The last 10 years has witnessed enormous advances in the structural enzymology of enzymatic glycosyl transfer. Developments in molecular biology, coupled with the possibility for “rapid-throughput” protein crystallography have led to a plethora of 3-D structures for glycoside hydrolases. Synergistic application of synthetic chemistry to structural biology has led to the development of numerous oligosaccharide mimics and mechanistic probes. Underpinning much of this work is a classification of carbohydrate-active enzymes (“CAZymes”) which has now been extended to include glycoside hydrolases, carbohydrate esterases, polysaccharide lyases and glycosyltransferases, together with their associated non-catalytic modules. Recently, there have been exciting developments in our understanding of the structures and catalytic mechanisms involved in the synthesis of glycosidic bonds both by Nature's own catalytic apparatus and by mutant glycoside hydrolases termed “glycosynthases”. Here we review those developments including the structure of the Bacillus subtilis SpsA, a “family GT-2” glycosyltransferase, and discuss its relevance to the synthesis of biopolymers such as cellulose, chitin, hyaluronan and the Nod factors.
Glucuronyltransferase I (GIcAT-I) is a key enzyme in heparan/chondroitin biosynthesis (1). The X-ray crystal structure of human GlcAT-1 has been solved in the presence of both UDP and substrate analog. The structure reveals a two subdomain structure known as the SGC domain. Donor substratebinding site resides in the N-terminal sub-domain, while acceptor substrate-binding site is located in the C-terminal sub-domain. In addition to conserved residues responsible for the donor binding, various residues that interact with the acceptor molecule have now been identified. The GlcAT-1 provides the structural basis for understanding the structure and function of glycosyltransferases.
Galactosyltransferases are among the most represented enzymes in the glycosyltransferases superfamily. Galactosyltransferases bind UDP-α-D-galactose as the donor substrate and transfer the galactose to acceptor substrates that are as different as glycoproteins, glycosaminoglycans, glycolipids and, to some extent, small lipophilic molecules such as plant hormones. For the past decade, mammalian galactosyltransferases have been the center of interest for glycobiologists focused primarily on the discovery of their genes and on the characterization of their enzyme activity. In 1999, the first picture of the crystallographic structure of bovine β4GalT1 catalytic domain was revealed at 2.4Å resolution. This study was the beginning of a new structural approach to glycosyltransferases that focused on the description of their three-dimensional structure at atomic resolution with the objective of interpreting former enzymatic properties and understanding their reaction mechanism. This minireview is an attempt to describe the current status of the galactosyltransferase structure-function relationship in the context of the glycosyltransferase superfamily. The crystallographic structures of two bovine galactosyltransferases catalytic domains, those of β4GalT1 and α3GalT, will be discussed as well as their reaction mechanism.
In the past few years, rapid advances have been made in sequencing the genomic DNA of human, Caenorhabditis elegans, and so on. As a result, a large number of novel glycosyltransferase genes have been discovered from those genome sequences. How did they increase their family members during the genome evolution? To presume the evolutionary pathway of glycosyltransferases, we have used the molecular evolutionary analysis (1). In that study, we conducted molecular evolutionary analyses on 55 glycosyltransferase genes and mainly discussed about glycosyltransferase genes for N- or O-glycan synthesis. The phylogenetic trees showed the glycosyltransferase genes increased their numbers through gene duplications. We also estimated the divergence time of each branch root and suggested that the glycosyltransferase genes increased their numbers through gene duplications and genome duplications. Comparison of evolutionary rates indicated that the glycosyltransferases tend to evolve more slowly than other genes, and the evolutionary rates changed within each of the glycosyltransferase gene families. These results indicate that the increase in glycosyltransferase genes allows the amino acid change and permits, the creation of the variety of specific activity of the enzyme. Here, we would like to introduce the essence of the evolutionary history of glycosyltransferase genes.
The Notch protein is a cell surface receptor that plays a key role in a number of developmental cascades. Its extracellular domain is composed of 36 tandem epidermal growth factor-like repeats, and recent work has demonstrated that many of these repeats are modified with two unusual forms of glycosylation: O-fucose and O-glucose. The large number of consensus sites for these types of glycosylation and the conservation of those sites across species suggest that the sugars play an important role in Notch function. A clue to the function of the O-fucose modifications was provided by the demonstration that the Fringe protein is a β1, 3-N- acetylglucosaminyltrans-ferase that modifies the O-fucose residues on Notch. Fringe is a known modifier of Notch function, altering the response of the receptor to its ligands. These results provide a clear example of how alteration of a specific carbohydrate structure on a specific protein mediates a specific biological event.
