Antibodies used in cancer therapy are only as effective as the sugar molecules attached to them, and current methods for controlling those sugars are slow and imprecise. The problem is that the protein surface of an antibody partially shields its sugar chains, blocking the enzymes needed to modify them. This makes it difficult to manufacture antibodies with specific sugar patterns—such as low fucose or high galactose—that boost their ability to kill cancer cells. The researchers aim to solve this by designing new enzymes that combine a sugar-modifying catalytic domain with a protein-binding domain that locks onto the antibody, bringing the enzyme close to its target. They will use structural biology and computational protein design to create three such enzymes: a fucosyltransferase, a galactosyltransferase, and a sialyltransferase. If successful, this work would enable the efficient production of antibody glycoforms that are currently inaccessible. That could allow precise tuning of immune responses in existing therapeutics—for example, making cancer antibodies more potent or creating anti-inflammatory antibodies—without overhauling manufacturing processes. The project is primarily fundamental science, but the tools it generates could directly improve the performance of drugs already generating hundreds of millions in revenue.
View original technical description
The immune effector functions of IgG antibodies can be fine-tuned by modification of conserved moieties, called N-linked glycans, within their Fc region. Antibody glycoengineering is of considerable biotechnological interest and has been utilised in several clinically approved antibodies, such as the cancer therapeutic Mogamulizumab (generated ~200 million USD revenue in 2023). Current methods of glycoengineering (which involve enzymatic modification of glycans in vitro, or manipulation of the antibody production cell line), are inefficient and yield heterogenous product. To solve this, we propose a structural and computational approach to develop novel enzymatic tools which specifically process IgG N-linked glycans, thereby increasing enzyme activity by proximity-driven effects. This will enable the manufacture of currently inaccessible IgG glycoforms. IgG glycans contain variable levels of fucose, galactose and sialic acid monosaccharide units, all of which influence antibody effector functions. The absence of fucose, and presence of galactose, potentiate activation of cell killing pathways by IgG antibodies, while sialic acid appears to impart anti-inflammatory activity. Conversely, elevated fucosylation has been established to be a desirable feature of antibodies with low cell killing properties (e.g. Simulect; Takeda UK Ltd v. F Hoffmann-La Roche AG [2019], High Court of Justice case no. HP-2018-000008). However, production of antibody glycans highly saturated in these units is impeded by the partial shielding of glycans by the antibody protein surface, which hinders glycan processing enzymes from accessing their glycan substrates. We aim to generate novel enzymatic tools for improved fucosylation, galactosylation and sialylation of IgG antibodies, using an approach divided into two stages: Structural investigations into existing glycosyltransferase enzymes which modify IgG glycans (Aim 1) Structure-guided, computational design of IgG-specific glycosyltransferase enzymes (Aim 2) Within Aim 1, we will investigate the structure and dynamics of enzyme targets catalysing addition of fucose, galactose or sialic acid units to IgG Fc N-linked glycans, using structural biology and computational modelling techniques. The output structural data will inform the computational protein design efforts in Aim 2, to harness the full potential of these natural biological systems. We will subsequently design three new IgG-specific, glycan-active enzymes (a galactosyltransferase, a sialyltransferase and a fucosyltransferase). Our approach to potentiate the activity of natural enzyme targets investigated in Aim 1 exploits bacterial endoglycosidases (EndoS and EndoS2 from Streptococcus pyogenes), which remove N-linked glycans from IgG. These enzymes specifically target IgG as their sole substrate, using a non-catalytic domain which binds the Fc protein surface. A strategy of proximity-driven enzyme enhancement can therefore be achieved by fusing our glycosyltransferase targets with an IgG-specific, non-catalytic Fc binding domain from EndoS/EndoS2. Structural information of enzyme targets from Aim 1 will be combined with our previous EndoS/EndoS2-IgG Fc crystal structures to inform the computational design of chimeric enzymes, using established deep-learning computational tools to generate suitable linker regions between catalytic and Fc-binding domains. The activity of novel enzymes against IgG glycans will subsequently be tested relative to the native enzymes, using established glycan analysis protocols. This project will combine expertise in structural biology, glycobiology and computational protein design to deliver the proposed aims. Our ambition is to provide new tools for research into fundamental biology of IgG glycosylation and translate this knowledge into the generation of novel tools for efficient manufacture of antibody glycoforms, which will find application in modification of existing therapeutics for precise modulation of their immune effector functions.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know