Glycosyltransferases, Chemistry of
Joel T. Weadge and Monica M. Palcic, Carlsberg Laboratory, Valby, Copenhagen, Denmark
Glycosyltransferases (GTs) form glycosidic bonds by catalyzing the transfer of saccharides from a donor to a wide variety of acceptors. The donors used by GTs are sugars conjugated to nucleotides, phosphates, or lipid phosphates;whereas acceptors consist of carbohydrates, proteins, lipids, DNA, and numerous small molecules such as antibioticonols, steroids, and so on. Together, the products of these reactions comprise the most diverse and abundant class of natural compounds found in nature. Numerous GTs are needed to synthesize these compounds because the formation of each distinct glycosidic linkage requires a different enzyme. The abundance of these enzymes is emphasized by the fact that GTs constitute 1% of the genes in all genomes sequenced. They are ubiquitous in every kingdom of life and in all compartments of the cell. Currently, over 32,000 GT ORFs have been classified into 90 families on the basis of amino acid sequence similarity. The structures for 64 GTs have been determined to date and generally reveal conserved architectures of a GT-A or GT-B fold, although other folds have been observed and are predicted. These crystal structures, together with biochemical data, have provided insight into the catalytic mechanism. GTs generally exhibit strict regio/stereospecificity and transfer with either retention or inversion of configuration at the anomeric carbon of the donor sugar. The importance of characterizing the precise activity of these enzymes is exemplified by the many genetic disorders that have been linked to aberrant glycosylation.
Glycosyltransferases (GTs) comprise a group of enzymes that catalyze the synthesis of glycosidic linkages by the transfer of a sugar residue, which generally is a monosaccharide, from a donor to an acceptor substrate (1-3). The products of these reactions are oligosaccharides, polysaccharides, and also numerous glycoconjugates, which consist of carbohydrates linked to other noncarbohydrate molecules (Fig. 1). Together, these reaction products comprise the most diverse and abundant class of natural compounds found in nature. A different GT is required for each distinct sugar that is transferred and for each unique linkage that is formed. Because thousands of different linkages exist, this class of enzymes is enormous. In fact, GTs encode for 1% of the genes of all genomes sequenced to date. However, compared with well-characterized families of enzymes like the proteases or the protein kinases, considerably less is known about the GT enzyme family. Now the monumental task begins of identifying the precise function for each GT enzyme. This article will give an overview of GT mechanisms and structure/function relationships that have aided our understanding of the numerous biological roles of GTs. Furthermore, classification schemes as well as biotechnological and pharmaceutical aspects of GTs will also be discussed.
Figure 1. The diversity of natural structures produced by GTs. These structures include carbohydrate energy sources such as sucrose, lactose, and amylose, as well as oligosaccharide structural elements such as cellulose, chitin, and peptidoglycan. Secondary metabolites like the flavanoid delphinidin 3,5,3’-tri-O-glucoside are modified by GTs, whereas the steroid estrogen is inactivated by conversion to estrone 3-glucuronide. Glycoconjugates important in therapeutics and disease such as antibiotics (streptomycin and oleandomycin), sialoside receptors for viruses, and blood group antigens are also produced by GTs. The synthesis of glycolipids such as GM1 ganglioside and galactolipid, is also dependent on GTs. Cellulose, chitin, and amylose are synthesized by processive enzymes that repetitively transfer monosaccharides, whereas the remaining structures are synthesized by nonprocessive enzymes each carrying out single addition of a different sugar.
GTs catalyze the transfer of a monosaccharide or more rarely an oligosaccharide residue from a donor to an acceptor. The most common donors are sugar-nucleotides and GTs that use these donors are termed the Leloir-type GTs. These GTs comprise 90% of GT annotations, with UDP/TDP nucleotides accounting for more than 60% of these. However, dolichol-phospho-sugars, sugar-1-phosphates, and sugar-lipid-phosphates also serve as donors (Fig. 2). Acceptors cover every chemical class of compounds that include saccharides, polysaccharides, lipids, protein, DNA, and other natural products. GTs use these donor and acceptor substrates to form glycosidic linkages either through processive transfers with multiple additions of the same monosaccharide to the nonreducing end of a growing chain, as observed in the production of glycan biopolymers like cellulose and amylose, or through nonprocessive single transfer reactions, as exemplified by the blood group antigens, glycolipid GM1ganglioside and streptomycin (Fig. 1). Because a different enzyme is required to synthesize each distinct glycosidic linkage, GTs are probably the enzyme class with the greatest chemical diversity of substrates.
Figure 2. Examples of glycosyl donors. Nucleotide donors are represented by UDP-galactose (UDP-Gal), GDP-fucose (GDP-Fuc), CMP-N-acetylneuraminic acid (CMP-sialic acid, CMP-Neu5Ac), GDP-galactose (GDP-Gal), dTDP-rhamnose (dTDP-Rha), and dTDP-daunosamine; lipid phosphate donors are represented by Lipid II and dolichol-phosphate-glucose (Dol-P-Glc); sugar phosphate donors are represented by glucose-1-phosphate. The saccharides are transferred from these donors by GTs to form oligo/polysaccharides and glycoconjugate products.
With rare exceptions, the transfer of saccharides by GTs is regiospecific, such that the saccharide is transferred to only one of the many hydroxyl groups on acceptor molecules. The transfer reaction is also highly stereospecific. Thus, GT reactions follow two mechanistically distinct pathways that result in either inversion or retention of configuration of the anomeric configuration of the transferred sugar (Fig. 3). It is commonly assumed that the mechanisms of GT enzyme reactions are similar to those of the well-studied glycosidases that hydrolyze glycosides; one difference is that glycosidases transfer to a water molecule whereas GTs transfer to a hydroxyl group of an acceptor molecule. Although GT reactions in general favor the formation of glycosidic bonds biosynthetically, the reversibility of sucrose synthase and some natural-product enzymes have been useful for the production of nucleotide donors and natural product libraries (4).
Figure 3. GT reactions are regiospecific and stereospecific that occur with either inversion (a) or retention (b and c) of configuration at the anomeric center of the donor sugar. Inverting enzymes are thought to follow a nucleophilic displacement mechanism in which a general base deprotonates the acceptor, which renders it nucleophilic so it can attack the donor sugar. In contrast, retaining enzymes are thought to proceed through a double displacement reaction via the formation of a covalent intermediate (b) or by the SNi-like mechanism (c) where the nucleophilic acceptor and departing donor are on the same face of the sugar ring.
