Galantinic Acid 2

¡-C-Glycosides have also been prepared by the Lewis acid catalyzed addition of ketene acetals, derived from a-amino acids, to glycosyl bromides [34]. The acyl protecting group at C2 was critical in this reaction because it directed the 5 incorporation of the amino acid. However, benzoyl protecting groups were trapped by addition to the carbonyl rather than the anomeric carbon. This problem was overcome by using a pivaloyl (OCOf-butyl) protecting groups instead (Fig. 16).

V-Linked ¡-amino acid conjugates of sugars were reported by Kunz et al. [35]. Mannich reaction of bis(O-trimethylsilyl) ketene acetals and V-galactosylimines gave a-branched ¡-amino acids in high diastereomeric ratios (Fig. 17).

1-Amino-1-deoxyglucuronic acids have also been prepared. In one account, 1,2,3,4-tetra-O-acetyl glucuronic acid was treated with iodine at 0°C, followed by the addition of trimethylsilyl azide. The acetates were removed with hydrazine and the azide was reduced with 1,2-ethanedithiol [36]. Other azido sugars have been used in the preparation of sugar amino acids. For example, Fleet and coworkers converted the isopropylidene of d-glucuronolactone into an a-azido lactone, which was subsequently reduced to an a-amino lactone [37]. Removal of the acetal protecting group unmasked the aldehyde, which underwent reductive amination and hydrolysis of the lactone to give a trihydroxypipecolinic acid (Fig. 18). Fleet developed this elegant methodology as a rapid entry into several picolinic acid derivatives, which are naturally occurring l-amino acids with known biological activity. These molecules further illustrate the difficulty in defining the difference between sugar amino acids and a-amino acids. In this case, they are arguably indistinguishable.

C. Syntheses of Neuraminic Acid Derived Amino Acids

In his review on complex carbohydrates, Nathan Sharon stated: "Neuraminic acid is a nine carbon sugar acid, with an amino group in the molecule'' [38]. This simple and obvious declaration was an epiphany for Gervay-Hague, whose prior experience with NeuAc had been limited to the challenges of O-glycosylations. The realization that neuraminic acid is an amino acid presented new possibilities for its utilization in the production of novel materials. Sharon's writing inspired a program in the Gervay-Hague laboratories directed to the synthesis of amino acid equivalents derived from neuraminic acid.

Since j -acetyl neuraminic acid is the most abundant form of the sialic acids, it was important to first establish a method for removing the acyl group. This turned out to be remarkably difficult, since both acid and base hydrolyses led to retro aldol products rather than the desired amine. Borrowing from work published by Roy and Pon [39], the ¡-methyl glycoside of NeuAc was prepared and successfully V-de-acylated using 2 N sodium hydroxide at 100°C for 48 h. Realizing that these conditions would not be suitable for a wide variety of substrates, the investigators sought milder conditions. After much experimentation, it was found that treatment of the amide with ferf-butoxycarbonyl (Boc) anhydride followed by mild hydrolysis with sodium methoxide provided the Boc-protected sugar amino acid in high yield [40]. This general strategy provided a reliable route to several j -protected neuraminic acid analogs, including a- and ¡-O-methyl glycosides and a 2,3-dehydro derivative (Fig. 19).

2-Deoxy analogs of V-protected neuraminic acids were also prepared. The ¡-hydrido derivative was obtained by hydrogenation of j -acetyl-j -Boc-4,7,8,9-tetra-

Figure 16

0 0

Post a comment