (m, 2H, cis HCeCH) ; mass spectrum mle 348 (Mf -H,O)). The i.r. and mass spectra were identical with the corresponding spectra of the PGE, methyl ester obtained from natural PGE,. The two minor products were identified as PGA, methyl ester 18 (1 8%) (E,, ,9000 ; mass spectrum mle 348(M+)) which was contaminated by a small amount of the second compound, PGB, methyl ester 19 (2%) (h,,, 277 mp).

The 15-epi-PGE, methyl ester 16 was characterized from its mass spectrum. The fragmentation was the same as that obtained for PGE, methyl ester 17. Exact mass measurement of the fragment mle 348 was correct for C,,H,,O, or M+-H,O (calcd., 348.230045; found, 348.- 229694). The fragment mle 190, one of the most abundant fragments in the mass spectrum of PGE, methyl ester (29), was correct for an em pirical formula C1 3H2002 (calcd., 190.135758 ; found, 190.135990). Such an ion probably arises from C2,H3,04 by fragmentation ofthe ester side chain with hydrogen transfer to the ion fragment mle 208, followed by loss of H,O.

The R, values of synthetic PGE, methyl ester obtained from natural PGE, were compared by t.1.c. on silica gel and 10% silver nitrate impreg-

FERDINAND1 AND JUST: SYNTHESIS O F PROSTAGLANDIN METHYL ESTER

CEC-'-"-'~02~~3 -+

HO" HO' HO,' '

l5 HU' H 3 + *2cH3

H a ' HO'

17 18

+ k 2 C H 3

HO'

TABLE 2. The R, values on silica gel

Compound 10% AgN03/MIII* Ethyl acetate

PGE, Methyl ester from natural PGE, 0.31 0.27

PGE, Methyl ester 17 0.31 0.27 15-epi-PGE, Methyl ester 16 0.32 0.35 5-Dehydro-PGE, methyl ester 11 0.39 0.27

'MIII. Ethyl acetate - methanol -water (16:2.5:10).

nated silica gel. The R, values are listed in Table 2. The PGE, methyl ester 17 had the same R, value as that of the PGE, methyl ester derived from natural PGE,, confirming the stereochemical assignment at C-15 made with the 5-dehydro- PGE, methyl esters. The 15-epi-PGE, methyl ester 16 had a higher R, value in both systems.

The bioassay (30) on the isomerized product 17, carried out on the gerbil colon smooth muscle showed that it had very high biological activity,

and in the expected order of potency4 (Table 1). Although PGE, and PGE, activity has not been compared on gerbil colon, it was estimated4 that PGE, would be several-fold more potent. On the rabbit duodenum, the activity of PGE, was about three-fold PGE, and they were equal on guinea- pig ileum (31). Further, in tests on other prostaglandins, it has not been found that esterification with methyl has any great effect on activity in isolated baths.4 The increased activity after


epimerization indicated that isomerization had ture for 3 h. The mixture was then taken up in ether and taken place, and the gel isomer 17 (d,l-pG~, washed twice with 5 % aqueous potassium hydroxide,

methyl ester) had been obtained. followed by usual work-up procedure. The resulting tetra- hvdro~vranvl ether could be distilled at 82-84'12 mm as

Experimental The instruments used in this work to characterize the

compounds were: Perkin-Elmer 257 and 357 i.r. spectrometers, Varian A60 and 100 Hz n.m.r. spectrometers, MS9 high resolution mass spectrometer, Unicam u.v. spectrometer, and F&M 700 gas chromatograph. Ele- mental analyses were carried out by Beller, Microanalyt- ical Laboratory, West Germany and Dr. Franz Pascher Microanalytical Laboratory, Bonn, West Germany.

Unless otherwise mentioned. the term "worked-UD as

long as-the distillation apparatus was acid free, otherwise the tetrahydropyranyl ether would cleave. Purification was also achieved by column chromatography on silica gel using hexane, then hexane-benzene mixtures (2.5 to 20%) as the solvent system. The yield of 36 was better than 95 %: i.r. (CCI,) 1143,1130, 1038, 1028 (C-0-C), 914, 872 cm-' ; n.m.r. (CCI,) 6 4.53 (m, lH , H C ketal), 4.05-3.18 (m, 6H, 2(OCH,), CH,Br), 2.25-1.35 (m, 10H, CHz).

Anal. Calcd. for C9Hl,0zBr: C, 45.59; H, 7.17; Br, 33.75. Found: C, 45.33; H, 7.38; Br, 33.93.

usual" means extracting with a specified solvent, washing 7-Tetrahydropyranyloxy-2-heptyn-1-01 (4a) the combined extracts with water and saturated aqueous A 500 ml, three-necked, round-bottom flask was fitted sodium chloride, drying over anhydrous sodium sulfate with a Claisen adapter on the middle neck, and a stop- or magnesium sulfate, filtering, and finally evaporating cock adapter (nitrogen inlet), and dropping funnel on the the solvent in vacuo. other two necks. A mechanical stirrer was passed through

4-Bromobutyric Acid (2a) Compound 2a was prepared according to the procedure

of Avison et al. (4).

Methyl 4-Bromobutyrate (26) The crude 4-bromobutyric acid (2a) was treated with

ethereal diazomethane (5) at 0". The ether was then removed in oacuo and the product was distilled at 77"/15 mm: i.r. (CCI,) 1778, 1740 cm-' (C=O); n.m.r. (CC14) S 3.59 (s, 3H, 0CH3), 3.38 (t, 2H, J = 6.0 Hz, CH,Br), 2.63-1.96(m, 4H, CH,CH,CO).

Anal. Calcd. on CSH9O2Br: C, 33.15; H, 4.97; Br, 44.19. Found: C, 33.42; H, 5.06; Br, 43.98.

