In multi-cellular tissues and organisms, extracellular chemical stimuli are processed by cells via a network of interactions between protein molecules, that sequentially modify each other through post-translational modifications such as phosphorylation to transmit information throughout the cell. Mass spectrometry-based quantitative phosphoproteomics, which allows systematic identification of proteins modified by phosphorylation, has become an essential approach to study intracellular signalling. The work in this thesis relates to the bioinformatics analysis of phosphoproteomics data. The modification of proteins downstream of different receptor tyrosine kinase (RTK) signalling cascades is known to change their function. An ongoing issue in the field of phosphoproteomics is the lack of phosphorylated site-centric analysis methods as alternatives to widely used gene- or protein-centric methods. As a solution to this, I propose a method to relate phosphorylated sites to biological functions using multilayer networks. By applying a random walk on heterogeneous network (RWHN) algorithm, phosphorylated sites are related to functional descriptors. This method outperformed the standard enrichment analyses. Network-based bioinformatics approaches were further used to analyses complex phosphoproteomics datasets describing phosphoproteomic changes downstream of epidermal growth factor receptor (EGFR) and fibroblast growth factor receptor (FGFR) signalling cascades. We used a bespoke proximity-dependent biotinylation approach to generate a spatially resolved phosphoproteomics dataset describing global and recycling endosome-proximal phosphorylation events of activated FGFR2b. From this, we revealed that FGFR2b locally suppresses autophagy through the mTOR pathway from the recycling endosome, downstream of FGF10 activation. As a result of this work, I developed an application to visualise enrichment of GOCC terms. I also investigated the role of ligand concentration in regulating phosphorylation events and long-term proteome-level changes downstream of FGF1 and EGF stimulation. The analysis method utilised links short-term phosphorylation dynamics to long-term changes in protein abundance, downstream of different ligands and ligand concentrations, via modified transcription factors, to identify differential responses. The work in this thesis demonstrates the valuable role of integrating network-based methods in the analysis of phosphoproteomics datasets in order to better understand RTK signalling cascades.
- receptor tyrosine kinases
- cell signalling
- networks
- phosphoproteomics
- bioinformatics
Integration of Complex Phosphoproteomics and Network-Based Bioinformatics to Study Receptor Tyrosine Kinase Signalling
Watson, J. (Author). 14 Dec 2021
Student thesis: Phd