Inverting GT reactions are believed to follow a single displacement mechanism that involves nucleophilic attack of the OH-group of the acceptor on the anomeric center of the donor sugar (Fig. 3a). In this mechanism, a catalytic amino acid serves as a general base to deprotonate the reactive oxygen of the acceptor. Reaction occurs with formation of an oxocarbeniumon transition state and is concomitant with departure of the nucleotide leaving-group. X-ray structures of several inverting enzymes (β1,4-galactosyltransferase, α1,3-galactosyltransferase, and the bacterial α2,3-sialyltransferses CstI and CstII) in complex with their donor and acceptor substrates, have revealed an amino acid residue (E317, E317, H202, and H188, respectively) that is in position to deprotonate the acceptor molecule or to stabilize the transition state (5-8). Other residues have also been found that determine the substrate specificity. In the β1,4-galactosyltransferase, a tyrosine residue (Y289) facilitates the use of UDP-Gal as a substrate but not UDP-GalNAc (5). A histidine (H308) plays a similar role in β1,3-glucuronyltransferase I by determining its specificity for UDP-GlcA (9).
The catalytic mechanism for retaining GTs is unclear but is believed to occur either via a double-displacement mechanism or a SNi -like mechanism (Fig. 3b, 3c). The double-displacement mechanism (Fig. 3b) involves nucleophilic attack by an enzyme catalytic residue on the anomeric centre of the donor substrate. This mechanism leads to the formation of a covalent glycosyl-enzyme intermediate with inversion of configuration. In the second displacement, the glycosyl-enzyme intermediate is attacked by a hydroxyl group of the acceptor, after its deprotonation by a catalytic base. The configuration is inverted in the second step, which results in net overall retention of configuration in the product (Fig. 3b). For the SNi -like mechanism (Fig. 3c), catalysis involves nucleophilic attack and departure of the leaving group through a concerted, but asynchronous manner on the same side of the sugar ring. In early proposals, it was believed that a single catalytic residue on the retaining GT was required to act as the catalytic nucleophile by attacking the anomeric carbon of the donor. This mechanism was supported by the data obtained from structural elucidation of two retaining GTs, lipopolysaccharyl-α-1,4-galactosyltransferase C (LgtC) from Neisseria meningitidis (10) and α1,4-N-acetyl glucosaminyltransferase (EXTL2) (11). Amino acids Gln189 and Arg 293 of LgtC and EXTL2, respectively, were the only functional groups positioned to contribute in the mechanistic process. Thus, it seems that (at least for these two enzymes) a single catalytic residue is likely, but more experimental evidence is required to corroborate these findings. Molecular modeling of various reaction processes with LgtC (12) as well as structural complexes of trehalose-6-phosphate synthase with nontransferable analogs (13) suggested that the catalytic mechanism may actually be reminiscent of that proposed for glycogen phosphorylase (14). In this mechanism, it is thought that the active site constraints of the retaining GTs position the donor and the acceptor substrates in conformations that may not require a catalytic amino acid. When bound to the enzyme, the nucleotide sugar adopts a conformation where the sugar is folded over the pyrophosphate. The anomeric bond is elongated and weakened in this conformation thereby making the C1 position of the donor spatially accessible to direct attack by the OH nucleophile of the acceptor (13).
Given the vast structural and functional diversity of GT products, together with the divergent evolution of the enzymes, it may be expected that numerous possibilities exist for GT folds. Indeed, GTs display a high level of diversity in their primary sequences, which indicates that multiple solutions may exist to the problem of how a protein can catalyze glycosyl-transfer. Furthermore, a large number of folds have been identified for the glycosidase enzymes (15). Structural data indcate that GTs belong mainly in either the GT-A or the GT-B fold super families or variants thereof (Fig. 4a-e; 5, 8, 16, 17), although a lysozyme-like fold has been observed for peptidoglycan glycosyltransferase (18) (Fig. 4f), and a GT-C fold is predicted for integral membrane protein GTs that use lipid donors. Therefore, from an evolutionary standpoint, nature has settled on a limited number of protein folds to facilitate glycosylation events. It is noteworthy that the stereochemical outcome of glycosyl transfer reactions is not determined by fold because both retaining and inverting enzymes can belong to the same fold superfamily.
The first member of the GT-A superfamily fold was identified in 1999 when the three-dimensional structure of SpsA was reported (19). This enzyme from Bacillus subtilis is involved in the formation of the spore coat and is a member of the Leloir type of GTs given its apparent use of nucleotidyl-diphospho-donor sugars as a substrate. Enzymes with the GT-A fold have an N-terminal a/p/a sandwich motif that resembles a Rossmann motif and is involved in nucleotide donor binding. Most GT-A fold enzymes also have a characteristic Asp-Xxx-Asp (DXD) or equivalent motif (EXD or TDD) near the center of the protein that coordinates to the phosphate in nucleotide donors via a divalent metal cation (Mn2+ or Mg2+). These motifs are present in the structure of the bovine β1,4-galactosyltransferase I (5) (Fig. 4a).
Binding of the acceptor substrate does not have the same strong consensus character that has been noted for the donor site, but it has been noted that it takes place in the C-terminus. This acceptor-binding region generally contains motifs that consist of two flexible loops that undergo conformational changes after donor binds. In the retaining α1,4-galactosyltransferase LgtC, the central flexible loop was shown in crystal structures to interact with both the donor and the acceptor sugar substrates (10, 20). An additional flexible loop involved in acceptor binding is located at the extreme C-terminus of these enzymes (20). Isothermal titration calorimetry studies with α1,3-galactosyltransferase demonstrated that the disordered loop of the free enzyme has little affinity for the acceptor substrate (6). In this form, the enzyme possesses a wide-open active site cavity that serves to facilitate initial access of the donor sugar to the active site. Similar results have been noted for the crystal structure of β1,4-GalT in complex with donor substrate (5) (Fig. 4a). It seems that residues in the flexible loop are fixed by their interactions with the donor sugar that leads to a semi-closed conformation of the active site that can bind the acceptor. It has been speculated that this semi-closed form is important for excluding solvent from the donor-binding site and thereby for preventing unwanted hydrolysis of the nucleotidyl-sugar substrates in the absence of acceptor. Once the acceptor binds, the loop adopts a closed conformation where donor and acceptor are positioned properly for catalysis (5, 6, 20).
Figure 4b shows a modular variant of the GT-A fold as reported for polypeptide GalNAc transferase 10 (21). This enzyme has a C-terminal lectin domain that is thought to bind to GalNAc-containing peptides, which favors its substrate specificity for glycosylated peptides. Another variation of the GT-A fold is observed in the structures of α2,3-sialytransferases, CstI, and CstII, from Campylobacter jejuni (7, 8). These enzymes use a monophospho-nucleotidyl sugar CMP-Neu5Ac as a donor. Crystal structures of these enzymes with substrate indicate that this protein has an a/p/a sandwich motif reminiscent of the GT-A fold (Fig. 4c) and also contains flexible loops that undergo similar conformational changes on substrate binding.