4-Bromobutanol(3a) A slurry of lithium aluminum hydride (3.05 g) in an-

hydrous ether (250 ml) in a 1 1, three-necked flask was cooled to - 60". The bromo ester 26 (23.8 g), dissolved in anhydrous ether (200 ml), was added at such a rate that the temperature did not go above -50". Following the addition, the reaction was kept below -60" for 1.5 h, then slowly allowed to warm to - 10". The reaction was then quenched with 200 ml of 2 N sulfuric acid at 0" and worked-up as usual with ether. The resulting alcohol 3a was obtained in 95 % yield (b.p. 48-49"/3 mm): i.r.(CCI4) 3630, 3300 (OH), 1060 cm-' (C-0); n.m.r. (CC1,) S 3.68(t, 3H, J = 6.5 HZ, CHZO), 3.46(t, 3H, J = 6.0 Hz, CH,Br), 2.25-1.50(m, 4H, CH,CH2), 4.62(s, l H , exchangeable, OH).

Anal. Calcd. for C4H90Br: C, 31.37; H, 5.88; Br, 52.22. Found: C, 31.49; H, 5.83; Br, 51.98.

4-Tetrahydropyranyloxybutyl Bromide (36) Oneequivalent of the bromo alcohol 3awastreated (neat)

with a slight molar excess of 3,4-dihydropyran and three drops of phosphorous oxychloride. A vigorous exothermic reaction started almost immediately after the phosphorous ~xychloride had been added (in one instance, the reaction started without addition of catalyst, probably because the mixture was contaminated with some mineral acid). The reaction temperature was controlled with an ice-water bath. After the exothermic reaction had sub- sided, the reaction mixture was stirred at room tempera-

the middle neck and Dry Ice condenser was placed on the other arm of the Claisen adapter. The system was flushed with nitrogen gas and flame dried. The reaction flask was charged with 0.22 mol lithium amide, the flask and condenser were then cooled to - 80" (isopropyl alcohol - Dry-Ice), and 200 ml ammonia was distilled through the condenser into the flask. I t was not found necessary to redistill the ammonia from sodium metal. After the distillation, the system was again flushed with nitrogen and the dropping funnel was charged with 0.10 mol propargyl alcohol in 30 ml of anhydrous ether. The alcohol was added dropwise with stirring. The coolant from the flask was then removed and the solution was stirred for 2 h. After this time, the dropping funnel was charged with 0.09 mol of the bromide 36 in 20 ml of anhydrous ether under a positive nitrogen pressure. The bromide was added dropwise and then the reaction mixture was stirred for 6 h at -33" (refluxing ammonia). After this time the reaction was quenched by the addition of 0.05 mol of ammonium chloride and the ammonia was allowed to evaporate overnight. The resulting solid mass was then taken up in water and ether. The ether was separated and the aqueous phase was further extracted (six times) with ether. The work-up was then continued in the usual way. The crude reaction product was chromatographed on a silica gel column (30 g silica gel to 1 g mixture). The column was eluted with hexane containing increasing amounts of ethyl acetate (2.5 %-20%).

The compounds that were isolated from the reaction mixture are given below in the order that they were eluted off the column.

(1) 4-Tetrahydropyranyloxybutyl Propargyl Ether (46)

This compound was eluted with 2.5-5% ethyl acetate in hexane and obtained in 2.5-5% yield: i.r. (CCI,) 3310 (=CH), 2120 (C=C, terminal), 1124, 1105, 1079, 1038, 1023 cm-' (C-0-C); n.m.r. (CCl,) 6 4.53(m, lH , H C ketal) 4.09(d, 2H, J = 2.5 Hz, OCH,C=), 3.9-3.1(m, 6H, OCH,), 2.57(t, l H , J = 2.5 Hz, =CH), 1.8-1.4(m, 10H, CH,).

Anal. Calcd. for C1ZH20o3: C, 67.89; H, 9.50. Found: C, 68.10; H, 9.66.

FERDINAND1 AND JUST: SYNTHESIS OF PROSTAGLANDIN METHYL. ESTER 1079

(2) 4-Tetrahydropyranyloxybutyl-7-tetrahydropyranyloxy-2-heptynyl Ether (4c)

Compound 4c was eluted with 5-10% ethyl acetate in hexane and obtained in 2&25% yield: i.r. (CCI,) 2230 ( k c ) , 1145, 1130, 1085, 1075, 1045, 1030 cm-' (GO-C); n.m.r. (CCI,) 6 4.59(m, 2H, H C ketals), 4.09 (t, 2H, J = 2.0 Hz, =CCHzO), 4.&3.2(m, IOH, CHzO), 2.2(m, 2H, C H Z G ) , 1.9-1.3(m, 20H, CH2).

Anal. Calcd. for C21H3605: C, 68.44; H, 9.85. Found: C, 68.04; H, 9.74.

(3) 7-Tetrahydropyranyloxy-2-heptyn-1-01 (4a) This con~pound was eluted off the column with 20%

ethyl acetate in hexane and obtained in 5&60% yield. Distillation of this compound was unsatisfactory because of tetrahydropyranyl ether cleavage: i.r. (CCI,) 3610, 3450 (OH), 2220 ( G C ) , 1140, 1123, 1039, 1025 cm-I (C-0-C); n.m.r. (CC1,) 6 4.60(m, IH, H C ketal), 4.17 (t, lH, J = 2.0 Hz, 0CH2C+), 4.&3.2(m, 4H, CHzO), 2.5-2.1(m, 2H, C H z G ) , 2.&1.4(m, 10H, CHZ), 2.91 (s, lH, exchangeable, OH).

Anal. Calcd. for CL2HZ003: C, 67.89; H, 9.50. Found: C, 68.09; H, 9.63.

7-Tetrahydropyranyloxy-2-heptynyl Mesylate (50) A solution of n-butyl lithium (Alfa Inorganics) in

hexane (66 g of 22.6 % by weight) was transferred under nitrogen in a "glove bag" into a 1 1, three-necked flask. The flask was then fitted with a nitrogen inlet stopcock, mechanical stirrer, dropping funnel, and nitrogen outlet. Under a stream of nitrogen, the solution was cooled to 0' (ice-water bath) and the dropping funnel was charged with 45 g (0.212 mol) of the alcohol 4a in 100 ml of anhydrous ether. The ethereal solution of alcohol was slowly added with stirring over a period of 1 h, then stirred at 0'

ketal), 3.76(t, 2H, J = 2.0 Hz, CHzI), 3.9-3.2(m, 4H, CHzO), 2.30(m, 2H, CHzC=), 1.9-1.4(m, 10H, CH,).

