Carbohydrates in nature adopt incredibly diverse structures that vary in their stereochemistry, ring structures, linkages, and adornment by a range of functional groups. The diversity of carbohydrates is controlled by a palette of enzymes involved in their assembly, modification, and breakdown. The formation of glycosidic bonds is catalyzed by enzymes including glycosyltransferases, transglycosidases, and phosphorylases. The cleavage of glycosidic bonds is catalyzed by enzymes including glycoside hydrolases (glycosidases) and the closely related transglycosidases, lyases, phosphorylases, and lytic polysaccharide monooxygenases. Carbohydrate structures can be modified by a wide range of transferases and hydrolases that install, rearrange, or remove adorning groups such as sulfates, phosphates, acyl groups, and ethers. Various redox-active enzymes act on carbohydrate structures to alter their oxidation states at key sites. Carbohydrate epimerases catalyze changes in stereochemistry at specific sites in a carbohydrate structure.
Formation and Cleavage of the Glycosidic Bond
Leloir and non-Leloir glycosyltransferases catalyze glycosyl transfer with either net retention or inversion of anomeric stereochemistry. Structural and kinetic data for inverting glycosyltransferases are consistent with a one-step mechanism that occurs by a direct substitution reaction at the anomeric center and proceeds through an oxocarbenium ion-like transition state, with general base catalysis most commonly by enzymatic histidine or carboxylate residues and possibly amide groups. For retaining glycosyltransferases, two mechanistic alternatives are possible. A two-step mechanism involving sequential inversions by first an enzymatic nucleophile and then an alcohol or other external nucleophile and an SNi mechanism that involves the non-concerted loss of the leaving group and same-side attack by an alcohol.
Phosphorylases and Pyrophosphorylases
Phosphorylases cleave glycosidic bonds by reaction with phosphate (Fig. 1b). This is a reversible reaction under physiological conditions, which can either synthesize (from sugar-1-phosphates) or cleave glycosidic bonds (to form sugar-1-phosphates). Phosphorylases may utilize glycosyltransferase or glycoside hydrolase-type mechanisms (see below), involving either retention or inversion of anomeric stereochemistry. Glycoside hydrolase-type phosphorylases (e.g., maltose phosphorylase) operate through mechanisms and catalytic residues that are conserved with glycoside hydrolases, although typically without the need for a general acid residue to assist in departure of the phosphate group, and are sequence- and structure-related to glycoside hydrolases. Glycosyltransferase-like phosphorylases are sequence- and structure-related to glycosyltransferases. One notable glycosyltransferase-like phosphorylase is glycogen phosphorylase, which uses a pyridoxal phosphate cofactor that is believed to be a vestigial mimic of a nucleoside phosphate of an evolutionary precursor glycosyltransferase. Pyrophosphorylases, like phosphorylases, are reversible enzymes, which can form glycosidic bonds by reaction of sugar-1-pyrophosphates with a nucleophilic alcohol.
Glycoside Hydrolases and Transglycosidases
Glycoside hydrolases (glycosidases) catalyze the cleavage of glycosides through C-O scission of the exocyclic glycosidic bond (Sinnott 1990) (Fig. 1c,d). Glycoside hydrolases may be classified on the basis of whether the reaction occurs with retention or inversion of anomeric stereochemistry. After enzymatic cleavage, the sugar hemiacetal mutarotates to form a mixture of anomers. Glycoside hydrolase specificity is defined by the following characteristics: (a) the anomeric specificity (α or β), (b) the stereochemical outcome (retaining or inverting), (c) whether cleavage occurs within (endo-) or at the nonreducing end (exo-) of the sugar chain, and (d) the nature of the sugar that is cleaved.
Inverting glycoside hydrolases operate through a one-step reaction in which a catalytic dyad of enzymatic carboxylate residues act as a general acid and base, respectively, to deprotonate a nucleophilic water molecule and protonate the departing glycosidic oxygen, leading to a nucleophilic substitution reaction with inversion of anomeric stereochemistry. The general acid and base residues of inverting glycoside hydrolases are typically carboxylate groups as part of glutamate and aspartate residues.
Retaining glycoside hydrolases operate using a two-step mechanism through a covalent intermediate. Two main classes are known. The first utilizes a dyad of enzymatic residues, with one acting as a general acid/base and the other as a nucleophile. In the first step, the general acid promotes the departure of the anomeric leaving group, while the nucleophilic residue performs a substitution reaction at the anomeric center resulting in inversion of configuration. In the second step, the now deprotonated acid/base residue acts as a general base to deprotonate a water molecule and promote a second nucleophilic substitution, now by water, leading to an overall net retention of anomeric configuration. The second class of retaining glycoside hydrolases utilizes a variant of this reaction involving neighboring group participation. In this mechanism a nucleophilic site on the substrate acts as the nucleophile in the first step, forming a covalent intermediate which is not attached to the enzyme. In the second step this intermediate is hydrolyzed, in an overall net retention of anomeric configuration. Examples of retaining glycoside hydrolases that use a neighboring group participation mechanism include those with substrates bearing a 2-acetamido group, such as β-N-acetylhexosaminidases, chitinases, and hyaluronidases.
