Solutions to the glycosylation problem for low- and high-throughput structural glycoproteomics

N- and O-glycosylation profoundly affect the biological properties of glycoproteins, principally by influencing their structures and cellular trafficking, and by forming the recognition sites of carbohydrate-binding ligands. For crystallographers interested in studying the protein component of glycoproteins, the two most important aspects of glycosylation are (1) that it is often essential for the correct folding of a given protein and for ensuring its solubility, which generally necessitates expression of the molecule in eukaryotic cells, and (2) that there are now procedures for the efficient post-folding removal of N-linked glycans from glycoproteins and for minimizing the effects of O-glycosylation, which will generally benefit crystallogenesis. We provide an overview of how glycans influence glycoprotein folding and then identify the sources of structural heterogeneity at the heart of the 'glycosylation problem'. We then discuss the options available to structural biologists for circumventing the problems associated with protein N- and O-glycosylation. Our emphasis is on methods for producing glycoproteins with homogeneous and/or removable N-glycosylation in mammalian cells that can be implemented in both very high yield, stable expression systems and in a high throughput format based on transient expression protocols. We also consider whether deglycosylation reduces protein stability and end by emphasizing the importance of using rigorous stereochemical and biosynthetic data when building glycosylation into partial or complete electron density.