I-Bromo-7-tetrahydropyranyloxy-2-heptyne (5b) The mesylate 5a (1.55 g) was dissolved in anhydrous

acetone under nitrogen. A slight molar excess of anhydrous lithium bromide was added and the solution was heated at 80' for 1 h. After cooling, the solution was concentrated in vacuo, taken up in ether, and worked-up as usual. The bromide 56, obtained in 85 % yield, was purified by column chromatography on alumina 111-IV using hexane as the solvent system; i.r. (CCI,) 2245 ( G C ) , 1142-1025 cm-' (C-0-C); n.m.r. (CCW 6 4.43(m, lH, H C ketal), 3.79(t, 2H, J = 2.0 Hz, B r C H Z G ) , 3.7-3.1 (m, 4H, CH20), 2.21(m, 2H, CHzC=), 1.8-1.3(m, 10H, CH,) . - --

G a l . Calcd. for C,zH1,02Br: C, 52.36; H, 6.91 ; Br, 29.09. Found: C, 52.13; H, 7.10; Br, 28.89.

I-Iodo-2-heptyn-7-01 (5d) The iodide 5c (0.39 mol) was dissolved in 500 ml of

methanol in a 1 1, three-necked flask containing a magnetic stirrer. Under a stream of nitrogen, 1 ml of concentrated hydrochloric acid was added and the solution was stirred at room temperature for 4 h. Most of the methanol was then removed in vacuo at room temperature, and the residue was diluted with ether, and washed with 5% sodium carbonate. The aqueous fraction was extracted with ether in the usual manner. The crude hydroxy iodide 5d was not purified but oxidized directly with Jones reagent: i.r. (CCI,) 3635, 3460 (OH), 2240 ( k c ) , 1063 cm-' (C-0); n.m.r. (CCI,) 6 3.73(t, J = 2.0 Hz, I C H Z G ) , 3.9-3.4(m, CH20), 2.23(m,CHzk), 1.8-1.2(m, CH,), 3.99(s, exchangeable, OH).

for an additional hour to ensure complete salt formation. Mesyl chloride (24.3 g) in 70 ml anhydrous ether was 7-10d0-5-heptyn0ate (6b)

placed into the dropping funnel and slowly added at 00. A The crude h ~ d r o x ~ iodide 5 4 dissolved in 1 1 acetone,

precipitate formed immediately. After addition (about Was placed in a three-necked flask fitted with a me- 1 h) the solution was allowed to stand an 3 h. chanical stirrer, dropping funnel, and thermometer.~he

~h~ reaction was then quenched with ice water and solution was cooled to 5' and stirred while molar excess worked-up as usual in ether. ~h~ of the Jones reagent was added at such a rate that the tempera- mesylate were better than 95 % and were used as such for ture did not go above 10" (the addition took about 2.5 h). conversion to the iodide loc. The mesylate was purified After the addition, isopropyl alcohol was added to destroy by column c~romatograp~y on silica gel and elution with excess chromic acid, and the reaction mixture was diluted 2.5-10% ethyl acetate - hexane. This purified material with water and ether. The aqueous ~ h a s e separated and was used for the alkylation: i.r. (CCI,) 2245 (ex), was extracted five times with ether. The combined ether

1390, 1188 (SOz), 1145-1025 cm-' (C-0-C). fraction was then extracted with 5 % sodium bicarbonate. This basic fraction was reacidified with 6 N hydrochloric

1-Iodo-7-tetrahydropyranyloxy-2-heptyne (5c) acid and back extracted with ether. The ethereal solution Sodium iodide (75 g), which had been heated in the was washed with brine, dried over sodium sulfate, filtered,

oven at 120" for 2 h, was placed in a 2 1, three-necked and concentrated. The crude acid was then again dis- flask fitted with a nitrogen gas inlet and outlet and me- solved in anhydrous ether, cooled to Oo, and treated with chanical stirrer. The sodium iodide was dissolved in 500 an ethereal solution of diazomethane. Concentration of ml anhydrous acetone and the resulting solution was this solution gave a brown oil which was chromatographed cooled to 0' under nitrogen. The crude mesylate 5a on a silica gel column using 10% ethyl acetate in hexane (1 13 g) dissolved in 500 ml anhydrous acetone was added as the solvent system. The pure iodo ester 66 was ob- at once with stirring. The solution became cloudy, and tained as a colorless liquid in 20% overall yield from 4a: finally turned into a thick, light brown slurry. This was i.r. (CC1,) 2240 ( k c ) , 1740 (C=O), 1163 cm-' stirred at &20° for 1.5 h, then diluted with ether, and (C-0-C); n.m.r. (CCI,) 6 3.70(t, J = 2.0 Hz, ICHz- worked-up the usual way. The crude iodide 5c, obtained k), 3.64(s, OCH,), these peaks overlap but integrate in 85-90% yields, was purified by column chromatogra- together to 5H, 2.5-2.1(m, 4H, C H z k ) , 1.83(m, 2H, phy on silica gel using 2.5-10% ethyl acetate in hexane as CH2); mass spectrum: M + 266; mle 235, M+-OCH3; the solvent system: i.r. (CC1,) 2240 ( k c ) , 1155-1032 207, M+-COzCH3; 179, M+-CH2CH2COZCH3; 139, cm-' (C-0-C); n.m.r. (CC1,) 6 4.60(m, lH, H C M+-1; 193, M+-CH2C02CH3.


6-Exo-(I '-heptenyl) -2-a and !3-(6"-Carbomethoxy-2- hexynyl-bicyclo[3.1.O]hexan-3-one (9)

Iodo ester 6b (0.0752 mol) and the bicyclic ketone 77 (0.026 mol) were dissolved in 400 ml anhydrous freshly distilled tetrahydrofuran (THF).

