analog. This lack of specificity was also seen with the purified Neu5Ac-synthase from Neisseria meningitidis [138]. Since its initial identification in E. coli, the Neu5Ac-synthase gene has also been cloned from numerous microbial organisms including Helicobacter pylori. The availability of these enzymes promises to facilitate the enzymatic synthesis of sialic acid analogs.

In contrast to the eukaryotic CMP-Neu5Ac-synthetases, the bacterial counterparts have not been extensively characterized. Therefore, it would be interesting to see whether the bacterial enzymes exhibit the same broad substrate specificity. Thus far, only 9-N3 and 9-NH3 modified sialic acids have been tested with the E. coli and N. meningitidis CMP-Neu5Ac-synthetase [138,139]. While the 9-N3 analog was similar to Neu5Ac as a substrate for the enzyme, the 9-NHt derivative was a poor substrate. Unfortunately, many of the sialic acid analogs that were active as substrates for the mammalian enzyme have not yet been examined by means of bacterial enzymes. A few C5-modified analogs have been tested, namely, N-propanoyl sialic acid (SiaProp), N-glycolylneuraminic acid (Neu5Gc), and N-carbomethoxyneuraminic acid (Neu5CMe), and these were tolerated by the enzyme. 5-Azidoneuraminic acid was not a substrate for the enzyme, and N-carbobenzyloxyneuraminic acid (Neu5Cbz) coupled at a much slower rate than Neu5Ac [140]. While it would appear that the bacterial enzyme is more restrictive in its substrate specificity, it does show some tolerance for unnatural C5 and C9 substituted sialic acid derivatives.

Sialylated glycoconjugates found in bacteria often mimic oligosaccharides found on the surface of mammalian cells [141-143]. The similarity in structure between the glycoconjugates of mammalian and bacterial cells might reflect a mechanism by which bacterial cells evade the host immune response. Given the similarity of bacterial and mammalian sialoglycoconjugate structures, the bacterial sialyltrans-ferases might have substrate recognition properties similar to those of their mammalian counterparts. Unfortunately, few bacterial sialyltransferases have been cloned or characterized [144]. However, it is known that similar to mammalian sialyltrans-ferases, bacterial sialyltransferases are membrane-associated proteins that are responsible for terminating glycans by transferring a sialic acid residue to a terminal

GlcNAc, GalNAc, or Gal residue. Like their mammalian counterparts, they too are capable of producing a2,3-, a2,6-, or a2,8-linked sialosides. An a2,6-sialyltransfer-ase isolated from the bacterium Photobacterium damsela is capable of sialylating 3'-sialylLacNAc oligosaccharides, generating a rare 3',6'-disialylated structure [145]. Bacterial sialyltransferases are also capable of forming a2,9-linked structures. The a2,8- and a2,9-linked sialic acids are generally found as homopolymers that make up the capsule of pathogenic bacteria like N. meningitidis, E. coli K1, and E. coli Bos 12.

Of the few bacterial sialyltransferases cloned, even fewer have been studied with respect to donor-substrate specificity. CMP-Neu5Ac, CMP-Neu5Gc, CMP-SiaProp, and CMP-9-fluoro-Neu5Ac have been tested as substrates for the N. meningitidis a2,3-sialyltransferase [146-148]. While these substrates were effectively recognized, the small panel does not provide enough information to generalize the specificity of the enzyme. If the enzyme is to be useful in chemoenzymatic synthesis or in studying cellular processes, knowledge of its specificity with a broader panel of substrates would be useful.

0 0

Post a comment