Carbohydrate Structure and Diversity

The structural diversity characteristic of the oligosaccha-rides found in nature stem principally from three sources: (1) a large number of monosaccharide types, (2) the multiple ways in which the monosaccharides can be linked

Figure 1 Structural representation of the sulfated sialyl Lewis x tetrasac-charide, NeuAca2-3Galpi-4[Fuca1-3](6-sulfo)GlcNAc. NeuAc, Gal, GlcNAc, and Fuc label the monosaccharide moieties N-acetylneuraminic acid, galactose, N-acetylglucosamine, and fucose, respectively.

together, and (3) the fact that oligosaccharides can be further modified chemically (e.g., sulfate, phosphate, and acetyl). The basic themes are illustrated in Fig. 1, which shows the structure of sulfated sialy Lewis x, a tetrasaccharide important in selectin-mediated recognition. Because most monosaccharides have more than one hydroxyl group available for glycosidic bond formation, oligosaccharides, unlike their peptide counterparts, can form branched structures. The oligosaccharide structures linked to lipid or to protein, through Ser/Thr (O-linked) or Asn (N-linked), typically contain between 1 and 20 monosaccharide moieties and may be branched or linear. The much longer linear glycosaminylglycans, either in isolation or as the oligosaccharide chains of proteoglycans, are found on cell surfaces and in the extracellular matrix.

In vivo, oligosaccharides are synthesized by glycosyl-transferases, each of which typically has a unique donor, acceptor, and linkage specificity. As such, a very large number of glycosyltransferases and related enzymes are required to generate the oligosaccharide diversity seen in nature. Although the basis for this diversity is not fully understood, general themes are beginning to emerge. The so-called terminal elaborations (e.g., sialic acid, galactose, and sulfate) typical of the N-linked oligosaccharides of multicellular organisms, for example, seem to have appeared as part of the machinery required to mediate cell-cell and cell-matrix interactions [15]. In addition, it seems likely that oligosaccharide diversity has also been driven by evolutionary pressures arising from the need to differentiate self from nonself [16].

Lectins and Carbohydrate Recognition

Carbohydrate-binding proteins or lectins, like their saccharide counterparts, are also found in organisms ranging from microbes to humans [1]. The canonical carbohydrate recognition domain (CRD), characteristic of a given lectin type, can be found either in isolation or in conjunction with other protein domains, including coiled-coil domains and membrane-spanning motifs. Although many of the known CRD types are completely unrelated at the protein structural level [17-19], they can be grouped into two broad classes [20]. The type I CRDs are typified by the bacterial carbohydrate transporters and are characterized by deep carbohydrate-binding sites that essentially envelop their small saccharide ligands. In type II CRDs, the carbohydrate-binding sites are more shallow in nature and the saccharide remains relatively exposed to solvent, even when bound to the CRD. As a result, the dissociation constants (Kd) for small mono- or disaccha-rides can approach 0.1 ^M for the type I CRDs, while the type II CRDs tend to bind small saccharides with Kd in the range of 0.1 to 1.0 mM.

Despite their relatively weak affinities for small saccharides, type II CRDs often show a strict mono- or disaccharide binding specificity. From a structural standpoint, this is achieved by a complementarity of fit between the CRD and the saccharide moiety which includes both hydrogen bond and van der Waals interactions. The structural and thermo-dynamic basis for this specificity has, in fact, been well studied and reviewed in detail elsewhere [17-23].

Given that the type II CRDs bind small saccharides relatively weakly, most of these lectin types have employed multivalency as a means of conferring additional affinity and specificity on their binding interactions with larger oligosaccharides [17]. In addition to the monosaccharide in the primary site, the CRD may possess subsites for interaction with other monosaccharides of the oligosaccharide. Alternately, many lectins cluster their CRDs as a means of making multivalent interactions with larger oligosaccharides or other extended structures such as cell surfaces. Members of the C-type lectin family, for example, are known to form monomers, trimers, tetramers, pentamers, and hexamers, as well as higher order oligomers, and in some cases a single polypeptide chain will possess more than one canonical CRD.