Freshly prepared potassium t-butoxide (0.0395 mol) in 900 ml THF was slowly added to the stirred solution of the ketone 7 and iodide 6b in a dry nitrogen atmosphere over a period of 6 h with the temperature being main- tained between 0-5". After about 30 min of addition, the solution became cloudy and assumed a beige color which did not change throughout the addition although the solution became thicker. This was probably due to potassium iodide precipitating out. After the base had been added, the reaction was quenched with 50 ml of 5% hydrochloric acid. The solution was then concentrated on an aspirator to about 300 ml, diluted with water, and extracted with ethyl acetate (five times, 100 ml). The combined dark organic fraction was then washed with 5% sodium thiosulfate to remove iodine, and with brine, and finally dried over sodium sulfate. The solution was then filtered and concentrated. A brown oil (19.1 g) was obtained.

Attempts to strip off unreacted ester and ketone failed at 100°/5 p. A more efficient vacuum was required but was not available. The material was then chromatographed on a 1 kg silica gel column, eluted with hexane followed by 2.540 % ethyl acetate in hexane. The components of the mixture were eluted in the order given below.

(I) Unreacted 7 and 6b ( 5 % Ethyl Acetate (EtOAc)) These compounds were identified by t.1.c. comparison

with the pure compounds.

(2) Monoalkylated Ketone 9 (5-10 % EtOAc) This ketone was obtained in 23 % yield: i.r. (film) 1740

( L O ) ketone and ester, 1165 cm-' (C-0-C); n.m.r. (CDCI,) 6 5.6-4.6(m, 2H, HC=CH), 3.62(s, 3H, OCH3). Mass spectrum: M+ 330; m/e 299, M+-0CH3; 271, M+- COzCH3; 302, M+-CO; 243, M+-(CHz)2COzCH3 ; 191, M+-CH2C=C(CH2)3CO2CH3; 273, M+-CaHg.

(3) Dialkylated Ketone (1 0-20 % Et OAc) This ketone was obtained in 8 % yield; i.r. (film) 1740

( ~ = 6 ) ketone and ester, 1168 cm-I (C-0-C); n.m.r. (CDCI,) 6 5.6-4.7(m, 2H, HC==CH), 3.66(s, 6H, 0CH3). Mass spectrum: M+ 468; m/e 437, M+-0CH3; 409, M+-COZCH3; 440, M+-CO; 329, M+-CHZCEC (CHZ)3C02CH3; 425, M+-C3H7; 411, M+-CaHg.

(4) Trialkylated Ketone (20-40 % EtOAc) This ketone was isolated in 11 % yield: i.r. (film) 1740

(C=O) ester and ketone, 1168 cm- ' (C-0-C); n.m.r. (CDCI,) 6 5.6-4.6(m, 2H, HC=CH), 3.68(s, 9H, OCH3). Mass spectrum: M+ 606; m/e 575, M+-OCH3; 547, M+- C02CH3; 578, M+-CO; 549, M+-C4Hg; 467, M+- CHzC=C(CHz)sCOzCH3.

(5) Tetraalkylated Ketone (20-40 % EtOAc) This ketone was isolated in 5 % yield: i.r. (film) 1740

(C=O) ester and ketone, 1168 cm-I (C-0-C); n.m.r. (CDCI,) 6 5.6-4.6(m, 2H, H G C H ) , 3.79(s, 12H, 0CH3). Mass spectrum: M+ 744; mle 713, M+-0CH3; 685, M+-C02CH3; 716, M+-CO; 699, M+-C4H9; 605,

7This compound was generously supplied by the Up- john Company, Kalamazoo, Michigan.

M+-CH2C=C(CHZ),CO2CH3. The four alkylated products were purified by preparative t.1.c. on silica gel HF 254 using benzene+ther 9:l as the solvent system. The yields were calculated on the total amount of 12 used in the reaction.

The g.1.c. of the monoalkylated product (6', 3 % SE30 on chromosorb W, 200") showed two peaks (R, 2.9 and 4.75 min) in the ratio 3:7. On the same column the retention times of the analogous compounds ia and b3 were 3.05 and 4.67 min, respectively. Also on the same column (248") the dialkylated ketone showed two peaks (R, 3.50, 5.25) in the ratio 1 :5. The tri- and tetraalkylated products did not come off the column at this temperature.

Alkylation of the Ketone 7 with I-Zodo-7- tetrahydropyranyloxy-2-hexyne (5c)

(a) The procedure was used as described above except that the molar ratio of ketone to iodide was 1 :l. Column chromatography of the product on silica gel (hexane and ethyl acetate solvent system) gave only one alkylated product in 30% yield. Spectral and elemental analysis showed this to be dialkylated: i.r. (CCI,) 1750 (C=O), 1145-1025 cm-' (C-0-C) tetrahydropyranyl group; n.m.r. (CCI,) 6 5.64.6(m, 2H, HC=CH), 4.52 (m, 2H, H C ketals), 4.1-3.l(m, 8H, CH20).

(b) The procedure was the same as described above except that the molar ratio of ketone to iodide was 1 :2. Column chromatography on silica gel using hexane and hexane - ethyl acetate mixtures as eluantsave the monoalkylated ketone 8 in 19% yield: i.r. (CCla) 1775 (C=O); 1138-1025 cm-' (C-0-C) tetrahydropyranyl ether; n.m.r. (CCI4) 6 5.5-4.6(m, 2H, HC=CH), 4.50(m, lH, H C ketal), 4.0-3.0(m, 4H, CHZO). Mass spectrum: m/e 302, M+-dihydropyran (33); 84, C5H80+ (dihydropyran); 274, 302-CO ; 245, 302-C4Hg; 243,302-CjH7O ; 272,302- C H 2 0 ; 191,302-side chain. N o serious attempt was made to isolate polyalkylated materials.

