COMBINED CHEMICAL AND ENZYMATIC APPROACHES TO PROTEIN GLYCOSYLATION

Abstract

Glycosylation is a key post-translational modification of proteins and influences the structure and biological functions of proteins. Glycoproteins are significant in treating a variety of diseases and make up a large fraction of biotherapeutics. The carbohydrate structures on the proteins regulate biological activity and pharmacokinetic properties, thereby dictating the efficacy and cost of glycoprotein drugs. However, glycoproteins expressed in biological systems are heterogeneous in nature and impose a challenge to structure-function studies as well as design of potent therapeutics. Thus, developing tools to modulate the glycan structures on proteins is highly significant. In my thesis, we have explored biological and chemoenzymatic methods to generate homogeneously glycosylated therapeutic proteins. First, we designed a glycosylation machinery in Escherichia coli (E. coli) using an N-glycosyl transferase enzyme to transfer a sugar handle onto a model protein. The protein was then elaborated with a homogeneous glycoform using in vitro chemoenzymatic transglycosylation. Using this methodology, we produced a fully glycosylated human interferon alpha-2b that was biologically active and displayed significantly enhanced proteolytic stability. Next, we focused on expanding the toolbox of enzymes available to perform the chemoenzymatic glycan remodeling of proteins. Specifically, we compared the substrate specificities of the human α-L-fucosidase (FucA1) and two bacterial α-L-fucosidases (AlfC and BfFuc) with a panel of structurally well-defined core-fucosylated substrates. FucA1 was the only α-L-fucosidase to display hydrolytic activity towards full-length core-fucosylated glycopeptides and glycoproteins. Moreover, FucA1 showed low but apparent activity to remove core fucose from intact monoclonal antibodies. This finding reveals an opportunity to employ FucA1 to remove core fucose from therapeutic antibodies to improve their antibody-dependent cellular cytotoxicity. Finally, we explored modulation of core fucosylation of monoclonal antibodies through metabolic glycoengineering. We designed L-fucose analogs to potentially incorporate functionalized fucose into IgG-Fc glycan. We showed incorporation of a few fucose derivatives into antibodies and identified a concentration-dependent effect of some of the previously known analogs. While some of the novel compounds did not show effect, the study supplements the existing tools available for metabolic modulation of antibodies. In summary, these studies present feasible new approaches to produce therapeutic eukaryotic glycoproteins with desired, homogeneous glycosylation.