The GT-B fold family includes most prokaryotic enzymes that produce secondary metabolites, like the antibiotics streptomycin, oleandomycin (Fig. 1) and vancomycin, and important bacterial cell wall precursors. It is also predicted to contain the vitally important O-GlcNAc transferase that modifies many nuclear and cytoplasmic proteins and influences gene transcription. The first glycosyltransferase structure reported in 1994 was for the GT-B fold enzyme, β-glucosyltransferase (BGT) from bacteriophage T4 (22). This enzyme attaches glucose to modified cytosine bases on duplex DNA. Its low-sequence homology to other GTs made it difficult to draw comparisons between it and other GT-B enzymes. Only after the structures of two additional enzymes, MurG and GtfB, became available could consensus domains within the GT-B fold be verified (23, 24). MurG is an enzyme from Escherichia coli that catalyzes the transfer of GlcNAc from UDP-GlcNAc to the nonreducing end of a lipid-linked N-acetylmuramic acid acceptor to form the repeating unit of peptidoglycan. GtfB is a UDP-glucosyltransferase involved in the biosynthesis of chloroeremomycin (or vancomycin). More recently, the structure of another GT-B folded enzyme, oleandomycin glucosyltransferase from Streptomyces antibioticus that is involved in antibiotic synthesis was also reported (16) (Fig. 4d). All these enzymes, OleD, MurG, GtfB, and BGT, exhibit almost identical topology despite very little sequence similarity. These enzymes and other members of the GT-B fold super family adopt a two-domain structure with a Rossmann-like fold in either domain. The predicted active site of the GT-B folded enzymes is located between the two Rossmann folds. Within the C-terminal portion of the cleft between the enzymes, another subdomain consists of an α/β/α structure that has a consensus glycine-rich pattern. This subdomain compromises the donor-binding domain. A crystal structure of MurG complexed with UDP-GlcNAc revealed that the first α-helix of this sub-domain makes contact with the ribose moiety of the nucleotide, whereas the second a-helix interacts with the pyranose residue. Enzymes that use diphospho-containing sugar donors have to contend with the negative charge of the phosphates. The GT-B enzymes, unlike the GT-A enzymes, do not seem to use divalent metal cations for this purpose. Instead, the α/β/α subdomain of the Rossmann fold may account for this by creating a positively charged helix dipole with the first a-helix that acts to stabilize the a-phosphate of the nucleotidyl donor (2, 25), whereas other enzymes use basic residues to counteract the positive charge of the diphosphate moiety.
In addition to the conserved donor binding domains, the acceptor-binding site for GT-B folded enzymes has also been identified near the N-terminus. Analysis of GtfB-related sequences aided the identification of this site because all these enzymes are highly homologous at the primary sequence level; but functionally, they glycosylate different acceptors (2, 24). Because the acceptors are all from a similar group of peptide antibiotics, the variation in the sequence was presumed to reflect structural adaptations in the binding site to accommodate the slightly different acceptors.
A variant of the GT-B fold is shown in Fig. 4e for an α1,6-fucosyltransferase (17). The structure shows an N-terminal coiled-coil domain, a catalytic domain that is similar to GT-B fold enzymes, and a C-terminal SH3 domain whose biological significance is currently uncertain.
Figure 4. Ribbon diagrams of glycosyltransferases that demonstrate the different structural folds. When possible, nucleotide-donor sugars are depicted as stick models, manganese is depicted by a magenta ball, and N- and C-terminals are labeled. GT-A fold members are represented by bovine β1,4-galactosyltransferase I complexed with UDP-Gal (5) (Fig. (4)a, PDB accession number 1O0R) and by human polypeptide α-N-acetylgalactosaminyltransferase 10, ppGalNAcT-10 (19) (Fig. (4)b, PDB 2D7 R). The latter is shown in complex with UDP and free GalNAc presumably which comes from the hydrolysis of UDP-GalNAc in the catalytic domain. The carbohydrate-binding lectin domain at the C-terminus is in complex with GalNAc serine. A variation of the GT-A fold is represented by CstII, which is a sialyltransferase from Campylobacter jejuni, complexed with CMP-3-fluoro-N-acetylneuraminic acid (8) (Fig. (4)c, PDB 1RO7). The GT-B fold is represented by oleandomycin glucosyltransferase from Streptomyces antibioticus complexed with UDP at the C-terminus and erythromycin at the N-terminus (16) (Fig. (4)d, PDB 2IYF). A variant of the GT-B fold is shown in Fig. (4)e for FUT8, human α1,6-fucosyltransferase, which has an N-terminal coiled-coil domain and a C-terminal SH3 domain (17) (PDB 2DE0). The possibility of other GT folds is represented in Fig. (4)f by a penicillin-binding protein from Staphylococcus aureus that is complexed with moenomycin in the peptidoglycan glycosyltransferase domain (PDB 2OLV). This structure is intriguing because the GT domain has a lysozyme-like fold (18). Diagrams were prepared with the PyMOL Molecular Graphics System (2002) (http://pymol.sourceforge.net).
Other glycosyltransferase folds
In addition to the variations of the GT-A and GT-B fold described above for sialyltransferases (7, 8) and fucosyltransferases (17), a distinct GT-C fold family has been predicted for GTs that use lipid linked donors. The crystal structures of the GT domain of the peptidoglycan glycosyltransferase from Staphylococcus aureus (18) (Fig. 3f) and Aquifex aeolicu s (26) show structural similarity to the bacteriophage T-lysozyme. These novel structures demonstrate the possibility of additional folds.
Classification of Glycosyltransferases
The classic enzyme commission (EC) classification for GTs is on the basis of their donor and acceptor specificity as well as the product formed. Currently, 295 entries are in this database (http://www.chem.qmul.ac.uk/iubmb/). The distinction between these enzymes is noted by their ability to catalyze the transfer of hexoses (EC 2.4.1.y, hexosyltransferases), pentoses (EC 2.4.2.y, pentosyltransferases), or other glycosyl groups (2.4.99.y, sialyltransferases). This classification is restricted to enzymes that are fully characterized, and it can be problematic for enzymes that act on several distinct acceptors but at different rates. It also does not take into account the origin of the enzyme or its three-dimensional structure.