Exo-6-(I ',2'dihydroxyheptyl) -bicyclo[3.l .O l - hexan-3-one (20)

Exo-6-(l'-heptenyl)-bicycl0[3.l.O]hexan-3-one 7 (5.0 mmol) was dissolved in 25 ml ice cold 97-98 % formic acid buffered with 25 mmol sodium carbonate (26) in a 100 ml three-necked flask fitted with a magnetic stirrer. The solution was cooled to O" in an ice water bath and 5 mmol hydrogen peroxide as a 30% solution was added with stirring. The solution was brought to room temperature and stirred for 30 to 40 min under a stream of nitrogen. The formic acid was then stripped off in uacuo and the residue was dissolved in 80 ml of methanol. To this was added with stirring 18 g sodiurn carbonate dissolved in 50 ml water and the suspension was stirred at room temperature for three hours. The mixture was then acidified with 3 N hydrochloric acid, diluted with water, and extracted with methylene chloride in the usual manner. The yield of the diol 20 was 85 %: i.r. (CCI,) 3620,3590 (OH), 3030 (CH, cyclopropyl), 1745 (C=O), 1060 cm-' (C-0).

Exo-6- (1',2'ditrichloroacetoxyl1eptyl) -bicyclo[3.1.0]- hexan-3-one (21)

Anhydrous pyridine (20 ml) was placed into a 50 ml three-necked flask fitted with a magnetic stirrer, gas inlet and outlet, and a thermometer. The pyridine was cooled to 0" in an ice-water bath and trichloroacetyl chloride (20 mmol) was added. The diol20 (4 mmol), dissolved in


(30% water) was added to hydrolyze the isocyanate. The Solvolysis of the Dimesylate (lob) solution was then cooled to -8" and sodium nitrite (4.7 A solution of the dimesylate lob (380 mg) in 20 ml mmol) was added in 1 ml water followed by formic acid acetone-water (2:l) was stirred in a nitrogen atmosphere (4.7 mmol). The solution was stirred at 0" for 1 h and at room temperature for 36 h.8 The mixture was then then 3.5 h at room temperature. The reaction mixture diluted with water and extracted with methylene chloride. was then diluted with water and extracted with ether in The methylene chloride extract was dried over anhydrous the usual way. Purification by preparative t.1.c. (benzene- sodium sulfate, filtered, and concentrated. The reaction ether 7:3) gave the iodohydrin 27b in approximately 60% products were separated by preparative t.1.c. on silica gel yield: i.r. (CCI,) 3545, 3520, 3460 (OH), 3035 (CH, using ethyl acetate as the solvent system. Thecomponents cyclopropyl), 1743 (C=O), 1137 (C-C-C, ketone), of the mixture were detected on the t.1.c. plates in the 1060 cm-' (C-0); n.m.r. (CCI,) 100 MHz 6 4.19(m, lH, following manner: silica gel H F 254 (E. Merck) was used CH-0), 3.22(m, lH, CHI), 0.59(m, CH, cyclopropyl). which enabled the detection of the hydroxy mesylate bands Mass spectrum: M + 336; m/e 318, M+-H20); 308, M+- on the plates by means of short wavelength U.V. light. The CO; 209, M+-I; 191, 318-1; 181, 308-1. prostaglandin bands could be detected only by spraying

Exo~~-(2~-~odo-~~-oxohep~any~)~b~cyc~o[3~~~0]hexan-3- with 10% P ~ ~ ~ P ~ ~ ~ ~ ~ Y ~ ~ ~ ~ acid in ethanol, a destruc-

one (27bl tive process. In order to recover the prostaglandins, a small part of the mixture was sacrificed for detection

The jodoh~drin 26b (60 mg), obtained from the hy- purposes and applied on a 4 x 10 cm plate. The re- drolysis and deamination of the iodoisocyanate 24, was mainder was spread on 20 20 cm plates and separated oxidized with ones reagent by the same procedure as by measuring the R, obtained from the small plate. This described for the oxidation of the bromohydrin (26a). made the separation difficult since the bands did not The resulting diketoiodide 27b was obtained in 80% travel uniformly. yield: i.r. (CCI4) 3030 (CH, c ~ c l o ~ r o ~ ~ l ) , 1745 (C=O, The silica gel was extracted a few times with methanol. ring 1695 cm-l (C=O, a cyclO- The methanol was then evaporated and the residue was propyl group); n.m.r- (CC14) loo MHz 4.48(t3 lH9 = taken up in methylene chloride, dried over sodium sulfate, 4.0 Hz, CHI adjacent to CH2). and filtered through celite. The recovery by this procedure

was better than 95 %. Exo-6-(1',2'-dihydroxyheptyl)-2a and 0-(6"- The components of the reaction mixture separated on

Carbomethoxy-2"-hexynyl) -bicyclo[3.l.O]- the t.1.c. plates are listed below. hexan-3-one (IOU)

The bicyclic keto ester 9 was treated with performic (I) Ex0-6-f1'-h~drox~-2'-mes~lox~he~t~1)-2-

acid in buffered formic (26) in the same way as the (6'-carbomethox~-2''-hex~n~l)-bic~clof3.1.01-

unalkylated ketone 7. The diol 1Oa was obtained in 87 % hexan-3-0ne (I3)

yield: i.r. (film) 3440 (OH); 3035 (CH, cyclopropyl), 1740 This compound was obtained in 7c80% yield from

ketone and ester), 165 (C-O-C, ester) 1065 the starting dimesylate lob. Three hydroxymes~lates were

cm- 1 (C-0, alcohol); n.m.r. (CDCI,) 6 3.69(s, 3H, obtained (R1 0.62, 0.48, 0.37). The latter (R1 0.37) was

0CH3), 3.7-2.7(m, 4H, CH-O and OH). Mass spec- obtained only in small amounts and could not be ob-

trum: M + 364; mle 333, M+-OCH3; 305, M+-C02CH3; tained pure for proper characterization. The first two

