This page has been approved by the Responsible Curator as essentially complete. CAZypedia is a living document, so further improvement of this page is still possible. If you would like to suggest an addition or correction, please contact the page's Responsible Curator directly by e-mail.
The glycoside hydrolases of this family are exo-acting α-fucosidases from archaeal, bacterial and eukaryotic origin. No other activities have been observed for GH29 family members. So far the only other CAZY family containing α-fucosidases is family GH95. The human enzyme FucA1 is of medical interest because its deficiency leads to fucosidosis, an autosomal recessive lysosomal storage disease [1].
Kinetics and Mechanism
GH29 α-fucosidases are retaining enzymes following a classical Koshland double-displacement mechanism, as first proposed in 1987 for human liver α-fucosidase via burst kinetics experiments and using methanol as an alternative glycone acceptor to produce methyl α-L-fucoside [2]. This has been further confirmed by 1H NMR monitoring of the reaction catalyzed by an α-L-fucosidase from Thermus sp. [3], and a α-L-fucosidase from the marine mollusc Pecten maximus [4], as well as by COSY and 1H-13C NMR spectroscopy analysis of the interglycosidic linkage of disaccharides formed by the transglycosylase action of Sulfolobus solfataricus α-L-fucosidase, Ssα-fuc [5]. GH95 α-fucosidases, in contrast, operate with inversion of the anomeric configuration.
Catalytic Residues
The catalytic nucleophile in GH29 was first identified in the Sulfolobus solfataricus α-L-fucosidase, Ssα-fuc, as Asp242 in the sequence VYFDWWI via chemical rescue of an inactive mutant with sodium azide [6]. Concomitantly the catalytic nucleophile of Thermotoga maritima α-L-fucosidase, Tmα-fuc, was confirmed to be Asp224 in the sequence LWNDMGW through trapping of the 2-deoxy-2-fluorofucosyl-enzyme intermediate and subsequent peptide mapping via LC-MS/MS technologies, as well as by chemical rescue of an inactive mutant [7]. The trapping of the 2-deoxy-2-fluorofucosyl-enzyme intermediate in Tmα-fuc was corroborated by crystallographic studies [8]. The catalytic nucleophile of the human enzyme FucA1 has recently been identified as being Asp225 [9].
Whereas the catalytic nucleophile in GH29 has been shown to be a conserved aspartate residue, the identity of the general acid/base is still controversial. Structural and mutagenesis studies of Tmα-fuc provided strong evidence for the variant Glu266 being the general acid/base [8]. In the crystal structure the carboxyl function of this residue is 5.5 Å away from that of the catalytic nucleophile Asp224, a distance commonly observed in retaining glycosidases proceeding via a classical Koshland double-displacement mechanism. Although multiple sequence alignments show that Glu266 is not conserved within GH29, the residue is structurally conserved in two 3-D structures of α-L-fucosidases from Bacteroides thetaiotaomicron VPI-5482, recently deposited in the Protein Data Bank (PDB accession numbers 3eyp and 3gza). Studies of Ssα-fuc demonstrated that mutation of the Glu residue corresponding in sequence to Tmα-fuc Glu266 barely impaired the catalytic activity of the enzyme, whereas a Glu58Gly mutant had a 4000 fold lower kcat/KM and could be chemically rescued [10]. In the crystal structure of Tmα-fuc in complex with fucose [8], the residue corresponding to Ssα-fuc Glu58, Glu66, is found 7.5 Å away from the catalytic nucleophile Asp224 and hydrogen bonded to the C-3 hydroxyl group of fucose, which altogether makes this residue an unlikely candidate for the function of the general acid/base. A recent study of the human α-L-fucosidase FucA1, carefully done combining bioinformatics, structural modelling, mutagenesis, chemical rescue with azide and 1H NMR spectral analysis, identified Glu289 as the general acid/base [9]. The equivalent residue in Tmα-fuc, Glu281, as inferred from sequence alignment of FucA1 and Tmα-fuc, points to the interior of the (β/α)8 barrel and lies about 15 Å apart form the catalytic centre.
Altogether it appears that family GH29 is a quite exceptional CAZy family in that multiple sequence alignments do not allow an unambiguous assignment of the general acid/base.
Three-dimensional structures
The first crystal structure to be solved is that of the α-L-fucosidase from T. maritima, Tmα-fuc (PDB ID 1hl8). The simultaneous solution of the structures of an enzyme-product complex (PDB ID 1odu) and of a glycosyl-enzyme intermediate (PDB ID 1hl9) allowed the unambiguous identification of the general acid/base [8], as described above. Tmα-fuc assembles as a hexamer and displays a two-domain fold, composed of a catalytic (β/α)8-like domain and a C-terminal β-sandwich domain. The two key active site residues are located at the C-terminal ends of strands β-strands 4 (nucleophile) and 6 (acid/base).
Crystallization experiments for the S. solfataricus α-L-fucosidase, Ssα-fuc, were not very fruitful, but a small angle scattering study has been reported [11], which suggests a nonameric assembly of the enzyme in solution. Two crystal structures, arising from Structural Genomics initiatives, have been deposited in the Protein Data Bank for α-L-fucosidases from Bacteroides thetaiotaomicron VPI-5482, with accession numbers 3eyp and 3gza.
The catalytic domain of Tmα-fuc does not adopt the canonical TIM-barrel (β/α)8 fold, as it lacks helices α5 and α6. Helix α5 is also missing in the structure of one of the B. thetaiotaomicron VPI-5482 α-L-fucosidases, BT3798 (PDB ID 3gza), whereas α-L-fucosidase BT2192 (PDB ID 3eyp) from the same organism adopts the canonical TIM-barrel fold. The three structures differ furthermore by the insertion/deletion of a considerable number of additional α-helices, 310 helices, and extended surface loop regions. The closest structural homologues of GH29 enzymes within the CAZy classification can be found in GH107, which together with GH29 forms Clan GH-R. GH29 also bears some structural similarity to families GH13 (Clan GH-H) and GH27 (Clan GH-D).
Transglycosylation and Glycosynthases
Transglycosylation activity had been observed in 1987 for human liver α-fucosidase [2]. The first successful transformation of an α-fucosidase into an α-transfucosidase by directed evolution has been reported for Thermotoga maritima α-fucosidase [12]. α-Fucosidases mutated in the catalytic nucleophile from both Sulfolobus solfataricus and Thermotoga maritima were successfully transformed into a type of synthetic enzyme termed a 'glycosynthase', in this case a fucosynthase, which use β-L-fucopyranosyl azide as donor substrate [13]
Family Firsts
First stereochemistry determination
Retention of anomeric stereochemistry suggested for human liver α-fucosidase by the formation of methyl α-L-fucoside using methanol as an alternative glycone acceptor [2]. Later confirmed by 1H NMR for α-L-fucosidase from Thermus sp. [3].
Sulfolobus solfataricus α-L-fucosidase by azide rescue of an inactivated mutant [6] and confirmed shortly thereafter by labeling of the nucleophile and peptide mapping [7].