A powerful, comprehensive classification of GT enzymes has been proposed by Henrissat and coworkers (http://www.cazy.org/). This classification scheme has been termed CAZy for carbohydrate active enzymes and delineates the enzymes into families based solely on amino acid similarity. This scheme has also been applied to the classification of other groups of enzymes, such as the glycosyl hydrolases, polysaccharide lyases, and carbohydrate esterases. These families are updated continually; currently, over 32,000 GT ORFs have been identified and classified into 90 families. Most enzymes in these families remain biochemically uncharacterized ORFs, but it has been demonstrated with both the GT and the glycosyl hydrolase families that this method of classification often leads to the grouping of enzymes with similar three-dimensional structures. However, as outlined in the structural fold section above, these three-dimensional structures do not necessarily translate into a predictable catalytic mechanism, substrate specificity, or the origin of the GT. For example, a few of the larger CAZy families, like GT2 with 8800 members, contain sequences that originate in bacteria, yeast, plant, viral, archaeal and animal species. This family also has at least 12 distinct GT activities, which include cellulose and chitin synthases, mannosyltransferases, rhamnosyltransferases and galactosyltransferases. Family 63 is mono-functional with a single entry for β-glucosyltransferase from bacteriophage T4. However, these families are not static. An example is the introduction of family GT78 when the three-dimensional structure of α-mannosylglycerate synthase (Mgs), from Rhodothermus marinus was reported (27). Despite the fact that this enzyme has high sequence similarity to members from GT2, a new family was generated because Mgs is a retaining enzyme whereas members of GT2 are typically inverting enzymes.
Glycosyltransferases in Biological Systems
The chemical diversity of glycosyltransferase reaction products is also reflected in their numerous biological roles (28). They serve as sources of energy exemplified by sucrose, lactose, and amylose (Fig. 1). They are also structural elements in cell walls and extracellular matrices. Oligosaccharides found on glycoproteins and glycolipids can be involved directly in the structure and the function of these molecules by affecting their stability, half-life, and activity within the cell. These oligosaccharides can also serve as receptors in biological recognition events such as signaling, development, and cell-cell adhesion or provide receptors for hormones, bacterial toxins and viruses. In bacteria and plants, oligosaccharides are also known to modulate the activity of secondary metabolites. Given these numerous biological roles of glycan products, it is important to understand the structure and the function of the GT enzymes that synthesize them. It is also pertinent to understand their overall function in the context of the biological system to which they belong. An overview of some important biological glycosylation reactions performed by GTs is described below. They are separated by taxonomy, but many occur throughout the kingdoms of life.
The best characterized eukaryotic glycosyltransferases are those from mammals. These GTs are important for the synthesis of N-/O-linked glycoproteins and glycolipids and are integral to the process of storing energy in the form of glycogen. Mammalian GTs use only nine nucleotide donor sugars, UDP- D-Galactose, UDP-D-Glucose, UDP-D-N-acetylglucosamine, UDP-D-Glucuronic acid, UDP-D-N-acetylglalactosamine, UDP-D-xylose, GDP-D-mannose, GDP-L-Fucose, or CMP-D-N-acetylneuraminic acid. Mammalian GTs also use dolichol-diphosphate-GlcNAc2Man9Glc3, dolichol-phosphate-mannose and dolichol-phosphate-glucose (Fig. 2).
In comparison with mammals, plants contain considerably more GTs because, in addition to the reactions carried out by mammalian GTs, they are required to convert the products of photosynthesis into diverse cell carbohydrates. For example, these GTs synthesize cell wall polysaccharides as well as secondary metabolites and xenobiotics. Plant GTs differ from mammalian GTs even more by their diversity of nucleotide donors. They use not only eight of the nine mammalian nucleotide donors, but numerous others such as UDP-L-rhamnose, GDP-L-glucose, GDP-L-galactose, UDP-L-arabinose, UDP-D-galacturonic acid, UDP-D-apiose, and so on. (29).
Despite the differences in roles, nucleotide donors, and number of GTs between eukaryotes, striking similarities exist between the processes by which these organisms synthesize glycan products. For example, the GTs responsible for the synthesis of N-/O-linked glycoproteins and glycolipids may differ in their substrate specificity, but they are all primarily located in the endoplasmic reticulum (ER) and/or Golgi apparatus. Furthermore, they follow similar steps for the assembly of their respective glycans. Although fewer in number, both plants and mammals do possess some important GTs that are not localized to the ER-Golgi synthesis pathways. Examples have been noted in both the cytoplasmic and the nucleoplasmic compartments as well as associated with the plasma membrane. To elaborate on some prominent GT reactions, the following descriptions have been separated based on their cellular location.
The GTs located in the ER-Golgi produce glycolipids and glycoproteins that are found on cell surfaces, in the insoluble extracellular matrix, or in fluids throughout the organism. These enzymes are usually type 2 transmembrane proteins composed of a short N-terminal cytoplasmic tail, a transmembrane domain, a stem region of variable length, and a large catalytic domain that faces the luminal side (30). Proteolysis in the stem regions can generate soluble forms that are found in fluids such as milk and serum. It should also be noted that a crossover exists between many of these GTs, which means that they can act on both lipids and proteins. Protein glycosylation events in the ER-Golgi are divided into two groups that consist of either Nor O-linked glycosylation.
Synthesis of N-linked glycoproteins begins by the precursor glycan first being assembled by GTs on dolichol-pyrophosphate. In mammals, successive addition of monosaccharides from UDP-GlcNAc, GDP-Man, dolichol-phosphate-Man, and dolichol-phosphate-Glc results in a 14-sugar chain product (GlcNAc2Man9Glc3) (31). The next step involves the activity of a membrane-associated enzyme complex, which is termed oligosaccharyl transferase (OST). This complex catalyzes the en bloc transfer of the assembled sugar chain from the dolichol-pyrophosphate donor to an asparagine residue of a nascent protein. OST specifically modifies surface-exposed asparagines within the consensus sequence Asn-Xaa-Ser or Asn-Xaa-Thr. After the en bloc transfer, the N-glycans are trimmed by ER glucosidases and ER-Golgi mannosidases and then converted to more complex structures by different GTs that use various donor sugars. By this process, it is possible for N-linked glycans to exhibit heterogeneity even on the same polypeptide. These glycans are called glycoforms, and well over 500 different structures have been chemically characterized (32). Among these glycans are the ABO blood group structures that consist of different glycan chain terminating sugars. In the case of the O blood group, the structures terminate in the H-antigen, which consists of Fuc-α1,2-Gal. The A and B blood groups are H-antigen structures that have been modified even more with either GalNAc or Gal, respectively, through the action of two separate but highly homologous enzymes (33). N-linked glycoprotein synthesis in plants is similar to mammals with the exception that the oligosaccharide structures contain α1,3-fucose linked to internal GlcNAc residues and β1,2 xylose linked to β-mannose (34).
