ChemInform Abstract: The Synthesis of Polyoxygenated Natural Products via Fully Synthetic Branched Pyranose Derivatives: Application to the Erythronolide Problem

Abstract

Reiterative cyclocondensation reactions and highly stereoselective functionalization reactions have been coordinated to reach a derivatized version (5b) of the seco acid of 6-deoxyerythronolide B (1). BACKGROUND In 1983 our laboratory began to explore some new possibilities for synthesizing extensively oxygenated natural products with a particular focus on polypropionates and polyols (ref. 1). The polypropionate pattern is readily identified in the backbone functionality of the macrolide aglycones (ref. 2). This pattern is also encountered, though in a less regular fashion, in various ionophores bearing pyranoid and furanoid substructures (ref. 3). The polyol functionality is widely encountered in carbohydrates. While the most common of the polyol arrangements are found in the pentoses and hexoses, we have been particularly concerned with the longer ensembles found in the complex higher order monosaccharides (ref. 4). The “carbohydrate-connection‘‘ in our synthesis of polypropionates is only slightly less obvious than is its involvement in our higher order monosaccharide efforts. Indeed, we treat the polypropionate targets in the context of more general issues in the synthesis of C-alkylated sugars. Our emphasis is on elaborating, by total synthesis, branched pyranose rings, For long chain polypropionate ensembles, the pyranose rings are disconnected at the 0-C1 (anomeric carbon) bond. An aldehyde function, fashioned from C I , becomes the device f o r chain elongation. In the case of the complex monosaccharides again an aldehyde is employed for major extension. In these cases the aldehyde projects from the pyranose or furanose matrix (either with or without an intervening spacer). The strategies for these two synthetic goals are summarized in Fig. 1. In this lecture we will describe our progress in using this type of generalized protocol toward the synthesis of 6-deoxyerythronolide B (1) (ref. 5 ) . As matters transpired, we focused on the tetra-protected seco ester 5b corresponding in its array of relative stereogenic centers to macrolide 1 . The identification of this particular compound as a target was not based on any prior knowledge that it would be an ideal or even workable substrate for lactonization (ref. 6). Our objective was to demonstrate the feasibility of our strategy for dealing with the eleven stereogenic centers contained in this macrolide system. Presumably the synthesis could be modified toward products with other protective arrangements which might be more suitable for macrolactonization. During our efforts it was found that the series of compounds 3-5 could be prepared from the natural product itself. Reduction of 6deoxyerythronolide B (1) with sodium borohydride in the presence of alumina afforded the crystalline dihydro compound 2.