Carbohydrate-Mediated Signaling

Lectins as Receptors

Most of the current evidence for the biological roles of complex carbohydrates comes from systems where they act as ligands for membrane-bound receptors that are lectins. Typically, these receptors have one or more extracellular CRDs, a single transmembrane-spanning region, and a relatively short cytosolic tail. In most cases, they are probably activated by receptor cross-linking mechanisms.

L-, P-, and E-selectin are cell-surface, C-type lectins responsible for leukocyte homing [24]. Unlike other members of the family, they do not possess a monosaccharide binding specificity. They require at least a tetrasaccharide, sialyl Lewis x (Fig. 1) for binding, and specific sulfation further enhances binding to L- and P-selectin [25]. The crystal structures of P- and E-selectin, in complex with oligosaccharide/ glycopeptide ligands, have shown the importance of electrostatics in these interactions, a factor thought to be important in the rapid binding kinetics required for leukocyte rolling [26]. Moreover, the structures have provided a rationalization for the specificity differences that ensure that lymphocytes target to lymph nodes and neutrophils reach sites of inflammation. Although the selectins are not known to form oligomers, E-selectin-mediated clustering at contact points between interacting cells has been shown to activate the ERK1/2 signaling pathway [27].

DC-SIGN and DC-SIGNR are also C-type lectins, but in this case they are involved in dendritic cell/T-cell interactions [28], as well as the promotion of HIV-1 infection [29]. These lectins possess a mannose-binding specificity, but in addition show a marked increase in affinity for high mannose oligosaccharides [30]. The crystal structures of their CRDs in complex with a mannopentasaccharide show that the increased affinity arises from a further set of interactions in addition to those made with the mannose in the primary binding site [31]. Because these lectins also possess a-helical tetramerization domains, it seems likely that they would be capable of making high-affinity interactions with ICAM-3 and HIV gp120, two of their natural ligands. In fact, it has been suggested that the cross-linking of DC-SIGN tetramers, by the highly multivalent high mannose oligosac-charide containing HIV virus, provides the signal required to promote transport of HIV from the periphery to the T-cell-containing lymph nodes [29].

The hepatic asialoglycoprotein receptor, a member of the C-type lectin family, provides a well-characterized example of the interplay between structure, specificity, and receptor cross-linking. Although an isolated CRD of this receptor binds galactose with a Kd in the millimolar concentration range, the cell-surface form of the receptor can bind the appropriate triantennary N-linked oligosaccharide with nanomolar affinity. Cross-linking studies have shown that the HL-1 subunit forms trimers on the cell surface and that recruitment of an additional HL-2 subunit(s) generates the high-affinity receptor. The galactose terminii of the triantennary oligosaccha-rides (separated by 15 to 25 A) are found to interact with both the HL-1 and HL-2 subunits [32]. Linking receptor specificity to receptor cross-linking in this way may be important for both receptor uptake and signal transduction [33].

The targeting of lysosomal enzymes is also dependent on receptor-mediated endocytosis. In this case, the cation-dependent mannose 6-phosphate receptor (CD-MPR) and the insulin-like growth factor II/cation-independent man-nose 6-phosphate receptor (IGF-II/CI-MPR) specifically recognize the mannose-6-phosphate moiety on acid hydro-lases destined for lysosomes [34]. Again, multivalency is important; CD-MPR binds mannose 6-phosphate with a dissociation constant in the micromolar concentration range, while the dimeric receptor binds tetrameric P-glucuronidase with nanomolar affinity. Both dimeric and tetrameric forms of the receptor are found in the Golgi membrane, and, based on the crystal structure of the dimeric CD-MPR, a model for its high-affinity interaction with P-glucuronidase has been proposed [35]. The IGF-II/CI-MPR receptor contains two canonical CRDs presumably capable of promoting high-affinity interactions with multivalent lysosomal enzymes, and together with CD-MPR these receptors are responsible for targeting over 50 structurally distinct lysosomal enzymes. Dimerization of the IGF-II/CI-MPR receptor by

P-glucuronidase binding increases receptor internalization at the cell surface [36].

