Chapter  14:  Sialic Acids

HISTORICAL BACKGROUND

FIGURE 14.1

Two common “primary” sialic acids. (more...) Download Teaching Slide

FIGURE 14.1

.

Two common “primary” sialic acids. Shown are 5-acetamido-2-keto-3,5-dideoxy-D-glycero-D-galactonononic acid (N-acetylneuraminic acid, Neu5Ac; a) and 2-keto-3-deoxy-D-glycero-D-galactonononic acid (2-keto-3-deoxynononic acid, Kdn; b). The only difference is the substitution at the C-5 position. All other sialic acids are apparently metabolically derived from these two, with the exception of some bacterial molecules such as legionaminic acid (not shown; see text for discussion). Neu5Ac is more common than Kdn in most vertebrate cell types. The bonds of the anomeric center (C-2) are drawn to indicate mutarotation between α- and β-anomeric forms in solution. Glycosidically bound sialic acids in naturally occurring glycans are in the α form, and free sialic acids in solution are mainly in the β form.

Figure 14.1

Figure 14.1

DIVERSITY IN STRUCTURE AND LINKAGE

FIGURE 14.2

Diversity in the sialic acids. The nine-carbon backbone (more...) Download Teaching Slide

FIGURE 14.2

.

Diversity in the sialic acids. The nine-carbon backbone common to all known Sias is shown, in the α configuration. The following variations can occur at the carbon positions indicated:

R1 = H (on dissociation at physiological pH, gives the negative charge of Sia); can form lactones with hydroxyl groups on the same molecule or on other glycans; can form lactams with a free amino group at C-5; tauryl group.

R2 = H; alpha linkage to Gal(3/4/6), GalNAc(6), GlcNAc(4/6), Sia (8/9), or 5-O-Neu5Gc; oxygen linked to C-7 in 2,7-anhydro molecule; anomeric hydroxyl eliminated in Neu2en5Ac (double bond to C-3).

R4 = H; -acetyl; anhydro to C-8; Fuc; Gal.

R5 = Amino; N-acetyl; N-glycolyl; hydroxyl; N-acetimidoyl; N-glycolyl-O-acetyl; N-glycolyl-O-methyl; N-glycolyl-O-2-Neu5Gc.

R7 = H; -acetyl; anhydro to C-2; substituted by amino and N-acetyl in Leg.

R8 = H; -acetyl; anhydro to C-4; -methyl; -sulfate; Sia; Glc.

R9 = -H; -acetyl; -lactyl; -phosphate; -sulfate; Sia; OH substituted by H in Leg.

Figure 14.2

Figure 14.2

Figure 14.2

THE EXTENDED 2-KETO-3-DEOXYNONONIC ACID FAMILY

Figure 14.1

Also structurally related to Sias are eight- and seven-carbon 2-keto-3-deoxyoctonic acids and heptonic acids. Kdo belongs to the former group, although because of its different configuration it cannot be considered a true eight-carbon analog of the Sia Kdn. The biosynthetic pathways of Kdn and Kdo are also similar and they appear to share common ancestral genetic origins.

NOMENCLATURE AND ABBREVIATIONS

The complete chemical names of Sias are too cumbersome for routine use. A uniform and simple nomenclature system is being increasingly used, in which the abbreviation Neu denotes the core structure neuraminic acid, and Kdn denotes the core structure 2-keto-3-deoxynononic acid. Various substitutions are then designated by letter codes (Ac = acetyl, Gc = glycolyl, Me = methyl, Lt = lactyl, S = sulfate), and these are listed along with numbers indicating their location relative to the carbon positions. For example, N-glycolylneuraminic acid is Neu5Gc, 9-O-acetyl-8-O-methyl-N-acetylneuraminic acid is Neu5,9Ac₂8Me, and 7,8,9-tri-O-acetyl-N-glycolylneuraminic acid is Neu5Gc7,8,9Ac₃. If one is uncertain of the type of the Sia present at a particular location, the generic abbreviation Sia should be used. If other partial information is available, this can be incorporated, for example, a Sia of otherwise unknown type with an acetyl substitution at the C9 position could be written as Sia9Ac.

OLIGOSIALIC AND POLYSIALIC ACIDS

FIGURE 14.3

Terminal sialic, oligosialic, and polysialic acids, (more...) Download Teaching Slide

FIGURE 14.3

.

Terminal sialic, oligosialic, and polysialic acids, and the enzymes that can degrade them. (Arrows) Typical cleavage points for the action of the enzymes. Bacteria can also express some forms of sialic acids, but the linkage to the underlying core region is not always the same; in some instances, it is unknown (e.g., colominic acid, a bacterial polysialic acid that can also be cleaved by endosialidase).

