CHEMOENZYMATIC SYNTHESIS OF FUCOSYLATED OLIGOSACCHARIDES AND ANTIBODY GLYCOFORMS FOR ELUCIDATING BIOLOGICAL FUNCTIONS

Abstract

Fucosylation is critical for molecular recognition events in the immune system, microbial interactions, and cancer metastasis. Terminal fucosylation (α1,2, α1,3, and α1,4) is characteristic of the histo-blood group antigens (ABO, H, and Lewis) and is expressed on many N- and O-glycans and most human milk oligosaccharides (HMOs). Core fucosylation (α1,6) exists primarily on the sixth position of the core GlcNAc of N-glycans and plays a profound role in modulating the biological functions of IgG antibodies. Due to their diverse biological properties, structurally well-defined fucosylated oligosaccharides and glycoconjugates are highly demanded for detailed structure-function relationship studies and translational applications.

Three main approaches exist to meet this end: metabolic engineering, traditional chemical synthesis, and chemoenzymatic synthesis. Metabolic engineering has unmatched scalability but can suffer from limitations in many factors, including the availability of various sugar nucleotides, genetic instability, and inherent microheterogeneity due to the non-template-driven multiple-step assembly of glycans. Traditional chemical synthesis is ideal for producing a diverse array of pure targets due to its flexibility but may require prohibitively tedious multistep protocols. Alternatively, chemoenzymatic synthesis harnesses the precision of traditional chemical synthesis and marries it to the usually regio- and stereo-selective enzymatic transformations.

Glycosyltransferases (GTs), the natural enzymes for synthesizing glycosidic bonds, are most widely applied in chemoenzymatic routes yet some glycosidases, which naturally hydrolyze glycosidic bonds, can be used in transglycosylation mode for chemoenzymatic synthesis. Generally, glycosidases can be easier to express, have a more relaxed acceptor substrate specificity, and utilize simpler synthetic donor substrates than GTs. However, glycosidases possess an inherent propensity toward product hydrolysis. This has led to the design of mutant glycosidases, known as glycosynthases and glycoligases, which have diminished hydrolysis and enhanced transglycosylation activity.

A glycosynthase is a catalytic nucleophile mutant that inverts the anomeric configuration of the donor substrate (α→β or β→α) in glycosylation and a glycoligase is a mutant at the general acid/base residue that retains the anomeric configuration (α→α or β→β) in its catalytic glycosylation. Both require an activated glycosyl donor substrate with a suitable leaving group at the anomeric position. Readily synthesized glycosyl fluorides are frequently used. Given their superior stability in aqueous conditions, α-glycosyl fluorides are preferred over β-glycosyl fluorides. Therefore, the glycoligase approach is favored in the synthesis of α-glycosidic bonds, like the α-fucosides found in mammalian systems.

Our group has successfully applied the glycoligase strategy for robust synthesis of core fucosylated (α1,6) N-glycans and glycoproteins. For my first project, I aimed to expand our fucoligase toolbox and designed an α1,3/4-fucoligase (AfcB E746A) for the synthesis of α1,3 and α1,4-fucosylated oligosaccharides. AfcB E746A very efficiently catalyzed the synthesis of the Lewis X (LeX) and A (LeA) trisaccharides and two HMOs: 3-fucosyllactose (3FL) and lacto-N-fucopentaose (LNFP) III with only slight excess (1.5 eq.) of the αFucF donor substrate. I concluded that AfcB E746A prefers to synthesize α1,3- over α1,4-fucosidic bonds, demonstrates a unique specificity for acceptors with a reducing end GlcNAc over glucose, and seems to require a free terminal Gal in its acceptor substrate.

In my second project, I aimed to characterize the enzyme activity of two of our lab’s fucoligases, AfcB E746A and BfFucH E277G, in the synthesis of novel difucosylated tetrasaccharides. I concluded that AfcB E746A requires a free terminal galactose in the acceptor substrate given that AfcB E746A cannot efficiently utilize BfFucH E277G’s monofucosylated products. Alternatively, BfFucH E277G utilizes AfcB E746A’s monofucosylated products. BfFucH E277G synthesized several difucosylated tetrasaccharides, which were characterized by mass spectroscopy and one- and two-dimensional NMR analysis. I concluded that BfFucH E277G catalyzes the synthesis of an α1,3-fucosidic bond on the terminal galactose of 3FL and LeA to yield two novel HMO/Lewis antigen-like structures: 3’3-lactodifucotetraose (LDFT) and 3’-fucosyl-Lewis A (3’-FLeA), respectively. This study further elucidates BfFucH E277G’s unique acceptor substrate driven regioselectivity. In my first two projects, I meticulously characterized the transfucosylation activity of AfcB E746A and BfFucH E277G to provide insight into how these fucoligases may be integrated into synthetic schemes. Enzyme-catalyzed synthesis is an indispensable synthetic strategy given its reliability in substrate specificity and stereo- and regioselectivity.

Pivoting from the study of fucoligases, in my third project I prepared highly pure and structurally well-defined IgG antibody glycoforms by chemoenzymatic remodeling with enzymes discovered and designed by our group. The objective of this work was to demonstrate our lab’s expertise in the chemoenzymatic remodeling of the IgG Fc N297 glycan. Furthermore, these antibodies will be applied in future experiments to more confidently characterize the effect of antibody core (α1,6) fucosylation on their binding affinity for Fcγ receptors (FcγRs) and demonstrate how the allosteric effect of immune complex formation contributes to this phenomenon.

Core fucosylation of the N297 glycan reduces antibody-dependent cellular cytotoxicity (ADCC) by decreasing the antibody’s affinity for the FcγIIIa receptor. In fact, by removing core fucose, binding can be enhanced up to 50-fold, leading to improved ADCC and enhanced therapeutic efficacy. However, the current data is considerably disparate with enhancements ranging from 3- to 53-fold. These discrepancies may be attributed to glycoform heterogeneity and contamination with afucosylated glycoforms. Additionally, some studies have reported that, in the presence of core fucosylation, sialylation negatively impacts ADCC. This is accompanied by only a modest decrease in FcγRIIIa binding affinity which cannot solely account for the large difference in ADCC. This phenomenon may be explained by conformational allosteric cooperativity where a conformational change in Fab, upon antigen binding, is transmitted to the Fc to alter its affinity for the FcγRs. We hypothesize this phenomenon explains the discrepancy between FcγRIIIa binding affinity and the degree of ADCC activated by the sialylated and core fucosylated IgG antibodies. These detailed studies on the IgG-FcγRIIIa interaction are important for basic research and translational science. Exceedingly pure glycoforms are required for these experiments. My final project demonstrates the Wang group’s exceptional position in the field of basic antibody research and the indispensable nature of our work for the improved therapeutic application of monoclonal antibodies.