Glycoproteomics incursions into the realm of proteoglycans

Abstract

The term proteoglycan encompasses a heterogeneous group of heavilymodified metazoan glycoproteins that are involved in fundamental biological processes. They are essential for embryonic development and play important roles in tissue organization and cell haemostasis. Proteoglycans are also linked to pathogenesis and modulate key processes related to microbial infection, cancer behaviour and cardiovascular dysfunction. To understand their impact on human health and disease, thorough studies of their structure-and-function relationships are required. However, this has been hampered by technical difficulties mainly related to the structural complexity of their modifying glycosaminoglycan (GAG) chains. In this thesis, we developed protocols for the purification and structural characterization of chondroitin sulfate (CS) and heparan sulfate (HS) proteoglycans from complex samples. Our approaches were enclosed within a glycoproteomics framework allowing for the simultaneous identification of the core-proteins and the glycan attachment sites. Additionally, they facilitated the characterization of the proteoglycan linkage region. Our workflows entailed multiple enzymatic degradation steps, chromatographic separation and high-resolution tandem mass spectrometry. Finally, we developed SweetNET, a bioinformatics platform to cope with the large amounts of data generated from these high-throughput experiments. In addition to the limited number of known mammalian proteoglycans (less than 50), we identified 21 novel human core proteins modified with CS chains. We found that several pro-hormones carry CS-modifications, defining them as a novel class of proteoglycans. We also identified novel glycan variations of the proteoglycan linkage region, close to the peptide attachment site, which included fucosylation and sialylation. The examination of the small CS proteoglycan bikunin across different human body fluids revealed an unforeseen heterogeneity of its linkage region, especially in urinary samples. In addition, we could determine the exact macromolecular architecture of the bikunin CS-chain within the inter-alpha trypsin inhibitor complex in serum and cerebrospinal fluid. Finally, we identified placental-type GAGs in induced pluripotent stem cells using a recombinant malaria protein probe. These GAGs displayed a stage-specific dependence and were associated with a heterogeneous group of core-proteins. The extent and biological implications of these findings for basic stem cell biology need further clarification. Taken together, we have established preparative and analytical protocols as well as bioinformatics tools for the structural characterization of native proteoglycans in complex samples. This led us to identify novel human proteoglycans as well as novel glycosaminoglycan modifications. Finally, we found that the proteoglycan landscape of pluripotent stem cells changes upon differentiation and can be specifically targeted using a unique protein probe.