Figure 14.3

Figure 14.3

TISSUE- AND MOLECULE-SPECIFIC EXPRESSION OF LINKAGES AND MODIFICATIONS

Certain linkages and modifications of Sias typically show tissue-specific and developmentally regulated expression. Some linkages and modifications are even molecule-specific, that is, they are found only on certain types of glycoconjugates in a given cell type. Even within a particular glycoconjugate group, a modification such as O-acetylation may be restricted to certain Sia residues at particular positions on a glycan. Such findings indicate the occurrence of specific enzymatic mechanisms for the generation and regulation of Sias (see below); they also suggest specific roles for these linkages and modifications. On the other hand, available evidence indicates substantial species-specific variations in the cell- and tissue-type distribution of different Sia linkages and modifications. Thus, at least some of this regulated expression may be unrelated to intrinsic functions of Sias. Rather, it may be the signature of the evolutionary history of a species in relation to the Sia-binding preferences of its pathogens and/or symbionts. Effectively, each species expresses a distinct “sialome,” a term defined as the total array of sialic acid types and linkages expressed by a particular organelle, cell, tissue, organ, or organism. Of course, unlike the genome, which is the same in every cell type of an organism and undergoes very few changes during the lifetime of the organism, the sialome differs among cell types and varies markedly with regard to time, space, and environmental cues.

METABOLISM

Synthesis of Sialic Acids

FIGURE 14.4

Genes and pathways involved in the biology of animal (more...) Download Teaching Slide

FIGURE 14.4

.

Genes and pathways involved in the biology of animal sialic acids. The general pathways for biosynthesis, activation, transfer, and recycling of the three common core sialic acids are shown in the context of two cells, one including the relevant organelles such as the Golgi apparatus, the nucleus, and the lysosome. Details on each gene product can be found by searching the gene name given for each reaction at the website of the Human Gene Nomenclature Committee, and related links. Pathways for Sia modification other than CMP-Neu5Gc production and O-acetylation/de-O-acetylation reactions are not shown (see text for discussion). (Question marks) Unknown or hypothetical pathways. (Modified and redrawn, with permission, from Altheide T.K. et al. 2006. J. Biol. Chem. 281: 25689–25702, ©American Society for Biochemistry and Molecular Biology.)

Neu5Ac and Kdn appear to be the metabolic precursors for all known animal Sias (see Figure 14.4). In vertebrate systems, they are derived by condensation of ManNAc-6-P (for Neu5Ac) or Man-6-P (for Kdn) with phosphoenolpyruvate. The ManNAc-6-P is produced by a bifunctional enzyme (encoded by GNE) that converts UDP-GlcNAc to ManNAc-6-P and UDP in two steps. Missense mutations in this gene give rise to hereditary inclusion body myopathy (HIBM) in humans (see Chapter 42), and inactivation causes embryonic lethality in the mouse. Condensation of the sugar phosphates with phosphoenolpyruvate yields the corresponding Sia-9-phosphates, which must be dephosphorylated by a specific phosphatase (encoded by NANP), giving free Sias in the cytoplasm. In contrast, Neu5Ac biosynthesis in prokaryotes involves condensation of ManNAc with phosphoenolpyruvate, giving nonphosphorylated Neu5Ac (Figure 14.4). Notably, various synthetic unnatural mannosamine derivatives can be utilized by the Sia biosynthetic machinery, allowing manipulation of the chemical structures of cell-surface sialic acids (see Chapters 49 and 50).

Activation to Form CMP–Sialic Acids

Free Sia derived from biosynthesis (or recycled/recovered from the lysosome; see below) can be used for glycan biosynthesis only after activation into the nucleotide donor CMP-Sia, a reaction catalyzed by CMP-Sia synthases (encoded by CMAS) using CTP as a donor. For reasons that are unclear, in all eukaryotic cells studied so far, this particular reaction takes place within the nucleus. The CMP-Sia products then return to the cytoplasm, where they are delivered into the lumen of Golgi compartments by the action of a specific antiporter (balanced by the export of CMP), which allows the generation of a higher concentration of CMP-Sias within the Golgi lumen than would be possible with passive transport (see Figure 14.4 and Chapter 4). These topological issues do not apply in prokaryotes, where CMP-Sias are synthesized in the cytoplasm and directly used in the coordinated assembly of cell-surface glycans, before their delivery to the surface. In eukaryotes, the levels of cytoplasmic free CMP-Sia can also cause feedback inhibition of UDP-GlcNAc 2′epimerase (encoded by GNE), the rate-limiting enzyme in the endogenous synthesis of the Sia precursor ManNAc. A genetic disease called “sialuria” arises from the failure of feedback regulation of this enzyme, which results in overproduction and excretion of sialic acids.