β1, 4-N-Acetylglucosaminyltransferase III (GnT-III) is known to be a key glycosyltransferase which plays an important role in regulating the biosynthesis of Asn-linked oligosaccharides on glycoproteins. The regulatory role of the enzyme is based on effects of a reaction product, namely the bisecting GlcNAc structure, on the biosynthetic process. This unique structure is not tolerated by other enzymes involved in the formation of the core structures, and, as a result, prevents further reactions which are catalyzed by these enzymes. This inhibitory regulation is the result of the broad specificity of GnT-III, as well as the properties of the bisecting GlcNAc. The overexpression and ectopic expression of GnT-III lead to a variety of significant alterations in the cellular functions. Although it is not known, except for a few cases, whether the direct involvement of the bisecting GlcNAc residue or the inhibition of the synthesis of a biologically important structure of the sugar chain results in these alterations, it seems certain that marked structural changes by GnT-III catalysis result in the biological alterations in the cells. These findings suggest that GnT-III and the bisecting GlcNAc play an important role in cellular functions and that N-glycans are associated with a variety of biological events.
Poly-N-acetyllactosamine is a unique carbohydrate composed of N-acetyllactosamine (LacNAc) repeats and provides the backbone structure for additional modifications such as sialyl Lewisx, It is attached to N-glycans, O-glycans, and glycolipids and synthesized by the alternate addition of β1, 3-linked N-acetylglucosamine (GlcNAc) and β1, 4-linked galactose (Gal) by i-β1, 3-N-acetylglucosaminyltransferase (iGnT) and a member of the β1, 4-galactosyltransferase (β4Gal-T) gene family. Poly-N-acetyllactosamines in mucin-type O-glycans can be formed in core 2- and core 4-branched oligosaccharides, which are synthesized by core 2 β1, 6-N-acetylglucosaminyltransferase (C2GnT) and core 4 β1, 6-N-acetylglucosaminyltransferase (C4GnT), respectively. β4Gal-TIV was found to be most efficient in the addition of a single Gal residue to core 2-branched oligosaccharides among the members of the β4Gal-T gene family and to synthesize poly-N-acetyllactosamine in core 2-branched O-glycans together with iGnT. On the other hand, β4Gal-TI was shown to be most efficient for poly-N-acetyllactosamine synthesis in N-glycans. In contrast to β4Gal-TI, the efficiency of β4Gal-TIV decreases dramatically as the acceptors contain more LacNAc repeats, consistent with the fact that core 2-branched O-glycans contain shorter poly-N-acetyllactosamines than N-glycans in many cells. Poly-N-acetyllactosamines in core 4-branched O-glycans were found to be synthesized most efficiently by iGnT and β4Gal-TI although the synthesis in core 4 branches is less efficient than in core 2 branches because of inefficient addition of GlcNAc to core 4 branches by iGnT. Thus, poly-N-acetyllactosamine extension in core 2- and core 4-branched O-glycans is differentially controlled by iGnT and different members of the β4Gal-T gene family.
Over the past decade glucuronyltransferases (GlcATs) involved in the biosynthesis of the HNK-1 carbohydrate epitope have been identified and cloned successively. The structure of the HNK-1 carbohydrate epitope is characterized by terminal sulfated glucuronic acid, and GlcATs are thought to be the key enzyme in the biosynthesis pathway of this epitope. Following isolation of several GlcAT cDNAs, structural comparison of GlcATs has proposed a new glucuronyltransferase family characterized by four conserved modules in the catalytic region. In mammals GlcAT-P and GlcAT-S are involved in the biosynthesis of the HNK-1 epitope, and their expression were mainly localized in the neural tissues. These two enzymes however expressed in the different region of the neural tissues and exhibit different acceptor specificity. Cloned cDNAs of these GlcATs are able to induce expression of the LINK-1 epitope on the surface of HNK-1 negative cells. GlcATs are powerful tools for investigating the biological function of the HNK-1 carbohydrate epitope.