The siglecs are a family of sialic acid binding lectins whose canonical CRD is a member of the immunoglobulin (Ig) superfamily. They are particularly important in the immune system, where they function in processes ranging from leukocyte adhesion to hemopoiesis [37]. Members of the family show specificity differences for a2,3- versus a2,6-linked sialic acids, as well as for sialic acids modified with respect to O-acetylation. The crystal structure of the CRD of siaload-hesin in complex with 3' sialylactose shows that interactions with the bound oligosaccharide are mediated primarily with the terminal sialic acid moiety [38]. Of particular interest are the roles played by cis interactions. CD22 (Siglec-2), for example, is a B-cell-specific receptor which, through interaction with a2,6-linked sialic acid containing glycoproteins on its own cell surface, inhibits B-cell receptor signaling. This stable inhibition can be broken by the addition of external competing saccharide and in vivo may be controlled by the regulation of sialytransferases and/or sialidase expression levels [39]. The cloning of several CD33-related receptors expressed on myeloid cell progenitors suggests new insight into the significance of their sialic acid binding properties. In all cases, these receptors possess cytoplasmic immunorecep-tor tyrosine-based inhibitory motifs (ITIMs), elements now known to be hallmarks of inhibitory receptors central to the initiation, amplification, and termination of immune responses [40]. Through interactions with sialic acid containing self determinants, these receptors may play roles in the control of innate immunity [41].

Serum mannose binding protein (MBP), a component of the vertebrate innate immune system, is also a C-type lectin. Although not membrane bound, it signals activation of the complement cascade though a conformational change initiated by binding the cell surface of a foreign pathogen [42]. Like the asialoglycoprotein receptor, the CRD of MBP also recognizes only a terminal monosaccharide moiety, in this case mannose. The CRDs are also found to form trimers; however, in MBP they are mediated by long, triple-helical, coiled-coil domains that in addition promote the formation of trimer clusters containing 18 CRDs in total [43]. The crystal structures of truncated forms of the trimer show that the mannose binding sites are separated by 45 and 53 A, respectively, in human [44] and rat [45] MBP. Thus, unlike the asialoglycoprotein receptor, which is designed to recognized the closely spaced galactose determinants of a single N-linked oligosaccharide, MBP is designed to bind the widely spaced mannose determinants typical of the cell surfaces of pathogenic microorganisms [46].

Glycoproteins as Receptors

It has long been known that certain multivalent, soluble plant lectins (e.g., PHA and Con A) can induce mitosis in lymphocytes and oxidative burst in neutrophils. The mechanism for initiation of these signals has generally been assumed to result from the cross-linking of cell-surface glycoproteins.

More recently, soluble animal lectins of the galectin type have also been found to induce a variety of signals, including, among others, apoptosis, oxidative burst, cytokine release, and chemotaxis in immune cells [47]. In structural terms, the galectins are either dimeric or contain more than one CRD on a single polypeptide chain and as such they are capable of cross-linking receptors [48]. Recent studies aimed at understanding T-cell homeostasis have suggested that CD45, CD43, CD7 [49], and the TCR-CD3 complex [50] are physiologically relevant cell-surface receptors for galectins-1 and -3, respectively.

Glycolipids as Receptors

The role of glycolipids as receptors for microbial lectins has been well studied. Bacterial AB5 toxins possess a pen-tameric arrangement of B-subunit lectins which, through multivalent interactions, promote high-affinity binding with host cell-surface gangliosides [51]. In the case of cholera toxin, binding to GM1 on the cell surface is followed by retrograde transport and translocation across the ER membrane [52]. Once in the cytosol, the A1 fragment of the A subunit catalyzes the ADP ribosylation of the heterotrimeric Gas protein, leading to the characteristic chloride and water efflux. In what is a fundamentally different type of interaction, the lectin subunits of the Escherichia coli P-fimbriae bind gly-colipids in uroepithelial cells leading to ceramide release, activation of ceramide signaling pathways, and ultimately cytokine release through a process that also appears to involve activation of the TLR-4 receptor pathway [53-55]. Although not yet fully characterized, the interactions of glycosphin-golipids with various adhesion and signaling receptors found in cell-surface microdomains are being found to mediate signaling events important in cell-cell interactions [56].