The synthesis of glycolipids also occurs by successive GT reactions associated with the ER-Golgi. Two common classes of glycolipids synthesized by this method are the glycosphin-golipids and the glycophosphatidylinositols (GPI). Within the glycosphingolipid class, Glc-ceramide synthesis is initiated by the addition of β-Glc from a UDP donor by glucosylceramide synthase (35). For GPIs, the first sugar attached to the lipid is GlcNAc, that is deacetylated to GlcN (36). These initial steps occur on the cytoplasmic face of the ER followed by flipping to the luminal face. The GPIs are then conjugated to proteins through N-linkages, whereas glycosphingolipids are modified by the addition of Gal, GalNAc, and NeuAc by Golgi GTs. This method of glycosylation and flipping across a membrane has been noted in plants during galactolipid synthesis on the chloroplast’s thylakoid membrane. The GTs that form these galactolipids are particularly interesting because these galactolipids are the most abundant class of lipids in the biosphere (37).
Although still associated with the ER-Golgi, the synthesis of O-linked glycoproteins differs greatly from the N-linked processes described above. Typically O-linked glycosylation of proteins occurs on serine and threonine residues, but it has also been identified on hydroxylated lysines (38). Serine/threonine glycosylation is initiated in the Golgi by the action of retaining polypeptide GalNAc transferases (ppGalNAcTs, Fig. 4b) (21, 39). The ppGalNAcTs represent one of the largest human GT families because they account for over 20 of the 250 known human GTs. Unlike N-linked protein glycosylation, the specificity of O-GalNAcTs does not seem to be determined by a known consensus sequence. This enzyme family is rather unique among GTs, in possessing a C-terminal, ricin-type lectin domain (39). This additional domain is believed to endow certain ppGalNAcTs with an even greater capacity to adapt to and capture glycosylated substrates, which ensures the high density of glycosylation characteristic of mucin domains (39). After pp-GalNAcTs initiate glycosylation, other nucleotide GTs produce core 1 and core 2 glycan structures that can in turn be elaborated even more. These glycan structures impart unique physicochemical features to the proteins. For example, heavily glycosylated mucins facilitate the retention of water along environmentally exposed surfaces of the body.
Mammalian proteoglycans represent another class of heavily O-glycosylated proteins; however, they are synthesized by an alternative ER pathway. The protein attachment site for glycan synthesis initiation contains a Ser-Gly motif that is β-xylosylated by the action of a xylosyltransferase (40). The β-xyloside is extended by the action of additional galactosyl-transferases to produce a core common to proteoglycans. This core is decorated by GTs even more through the sequential addition of alternating sugars of disaccharide repeating units to produce heparin, keratin, dermatin, and chondroitin (41). The biosynthesis of fully glycosylated proteoglycans involves different GT reactions that results in a heterogeneous product. Despite this heterogeneity, the glycans all contain the reproducible structural elements that allow them to fulfill their functions, such as in load-bearing portions of the joint that release water slowly under pressure then reuptake water when pressure is reduced.
Apart from GTs associated with the ER-Golgi, eukaryotes also possess biologically significant GTs that are associated with the plasma membrane. In particular, cellulose synthase in plants is a complex of GTs that are responsible for the synthesis of the most abundant polymer on earth, cellulose. These GTs range in size from 988 to 1088 amino acids and have approximately eight transmembrane domains with a large central cytoplasmic domain (42). Research on cellulose synthase has focused on understanding how these enzymes associate and how they coordinate the processive addition of β1,4-glucose to a growing polymer. This research is complicated by the fact that the formation of cellulose depends on several noncellulosidic cell wall polysaccharides as well. Hemicellulose and pectin form the cell wall backbone on which cellulose is assembled, but these polysaccharides are synthesized within the Golgi by the activity of glycan synthases. Taking this into account, the synthesis of the final cell-wall polysaccharide product is astoundingly complex. Not only do they require the coordinated activity of over 50 GTs, but also the action takes place in two different cellular locations.
Given the above processes, it is evident that the synthesis of glycans and glycoconjugates are often associated with a membrane in some fashion. Presumably, this process helps to organize or sequester the correct GTs for a particular function. Some important glycosyltransferase reactions have also been noted in other areas of the cell. For example, the energy storage enzyme glycogen synthase and a superfamily of CAZy family 1 glycosyltransferases, which is involved in the biotransformation of drugs and xenobiotics via glucuronylation in mammals and glucosylation of small molecular weight acceptors in plants (44), are known to be located in the cytoplasm. The protein-modifying O-GlcNAc transferase is particularly interesting in that it uses UDP-GlcNAc to glycosylate serine or threonine residues on specific proteins. Glycosylation of these target proteins is balanced by the action of a soluble N-acetylglucosaminidase and seems to be reciprocal with protein phosphorylation. This dynamic balance in glycosylation levels is predicted to be involved in numerous processes, which include glucose metabolism, chaperone folding, nuclear pore protein translocation, and transcription factor regulation (43). The importance of O-GlcNAc transferase is clearly apparent not only because of its numerous activities in the cell, but also because of its discovery within the nucleus and cytoplasm of all metazoans.
Glycans in prokaryotic cells are integral to several cellular roles. Peptidoglycan is a determinant of cell shape and helps eubacteria withstand the pressures of the external environment. Lipopolysaccharide (LPS), lipooligosaccharides (LOS), capsules, and slime layers serve as attachment sites/receptors for cellular interactions, provide protection from environmental factors and are involved in immune system modulation/evasion. In addition to these cell wall polysaccharides, the list of other glycosylated natural products, which include antibiotics and other xenobiotics, continues to grow. In fact, the only GTs that prokaryotes were believed to lack were those for the synthesis of glycoproteins. However, this group has recently been revised given increasing demonstrations of the structure, function, and biosynthesis of glycoproteins in prokaryotes (45-47). The diversity of prokaryotic carbohydrate structures, and the corresponding GTs that synthesize them, is overwhelming. Moreover, striking similarities seem to exist between the glycosylation events of the prokaryotes and the eukaryotes. For example, both kingdoms use nucleotide-donor sugars to assemble oligosaccharide chains, trimming reactions are present in both, and they both use lipid-bound intermediates. Studies with S-layer glycoproteins have also demonstrated that prokaryotes are O-glycosylated at serine/threonine residues and N-glycosylated at asparagines residues (45).