Modification of Sialic Acids

The remarkable chemical diversity of Sias is generated by multiple enzymatic mechanisms. The synthesis of Neu5Gc occurs by conversion of CMP-Neu5Ac to CMP-Neu5Gc in the cytoplasm (Figure 14.4). The CMP-Neu5Ac hydroxylase (encoded by the CMAH gene) responsible for this reaction is a cytoplasmic, iron-dependent enzyme that uses molecular oxygen and the common electron transport chain of cytochrome b5 and b5 reductase. Alternative pathways for generation of Neu5Gc are being explored, because Neu5Gc has been found in low quantities even in species such as humans that lack the hydroxylase (but see discussion below). Once a Neu5Ac residue has been converted into Neu5Gc, there is no known way to reverse the reaction, perhaps accounting for the accumulation of Neu5Gc in cells that do express it (for the cellular pathways involving Neu5Gc, see Figure 14.4). In contrast to this cytoplasmic conversion reaction, the addition of O-acetyl esters and other hydroxyl group modifications seem to occur mostly in the lumen of the Golgi or in Golgi-related organelles, either onto the CMP-Sia precursor or after the transfer of Sias to glyco-conjugates. Regarding O-acetyltransferases, there is evidence for distinct enzymatic activities catalyzing O-acetylation of specific positions on Sias (e.g., C-4 vs. C-9), as well as specificity for O-acetylation of Sias on different linkages on different classes of glycoconjugates (e.g., gangliosides vs. N-glycans). Side-chain (C-7/8/9) O-acetyl groups appear to be initially added to C-7, followed by nonenzymatic migration to C-9 under physiological conditions, perhaps assisted by a “migrase” enzyme. The purification and cloning of these labile eukaryotic O-acetyltransferases has proven to be an intractable problem. O-Acetyltransferase genes from a few microorganisms were recently identified, but they show no homology to any eukaryotic gene. In mammalian systems, a protein complex in Golgi membranes is thought to be involved.

Other substitutions of the hydroxyl groups arise from use of the appropriate donors (e.g., S-adenosylmethionine for methylated Sias or 3′-phosphoadenosine 5′-phosphosul-fate for sulfated molecules). With 9-O-lactyl groups, even the donor is still unknown. Appropriate enzymes should also exist to permit the turnover of each of these substitutions. Notably, with the exception of Neu5Gc, the other modified Sias studied so far do not appear to be effective substrates for reactivation by most CMP-Sia synthases. Thus, O-acetyl esters need to be removed at some point in the life cycle of the parent molecule, either for terminal degradation or as part of an acetylation/ deacetylation cycle (see below).

The de-N-acetylated form of Neu5Ac (neuraminic acid, Neu) is unstable in the free state and thus had been assumed not to exist in nature. However, the glycosidically bound form of Neu is stable, and there is evidence that small amounts do exist in nature and that these molecules can be re-N-acetylated. The search is underway for enzymes that presumably remove and add back the N-acetyl group. In some instances, such a free amino group can react with the carboxylate at C-2, giving an intramolecular lactam ring. Various dehydrated or unsaturated Sias also occur in nature, including 2,7-anhydro Sias released following cleavage of bound Sias by certain unusual sialidases; 4,8-anhydro compounds formed during release or deacetylation of 4-O-acetylated Sias; and the 2-deoxy-2,3-didehydro Sias resulting from mild alkali-catalyzed breakdown of CMP-Sias or as products from sialidase reactions. Although many of these substances have been detected in free form in biological fluids, their biological significance is not known. Interestingly, the 2,3-didehydro forms are inhibitors of microbial sialidases and led to the development of potent anti-influenza drugs (see below and Chapter 50).

Release of Sialic Acids

Sialic acids attached to a glycoconjugate must eventually be removed at some point in the life cycle of the molecule (Figure 14.4). In eukaryotic systems, this occurs by the action of specific sialidases (encoded by NEUs; the term “sialidase” is now preferred over the older term “neuraminidase,” which is now used only in reference to viral enzymes, for historical reasons). Glycoconjugates are desialylated in endosomal/lysosomal compartments during recycling of cell-surface molecules and can sometimes return to the Golgi to undergo re-sialylation. In addition to endosomal/lysosomal sialidases, mammalian cells also have cell-surface (plasma membrane) and cytoplasmic sialidases. Cell-surface sialidases were originally thought to be involved in the abrupt shedding of cell-surface Sias that occurs upon activation of certain cell types (e.g., leukocytes). However, direct evidence for this is still lacking, and the plasma membrane sialidase appears to be specific for gangliosides, with claims for its involvement in signalling processes, apoptosis, and cell–cell contacts. The functions of cytoplasmic sialidases also remain quite obscure, because there is as yet no convincing evidence for glycosidically bound Sias in the cytoplasm nor on the cytoplasm-facing leaflet of cellular membranes. Recently, another sialidase was reported in human mitochondria. These enzymes are claimed to be involved in many cell biological processes, such as differentiation and cancer cell metastasis.

