During the work described in this thesis, different multidimensional liquid chromatography - tandem mass spectrometry workflows have been developed to perform glycoprotein characterization.

Glycoproteins are of major importance in health and disease, either as biomarkers or as biotherapeutics, but glycosylation is one of the most complex and heterogeneous post-translational modifications, involving two different chemical groups: a glycan and a protein. Glycoprotein characterization can be done using bottom-up proteomics, which is the analysis of peptides and glycopeptides resulting from enzymatic glycoprotein digestion, and mass spectrometry is the gold standard for such characterization.

Ion mobility is beneficial to enhance glycopeptide and glycoprotein separation and characterization. Thus, several workflows were developed using Differential Mobility Spectrometry (DMS). In the first section, a mix of three isomeric O-glycopeptide standards and the corresponding non-glycosylated peptides were separated using the combination of DMS and RPLC, while neither DMS nor microflow Reverse Phase Liquid Chromatography (RPLC) alone provided enough separation for all peptides in the mix. Characterization was then achieved using Collision Induced Dissociation (CID) and Electron Activated Dissociation (EAD) at 6-7eV, i.e., hot Electron Capture Dissociation (hECD). In the second section, a bottom-up analysis of a standard glycoprotein, fetuin A, was done, and DMS analyses were performed, showing that DMS separation is orthogonal to RPLC, and that peptides and glycosylated peptides are gas-phase separated. Moreover, the use of DMS allowed separation of co-eluting peptides and glycopeptides with overlapping isotope distributions and enabled further characterization through CID and EAD at 6-7eV (hECD) fragmentation.

Then, an innovative column switching workflow integrating a trapping RPLC column and an analytical microflow Hydrophilic Interaction Liquid Chromatography (μHILIC) column was introduced. While RPLC separation is known to be mainly driven by the peptide backbone, HILIC separation is mostly driven by glycan hydrophilicity, and the peptide has a minor influence. The RPLC-μHILIC setup allowed reproducible sample cleaning and preconcentration on the trapping column, followed by peptide and glycopeptide separation on the analytical column. Moreover, separation of up to six glycopeptide isomers was obtained, and focused analysis of an N-glycopeptide clearly showed that different CID fragmentation patterns are obtained at the different retention times.

Lastly, the current status, needs, and opportunities of glycoproteomic software are discussed. Glycosylation is a heterogeneous modification, and data processing to characterize intact glycopeptides is challenged by the use of a wide range of instruments, file formats, acquisition methods, fragmentation techniques, and databases. The GlycansToGraphs tool is introduced and shown to be beneficial for N-glycopeptide MS and MS/MS spectra visualization, identifying already annotated N-glycopeptides, potential new N-glycopeptides, and suggesting the presence of their adducts.