Mapping Protein Glycosylation by High-Resolution Single Molecule Imaging

When we think about proteins, the major players of cell function, we tend to visualise round or elongated structures that move throughout the body and operate by interacting with other biomolecules. Fundamentally, this is true, however, a single protein can have an enormous number of structures, each of which interacts with other proteins or biomolecules in different ways and fine-tunes how a protein functions or where it moves in the cell or even how long it lasts in the body. This structural diversity is derived from chemical modifications to the protein and the most common and complex type is called glycosylation - an intricate, non-template driven process that adds complex sugars, termed glycans, to specific sites on a protein. The majority of human proteins are modified by glycans and minor faults in the machinery that adds glycans to proteins leads to a rare but severe condition in humans called Congenital Disorders of Glycosylation, with nearly all those afflicted dying before adulthood. The inability to glycosylate proteins is lethal in all animals including the lowest forms of life, such as yeast. Critically, changes in glycosylation are observed in nearly all human disease states including cancer, diabetes as well as aging.

Glycosylation is increasingly important in the biopharmaceutical industry, as most protein-based drugs, like monoclonal antibodies, are glycosylated and glycans can significantly influence their safety and efficacy. Moreover, viruses use glycosylation to hide their surface proteins, which are required for host binding, under so-called "glycan shields" as these protein "spikes" are the principle focus for immune detection and antibody targeting. As a final point, COVID-19 vaccines carry the genetic message to encode the viral spike - a glycoprotein with 66 glycans that has been designed to mimic those found on the surface of SARS-CoV-2.

Glycoproteins are prevalent and central in human health and disease progression but we know surprisingly little of how glycans control the function of proteins. This is because we lack the tools capable of dealing with the complexity of glycoproteins. More specifically, a glycoprotein can have many glycosylation sites, each of which can be occupied by one of hundreds of various branched glycan structures - the possible combinations is therefore enormous. Because of this structural complexity we really only fully know the structure of a handful of glycoproteins. Francis Crick, who co-discovered the structure of DNA, once said, "If you want to understand function, study structure" - a statement that is as revenant for glycoproteins today as it was for DNA nearly 70 years ago.

Therefore, researchers across many areas of biological and medical research need new tools to study glycoprotein structure to better understand and treat disease. Our research project aims to solve this problem in structural biology by creating a completely new way to characterise glycoproteins by imaging sections of a given glycoprotein (termed a glycopeptide) at the single molecule level. We can then use this information to map what glycan structures are at specific sites on a protein with exact atomic detail of both parts. To make our new method possible, we will advance the molecular "probes" used for taking these single-molecule images - akin to a stylus that translates vinyl etchings into music on a record player.

To demonstrate the potential of our new method, we will characterise several glycoprotein-based drugs as well as a lead HIV vaccine candidate glycoprotein, which will add much needed insight into how antibodies interact with glycans on the surface of HIV. This is a key issue for understanding the function and efficacy of structure-based vaccines, including those used for COVID-19.

Protein function is fine-tuned by glycosylation, the addition of glycans at a specific site along the protein sequence. Glycans are remarkably complex and vary significantly in their topology, branching, and linkage between monosaccharide residues. The greatest problem concerning glycoproteins is the fact that high-resolution structural methods, including x-ray crystallography and cryo-EM, cannot resolve glycan structures. Currently only mass spectrometry-based methods allow for the identification of site-specific glycosylation, but provide rudimentary information, namely monosaccharide composition.

Here, we aim to solve this problem by combining mass spectrometry and single-molecule imaging to determine the structure of glycoproteins with atomic detail. We will use electrospray ion beam deposition (ES-IBD) of mass-selected glycopeptides to prepare samples for high-resolution scanning probe microscopy, namely scanning tunnelling microscopy (STM) and non-contact atomic force microscopy (nc-AFM) to reveal the precise structure of glycans attached to peptides isolated from glycoproteins.

STM is proficient in imaging synthetic oligosaccharides and nc-AFM probes, modified with CO molecules, can provide atomically resolved images of molecules adsorbed on surfaces. ES-IBD allows for individual glycopeptide species to be placed on an atomically flat metal crystal under ultrahigh vacuum. Together with imaging at low temperature to inhibit agglomeration and conformational changes, these conditions will enable us to resolve structural details of glycopeptides at the atomic level.

These images will reveal the structural detail missing from the current view of glycoproteins. The development of such a robust technology will allow us to characterise a number of complex biological targets, including the HIV spike glycoprotein. This insight will shed light on glycan-dependent interactions of broadly neutralising antibodies - a key factor in vaccine design.