Many microorganisms also express sialidases, several of which have been cloned and characterized. Whereas the viral sialidases represent two distinct families, the bacterial, fungal, and invertebrate enzymes are evolutionarily related to mammalian families (in this instance, horizontal gene transfer between animals and pathogens seems possible). Most sialidases share a set of common “Asp boxes” (Ser-X-Asp-X-Gly-X-Thr-Tyr) that are probably involved in the maintenance of the enzyme protein conformation, together with a number of other highly conserved amino acids. The three-dimensional structures of several viral and bacterial sialidases have been elucidated, some in a complex with their substrates or with transition-state analogs. Interestingly, some have additional lectin domains that recognize underlying sugar chains and appear to direct the action of the enzyme.

Most sialidases exhibit substrate specificity regarding Sia linkage or the presence of substituents. Generally, α2-3 linkages are hydrolyzed more easily than α2-6 bonds, with the hydrolysis rate of α2-8-Sia being intermediate. A known exception is the enzyme from Arthrobacter ureafaciens, which acts best on α2-6 bonds. O-Methylation and O-acetylation of Sias can hinder (or even prevent, in the case of 4-O-acetyl groups) hydrolysis of the glycosidic bond by sialidases. These properties are both biologically and practically significant. The only known β-sialidase is CMP-Sia hydrolase, a poorly studied enzyme of unknown function, localized in the plasma membrane of some cell types.

A different type of sialidase is the “trans-sialidase” expressed by certain pathogenic protozoa (e.g., trypanosomes). These novel enzymes remove Sias from mammalian cell surfaces and transfer the sugar directly onto the parasite’s own cell-surface acceptors, apparently providing protection from the host immune system (see Chapter 40). Microbial sialidases and trans-sialidases are powerful virulence factors that may assist invasion, unmask potential binding sites, and, in addition, provide nutrients for some bacteria. Viral neuraminidases (“receptor-destroying enzymes”) are thought to assist viral entry by cleaving interfering sialic acids on inappropriate targets. They also facilitate release and spreading of newly formed viruses. Neuraminidase inhibitors are already in practical use as antiviral drugs (see Chapters 50 and 51).

Recycling of Sialic Acids

Once a Sia is released into the lysosome of a vertebrate cell, it is delivered back to the cytoplasm by a specific exporter called “Sialin” (see Chapter 4). This allows Sias to be either efficiently reutilized or degraded (Figure 14.4). Genetic defects in Sialin cause Salla disease and infantile sialic acid storage disease, resulting in accumulation of Sia in lysosomes and excretion of excess Sia in the urine. Some microorganisms can also directly scavenge Sias from the extracellular space, using high-efficiency transporters. In contrast, there is no evidence for plasma membrane Sia transporters in eukaryotic cells. However, free Sias can be relatively efficiently taken up into mammalian cells via fluid-phase macropinocytosis, eventually arriving in the lysosomes, from which they are exported into the cytoplasm by Sialin. Sialic acids that are glycosidically bound to soluble extracellular glycoproteins can be similarly transported to the lysosomes, where lysosomal sialidases can release them for delivery to the cytoplasm and eventual utilization by the cellular CMP-Sia synthase. The extent to which various eukaryotic cell types rely on such exogenous sources of Sias and/or on internal recycling is unknown. As discussed above, O-acetylated Sias probably need to be de-O-acetylated by specific 9-O-acetylesterases before they can be reutilized by cells. The acyl-mannosamines derived from the degradative activity of lyases (see next section) may also be reused for Sia synthesis. At the whole-body level, free Sias in the bloodstream (derived from cellular sources or digestive processes in the intestine) are rapidly excreted in the urine.