225, M+-CH2CkC(CH2),C02CH3 ; 263, M +-C5Hl (RI 0.62, 0.48) had identical i.r. and n.m.r. spectra. Both

OH; 293, M+-C5Hll; 289, M+-C4H8; 346, M+-H20; gave the same prostaglandi~! mixture on recyclizing,

315, 346-OCH,. therefore no stereochemical assignment could be made: i.r (film) 3475 (OH), 3030 (CH, cyclopropyl), 1740

Exo-6-(I',2'-dimesyloxyheptyl)-2a and 0-(6"- (C==O, ketone and ester), 1355, 1180 (SO2), 1070 cm-' Carbomethoxy-2"-hexynyl) -bicyclo[3.1.0]- (C-0); n.m.r. (CDCI,) 100 MHz 6 4.72(m, lH, CH, hexan-3-one (lob) mesylate), 3.68(s, 3H, OCH,), 3.34(m, lH, CH, alcohol),

A 50 ml round bottom flask was fitted with a stopcock 3.09, 3.07(s, 3H, SOZCH,), 0.64(m, CH, cyclopropyl). adapter, thermometer, gas outlet, and a magnetic stirrer. (2) 8a and 8fl-5-Dehydro-15-epi-PGE2 Methyl Anhydrous pyridine (10 ml) was added to the flask and it Ester 12 (RI 0.29) was cooled to - 15" in an ice-salt bath. A molar excess of This was obtained in 5-8 % yield from the starting methanesulfonyl chloride (3.26 mmol) was added to the dimesylate lob: i.r. (film) 3440 (OH). 1740 (C---O, ester stirred solution under nitrogen. After the temperature and ketone), 1170 (C-0-C, ester), 1070 cln-' (C-0); was again - 15", the diol 10a (1.20 mmol) in 5 ml pyridine n.m.r. (CDCI,) 100 MHz 6 5.9-5.2(m, 2H, HC=CH), was added. The reaction was stirred in the cold for 3 h 4.l(m, 2H, CH-0), 3.66(s, 3H, OCH,). Mass spectrum: and then kept at - 10" overnight. The cold reaction mix- mle 346, M+-H20; 328, M+-2H20; 315, M+-H20, ture was then poured onto crushed ice, saturated with OCH,; 297, M+-2H20, OCH,; 317, M+-H20, HCO; sodium chloride, and extracted with ethyl acetate in the 293, M+-C5Hl1; 285, M+-2H20, C3H7; 275, M+-H20, usual way. The ethyl acetate was evaporated in vacuo at C5Hl,; 207, M+-H20, C8Hl1O2 (ester side chain). room temperature and traces of pyridine were removed (3) aa and 8 ~ - 5 - ~ ~ h ~ , . ~ p ~ ~ ~ ~ h ~ l E~~~~ 11 on a vacuum pump. The yield of the dimesylate 10a was (Rr 0.19) 78-85%: i.r. (film) 3030 (CH, ~ Y ~ ~ o P ~ o P Y ~ ) , 1740 (CEO, This was obtained in 5-8% yield from the starting

and ester), 1350t 1180 cm-l n.m.r. dimesylate lob: i.r. (film) 3400 (OH), 1740 (C=O ketone, (CDCI,) 6 4.83(m, lH, C,H), 4.30(m, lH, C6H), 3.65 6 , 3H, OCH,), four singlets: 189.8, 189.1, 188.0, 187.2 another run lasting 24 h, similar results were ob- Hz (6H, S02CH3). tained.

FERDINAND1 AND JUST: SYNTHESIS

ester), 1163 (C-0-C, ester), 1080 cm-' (C-0); n.m.r. (CDCI,) 100 MHz 6 5.9-5.2(m, 2H, HC=CH), 4.1 (m, 2H, CH-O), 3.68(s, 3H, OCH,). Mass spectrum: mle 346, M+-HZO. The fragmentation was identical with that of the 15-epi-5-dehydro PGE2 methyl ester 12.

Anal. Calcd. for CZ1H3205: C, 69.12; H, 8.79. Found: 69.02; H, 8.83.

PGE, and 8-iso PGE, Methyl Ester 15 5-Dehydro PGEz and 8-iso PGEz methyl ester 11

(10.3 mg) was placed in a sample cup with a small amount of ethyl acetate. Lindlar catalyst (32) (75 mg) was placed in a micro-hydrogenation flask and connected to a low pressure hydrogenation apparatus (35). Ethyl acetate (3 ml), containing 1 mg quinoline? was added to the flask, and the mixture was hydrogenated for 4 h. The solution was filtered through celite and the solvent was evaporated in vacuo. The crude products were chromatographed on silica gel (t.1.c.) using ethyl acetate solvent system. The PG band was detected and extracted as was previously described. The isomeric mixture of 15 was obtained in 63 % yield: n.m.r. (CDCI,) 100 MHz 6 5.62 (m, 2H, trans HC==CH), 5.38(m, 2H, cis HC=CH), 4.1 (m, 2H, CH-0), 3.70(s, 3H, OCH3).

Mass spectrum: mle 348, M+-HzO; 330, M+-2H20; 317, M+-HzO, OCH3; 316, M+-3Hz0, CH3OH; 299, M+-2Hz0, OCH3; 298, M+-2Hz0, CH30H; 299, M+- CsHll ; 277, M+-HzO, CsH1l; 208, M+-Hz0, CaHlzOz (ester side chain with hydrogen transfer); 190, M+-2HZ0, CaHlzOz.

15-Epi-PGEz and 15-Epi-8-iso PGEz Methyl Ester 16 The hydrogenation on 15-epi-5-dehydro PGE2 and 15-

epi-8-iso-5-dehydro PGE, methyl ester 12 was carried out in the same way as described for 11. Because of lack of material, an n.m.r. spectrum could not be taken. Mass spectrum: mle 348, M+-H20. Mass measurement found: 348.229694; calcd. for C2 1H3204 : 348.230045 ; other possible empirical formulae for this mass are: C16H3ZN206, C19H30N303, and C24H30N0. These are ruled out because 16 could not contain any nitrogen. Mass measurement of mle 190: 190.135990; calcd. for C13H180, 190.135758; another possible empirical formula for this mass is CllH16N3 which is again ruled out because 16 contains no nitrogen.

