Synthesis from gluconolactone

133 in 86% yield. Oxidation of 133 gave the corresponding keto amide 134, which was treated with formic acid and sodium cyanoborohydride, in a one-pot reaction, to provide the lactam 67 in 58% overall yield from 133. Debenzylation of 67 afforded gluconolactam 10, whereas its redaction with LiAlH4 afforded the tetra-0-benzyl-1-deoxynojirimycin, which was hydrogenated to produce 2. 

---

Proof of Constitution.—The method of synthesis from methyl 6-trityl-ajS-D-glucofuranoside furnished strong evidence that the trimethyl-D-glucose isolated by Smith121 was substituted at positions 2, 3 and 5. This conclusion was confirmed by the observations (a) that the trimethyl-gluconolactone, produced when the sugar was oxidized with bromine... 

Synthesis from o-gluconolactone D-Gluconolactone was converted to 1-deoxymannojirimycin (2) via introduction of an azide group at C-2, with retention of configuration as in (Scheme 9). Reduction of 47 followed by protection of the... 

We did not explore the first of these two approaches too deeply. The synthesis of the epoxide 76, an ideal partner for the alkylation of any trisaccharide amine, was daunting and seemingly difficult but there was available in the literature an excellent route to the enone 77 from tetra-O-benzyl-D-gluconolactone 38 [53]. However, earlier work by Kuzuhara suggested that any reductive amina-tion of the enone 77 would probably proceed in low yield and certainly give a mixture of diastereoisomeric amines [54]. We did prepare the amine 78, via the azide 79, but the amine 78 would not condense with the enone 77 to give the... 

Ramage and co-workers have given full details of their Wittig approach to 2-ulosonic acids (see Vol. 22, p.l61) and have extended this route to the synthesis of 3-deoxy-D-eryt/iro-2-hexulosonic acid (KDG, 30), albeit in low yield in this case.33 A French group have developed two routes to KDG (30) (Scheme 5) involving, respectively, a Wittig synthesis of (28) 4 or P elimination from a gluconolactone derivative to yield (29) 35 the second approach could be modified to make 5- and 6-(2-methyl ethers of (30). 

---