MS/MS ANALYSIS OF IgG3 DISULFIDE BONDS AND DEVELOPMENT OF A NOVEL TOOL TO ASSESS ALGORITHMS THAT ASSIGN GLYCOPEPTIDE CID DATA

Abstract

With the rapidly increasing use of proteins as biotherapeutics to treat diseases, the characterization of these large molecules using mass spectrometry has become a highly attractive field of research. A particular area of research is the identification and characterization of protein post-translational modifications. Disulfide bonds and glycosylation are among the most critical protein post-translational modifications (PTMs), as they play vital roles in maintaining the proper protein folding, structure, and functions. These two PTMs are particularly important in the development and characterization of monoclonal antibody-based drugs, which are the most prevalent protein therapeutics in the market. Among the four classes of immunoglobulins (IgG’s), the disulfide connectivity of IgG1, IgG2 and IgG4 have been effectively studied, and IgG2 and IgG4 have been shown to have disulfide bond-mediated isomers due to alternative disulfide bond connectivity. However, no studies to investigate the presence of disulfide related isomers in IgG3 have been done. In this dissertation, high resolution mass spectrometry is used map the disulfide bond connectivity in IgG3 in order to investigate the presence of disulfide-mediated isomers. The results indicate that no such isomers exist for endogenous IgG3 antibodies. The development of a novel glycoproteomics software, Glycopep Decoy Generator (Tool 1), and the generation of a large dataset of manually assigned CID spectra (Tool 2) from diverse glycopeptide compositions also are described herein. The decoy generator generates abundant decoys for any target glycopeptide composition, and when it is used along with the dataset of CID spectra, the accuracy of glycopeptide scoring algorithms can be readily determined. The tools were used to assess GlycoPep Grader, a scoring algorithm that assigns glycopeptides to CID spectra. The results indicate that GlycoPep Grader has some weaknesses in scoring spectra from fucosylated glycopeptide compositions. These weaknesses could not be easily identified without the aforementioned tools. In order to address GlycoPep Grader’s limitations, a thorough investigation of the root cause of its weaknesses is carried out, and potential updates that could improve the software are proposed.