Biotherapeutics are a rapidly increasing part of the total pharmaceutical market, and most of these are glycoproteins. The glycan moieties covalently attached to a therapeutic protein affect its biological activity, stability, and safety. As a consequence, N-glycosylation is considered as a critical quality attribute (CQA) that requires an adequate analytical approach to certify product quality. It is crucial to analyze glycoforms to monitor the batch-to-batch variability in therapeutic glycoprotein production and compare biosimilars with their originators. The isomeric complexity and branched structure of N-glycans still remains a central analytical challenge. Modern hyphenated techniques combined with cryogenic infrared spectroscopy show potential to distinguish glycans unambiguously.

In this work, we demonstrate a workflow for N-glycan profiling of a biotherapeutic protein. We first used PNgase F to deglycosylate Etanercept or TNFR:Fc (Tumor necrosis factor receptor linked to the Fc portion of human IgG1), which is used in treating rheumatoid arthritis. We monitored the efficiency of the enzymatic digestion by UPLC and purified the N-glycans from salts, detergents, and peptides, using C18 and porous graphitic carbon cartridges. We then characterized the N-linked glycans using a multidimensional approach that combines ultrahigh-resolution ion mobility spectrometry (IMS) and cryogenic, messenger-tagging, vibrational spectroscopy. One experiment provides the mass, drift time, and IR fingerprint spectrum of each of N-glycans. The recorded vibrational spectra demonstrate well-resolved transitions in a wavenumber region where free and weakly hydrogen-bonded OH oscillators appear, and these can be used as a fingerprint to identify a particular glycan. We identify the cleaved glycans by comparing their spectra and arrival time distributions (ATDs) to a targeted database that we created based on standards of the above-mentioned glycans. The advantage of IMS over current techniques is that it provides all the information about the possible isomers of N-glycans and has no need for a chromophore. Apart from making sample preparation time shorter and simpler, our technique allows glycan identification in as little as 3 minutes.

This work represents an important advance in the gas-phase analysis of N-glycans cleaved from biopharmaceutical proteins, which can be used as fingerprinting tool for monitoring glycoforms.