Chiral syntheses of the antibiotics anisomycin and pentenomycin from carbohydrates

Syntheses of the natural, optically active forms of the anti— biotics anisomycin and pentenomycin are described using simple glucose derivatives as chiral starting materials. The key step in the synthesis of anisomycin is the formation of a pyrrolidine ring embodying all three of the desired asymmetric centers via intramolecular nucleophilic displace— ment of a 3—0—tosyl function in an appropriately substituted 6—amino—6— deoxy——L—talofuranose derivative. Further elaboration involves the liberatin of an aldehyde group by periodate oxidation of the l,2—diol, Grignard coupling and removal of extraneous functions and blocking groups by several reductive steps. The syntheses of pentenomycin, and its simpler analog 6—deoxypentenomycin, proceed !i:: conversion of 3—benzyloxymethyl and 3—methyl derivatives of 1,2:5 ,6—di—O—isopropylidene—a—D—allofuranose, respectively, into the respective 3—substituted methyT 2 ,3—cycloalkylidene—5—deoxy—D—ythro—pent— 4—enofuranosides. Acidic hydrolysis of the latter gives the related 4—ketoaldehydes that undergo intramolecular aldol condensations to form the appropriately substituted 2—cyclopenten—l—ones that can be converted into the target compounds. It is axiomatic in organic chemistry that the synthesis of an optically active product must make use of chiral substances, either as resolving agents or as synthetic intermediates. From the point of view of overall efficiency, the use of an appropriate chiral starting material is of particular attraction. This approach has found extensive use over the years, and syntheses of a great variety of optically active natural products have been described starting from readily available chiral substances such as amino acids and simple terpenes (Ref. 1). In spite of the fact that carbohydrates constitute one of Nature's richest sources of chirality, only in recent years have these substances found wide use as starting materials for asymmetric syntheses. This stems, to a considerable degree, from a long standing aversion on the part of many organic chemists to work with carbohydrates, which were considered to be mainly intractable syrups. With the advent of improved and more selective protecting groups and, in particular, the extensive application of contemporary instrumental analysis, carbohydrate chemistry is now enjoying a new image. As synthetic intermediates, carbohydrates are now becoming recognized as readily available, polyfunctional molecules containing several sites of predictably controllable asymmetry. In recent years this has led to the use of carbohydrate derivatives as intermediates in the synthesis of a number of optically active natural products of widely divergent structures. By way of example one can mention notable syntheses in the prostaglandin (Ref. 2a) and thromboxane (Ref s. 2b—e) series, of biotin (Ref. 3), of the antifungal agents avenaciolide (Ref. 4) and isoavenaciolide (Ref. 5) and the imaginative approach to erythronolide devised by Hanessian and Rancourt (Ref. 6). In the present paper we further extend this general approach through a description of the syntheses of two optically active antibiotics, anisomycin (1) and pentenomycin (2), starting from D—glucose. Anisomycin (1) was isolated from two strains of Streptomyces by Sobin and Tanner of Chas. Pfizer and Co. in 1954 (Ref. 7). Its gross structure was initially deduced by Beereboom et al. (Ref. 8), but the relative and absolute stereochemistry were subsequently modified on the Note a. Contribution No. 137 from the Institute of Molecular Biology. Note b. Syntex Postdoctoral Fellows. Note c. On sabbatical leave from Queen Elizabeth College, University of London. 1364 J. P. H. VERHEYDEN et al.