METHODS FOR STUDYING SIALIC ACIDS

Linkage-specific sialidases, esterases, Sia lyases, and/or lectins can all help to define some aspects of the Sias on a given glycan of a glycoconjugate. Monoclonal antibodies, lectins, and combinations of mild periodate oxidation with saponification have also been used to identify Sias and/or O-acetyl groups histochemically on tissue sections. A recombinant soluble form of the 9-O-acetyl-specific hemagglutinin of influenza C virus can probe for such molecules on thin-layer chromatograms, microwells, cells, and tissues. In some instances, information derived from such simple analyses is sufficient to reach biologically relevant conclusions. Various mass spectrometric (MS) and nuclear magnetic resonance (NMR) methods allow Sias to be more precisely characterized while they are still attached to the underlying glycan. The most accurate analysis of Sias from biological sources requires complete release and purification, with their modifications intact. However, some methods used to release, purify, or characterize the glycans can result in loss of labile Sia modifications (see below). Released and purified Sias can be analyzed by spectrophotometry, thin-layer chromatography (TLC), gas-liquid chromatography, MS, or NMR spectroscopy. Derivatization with 1,2-diamino-4,5-methylenedioxybenzene dihydrochloride (DMB) followed by high-pressure liquid chromatography analysis with fluorescent detection has proven to be particularly sensitive, specific, and applicable to most Sias. The adaptation of this technique to on-line electrospray mass spectrometry has been a powerful enhancement. Several techniques have also been developed for the detailed analysis of substitutions on metabolically labeled Sias.

For technical reasons, studies of sialoglycoconjugates continue to miss the extent of naturally occurring Sia structural complexity. Some Sia linkages may be partially or completely resistant to certain sialidases. Some substitutions are particularly labile (e.g., O-acetylation) and/or can alter the behavior of Sias during release, purification, and analysis. In addition, substitutions can slow down or even completely prevent release of Sias by commonly used sialidases or by acid hydrolysis. On the other hand, when stronger acidic conditions are used, destruction of some substitutions and of Sias themselves occurs. Furthermore, many methods used in structural analysis of intact glycans (e.g., alkaline conditions) cause the destruction of Sia modifications. Additionally, the presence of sialidases or esterases in crude cell extracts can alter the natural spectrum of sialoglycoconjugates. Because Sia modifications can affect size, shape, hydrophilicity, net charge, and biological properties of a glycoconjugate, a careful analysis for their presence is worthwhile in situations in which Sias are thought to have biological roles. With regard to side-chain O-acetylation, chemical and enzymatic improvements now allow near-quantitative release and purification of such molecules, without loss or migration of the ester groups. With rarer molecules such as O-lactylated, O-methylated, or sulfated Sias, much less is known about their susceptibility to sialidases or their optimal release with acid, and other methods for their direct detection are not available. It is evident that much needs to be done to improve methods for the detection, release, and purification of Sias from biological sources.

GENERAL FUNCTIONS OF SIALIC ACIDS

The high expression of Sias on outer cell membranes (e.g., more than 10 million molecules per human erythrocyte) on the interior of lysosomal membranes and on secreted glyco-proteins (such as blood proteins and mucins) suggests that they have roles in the stabilization of molecules and membranes, as well as in modulating interactions with the environment. Some functions arise from the relatively strong electronegative charge of Sias, for example, binding and transport of ions and drugs, stabilizing the conformation of proteins including enzymes, and enhancing the viscosity of mucins. Sias can also protect molecules and cells from attack by proteases or glycosidases, extending their lifetime and function. Furthermore, Sias can regulate the affinity of receptors and are reported to modulate processes involved in transmembrane signaling, fertilization, growth, and differentiation. In one system, apoptosis was reported to be inhibited by Sia O-acetylation. A recently described general property of Sias seems to be their free-radical scavenging antioxidative effect, which could be particularly significant on endothelia of blood vessels.

Another prominent role of Sias is dualistic; they act either as masks or recognition sites. In the first case, they mask antigenic sites, receptors, and, most importantly, penultimate galactose residues. After Sia loss, molecules and cells can be bound, for example, by macrophages and hepatocytes, via Gal-recognizing receptors, and can even be taken up and degraded. This phenomenon has been most extensively studied with serum glycoproteins and blood cells. On the other hand, Sias themselves can serve as ligands for a variety of microbial and animal lectins, as is discussed in the following section.

Chemical modification of Sias can strongly influence all of these properties, in particular ligand functions. For example, 9-O-acetylation or N-acetyl-hydroxylation of Neu5Ac can create new receptor functions or decrease the affinity of binding.

SIALIC-ACID-RECOGNIZING LECTINS

TABLE 14.1

Examples of naturally occurring sialic-acid-binding (more...)

TABLE 14.1

Examples of naturally occurring sialic-acid-binding lectins

Vertebrate

C-type: Selectins (see Chapter 31)

I-type: Siglecs (see Chapter 32)

Unclassified: Complement factor H

Arthropod

Crab lectins: Limulin (American horseshoe crab, Limulus polyphemus)

Lobster and prawn lectins: L-agglutinin (lobster, Homarus americanus)

Scorpion lectins: Whip scorpion lectin (Mastigoproctus giganteus)

Insect lectins: Allo A-II (beetle lectin, Allomyrina dichotoma)

Mollusk

Slug lectins: Limax flavus agglutinin (LFA) (Limax flavus)

Mussel and oyster lectins: Pacific oyster lectin (Crassostrea gigas)