Anal. Calcd. for CZlH3,O5: C, 68.85; H, 9.23. Found: C, 70.12; H, 9.94.

Epimerization of the Isomeric Mixture of 15 The PGEz and 8-iso PGE, methyl ester mixture 15

(6.3 g) was dissolved in 3 ml of a 3 % ethanolic potassium acetate solution and stirred at room temperature for 94 h. The reaction mixture was then diluted with water and extracted with pure methylene chloride. The organic layer was dried over anhydrous magnesium sulfate, filtered, and concentrated in vacuo. Preparative t.1.c. separation of the crude mixture on silica gel, using ethyl acetate as the solvent system, gave two compounds:

(I) PGAz methyl ester18 made up 18% (0.7 mg) of the reaction mixture: U.V. (MeOH) 217 mp, E 9000; mass spectrum: M+ 348, mle 330, M+-H20; 317, M+-OCH,;

gMore than 3 % quinoline, by weight, to the amount of catalyst makes the reaction very slow. Best results were obtained using 1-2 % quinoline.

OF PROSTAGLANDIN METHYL ESTER 1083

299, 330-OCH3; 298, 330-CH30H; 287, 330-C3H7; 190, 330-C7HloCOzCH3 + H-transfer); 207, M+-C7H,,CO2- CH3; 141, [C7HloC02CH3]+. The PGA2 methyl ester was contaminated with about 2% PCB2 methyl ester 19: u.v. (MeOH) 277 mp, E 2200.

(2) PGE2 methyl ester 17 made up 80% (2.9 mg) of the reaction mixture: 100 MHz n.m.r. (CDCI,) using 109 scans which were analyzed with a computer of average transients (c.a.t.): 6 5.6(m, 2H, trans HC==CH), 5.3(m, 2H, cis HC=CH),4.l(m, 2H, CH-0) 3.6(s, 3H, 0CH3). Accurate chemical shifts were not obtained because of drifting in the spectrometer. Mass spectrum: mle 348, M+-HzO; the spectrum was identical to the PGEz methyl ester spectrum taken before epimerization. Because of the small amount of material, a good elemental analysis could not be obtained. However, from the spectral data and bioassay results, there was no doubt that this compound was d,l-PGEZ methyl ester.

We wish to thank the Upjohn Company, Kalamazoo, Michigan for generous samples of 6-exo-(lf-heptenyl)- bicyclo[3.1.0]hexan-3-one (7) and comparison samples of a- and p-isomers of 6-exo-(1'-hepteny1)-2-(6"-carbo- methoxyhexy1)-bicyclo[3.1.O]hexan-3-one (iiia and b) and natural PGE2.

We wish to thank Dr. L. Wolfe of the Montreal Neurological Institute and Dr. J. R. Weeks of the Upjohn Company for carrying out the bioassays.

Financial assistance by the National Research Council of Canada and the Upjohn Co. is gratefully acknowl- edged.

I. (a) G. JUST and CH. SIMONOVITCH. Tetrahedron Lett. 2093 (1967); (b) W. P. SCHNEIDER, U. AXEN, F. H. LINCOLN, J. E. PIKE, and J. L. THOMPSON. J. Amer. Chem. Soc. 90, 5095 (1968); (c) G. JUST, CH. SIMONOVITCH, F. H. LINCOLN, W. P. SCHNEIDER, U. AXEN, J. B. SPERO, and J. E.PIKE. J. Amer. Chem. Soc. 91,5364 (1969); (d) E. J. COREY, N. H. ANDERSON, R. M. CARLSON, J. PAUST, E. VEDEJS, I. VLATTAS, and R. E. K. WINTER. J. Amer. Chem. Soc. 90, 3245 (1968); (e) E. J. COREY, I. VLATTAS, N. H. ANDERSON, and K. HARDING. J. Amer. Chem. Soc. 90, 3247 (1968); Cf) E. J. COREY, I. VLATTAS, and K. HARDING. J. Amer. Chem. Soc. 91, 535 (1969); (g) W. P. SCHNEIDER. Chem. Commun. 304 (1969); (h) E. J. COREY, N. M. WEINSHENKER, T. K. SCHAAT, and W. HUBER. J. Amer. Chem. Soc.91,5675 (1969);(i)E. J. COREY, T. K. SCHAAF, W. HUBER, U. KOELLIKER, and N. M. WEINSHENKER. J. Amer. Chem. Soc. 92, 397 (1970); ( j ) E. J. COREY, Z. ARNOLD, and J. HUTTON. Tetrahedron Lett. 307 (1970); (k) E. J. COREY and R. NOYORI. Tetrahedron Lett. 31 1 (1970); (1) U. AXEN, J. L. THOMPSON, and J. E. PIKE. Chem. Commun. ti02 f 1 9701.

\ - - . -,. 2. (a)J. F. BAGLI, T. BOGRI, R. DEHENGHI, and K.

WIESNER. Tetrahedron Lett. 465 (1966); (b) J. F. BAGLI and T. BOGRI. Tetrahedron Lett. 5 (1967); (c) P. F. BEAL, 111, J. C. BABCOCK, and F. H. LINCOLN. J. Amer. Chem. Soc. 88, 3131 (1966); (d) E. HARDEGGER, H. P. SCHENK, and E. BROGER. Helv. Chim. Acta, 50, 2501 (1967); (e) J. FRIED, S. HEIM, S. J. ETHEREDGE, P. SUNDER-PLASSMANN, T. S. SANTAHANA KRISHANAN, J. I. HIMIZER, and C. H. LIN. Chem. Commun. 634 (1968); (f) J. F. BAGLI and T. BOGRI. TetrahedronLett. 1639 (1969).