Despite their similarities, glycosylation events in prokaryotes and eukaryotes have critical differences. The absence of intracellular organelles means that polymerization of prokaryotic oligo/polysaccharides takes place entirely in the cytoplasm or in controlled extracellular reactions. It is important to note that the extacellular glycosyltransferase reactions must be coordinated tightly given the lack of a membrane to prevent the loss of substrates/products combined with the lack of nucleotide donors to fuel the reaction process. The following description outlines the processes by which prokaryotic GTs synthesize carbohydrates destined for either protein glycosylation or macromolecular structures. These polysaccharides are typically synthesized by similar mechanisms. However, it should be noted that in bacteria, N-glycosylation occurs independently of protein translocation (47). Biosynthesis begins in the cytoplasm on a lipid-linked carrier. Undecaprenyl is used in the case of peptidoglycan and LPS, whereas a C60-polyprenol is used for protein glycosylation (45-48). Sugars are added to these carriers by successive transfers of either monomeric sugars from nucleotide precursors or by en bloc transfer of assembled oligosaccharides from other lipid-monophosphate carriers. The sialyltransferases, CstI and CstII from C. jejuni, are examples of GTs involved in this process. These enzymes aid in the production of LPS with mammalian-like structures which help the bacterium mimic host cell polysaccharides for the purpose of evasion/suppression of the host immune response (7, 8). When the oligo/polysaccharide precursors are assembled, they either remain in the cytoplasm or are transported across the membrane(s). However, the mechanism of transfer across the membrane(s) is not understood clearly. At least for capsular and lipopolysaccharides from E. coli and Salmonella enterica, it seems to involve the coordinated process of several membrane spanning and/or ABC transporter proteins (48, 49). Electron microscopy and X-ray crystallography studies have allowed the visualization of the core Wza-Wzc complex for capsular polysaccharide export in E. coli (49). When on the outer-surface of these cells, the oligo/polysaccharides are released into the external environment or transferred to their target proteins/macromolecular structures by the action of external glycosytransferases. The recently solved three-dimensional structure of the peptidoglycan glycosyltransferase domains of the penicillin-binding protein (18, 26) represent these final GTs that are involved in peptidoglycan biosynthesis. As for LPS, it seems that WaaL, which is a ligase enzyme, may be the only enzyme required for the transfer the O-polysaccharide to the Lipid A-core molecule (48).
The number of GT-encoded genes in viruses is relatively low compared with the other taxonomic groups discussed. However, they do have four prominent roles that contribute to the survival of viruses. First, some bacteriophages can glycosylate their own DNA to avoid host restriction endonucleases. This function became particularly apparent when the first structurally solved GT, T4 bacteriophage-encoded P-glucosyltransferase (22) was crystallized more recently in the presence of a short DNA segment (50). An a-glucosyltransferase with the same fold and similar function has since been described in the same T4 virus (51). Second, viral GTs can modify host structures that result in evasion of the immune response or promoting the transmissibility of the released virons. Modification of host structures in bacteria has been reported to aid in serotype conversion and immunity to infection by other viruses. This phenomena has been studied extensively with respect to LPS modification in Salmonella and Shigella bacterial species (reviewed by Reference 52). Third, viral GTs can alter host metabolisms to promote the release of increased numbers of progeny viruses. This theory is perhaps best exemplified by the modification of ecdysteroid by a baculovirus-encoded glucosyltransferase. Ecdysteriod is an insect hormone involved in the development of Lepidoptera sp. from the larval to the molting or pupating stages. Inactivation of ecdysteriod through the addition of glucose by virally-encoded UDP-glucosyltransferase, EGT, prevents this development. It allows the virus to monopolize insect resources and facilitates easier spread of the virus when the larva disintegrate (reviewed by Reference 52). Last, some viruses possess GTs that assemble unique virus-encoded products that aid in immune evasion or infection of surrounding hosts. Typically, viral glycoproteins are produced by “hijacking” the host glycosylation pathways associated with the ER-Golgi apparatus. However, the synthesis of the major capsid protein of the Chlorella virus has been found to differ from this paradigm. Paramecium bursaria chlorella virus-1 encodes at least five putative GTs, of which one of the gene products, A64R, has been characterized structurally by X-ray crystallography (53). Four of the five GTs are predicted to be localized in the cytoplasm, whereas the last GT is membrane associated. The coordinated activity of these GTs is proposed to assemble the major capsid protein, Vp54, which is independent of the host ER-Golgi apparatus.
Virus-encoded GTs may actually be more common than believed previously because the list of known viral GTs is far from being complete. Viruses are constantly coevolving with their hosts and many can extract genes, which include those for glycosyltransferases, from the host genome for incorporation into their own. If they can confer a selective advantage to the virus, they could be incorporated permanently into their genomes, which lead to new virally encoded GT variants.
Although many important GTs in glycobiology have been discussed, a need exists for even more exploration of these enzymes, because specific roles are often not defined. Furthermore, it is increasingly apparent that many GTs are related directly to numerous acquired and inherited diseases. For example, the potent B cytotoxin from Clostridium difficile is a GT that inactivates Rho-GTP through glucosylation (54). This GT leads to diarrhea, inflammation, and damage to colonic mucosa. In inherited diseases, aberrant glycosylation of O-linked mannose and GlcNAc on dystroglycan can lead to Walker-Warburg syndrome or muscle-eye-brain disease, respectively (55). Sequence variations in the coding genes of xylosyltransferases, XT-I and XT-II, have also been demonstrated to be responsible for altered proteoglycan metabolism. These variations have thus been identified as risk factors for diabetic nephropathy, osteoarthritis, or pseudoxanthoma elasticum (56). Although it is certainly an abbreviated list, these findings point to the important role of these GTs as disease modifiers in several different pathologies. Therefore from a disease standpoint, it is prudent that we continue to outline the roles for existing and unknown GTs within a biological context. Then, the potential to cure these diseases, by genetic or drug administered therapy, can be assessed properly.
GTs are also promising for the generation of pharmaceutically relevant glycoconjugates. Indeed, the discovery of the catalytic reversibility of AveB from Streptomyces avermitilis proved to be useful because it led to the production of a variety of avermectin variants that may have commercial use in controlling nematodes, insects, and arachnids (4). These results are particularly intriguing given the recently solved three-dimensional structures of two other antibiotic synthesis enzymes: the vancomycin and oleandomycin GTs (16, 24) (Fig. 4d). The characterization of new GTs also holds medical relevance as a way to greatly expand existing candidates for vaccine development, because many of these are known to be natural glycoconjugates synthesized by GTs.
Characterization of GTs also has industrial implications. Glycosylation of biopharmaceuticals such as antibodies can have profound effects on their half-life, stability, and activity, which makes GTs attractive tools for engineering proteins that are ideal for prolonged or specialized industrial processes as exemplified in the production of human glycoproteins in yeast (57). Furthermore, the incorporation of several GTs in a single reaction mixture can lead to the synthesis of many and often complex polysaccharides as opposed to lengthy and laborious chemical methods that are less environmentally benign. This mechanism includes the use of metabolically engineered bacteria for large-scale synthesis of complex glycans (58, 59).
Ultimately, all processes would be enhanced greatly if the precise catalytic mechanisms of GTs were understood better. To accomplish this task, new enzyme inhibitors and/or substrate analogs are needed. Elucidation of these mechanisms will certainly not only lend insight into the structure-function relationship of these enzymes, but also provide an increased knowledge of what makes GTs uniquely suited to perform certain functions.
1. Breton C, Snajdrova L, Jeanneau C, Koca J, Imberty A. Structures and mechanisms of glycosyltransferases. Glycobiology 2006; 16:29-37.
