key: cord-0037867-uhi4vd3a
authors: Shimano, Kazunori; Ge, Yuting; Sakaguchi, Kazuhiko; Isoe, Sachihiko
title: Synthesis of both enantiomers of halitunal
date: 1996-03-25
journal: Tetrahedron Lett
DOI: 10.1016/0040-4039(96)00268-7
sha: 03392f2a57880b8d42352902097600682c16fd9f
doc_id: 37867
cord_uid: uhi4vd3a

Both enantiomers of halitunal (1), a novel diterpene aldehyde having an iridoid carbon framework with a heteroaromatic 10π-system, have been synthesized from (+)-genipin (2).

Halitunal (1), a diterpene aldehyde isolated from the marine algae Halimeda tuna, exhibits significant in vitro activity against a virus, namely, mourn coronavirus strain A59. t From the structural point of view, halitunal has an iridoid carbon framework with a 10n-aromatic cyclopentadieno[c]pyran ring system and one asymmetric carbon center at C12, the absolute configuration of which has not yet been assigned. ~ We wish to report herein the first synthesis of optically active halitunal ((R)-I and (S)-I, respectively). Our synthetic plan was the use of (+)-genipin (2) as the chiral starting material, which has the requisite iridoid carbon framework and is abundantly available from the water extract of Gardeniajasminodes Eills. 26 The present study, which targeted the synthesis of both enantiomers of 1, required the introduction of the C12 hydroxyl group with an unambiguous absolute configuration. Therefore, the synthesis of a mixture of allylic alcohols A and a kinetic resolution of the mixture using the Sharpless asymmetric epoxidation method 7 was carried out first (Scheme 1). The presence of the chiral centers in the iridoid skeleton would be useful to certify the diastereomeric purity of the C12 hydroxyl group of the products. Thus, we started the synthesis of A from 2. Reduction of (+)-genipin bis(silyl ether) (3), obtained by silylation of both hydroxyl groups of 2, with diisobutylaluminum hydride (DIBAH) gave allylic alcohol 4 in good yield (Scheme 2). After mesylation of the hydroxyl group of 4, the resulting mesylate 5 was treated with the carbanion of cyano ether 148 to afford 6 as a mixture of diastereomers (1 : 1). This was converted to the enone 7 by the following sequence of reactions: (1) hydrolysis of the ethoxyethyl ether moiety, (2) dehydrocyanation, and (3) selective removal of the silyl ether moiety on the primary position. Reduction of 7 with NaBH 4 in the presence of cerium(III) chloride afforded a diastereomeric mixture of allylic alcohols (Sa and 8b (1 : 1)), which were separated by silica gel column chromatography to give the less polar isomer 8a and the more polar isomer 8b, respectively. 9 It is well-known that a mixture of secondary aUylic alcohols can be kinetically resolved under the Sharpless asymmetric epoxidation conditions where the rate of the formation of (R)-epoxide is much faster than that of (S)-epoxide when (+)-diisopropyl tartrate (DIFr) was employed. 7 Thus, treatment of the mixture of the allylic alcohols 9a and 9b, prepared by silylation of 8, with tert-butyl hydroperoxide (TBHP), (+)-DIPT, and titanium tetraisopropoxide (Ti(O-i-Pr)4), gave a mixture of the unreacted allylic alcohol and the diastereomerically homogeneous epoxy alcohol 10 (Scheme 3). Thus, the absolute configuration at C12 of the recovered allylic alcohol 9a was assigned to be R and that of the epoxide 10 to be S based on the above empirical method 7,~°'H Since the recovered 9a corresponded to the less polar isomer 8a, the absolute configuration at C12 was assigned to be R.

Finally, we examined the conversion of each isomer of 8 into optically pure (R)-I and (S)-I (Scheme 4). Acetylation of 8a followed by removal of the silyl group gave the desired diacetate lla in quantitative yield. Dehydration of lla was carded out by treatment with 1,1'-thiocarbonyldiimidazole in benzene at room temperature to afford 12a. Treatment of 12a with 2,3-dichloro-5,6-dicyano-l,4-benzoquinone (DDQ) in benzene at room temperature underwent dehydrogenation to give the desired (R)-I as a yellow oil. The conversion of lla to (R)-I was achieved in a one-pot process in 43% yield. The 12S enantiomer of 1 was synthesized from 8b in the same manner as (R)4. Synthetic (R)-I and (S)-I were completely identical with natural halitunal in all respects (IH-NMR, t3C-NMR, IR, UV, and high resolution mass spectra).~'12 Unfortunately, the optical rotation of natural halitunal was not described in Koehn's report I and its authentic sample was not available) 3 However, it is obvious that either (R)-I or (S)-I is the natural product.

Thus, we succeeded in the synthesis of both enantiomers of halitunal ((R)-I and (S)-I) from (+)-genipin (2) in 10 steps (5% overall yield). 

92 (s, 1 H), 6.67 (d, 1 H, J = 3.1 Hz), 6.02 (ddd, 1 H

We, at first, measured the specific rotation of (R)-I and (S)-I, though all [Ix]~ values of (R)-I and (S)-I were almost -0 ° (~, = 589, 577, 546, 435, and 365 nm, c = 1, CHC13). The signs of both enantiomers of 1 were indistinguishable, therefore, we measured CD spectra