1084 CANADIAN JOURNAL OF

3. M. D. D'ENGENIERES, M. MIOCQUE, and J.-A. GAUTIER. Bull. Soc. Chirn. Fr. 1370, 2477 (1964).

4 . A. W. D. AVISON and A. L. MORRISON. J. Chern. SOC. 1473 (1950).

5. A. I. VOGEL. Practical organic chemistry. 3rd Edi- tion. Longrnans.

6. L. CROMBIE and A. J. JACKLIN. J. Chem. Soc. 1622 (1957).

7 . D. ROSENTHAL. P. GRABOWICH. E. M. F. SABO. and J. FRIED. J. ~ r n e r . Chem. SOC.' 85, 3971 (1963):

8. P. R. BIRD and J. S. CHADHA. Tetrahedron Lett. 4541 (1966). . - . - , - - - - , .

9. CH. SIMONOVITCH. Ph.D. Thesis, McGill Univer- sity, Montreal, Quebec, 1967.

10. (a) K. B. WIBERG and A. J. ASHE, 111. Tetrahedron Lett. 1553 (1965); ( b ) K. B. WIBERG and A. J. ASHE, 111. J. Amer. Chem. Soc. 90, 63 (1968).

11. K. G. HOLDEN, B. HWANG, K. R. WILLIAMS. J. WEINSTOCK, M. HARMAN, and J. A. WEISBACH. Tetrahedron Lett. 1569 (1968).

12. (a) D. D. ROBERTS. J. Org. Chem. 30, 23 (1965); (b) W. D. CLOSSEN and G. T. KWAITKOWSKI. Tetrahedron, 21, 2779 (1965); (c) E. F. Cox, M. C. CASERIO, M. S. SILVER, and J. D. ROBERTS. J. Arner. Chem. Soc. 83, 2719 (1961); ( d ) H. HART and J. M. SANDRI. J. Arner. Chem. Soc. 81, 320 (1959); J. Amer. Chern. Soc. 84, 2462 (1962); (e) K. L. SERVICE and J. D. ROBERTS. J. Arner. Chern. Soc. 86,3773 (1964); (f) M. H A N A C K ~ ~ ~ H.-J. SCHNEIDER. Tetrahedron, 20, 1863 (1964); ( g ) M. HANACK and H.-J. SCHNEIDER. Angew. Chern. Int. Ed. 3, 698 (1964); (h) M. HANACK and H.-J. SCHNEIDER. Ann. 686, 8 (1965).

13. (a) M. HANACK and H.-J. SCHNEIDER. Angew. Chern. Int. Ed. 6 , 666 (1967); (b) H. M. WALBORSKY and L. PLONSHER. J. Arner. Chern. Soc. 83, 2138 (1961).

14. J. T. LUMB and G. H. WHITHAM. J. Chem. Soc. (C), 216 (1967).

\ - ~ ,~ 15. L. %. FIESER and M. FIESER. Reagents for organic

synthesis. John Wiley and Sons, Inc., New York, N.Y., 1967. p. 967.

16. C. 0. Guss and R. ROSENTHAL. J. Amer. Chern. SOC. 77, 2549 (1955).

CHEMISTRY. VOL. 49, 1971

17. K. BOWDEN, I. M. HEILBRON, E. R. H. JONES, and B. C. L. WEEDON. J. Chem. Soc. 39 (1946).

18. E. S. GOULD. Mechanism and structure in organic chemistry. Holt, Rinehart, and Winston, New York, N. Y. 1959. p. 261.

19. R. H. MAZUR, W. N. WHITE, D. A. SEMENOW, C. C. LEE, M. S. SILVER, and J. D. ROBERTS. J. Arner. Chem. Soc. 81, 4390 (1959).

20. A. HASSNER and C. HEATHCOCK. J. Org. Chem. 30, 1748 (1965).

21. P. V. R. SCHLEYER and G. W. VAN DINE. J. Amer. Chern. Soc. 88,2321 (1966) and references therein.

22. S. WINSTEIN and M. KOSOWER. J. Amer. Chem. Soc. 81, 4399 (1959).

23. K. B. WIBERG and J. G. PFEIFFER. J. Amer. Chem. SOC. 90, 5324 (1968).

24. U. AXEN, F. H. LINCOLN, and J. L. THOMPSON. Chern. Cornmun. 303 (1969).

25. E. J. COREY, M. OHNO, R. B. MITRA, and P. A. VATAKENCHERRY. J. Arner. Chem. Soc. 86, 478 (1964).

26. D. SWERN, G. N. BILLEN, T. W. FINDLEY, and J. T. SCANLAN. J. Arner. Chern. Soc. 67, 1786 (1945).

27. J. W. HINMAN. Biol. Sci. 17, 779 (1967). 28. E. G . DANIELS, W. C. KRUGER, F. P. KUPIECKI, J. E.

PIKE, and W. P. SCHNEIDER. J. Amer. Chern. Soc. 90, 5894 (1968).

29. S. BERGSTROM, F. DRESSLER, R. RYHAGE, B. SAMU- ELSSON, and J. SJOVALL. Arkiv. Kemi, 19,563 (1962).

30. J. R. WEEKS, J. R. SCHULTZ, and W. E. BROWN. J. Appl. Physiol. 25, 783 (1968).

31. J. R. WEEKS, N. C. SEKHAR, and D. W. DUCHARME. J. Pharrn. Pharrnac. 21, 103 (1969).

32. (a) H. LINDLAR. Helv. Chirn. Acta, 35, 446 (1952); (b) H . LINDLAR and R. DUBUIS. Org. Syn. 46, 89 (1966).

33. H. BUDZIKIEWICZ, C. DJERASSI, and D. H. WILLIAMS. Mass spectrometry of organic compounds. Holden- Day, Inc., San Francisco, California, 1967. p. 479.

34. R. G. NEVILLE and J. J. MCGEE. Inorg. Syn., VIII, 23. 35. C. L. OGG and F. J. COOPER. Anal. Chem. 21, 1400

(1949).

Documents

Can. J. Chem. 49, 1071-1084 (1971)-Methyl 4-Bromobutyrate