Most commonly, the enzymatic nucleophile and acid/base of retaining glycoside hydrolases are carboxylate residues (aspartate and glutamate), but variations are seen. Myrosinases, which are retaining β-glucoside hydrolases that act on glucosinolates, lack an enzymatic acid/base; the substrate bears a good leaving group that does not require acid assistance to react in the first step; in the second step ascorbate acts as a cofactor in the role of general base. In retaining sialidases, the enzymatic nucleophile is a tyrosine. It has been proposed that a neutral tyrosine residue avoids the potentially destabilizing charge repulsion that would ensue between the anionic α-keto nonulopyranosidonate substrate and a carboxylate nucleophile. For both retaining and inverting glycoside hydrolases, there is strong evidence from kinetic isotope effect and linear free energy studies that the transition states possess substantial oxocarbenium ion character, with little participation by the nucleophile and nucleofuge.
A distinct set of retaining glycoside hydrolases operate through a nonclassical mechanism involving an NAD+ cofactor. This mechanism involves a three-step oxidation-elimination-reduction sequence that proceeds via anionic transition states. Initially, abstraction of hydride from C3 leads to oxidation to a 3-keto sugar (with concomitant reduction of the cofactor to NADH). Next, deprotonation of C2 facilitates an elimination reaction and forms an α,β-unsaturated intermediate. Conjugate addition of water at C1 of the α,β-unsaturated intermediate generates a sugar hemiacetal, and finally reduction of the 3-keto group by the NADH regenerates the 3-hydroxyl group and the final hydrolysis product.
Transglycosidases are enzymes that catalyze the transfer of a glycosyl group from one glycoside to another. These enzymes use essentially the same mechanisms as retaining glycoside hydrolases. Following formation of a glycosyl enzyme intermediate (or a covalent intermediate formed by neighboring group participation), this species reacts with another alcohol (often a sugar), instead of water, to form a new glycoside.
Installation and Removal of Carbohydrate Adornments
Sulfotransferases and Sulfatases
Carbohydrate desulfation is catalyzed by hydrolytic enzymes termed sulfatases (Fig. 3b). Most carbohydrate sulfatases possess a unique amino acid residue, formylglycine, which is critical for their ability to function as catalysts. Most evidence supports a role for the hydrated form of formylglycine as an enzymatic nucleophile, in a three-step mechanism. This involves the nucleophilic substitution of a sulfate ester by a hydrated formylglycine, to form a sulfated enzyme; elimination of hydrogen sulfate to give formylglycine; and finally addition of water to the aldehyde group to regenerate the hydrated formylglycine.
Phosphotransferases and Phosphatases
Phosphorylation of carbohydrates is commonly encountered in primary metabolism but is only rarely seen in complex carbohydrates in secondary metabolism. Phosphorylated carbohydrates occur as part of the mannose-6-phosphate group used as a lysosomal targeting signal, in phosphoglycans of protozoa and yeast, and in the (lipo)phosphoglycans lipoteichoic acid and wall teichoic acid of Gram-positive bacteria. Mannose-6-phosphate residues are formed in a two-step process through the initial transfer of an N-acetyl-D-glucosamine-1-phosphate (by UDP-N-acetylglucosamine phosphotransferase) onto acceptor mannose residues, followed by the cleavage of the N-acetylglucosamine residue (by N-acetylglucosamine-1-phosphodiester α-N-acetylglucosaminidase) (Fig. 3c).
Acyltransferases and Esterases
Carbohydrate alcohols can be modified as esters of acetic acid or longer-chain fatty and complex acids. For example, bacterial lipopolysaccharides such as lipid A are extensively esterified, and bacteria of the grouping corynebacteria produce a diverse range of esterified carbohydrates, both as part of the cell wall macromolecule and as extractable non-covalently associated glycolipids. Acetylation is often observed as a modification of sialic acids. Acyl transfer to carbohydrate alcohols is catalyzed by acyltransferases that use acetyl-coenzyme A or fatty acyl-coenzyme A. The hydrolysis of carbohydrate esters can be catalyzed by esterases that often possess an α/β hydrolase fold and a conserved catalytic triad, Ser-His-Asp, which is common to lyases (Fig. 3d). Other carbohydrate esterases are metal-ion dependent. The antigen 85 complex of the corynebacteria are transesterases that catalyze the transfer of fatty acyl groups between sugar residues.
Carbohydrate acids such as glucuronic acids within pectins are often modified as methyl esters. Pectin methyl esterases catalyze the hydrolysis of these methyl esters to release methanol and utilize a conserved Gln-Asp-Asp triad.