Proteoglycans and Glycosaminoglycans

Proteoglycans contain long linear oligosaccharide chains (glycosaminoglycans) made up of disaccharide repeats containing acidic monosaccharides and variable degrees of sulfation. They are found at the cell surface and in the extracellular matrix, where they interact with a wide variety of molecules, including, among others, signaling receptors, growth factors, chemokines, and various enzymes [57-59]. In the well-characterized fibroblast growth factor (FGF)-fibroblast growth factor receptor (FGFR) interaction, heparin/heparan sulfate serves as coreceptor. Two recent crystal structures of ternary complexes have begun to shed light on how the intrinsically multivalent oligosaccharide serves to promote receptor cross-linking in this system [60,61]. Recent evidence from studies on hepatocyte growth factor/scatter factor suggests that heparan and dermatan sulfate binding serves to promote a conformational change in the growth factor that promotes receptor binding [62]. In some cases, specific sulfation patterns appear to be important determinants of specificity [58,63]. The syndecans are cell-surface proteoglycans whose core proteins contain cytoplasmic signaling motifs. They have been implicated in the formation of focal adhesions, where interactions with heparin binding domains and other receptors are proposed to lead to adhesion, cross-linking, and signal transduction [64].

Small Soluble Saccharides

Small nutrient saccharides are often sensed by the receiving cells after entry through a transporter. In mammals, for example, glucose is sensed by an alteration in the adenosine triphosphate (ATP)/adenosine diphosphate (ADP) ratio resulting from glucokinase-initiated glucose metabolism. In microbes, small saccharides are often sensed by specific, non-enzyme cytosolic binding proteins that in turn regulate gene expression (e.g., the Lac-repressor of E. coli). In plants, nutrient sugars are also known to be important mediators of signal transduction [65], and the recognition of small soluble oligosaccharides by membrane and cytoplasmic receptors is important in plant host defense [66]. Although these examples are beyond the scope of this review, it is worth noting that these carbohydrate-mediated signaling mechanisms may be operative in systems yet to be characterized.

Carbohydrates and Lectins in the Nucleocytosolic Compartment

The O-linked glycosylation of serine and threonine residues of nuclear and cytoplasmic proteins by N-acetylglucosamine (O-GlcNAc) is involved in signal transduction in multicellu-lar organisms [67]. This dynamic modification occurs at sites of protein phosphorylation and may serve to transiently block sites of phosphorylation. Although its roles are not yet fully characterized, O-GlcNAc has been found to modulate a wide range of cellular functions, including transcription, translation, nuclear transport, and cytoskeletal assembly [68].

Galectins are also cytosolic and nuclear proteins, but they are not known to bind carbohydrates in these compartments; however, galectins 1 and 3 have been implicated in pre-mRNA splicing, a process inhibited by oligosaccharide binding [69]. The galectins are also secreted from the cytoplasm (by non-classical pathways), and it is at the cell surface that they perform the carbohydrate-mediated processes discussed previously.

For the sake of completeness it is worth noting that well-known second messengers such as cyclic AMP, GDP, GTP, etc. are ribose-containing glycoconjugates and that even more complex saccharide second messengers may be operative in insulin signaling [70].

Diabetes 2

Diabetes 2

Diabetes is a disease that affects the way your body uses food. Normally, your body converts sugars, starches and other foods into a form of sugar called glucose. Your body uses glucose for fuel. The cells receive the glucose through the bloodstream. They then use insulin a hormone made by the pancreas to absorb the glucose, convert it into energy, and either use it or store it for later use. Learn more...

Get My Free Ebook


Post a comment