Development of workflows for O-glycan profiling and structural characterisation of biopharmaceuticals

Biopharmaceuticals are a growing branch of drugs that are produced within a living organism, the majority of these drugs being glycoproteins; by replicating endogenous molecules, they are more target-specific than small molecule drugs. However, being of greater structural complexity than small molecules, they are prone to batch-to-batch heterogeneity, with glycosylation being the main cause. O-linked glycosylation is a complex post-translational modification of proteins. In the Golgi, proteins are decorated with a series of monosaccharides attached either via serine or threonine side chains or to growing carbohydrate chains. As O-glycosylation is non-templated, O-glycan structures are determined by the availability of competing biosynthetic enzymes; this contributes to the complexity and heterogeneity of glycoproteins. This heterogeneity may influence the serum stability, immunogenicity, and efficacy of biopharmaceuticals and is therefore a critical quality attribute for batch release. It necessitates O-glycan profiling workflows that can be used in the pharmaceutical sector for the routine quality control of biopharmaceuticals. Workflows currently employed to characterise biopharmaceuticals in industry typically entail intact and middle-up analysis, peptide mapping, and the analysis of released N-glycans. While O-glycan species may be deduced using some of the data generated using such workflows, the assignments are often ambiguous. The analysis of released O-glycans is not incorporated into these workflows as current approaches for O-glycan analysis are often not directly adoptable into industrial practices. This work details the intact and middle-up, peptide mapping and released N-glycan analyses of abatacept, belatacept and etanercept, and the development of a released O-glycan profiling workflow that is tailored to the practices and instrumentation of industry. The workflow development has been exemplified and challenged using standard glycoproteins and an antibody-glycoprotein fusion protein.