2. Hu Y, Walker S. Remarkable structural similarities between diverse glycosyltransferases. Chem. Biol. 2002; 9:1287-1296.
3. Taniguchi N, Honke K, Fukuda M. Handbook of Glycosyltransferases and Related Genes. 2002. Springer, Tokyo.
4. Zhang C, Albermann C, Fu X, Thorson JS. The in vitro characterization of the iterative avermectin glycosyltransferase AveBI reveals reaction reversibility and sugar nucleotide flexibility. J. Am. Chem. Soc. 2006; 128:16420-16421.
5. Ramakrishnan B, Balaji PV, Qasba PK. Crystal structure of β1,4-galactosyltransferase complex with UDP-Gal reveals an oligosaccharide acceptor binding site. J. Mol. Biol. 2002; 318:491-502.
6. Boix E, Zhang Y, Swaminathan GJ, Brew K, Acharya KR. Structural basis of ordered binding of donor and acceptor substrates to the retaining glycosyltransferase, α-1,3-galactosyltransferase. J. Biol. Chem. 2002; 277:28310-28318.
7. Chiu CPC, Lairson LL, Gilbert M, Wakarchuk WW, Withers SG, Strynadka NCJ. Structural analysis of the α-2,3-sialyltransferase Cst-I from Campylobacter jejuni in apo and substrate-analogue bound forms. Biochemistry 2007; 46:7196-7204.
8. Chiu CPC, Watts AG, Lairson LL, Gilbert M, Lim D, Wakarchuk WW, Withers SG, Strynadka NCJ. Structural analysis of the sialyltransferase Cstll from Campylobacter jejuni in complex with a substrate analog. Nat. Struct. Mol. Biol. 2004;11:163-170.
9. Pederson LC, Darden TA, Negishi, M. Crystal structure of β1,3-glucuronyltransferase I in complex with active donor substrate UDP-GlcUA. J. Biol. Chem. 2002; 277:21869-21873.
10. Persson K, Ly HD, Dieckelmann M, Wakarchuk WW, Withers SG, Strynadka NCJ. Crystal structure of the retaining galactosyltransferase LgtC from Neisseria meningitidis in complex with donor and acceptor sugar analogs. Nat. Struct. Biol. 2001; 8:166-175.
11. Negishi M, Dong J, Darden TA, Pedersen LG, Pedersen LC. Glucosaminylglycan biosynthesis: what we can learn from the X-ray crystal structures of glycosyltransferases GlcAT1 and EXTL2. Biochem. Biophys. Res. Commun. 2003; 303:393-398.
12. Tvaroska I. Molecular modeling insights into the catalytic mechanism of the retaining galactosyltransferase LgtC. Carbohydr. Res. 2004; 339:1007-1014.
13. Gibson RP, Tarling CA, Roberts S, Withers SG, Davies GJ. The donor subsite of trehalose-6-phosphate synthase. J. Biol. Chem. 2004; 279:1950-1955.
14. Klein HW, Im MJ, Palm D. Mechanism of the phosphorylase reaction. Utilization of D-gluco-hept-1-enitol in the absence of primer. Eur. J. Biochem. 1986; 157:107-114.
15. Bourne Y, Henrissat B. Glycoside hydrolases and glycosyltransferases: families and functional modules. Curr. Opin. Struct. Biol. 2001; 11:593-600.
16. Bolam DN, Roberts S, Proctor MR, Turkenburg JP, Dodson EJ, Martinez-Fleites C, Yang M, Davis BG, Davies GJ, Gilbert HJ. The crystal structure of two macrolide glycosyltransferases provides a blueprint for host cell antibiotic immunity. Proc. Natl. Acad. Sci. U.S.A. 2007; 104:5336-5341.
17. Ihara H, Ikeda Y, Toma S, Wang X, Suzuki T, Gu J, Miyoshi E, Tsukihara T, Honke K, Matsumoto A, Nakagawa A, Taniguchi N. Crystal structure of mammalian α1,6-fucosyltransferase, FUT8. Glycobiology 2007; 17:455-466.
18. Lovering AL, de Castro LH, Lim D, Strynadka NCJ. Structural insights into the translycosylation step of bacterial cell-wall biosynthesis. Science 2007; 315:1402-1405.
19. Charnock SJ, Davies GJ. Structure of the nucleotide-diphospho-sugar transferase, SpsA from Bacillus subtilis in native and nucleotide-complexed forms. Biochemistry 1999; 38:6380-6385.
20. Qasba PK, Ramakrishnan R, Boeggeman E. Substrate-induced conformational changes in glycosyltransferases. Trends Biochem Sci. 2005; 30:53-62.
21. Kubota T, Shiba T, Sugioka S, Furukawa S, Sawaki H, Kato R, Wakatsuki S, Narimatsu H. Structural basis of carbohydrate transfer activity by human UDP-GalNAc: polypeptide α-.-acetyl- galactosaminyltransferase (pp-GalNAc-T10). J. Mol. Biol. 2006; 359:708-727.
22. Vrielink A, Ruger W, Driessen HP, Freemont PS. Crystal structure of the DNA modifying enzyme JYglucosyltransferase in the presence and absence of the substrate uridine diphosphoglucose. EMBO J. 1994; 13:3413-3422.
23. Ha S, Walker D, Shi Y, Walker S. The 1.9 A crystal structure of Escherichia coli MurG, a membrane-associated glycosyl-transferase involved in peptidoglycan biosynthesis. Protein Sci. 2000; 9:1045-1052.
24. Mulichak AM, Losey HC, Walsh CT, Garavito RM. Structure of the UDP-glucosyltransferase GtfB that modifies the heptapeptide aglycone in the biosynthesis of vancomycin group antibiotics. Structure 2001; 9:547-557.
25. Hu Y, Chen L, Ha S, Gross B, Falcone B, Walker D, Mokhtarzadeh M, Walker S. Crystal structure of the MurG: UDP-GlcNAc complex reveals common structural principles of a superfamily of glycosyltransferases. Proc. Natl. Acad. Sci. U.S.A. 2003; 100:845-849.
26. Yuan Y, Barrett D, Zhang Y, Kahne D, Sliz P, Walker S. Crystal structure of a peptidoglycan glycosyltransferase suggests a model for processive glycan chain synthesis. Proc. Natl. Acad. Sci. U.S.A. 2007; 104:5348-5353.
27. Flint J, Taylor E, Yang M, Bolam DN, Tailford LE, Martinez-Fleites C, Dodson EJ, Davis BG, Gilbert HJ, Davies GJ. Structural dissection and high-throughput screening of mannosylglycerate synthase. Nat. Struct. Mol. Biol. 2005; 12:608-614.
