The stereocontrolled synthesis of ¡-arabinofuranosides has received little attention since the reports, over 30 years ago, that solvolyses of 2-O-benzylated a-d-arabi-nofuranosyl halides (e.g., 30 or 31) with methanol gave mixtures of glycosides in which the 5 isomer predominates [86,87]. It is of tangential interest to note that kinetic studies were done, and it was proposed that this reaction proceeds through an ion pair SN1, not SN2 mechanism.
More recently, this method has been used to synthesize octyl ¡-d-arabinofu-ranosides from 30 and 31 . As outlined in Figure 16, reaction of these labile chlorides in the presence of octanol without a promoter gave modest yields of glycosides 54 and 55. The reaction of 30 under these conditions gave only the desired product; however, in the case of 31, traces of the a isomer were also detected. Although the yields are modest, the ease by which the starting material can be prepared (two steps from commercially available 2,3,5-tri-O-benzyl arabinofuranose, 56) makes this an attractive route to these products. Unfortunately, while this method can be used for the synthesis of glycosides of simple primary alcohols, attempts to glycosylate secondary carbohydrate alcohols by this method failed, and thus the method appears to be of limited utility for the synthesis of oligosaccharides.
On the other hand, reaction of the pyridyl thioglycoside 34 with alcohol 57 has been shown  to provide good yields of the ¡-linked disaccharide 58 (Fig. 17). No mention was made of the formation of any a-glycoside product in this
glycosylation reaction. This disaccharide product was then used as a synthon for the preparation of a pentasaccharide fragment (62) of the hexasaccharide motif found at the nonreducing terminus of AG and LAM. Coupling of 58 with alcohol 59 followed by deprotection of the silyl group afforded trisaccharide 60. Glycosylation of 60, again with 58, provided the protected pentasaccharide 61, which was subsequently hydrogenated to afford the product. In this synthesis the first glycosylation proceeded stereoselectively to afford the a-linked product, but in the second case, a separable 3:2 a: j mixture was produced.
Unquestionably, the stereocontrolled synthesis of j-d-arabinofuranosides is an area ripe for investigation. Indeed, no general methods for their preparation currently exist. Although new strategies will likely appear, it is probable that many of the methodologies used for j-mannoside synthesis can be applied to this problem. Indeed, recent work [88,89] has shown that Ogawa's [90,91] intermolecular aglycone delivery method for j-mannoside synthesis can be used for the stereoselective synthesis of j-d-fructofuranosides (e.g., 65) (Fig. 18). Given the very close structural similarities between d-arabinofuranose and d-fructofuranose, it is likely that the approach can be used, but this has yet to be demonstrated.
Was this article helpful?