Snail lectins: Achatinin-H (Achatina fulica)

Protozoa

Parasite lectins: Merozoite erythrocyte-binding antigens (EBAs) (Plasmodium falciparum)

Plant

SN agglutinin (SNA) (elderberry bark lectin, Sambucus nigra), PS agglutinin (Polyporus squamosus), MA agglutinin (MAH) (Maackia amurensis), wheat-germ agglutinin (Triticum vulgaris)

Bacteria

Bacterial adhesins: S-adhesin (Escherichia coli K99), SabA and SabB (Helicobacter pylori)

Bacterial toxins: Cholera toxin (Vibrio cholerae), tetanus toxin (Clostridium tetani), botulinum toxin (Clostridium botulinum), pertussis toxin (Bordetella pertussis)

Mycoplasma lectins: Mycoplasma pneumoniae hemagglutinin

Viruses

Hemagglutinins: Influenza A and B viruses, primate polyomaviruses, rotaviruses

Hemagglutinin neuraminidases: Newcastle disease virus, Sendai virus, fowl plague virus

Hemagglutinin esterases: Influenza C viruses, human and bovine coronaviruses

FIGURE 14.5

Examples of terminal glycan sequences recognized by (more...) Download Teaching Slide

FIGURE 14.5

.

Examples of terminal glycan sequences recognized by some sialic-acid-binding proteins. N-acetylglucosamine (GlcNAc) or N-acetylgalactosamine (GalNAc) residues on glycoproteins and/or glyco-sphingolipids can be extended by several biosynthetic pathways, some examples of which are shown. The sialylated sequences recognized by various binding proteins are based on published literature and/or reasonable predictions based on known specificities. The sequences shown are the minimal structural motifs necessary for binding, and relative differences in binding strength are not shown. Natural high-affinity ligands may be more complex. Recognition can be affected by modifications of sialic acid other than O-acetylation or by sulfation of adjacent monosaccharides (not shown). (ST) Sialyltransferase; (OAT) O-acetyltransferase; (CD22) Siglec-2; (Sn) Sialoadhesin/Siglec-1; (SNA) Sambucus nigra agglutinin; (MAA) Maackia amurensis agglutinin; (LFA) Limax flavus agglutinin; (Inf A HA) influenza A hemagglutinin; (Inf C HA) influenza C hemag-glutinin. With Inf A HA, the relative preference for α2-3 and α2-6 Sia linkages can vary with the viral strain. Note that “MAA” is typically a mixure of at least two lectins, Maackia amurensis hemagglutinin (MAH) and Maackia amurensis leukogglutinin (MAL), each with somewhat different preferences for the glycan underlying the α2-3-linked Sia. The latter (MAL) can also recognize 3-O-sulfated LacNAc termini.

Figure 14.5

PRACTICAL USES OF SIALIC-ACID-BINDING LECTINS

Figure 14.5

EVOLUTIONARY HISTORY OF SIALIC ACIDS

Early studies suggested species specificity in the occurrence of different types of Sias. However, with improvements in detection and analysis techniques, it is evident that most Sia types are widely expressed and simply occur at differing levels of detectability. As a group, Sias became prominent late in evolution, primarily in animals of deuterostome lineage (see Chapter 25), which comprises the vertebrates and some “higher” invertebrates (such as echinoderms) that emerged at the Cambrian expansion (~530 million years ago). Indeed, with rare exceptions (some that remain controversial), Sias are not generally found in plants or in most prokaryotes or invertebrates. However, there have been a few credible reports of Sias in mollusks, such as octopus and squid, and insects such as Drosophila. Also, genes structurally related to those involved in vertebrate Sia metabolism have been reported in insects and plants, and even in Archaea. With improved analysis techniques, Sias are now often found in membrane macromolecules of microorganisms. Overall, it appears that Sias may be a more ancient Precambrian invention, but they were then either eliminated or used only sparingly in many lineages—finally “flowering” into prominence only in deuterostome lineage. In this regard, genetic evidence also suggests that the original invention of Sias may have derived from homologous gene products that synthesize keto-deoxyoctulosonic acid (Kdo). Meanwhile, certain strains of bacteria can contain large amounts of Sias or other 2-keto-3-deoxynononic acids in their capsular polysaccharides and/or lipooligosaccharides. Some of these bacteria are pathogenic and cell-surface sialic acids protect them from complement activation and/or antibody production. Thus, although definitive proof has not been obtained, the possibility of gene transfer from host eukaryotes exists. However, it does seem that many of the bacterial enzymes involved in synthesizing and metabolizing Sias have evolved independently, possibly being “reinvented” from the Kdo pathway. Interestingly, there is wide variation in Sia expression and complexity within deuterostome lineage, with the sialome of echinoderms appearing very complex and that of humans being more simple. However, expression of Neu5Gc and O-acetylated Sias is highly conserved in deuterostomes, although exceptions exist, such as the lack of Neu5Gc in man, chicken, and some other birds.