28. Varki A. Biological roles of oligosaccharides: all of the theories are correct. Glycobiology 1993; 3:97-130.
29. Seifert GJ. Nucleotide sugar interconversions and cell wall biosynthesis: how to bring the inside to the outside. Curr. Opin. Plant Biol. 2004; 7:277-284.
30. Colley KJ. Golgi localization of glycosyltransferases: more questions than answers. Glycobiology 1997; 7:1-13.
31. Silberstein S, Gilmore R., Biochemistry, molecular biology, and genetics of the oligosaccharyltransferase. FASEB J. 1996; 10:849-858.
32. Kornfeld R, Kornfeld S. Assembly of asparagine-linked oligosaccharides. Annu. Rev. Biochem. 1985; 54:631-664.
33. Patenaude SI, Seto NOL, Borisova SN, Szpacenko A, Marcus SL, Palcic MM, Evans SV. The structural basis for specificity in human ABO(H) blood group biosynthesis. Nat. Struct. Biol. 2002; 9:685-690.
34. Wilson IBH. Glycosylation of proteins in plants and invertebrates. Curr. Opin. Struct. Biol. 2002; 12:569-577.
35. Jennemann R, Sandhoff R, Langbein L, Kaden S, Rothermel U, Gallala H, Sandhoff K, Wiegandt H, Gorne HJ. Integrity and barrier function of the epidermis critically depend on glucosyl- ceramide synthesis. J. Biol. Chem. 2007; 282:3083-3094.
36. Kinoshita T, Ohishi K, Takeda J. GPI-anchor synthesis in mammalian cells: genes, their products and deficiency. J. Biochem. 1997; 122:251-257.
37. Benning C, Ohta, H. Three enzyme systems for galactoglycerolipid biosynthesis are coordinately regulated in plants. J. Biol. Chem. 2005; 280:2397-2400.
38. Myllyla R, Wang C, Heikkinen J, Juffer A, Lampela O, Risteli M, Ruotsalainen H, Salo A, Sipila L. Expanding the lysyl hydroxylase toolbox: new insights into the localization and activities of lysyl hydroxylase 3 (LH3). J. Cell. Physiol. 2007; 212:323-329.
39. Fritz TA, Raman J, Tabak LA. Dynamic association between the catalytic and lectin domains of human UDP-GalNAc: polypeptide α-N-acetylgalacosaminyltransferase-2. J. Biol. Chem. 2006; 281:8613-8619.
40. Wilson IBH. The never-ending story of peptide O-xylosyltransferase. Cell. Mol. Life Sci. 2004; 61:794-809.
41. Pedersen LC, Tsuchida K, Kitagawa H, Sugahara K, Darden TA, Negishi M. Heparan/chondroitin sulfate biosynthesis. Structure and mechanism of human glucuronyltransferase I. J. Biol. Chem. 2000; 275:34580-34585.
42. Somerville C. Cellulose synthesis in higher plants. Annu. Rev. Cell Dev. Biol. 2006; 22:53-78.
43. Zachara NE, Hart GW. Cell signaling, the essential role of O-GlcNAc! Biochim. Biophys. Acta 2006; 1761:599-617.
44. Lim EK, Baldauf S, Li Y, Elias L, Worrall D, Spencer SP, Jackson RG, Taguchi G, Ross J, Bowles DJ. Evolution of substrate recognition across a mutigene family of glycosyltransferases in Arabidopsis. Glycobiology 2003; 13:139-145.
45. Upreti RK, Kumar M, Shankar V. Bacterial glycoproteins: functions, biosynthesis and applications. Proteomics 2003; 3:363-379.
46. Weerapana E, Imperiali B. Asparagine-linked protein glycosylation: from eukaryotic to prokaryotic systems. Glycobiology 2006; 16:91-101.
47. Kowarik M, Numao S, Feldman MF, Schulz BJ, Callewaert N, Kiermaier E, Catrein I, Aebi M. N-linked glycosylation of folded proteins by the bacterial oligosaccharyltransferase. Science 2006; 314:1148-1150.
48. Raetz CRH, Whitfield C. Lipopolysaccharide endotoxins. Annu. Rev. Biochem. 2002; 71:635-700.
49. Collins RF, Beis K, Dong C, Botting CH, McDonnell C, Ford RC, Clarke BR, Whitfield C, Naismith JH. The 3D structure of a periplasm-spanning platform required for assembly of group 1 capsular polysaccharides in Escherichia coli. Natl. Acad. Sci. U.S.A. 2007; 104:2390-2395.
50. Lariviere L, Morera S. A base-flipping mechanism for the T4 phage β-glucosyltransferase and identification of a transition-state analog. J. Mol. Biol. 2002; 324:483-490.
51. Lariviere L, Sommer N, Morera S. Structural evidence of a passive base-flipping mechanism for AGT, an unusual GT-B glycosyltransferase. J. Mol. Biol. 2005; 352:139-150.
52. Markine-Goriaynoff N, Gillet L, Van Etten JL, Korres H, Verma N, Vanderplasschen A. Glycosyltransferases encoded by viruses. J. Gen. Virol. 2004; 85:2741-2754.
53. Zhang Y, Xiang Y, Van Etten JL, Rossmann MG. Structure and function of a chlorella virus-encoded glycosyltransferase. Structure 2007; 15:1031-1039.
54. Reinert DJ, Jank T, Aktories K, Schulz GE. Structural basis for the function of Clostridium difficile toxin B.J. Mol. Biol. 2005; 351:973-981.
55. Vajsar J, Schachter H. Walker-Warburg syndrome. Orphanet. J. Rare Dis. 2006; 1:29.
56. Gotting C, Kuhn J, Kleeseik K. Human xylosyltransferases in health and disease. Cell. Mol. Life Sci. 2007; 64:1498-517.
57. Hamilton SR, Bobrowicz P, Bobrowicz B, Davidson RC, Li H, Mitchell T, Nett JH, Rausch S, Stadheim TA, Wischnewski H, Wildt S, Gerngross TU. Production of complex human glycoproteins in yeast. Science 2003; 301:1244-1246.
58. Antoine T, Priem B, Heyraud A, Greffe L, Gilbert M, Wakarchuk WW, Lam JS, Samain, E. Large-scale in vivo synthesis of the carbohydrate moieties of gangliosides GM1 and GM2 by metaboli-cally engineered Escherichia coli. ChemBioChem. 2003; 4:406-412.
59. Endo T, Koizumi S. Large-scale production of oligosaccharides using engineered bacteria. Curr. Opin. Struct. Biol. 2000; 10:536-541.
Taylor ME, Drickhamer K. Introduction to Glycobiology. 2nd edition. 2006. Oxford University Press Inc., New York.
Glycan Biosynthesis in Mammals
Glycan, Synthesis, Key Reactions of,
Glycolipids, Synthesis of