Methylation of carbohydrates is widespread in bacteria and plants, but comparatively little is known about its roles. O-Methyltransferases use S-adenosylmethionine as the methyl donor. Methylation may occur at the level of the sugar diphosphonucleoside prior to glycosyl transfer or to intact glycans after assembly of the glycan structure. The 3-O-methylmannose-containing polysaccharides and 6-O-methylglucose-containing lipopolysaccharides of mycobacteria form stoichiometric complexes with long-chain fatty acyl-coenzyme A and may have roles in regulating its synthesis or transport.
Carbohydrate Redox Enzymes
A large group of copper-dependent oxidases can oxidize various positions on a sugar ring. Lytic polysaccharide monooxygenases oxidize the C1, C4, and C6 positions of sugars within polysaccharides such as chitin, starch, and cellulose, forming hemiacetals that eliminate to depolymerize the sugar chain (Fig. 4b). LMPOs contain a type 2 copper center and enzymes from fungi typically contain a posttranslational modification in which a histidine that coordinates the copper is N-methylated. A Cu(I) centre activates dioxygen to abstract hydrogen from carbon at either end of a glycosidic bond, leading to hydroxylation at this site and cleavage of the glycosidic bond by elimination of an alcohol and formation of a carbonyl compound. Galactose oxidases are copper enzymes catalyze the oxidation of galactose at the 6-position and convert O2 to H2O2. They possess a posttranslational modification in which a cysteine residue cross-links with a tyrosine. In its oxidized form, a free radical species is formed on the tyrosine residue that couples antiferromagnetically to the copper(II) center.
Carbohydrate reductases include enzymes involved in the biosynthesis of sugar polyols. Mannitol and sorbitol are compatible solutes that are formed by the NADPH-dependent enzymes mannose-6-phosphate and glucose-6-phosphate reductases.
Non-anomeric carbohydrate epimerases catalyze changes in stereochemistry at one or more non-anomeric position (He et al. 2000). These enzymes usually act on sugar nucleotides, prior to their transfer to form glycoconjugates, or simple mono- or disaccharides. At least four different mechanisms have been uncovered for non-anomeric epimerases. (1) For keto sugars, the existence of the keto group in the substrate allows a facile epimerization at the adjacent site, which is achieved by deprotonation on one face and reprotonation on the other, with the assistance of a divalent metal coordinating to the carbonyl group. (2) For non-acidic carbohydrate substrates, epimerization can occur through a mechanism involving a reduction-oxidation process via a transient keto intermediate (Fig. 4c). Oxidation of the hydroxyl group at the epimerization site to a keto group using an NAD+ cofactor, rotation of the keto intermediate in the active site, and reduction from the other face effects an epimerization reaction. (3) A retroaldol-aldol epimerization mechanism has been observed for L-ribulose-5-phosphate 4-epimerase. Elimination of the bond β to the sugar carbonyl of the open-chain sugar yields an aldehyde fragment that can rotate and undergo an aldol reaction to afford the epimerized sugar. (4) Epimerization through nucleotide elimination-addition occurs for certain sugar nucleotides. In this reaction ionization of the sugar-1-phosphonucleotide forms a transient oxocarbenium ion-like transition state, which is deprotonated to form the 1,2-alkene. A proton is delivered to the opposite face, and the phosphonucleotide adds to the same face to afford the epimerized sugar-1-phosphonucleotide.
Mutarotases are carbohydrate epimerases that operate on sugar hemiacetals to interconvert the α- and β-anomers. Mutarotases are believed to act through the same basic steps as uncatalyzed mutarotation, namely, cleavage of the endocyclic C-O bond to generate an acyclic sugar aldehyde, rotation about the C1–C2 bond, and closure of the ring to form the inverted anomer (Fig. 4d).
Modification of the Carbohydrate Backbone
Amino sugars are produced by transamination reactions (He et al. 2000). The biosynthesis of glucosamine occurs by transfer of ammonia, from the hydrolysis of glutamine, to fructose-6-phosphate (Fig. 5a). An imine is formed from an enzymatic lysine to the 2-keto group of fructose-6-phosphate, which undergoes transamination with ammonia and then tautomerization to afford glucosamine-6-phosphate. An alternative method used for the synthesis of amino sugars occurs at the level of the sugar phosphonucleotide. TDP-mycaminose, a 3-amino-3,6-dideoxy-hexose, is synthesized from glutamate and the corresponding 3-keto sugar, by the action of a pyridoxal-5′-phosphate-dependent transaminase.
The major sulfonated sugar is sulfoquinovose, which is produced by phototrophs and incorporated into plant sulfolipid. Sulfoquinovose is synthesized as a sugar nucleotide from UDP-glucose through the action of UDP-sulfoquinovose synthase, an NAD+-dependent enzyme, in three steps (Fig. 5c). In the first step, UDP-glucose is oxidized by NAD+ at the 4-position to afford a 4-keto sugar. In the second step, elimination of water gives a 4-keto-5,6-ene (Fig. 4). Finally, conjugate addition of sulfite, at the 6-position, and concomitant reduction of the keto group by NADH affords UDP-SQ.