EXPERIMENTAL AND COMPUTATIONAL METHODS FOR EXPLORING THE GLYCOME AND GLYCOPROTEOME

Glycomics;Glycoproteomics;not listed

Toghi Eshghi, ShadiGoodlett, David ;

Johns Hopkins University

关键词: Glycomics;    Glycoproteomics;    not listed;   

Others  :  https://jscholarship.library.jhu.edu/bitstream/handle/1774.2/59379/Supplementary-Table-5-1.xls?sequence=4&isAllowed=y

瑞士|英语

来源: JOHNS HOPKINS DSpace Repository

【 摘 要 】

Glycosylation – attachment of carbohydrates to proteins – is the most prevalent form of post-translation modification and responsible for protein structural and functional diversification. These carbohydrates, also known as glycans, are ubiquitous and play key roles in many biological functions such as signal transduction, protein folding and quality control, cell recognition and pathogen invasion. Aberrant glycosylation is associated with major human diseases like cancer, viral and bacterial infections and neurodegenerative diseases. The goal of this dissertation is to develop novel experimental and computational tools that could provide new insights into the pathological changes of glycosylation.A mass spectrometry based imaging technique is introduced for direct profiling of N-linked glycans from formalin-fixed paraffin-embedded tissue sections. Imaging of mouse brain coronal sections with this technique revealed significant differences between the glycomics profiles of the midbrain and brain cortex. Notably, fucosylation appeared to be more abundant in the cortex, while oligomannose structures and non-fucosylated glycans were more abundant in the midbrain. Moreover, mass spectrometry imaging of N-linked glycans was employed to differentiate glioblastoma tumor cells injected into a mouse brain from the surrounding normal tissue. To examine the protein hosts of glycans and the microheterogeneity of glycosylation, an algorithm and accompanying software tool were developed for site-specific identification of glycopeptides from mass spectrometry glycoproteomics data. Spectral library matching is introduced to assign the structures of intact glycopeptides based on their higher-energy collisional dissociation fragmented tandem mass spectra. Taking advantage of the power of spectral library matching, novel glycan modifications were exposed. Machine learning was applied to the spectral features of glycopeptides to predict their glycosylation type. The application of the developed software tool, named GPQuest, was verified on recombinant glycoproteins and employed to study complex samples like prostate cancer cell lysates. GPQuest, powered with an easy to use graphical user interface, is made available online for the glycobiology community.