LOSS OF N-GLYCOLYLNEURAMINIC ACID PRODUCTION IN HUMANS

The common mammalian Sia Neu5Gc was once thought to be an onco-fetal antigen in humans, being apparently absent from normal adult human tissues but expressed in fetal samples and certain human tumors and tumor cell lines. Indeed, upon human intravenous exposure to horse antiserum (still sometimes used in situations such as snake bite), the resulting “serum sickness” (Hanganutziu–Deicher or “HD”) antibodies are prominently directed against Neu5Gc. Spontaneously appearing HD antibodies were also reported in patients with cancer and certain infectious diseases, as well as in chickens with Marek’s disease, a malignant herpesvirus infection. In humans, the explanation is homozygosity for an inactivating exon deletion in the CMAH gene that occurred after our last common ancestor with the African great apes. Meanwhile, using sensitive techniques, traces of Neu5Gc have been found in normal human tissues. This, as well as the higher level of reexpression of Neu5Gc reported in malignant tissues, seems to represent incorporation from dietary sources such as red meats and milk products. However, an alternate pathway for Neu5Gc synthesis in tumor cells has not been conclusively ruled out. With the discovery that most or all healthy humans have some levels of circulating anti-Neu5Gc antibodies, the possibility has been raised that this might account for the high frequency of atherosclerosis and epithelial cancers in humans, diseases that seem uncommon in the great apes and have been correlated with red meat consumption in humans.

A potentially related observation is the suppression of CMAH/Neu5Gc expression in the brains of all animals studied, including those that have high levels expressed in other tissues. Because the loss of Neu5Gc in human lineage may have predated the appearance of the genus Homo, it is possible that the complete elimination of Neu5Gc may have somehow facilitated human brain evolution; however, no testable hypothesis has been advanced. Additional consequences for the evolution of humans may relate to the ancestral condition of many CD33-related Siglecs (see Chapter 32), which selectively bind to Neu5Gc. Thus, loss of Neu5Gc during human evolution would have caused a temporary loss of ligands for these inhibitory molecules of the innate immune system, a situation that has apparently been corrected by multiple human-specific changes in this family of receptors, leading to better binding to Neu5Ac. Other possible consequences include human resistance to veterinary microbial pathogens such as Escherichia coli K99, and the successful emergence of Neu5Ac-preferring pathogens such as the human-specific malarial parasite P. falciparum.

SIALIC ACIDS IN DEVELOPMENT AND MALIGNANCY

Cultured cell lines that are grossly deficient in sialylated glycans show generally normal growth patterns. Thus, more critical biological roles of Sias may only be evident in multicellular or intact vertebrate systems. Indeed, as already mentioned, Sias are critically required for early mammalian development. However, apart from the function of polySias in allowing “neural plasticity,” the exact roles of Sias during development remain uncertain. The role of Neu5Ac expression in the larvae of the insects Drosophila and the cicada Philaenus spumarius is also unknown. Several examples of Sia regulation have been reported in living animals. Certain classes of T lymphocytes have O-acetylated Sias, whereas others do not. The expression of polysialylation and O-acetylation in neural gangliosides varies with developmental stage and location, and differences in O-acetylation of brain ganglio-sides have been reported between cold- and warm-blooded species, and between awake and hibernating animals. Developmental regulation of Neu5Gc expression and O-acetylation expression in the gut mucosa may occur in response to microbial colonization and has been suggested to have a role in protecting against certain microorganisms. Similarly, although adult bovine submandibular glands produce large amounts of highly O-acetylated mucins, this Sia modification is scarcely expressed in the corresponding fetal tissue. The type and linkages of endothelial, plasma protein, and erythrocyte Sias can undergo marked changes in responses to inflammatory stimuli. Interesting abnormalities have also been reported in transgenic mice expressing influenza C 9-O-acetylesterase and following genetic inactivation of various sialyltransferases in the intact mouse. A variety of sialyltransferase-null mice have been produced that show interesting and specific phenotypes, ranging from altered Siglec-2/CD22 function (ST6Gal-I null) to defects in T-cell maturation (ST3Gal-I null) and changes in brain development (ST8Sia-II and ST8Sia-IV null).

