Fucosylation, a post-translational modification, plays significant roles in human biology and disease. α-fucosidases, enzymes that hydrolyze α-L-fucosyl linkages, belong to glycoside hydrolase families 29 and 95. GH29 enzymes are diverse, with varying substrate specificities, making their mechanism complex. AlfC, a GH29A α-fucosidase from *Lactobacillus casei*, shows a strong preference for α(1,6)-fucosyl linkages, particularly relevant to human core fucosylation of glycoproteins like antibodies. Previous work demonstrated that the AlfC E274A mutant functions as an efficient transfucosidase, transferring fucose. However, the underlying mechanism remains unclear. This study aimed to use a multi-faceted approach involving X-ray crystallography, mutagenesis, kinetics, mass spectrometry, molecular dynamics, and transfucosylation experiments to decipher the mechanism of AlfC and its transfucosidase mutants. Understanding AlfC's mechanism is vital for designing improved transfucosidases with applications in modifying antibody fucosylation, impacting their effector functions and clinical outcomes.

The literature highlights the importance of fucosylation in various biological processes, including ABO blood grouping, antibody function, cancer progression, and immune cell development. Alpha-fucosidases, categorized into GH29 and GH95 families, exhibit diverse substrate specificities and catalytic mechanisms. While some GH29B enzymes' mechanisms are understood, those of GH29A enzymes, particularly regarding substrate recognition and the role of the catalytic acid/base, remain elusive. The diversity of linkage specificities within GH29A points towards different substrate-binding mechanisms. Studies on specific GH29A enzymes, such as BT2970, have provided insights into their catalytic mechanisms; however, whether these mechanisms are conserved across all GH29A enzymes is unclear. Existing literature about AlfC points to its high activity towards α(1,6)-fucose and the potential of its mutants for modifying core fucosylation of glycoproteins, but its detailed molecular mechanism is lacking, necessitating this study.

This study employed a combination of techniques to investigate the structure and function of AlfC and its mutants. X-ray crystallography was used to determine the three-dimensional structures of wild-type AlfC and various mutants, both in the apo form and in complexes with different ligands (L-fucose, 4-nitrophenyl-α-L-fucopyranoside, Fuca(1,6)GlcNAc). Site-directed mutagenesis was performed to generate point mutations of specific residues within and around the active site. These mutants were characterized using kinetic assays with 4-nitrophenyl-α-L-fucopyranoside (4NP-fuc) and α-fucosyl fluoride, along with azide rescue experiments to probe catalytic residues. Hydrogen-deuterium exchange mass spectrometry (HDX-MS) was employed to investigate the dynamic changes in protein structure upon mutation. Molecular dynamics (MD) simulations provided insights into conformational changes, specifically focusing on the movement of potential acid-base candidates. Finally, transfucosylation experiments were conducted using both GlcNAc and fully glycosylated antibodies as acceptors, assessing the efficiency of different mutants to transfer fucose.

The study revealed that AlfC forms a tetramer in solution and lacks a C-terminal carbohydrate-binding module, which is present in many other α-fucosidases. The crystal structure of AlfC shows fucose binding in a conserved orientation, but no clear acid-base residue within the expected distance from the nucleophile (D200). Three candidates for the acid/base (E39, E274, D242) were investigated. Kinetic analysis and azide rescue experiments, though providing initially conflicting data, eventually implicated D242, supported by MD simulations, which revealed that it adopts open and closed conformations, with the closed conformation being catalytically active. The α(1,6)-specificity of AlfC was attributed to an aromatic subsite adjacent to the active site that specifically accommodates GlcNAc in α(1,6) linkages. The requirement of endoglycosidase treatment for defucosylating IgG is also explained by this subsite. Several transfucosylation mutants (E274A, N243A) displayed high efficiency, and HDX-MS and MD simulations indicated that these mutants shift the equilibrium towards the open conformation, facilitating transfucosylation over hydrolysis. Interestingly, a rationally designed mutant (F237A) also exhibited transfucosylation activity, supporting the proposed mechanism. However, a double mutant (E274A/N243A) lost transfucosylation activity, highlighting the need for fine-tuning enzyme dynamics. Finally, the study revealed that AlfC transfucosylation mutants could fucosylate fully glycosylated antibodies, creating a product neither the wild-type nor any other mutant could hydrolyze; the reason for this remains unclear.

This study provides a comprehensive understanding of AlfC's mechanism, revealing a balance between open and closed conformations crucial for its activity. The identification of D242 as the likely general acid/base, supported by structural and computational data, contrasts with the initially perplexing azide rescue results and offers a new perspective on the application and interpretation of this technique. The results provide insight into α(1,6)-specificity via a dedicated binding site, explain the need for endoglycosidase pretreatment of antibodies for defucosylation, and establish a generalizable method for engineering new transfucosidase mutants by manipulating the equilibrium between open and closed conformations. The finding that transfucosylation mutants create non-hydrolyzable products opens new avenues for exploring efficient glycosyltransferases. The mechanistic insights could facilitate the design of improved α-fucosidases for therapeutic applications, particularly in enhancing the activity of afucosylated antibodies.

This research significantly advances our knowledge of α-fucosidase mechanisms. The findings reveal the structural basis of AlfC's α(1,6)-specificity and explain the need for endoglycosidase pretreatment in antibody defucosylation. A generalizable method for designing efficient transfucosidases by shifting the conformational equilibrium is presented. The discovery of mutants producing non-hydrolyzable products suggests new avenues for generating efficient glycosyltransferases. Future research should focus on applying these findings to other α-fucosidases and developing new strategies for creating highly specific and efficient glycosyltransferases.

The study's limitations include the challenges in obtaining a crystal structure of AlfC in a closed conformation, and the inconclusive azide rescue results. While MD simulations suggested D242's role as the acid/base, direct experimental evidence in a closed conformation is lacking. The reason behind the inability of wild-type or mutant AlfC to hydrolyze fucose from fully glycosylated antibodies after fucosylation by transfucosylation mutants remains unclear. Further investigations are needed to clarify these points.