NeuSAc

Of course, with the drastic simplification inherent in the use of aldehyde 8 as the "real world" equivalent of the hypothetical threose 2, came the postponement of the goal of the synthesis of the natural enantiomer of NeuSAc. The only way to strive for enantiospecificity would involve replacing 7 with another diene equipped with a chiral auxilliary (111. At this exploratory stage, we preferred to accept a downscaling of goals, confident that lessons learned in the synthesis of racemic NeuSAc could be applied to the preparation of enantiomer 1 (synthesis of (-) NeuSAc has been accomplished DeNinno, M. P., Yale University, unpublished data). 

One way in which the Z-a,P-unsaturated carbonyl functionality could be exploited would be via its incorporation into lactone 17. It could be predicted with some confidence that external reagents would attack the bicyclic lactonic system from its convex face. Such an a attack by osmium tetroxide would provide the correct 7,8-erythro diol stereochemistry required to reach NeuSAc. This anticipation turned out to be well founded. 

The uniquely distinguished axial alcohol at C5 was activated by triflation and the resultant triflate was displaced by the action of tetra n-butylammonium azide in benzene at room temperature. Compound 25 was thus available in 86% yield from alcohol 24. Hydrogenolysis ((H24 d(OH)2C) followed by acetylation afforded compound 26 in 96% yield. TTie richly detailed NMR spectrum (490 MHz) of 26 was identical in every respect with that of the L-antipode derived from reaction of NeuSAc, first with acidic methanol and then with benzoyl chloride under the influence of DMAP. The infrared and mass spectra, as well as the chromatographic characteristics of the two materials, were identical. 

Finally, saponification of 26 followed by acid treatment (Dowex resin) afforded racemic NeuSAc, whose NMR spectrum (490 MHz) and chromatographic properties were identical with those of authentic NeuSAc. An efficient and stereoselective total synthesis of Neu5Ac had been accomplished as shown in Figure 6. 

A New Approach to Sialic Acid Glycosides. One of the key steps in our synthesis of NeuSAc was the addition of methanol to 10. The furan presumably stabilizes the cation formed by protonation of the [3-carbon of the glycal. Indeed, as discussed at the outset, the de facto nucleophilic character of the furan was one feature which recommended its use. 

The goal of a total synthesis of a biologically active natural product should never be regarded as fully realized until the relevant enantiomer is obtained in homogeneous form. In the case at hand, the attainment of the goal was a particularly urgent matter, since the synthesis of NeuSAc glycosides was the primary thnist of the effort. 

DANISHEFSKY AND DeNINNO Biologically Active Form of NeuSAc... 

The second finding was a disclosure from the laboratory of Paul Hopkins (2) which provided experimental protocols wherein either the S or R enantiomers of compound 5 could be obtained from ethyl lactate. Thus the combination of the discoveries of Pearson (4) and Hopkins provided the basis for a synthesis of both the naturally occurring (8R) NeuSAc and (7R) KDO. 

This synthesis was accomplished using selenoaldehyde 5R. Again, cyclocondensation with 6 gave ent 7. Using the same steps as were used for the NeuSAc synthesis, ent 7 was converted to ent 13. From there the steps to the naturally occurring (7R) antipode of KDO merely involved retracing steps followed in the synthesis of the racemate. 

N. meningitides and human ST6Gal-I, EC2.4.99.1) were exploited to add sialic acid in a2-3 and a2-6 linkages to the terminal galactose residue of mono-, di- and tri-LacNAc. The fusion of the bacterial ST3 with CMP-Neu5Ac synthetase [83] generated the donor substrate in situ from added NeuSAc and CTP. 

However, two recent papers suggest that DMF is also a good solvent for this type of condensation under appropriate activation conditions of the donor. A simple and efficient method for the preparation of thioglycosides of N-acetyl-neuraminic acid has been developed [30]. This procedure involves the selective in situ S-deacetylation and activation of the 2-S-acetyl NeuSAc (24a) and the displacement of primary bromide of methyl glycosides (28a, 28b). The desired a(2->6) thioglycosides (29a,29b) were obtained in 75 and 82% yield. Condensation of the sodium salt (24b), freshly derived from (24 a) by selective S-deacetylation with sodium methoxide [16], with (28c) in DMF at 45 °C also gave the expected compound (29c) in 76% yield (Scheme 8) [31]. 

Scheme 2.2.5.4 Biosynthesis of N-acetyIneuraminic acid (NeuSAc, 12) in bacteria.

C-C bond formation renders the addition essentially irreversible. Our first preparative screening efforts have shown that the enzyme exhibits a broad substrate tolerance, particularly by accepting variously C5-modified NeuSAc derivatives as substrates with complete stereocontrol [20, 21]. Thus, the enzyme seems to be practically equivalent to the commercial aldolase, except that reactions attain complete conversion without a need to drive an equilibrium by a large excess of substrate, which strongly simplifies product isolation. 

Fig. 2.2.5.4 Synthetic trisaccharides containing KDN and other NeuSAc derivatives prepared by enzymatic glycosyl transfer.

---