In addition to the accumulation of Neu5Gc, several other specific changes in Sias occur in malignancy (see Chapter 44). In general, the total amount of Sia increases and switches occur in linkages, with α2-6 linkages becoming particularly prominent. O-Acetylation at C-9 can either disappear (as occurs in colon carcinomas) or become prominent (as in 9-O-acetyl-G_D3, which is much increased in melanomas and basal cell carcinomas). With the exception of the role of Sia in selectin ligands (see Chapter 31), the precise mechanisms by which these Sia changes enhance tumorigenesis and/or invasive behavior remain uncertain. Increased sialylation may also enhance the masking effect of Sia on antigenic sites of tumor cells, which become more like “self” and therefore more invasive. Regardless of the mechanisms involved, certain sialylated molecules are specific markers for some cancers and potential ligands for targeted therapies (see Chapter 44).

SIALIC ACIDS IN PATHOLOGY AND PHARMACOLOGY

Because Sias are involved in so many cellular functions, disturbances of their biosynthesis or degradation can lead to medical problems. Because of their exposed position, Sias are vulnerable to the action of microbial esterases, sialidases, and lyases. The actions of these enzymes can affect the amount of ligand present, masking of antigenic sites, stabilization of membranes, and immunological and other functions of Sias. In this regard, microbial lectins, sialidases, and trans-sialidases are potent virulence factors. Many bacterial toxins (e.g., cholera, tetanus, and pertussis toxins) and species of virus (e.g., influenza viruses) bind to sialylated glycoconjugates (see Chapter 34). Bacteria may also create new binding sites by sialidase-mediated unmasking of penultimate galactose residues. Trans-sialidases of some pathogenic trypanosome species make these parasites fitter for survival in the vector or host, and they strongly disturb the host’s immune system by compromising the cytokine network and influencing signaling processes.

Changes in Sias have also been found to be involved in degenerative diseases such as artherosclerosis and diabetes as well as neurological disorders such as Alzheimer’s disease and alcoholism. Mucins also have to be properly and highly sialylated in order to exert their physiological functions as lubricants and in innate immunity. Selectins recognize sialyl Lewis^x glycans that generally contain a terminal Sia, and Sias are thus involved in rolling and extravasation of leukocytes during inflammation (see Chapter 31).

Several human genetic Sia disorders are known: for example, hereditary inclusion body myopathy (HIBM) (caused by missense mutations of the UDP-GlcNAc-2-epimerase/ManNAc kinase [GNE] gene), sialuria (a defect of GNE feedback inhibition by CMP-Neu5Ac), Salla disease (a defect in the lysosomal Sia transporter Sialin), and galactosialidosis (galactosi-dase-peptidase-sialidase complex deficiency). Many of the congenital disorders of glycosylation (CDGs) may also lead to altered sialylation (see Chapter 42), but less is known regarding the molecular basis of phenotypic consequences.

On the basis of these diverse pathophysiological roles of Sias, many efforts have been undertaken to create appropriate pharmacologically active agents. Best known are the competitive inhibitors of sialidases, derived from the natural sialidase inhibitor Neu2en5Ac (2-deoxy-2,3-didehydro-Neu5Ac), which hinder budding and spreading of influenza A and B viruses (see Chapter 50). Inhibitors of bacterial sialidases and trypanosomal trans-sialidas-es are urgently needed. Although sialyltransferase inhibitors could be useful in cancer, potent agents are not yet available. Many attempts have been made to generate agents derived from sialyl Lewis^x (sLe^x) to compete with the selectins and to affect inflammatory processes, reperfusion injury, or tumor cell metastasis. Antiadhesive sialylated molecules could also be potentially useful for the treatment of bacterial and viral infections; for example, corresponding sialylated glycodendrimers can inhibit binding of the influenza virus hemagglutinin. Sialylated milk oligosaccharides are claimed to fulfill this task in a natural way in the intestine and stomach, perhaps reducing H. pylori infection. Sulfated polysialic acid has been found to suppress human immunodeficiency virus (HIV). Strategies for attacking cancer cells include vaccination with sialoglycoconjugates (e.g., gangliosides) in the case of melanoma or vaccination with polysialic acid modified with unnatural Sia in other cancers.

Manipulation and control of sialylation levels are also important in biotechnology (see Chapter 51). Engineering of erythropoietin to a hypersialylated form gives better pharmacokinetic properties, especially a longer lifetime in the bloodstream. The same strategy is presently being investigated with other naturally carbohydrate-free peptide hormones such as insulin or cytokines by adding N-glycosylation sites. These procedures require sophisticated chemical, recombinant, and other methods in order to achieve proper and possibly variable sialylation with Neu5Ac (Neu5Gc should be avoided because of its antigenicity in man). The production of recombinant glycoproteins of pharmaceutical value in large quantities is most promising in yeast, insect, and plant cells that do not express Sia, but these are now being engineered by transfection of the appropriate enzymes so they can make “humanized” glycoproteins with complex sialylated N-glycans (see Chapter 51).