S IALIC ACIDS

A dense glycocalyx composed of glycolipids, glycoproteins, glycophospholipid anchors and proteoglycans covers the surface of all eukaryotic cells. More than 1%

of the genome participates in the generation of the developmentally regulated and tissue-specific glycosylation characteristics of each cell type in humans and other vertebrates. The biosynthesis of these glycan chains mostly occurs in stepwise reactions in compartments of the ER-Golgi pathway, involving specific glycosyltransferases, glycosidases and other glycan-modifying enzymes (Varki, A., 1999; Drickamer, K. and Taylor, M., 2006). During embryogenesis, cancer, injury and inflammation the expression of some of these gene products is altered, which results

in altered glycan patterns. Sialic acids (Sia) are mainly found in higher animals at the outermost position of glycan chains of glycoproteins and glycolipids; they act as important intrinsic factors and are required for a normal development (Schwarzkopf et al., 2002; Varki, A., 2007).

The term sialic acid describes a group of more than 50 nine-carbon monosaccharides that derive from the precursor N-acetyl-neuraminic acid. They are mostly found in higher animals on the outermost position of glycan chains of glycoproteins and glycolipids. The 1-carbon carries a carboxyl group that is responsible for the negative charge of the molecule (Traving, C. and Schauer, R., 1998). A variety of linkages to the underlying sugar chain from the 2-carbon as well as various types of substitutions at the 4, 5, 7, 8, and 9^th carbon position combine to generate their diversity (Varki, A., 1997; Schauer, R., 2000; Angata, T. and Varki, A., 2002). An N-acetyl group at the 5-carbon for example forms N-acetylneuraminic acid.

Hydroxylation at this position results in N-glycolyl-neuraminic acid. Other potential modifications are acetylation, methylation, phosphorylation or sulphatation (Varki, A., 1992). There are enzymes which catalyze the addition of sialic acids to the sugar chains; they are called sialyltransferases. They link sialic acids either to the 3-carbon or the 6-carbon hydroxyl group of galactose (-2,3- or -2,6-linkage) or to N-acetylglucosamine and N-acetylgalactosamine. An -2,8-linkage to other Sias is also possible (Nicholls et al., 2008). During the development and differentiation of a cell the sialic acid expression varies. The most common Sia is N-acetyl-neuraminic acid (Neu5Ac) which is thought to be the biosynthetic precursor for all other members of the family (Varki, A., 1992).

Sialic acids play an important role in many processes of health and illness. They repel each other because of their negative charge and, thus, result for example in the viscosity of mucus which is composed of glycoproteins, called mucins, or prevent thrombocytes from clumping. The negative charge of Sias shields the inner sugars of the oligosaccharide and therefore protects the glycoproteins from degradation. But the most important function of Sias is their role in cellular and molecular recognition.

Sias are employed by the immune system to distinguish between self and non-self-structures. Furthermore, the glycosylation pattern to a large extent determines the

different blood types (Traving, C. and Schauer, R., 1998). However, sialic acids are not only recognized by cells but also by different pathogens. For many viruses, such as influenza viruses, Sias serve as receptor determinants.

That Sias serve as a receptor determinants for influenza A and B viruses was the first discovered „function” of Sias (Klenk, E. and Stoffel, W., 1956). Most influenza viruses that infect and spread among wild and domesticated birds preferentially recognize Sias that are connected via an -2,3-linkage to the underlying glycan chains. This linkage type is most abundant in avian species, whereas most human isolates, on the contrary, prefer -2,6-linkages. It has been assumed for a long time, that infection of humans by an avian influenza virus strain would only be possible after its acquisition of the human-receptor binding preference. This might be achieved by single mutations within the influenza virus genome (antigenic drift) or reassortment within a cell which was infected with both avian- and human-type influenza viruses (antigenic shift). As it was shown for the pig that it expresses both -2,3- and -2,6-linkage types, it was supposed to serve as a „mixing vessel” for the generation of pandemic viruses (Horimoto, T. and Kawaoka, Y., 2005). Nevertheless, after the discovery that the distribution of Sia linkages is not as restricted as supposed, and several reports of direct transmission of avian influenza viruses to humans had emerged, this dogma had to be revised, (Horimoto, T. and Kawaoka, Y., 2005).

Both Sia linkage types, -2,3 and -2,6, were detected on the human respiratory epithelium, though -2,6-linked Sias predominate. Sias connected via an -2,3-linkage were found primarily in the lower respiratory tract, expressed by ciliated cells, while -2,6-linked Sias were mainly detected on the surface of non-ciliated cells (Matrosovich et al., 2004; Nicholls et al., 2007). Nevertheless, an unfavorable Sia linkage distribution might be overcome by a high local concentration of virus (Ha et al., 2002; Nicholls et al., 2008). Moreover, there might also be other important factors for receptor recognition, e.g. the type of linkage, the underlying sugar, as well as the type of carbon modification (Gambaryan et al., 2005; Gambaryan et al., 2008;

Nicholls et al., 2008).

For the detection of Sias in tissue culture, the sugar binding properties of different lectins are utilized. Although recombinant mammalian receptors and antibodies are also applied, most of these lectins are of microbial or plant origin. It is possible to discriminate between -2,3- and -2,6-linked Sia by the use of the agglutinins derived from Sambucus nigra (SNA) and Maackia amurensis (MAA). SNA binds in a highly selective manner to Sias connected via an -2,6-linkage to either galactose or N-acetylgalactosamine (GalNAc) (Shibuya et al., 1987). MAA is used for the identification of -2,3-linkages. There exist two isoforms of MAA: MAA-1 (also called MAL, the „leukoagglutinin”) and MAA-2 (MAH, a „hemagglutinin”). While MAA-2 binds only to one type of sialylated sugar chain with high affinity (Sia2,3Gal1,2(Sia2,6)GalNAc), MAA-1 additionally recognizes sugars terminating in other Sias as well as a glycan wherein the Sia is replaced by a sulfate ester at the 3-position of galactose (Bai et al., 2001). Identification of different glycans by these methods is severely limited, for this purpose, new additional tools for the identification of different Sia types and linkages need to be developed.