Function and evolution of genes in the human protein interaction network

Abstract

The research conducted for this thesis aims to elucidate how the human protein interaction network has evolved, and how protein interactions influence the spatial organisation of the metabolic network. The thesis presents compelling results suggesting that indirect protein-protein interactions between metabolic enzyme proteins, and non-metabolic proteins is a feature conserved from prokaryotes through to mammals. Indirect protein-protein interactions are shown to influence the structure of metabolic networks, very likely through a mechanism called metabolic channelling. This mechanism brings metabolic pathway enzymes into close proximity of each other so that reactions between the enzymes can occur. In E. coli and yeast, reactions that are linked to each other through mediator proteins have a much higher flux compared to reactions lacking them. The thesis presents indirect evidence that indirect protein-protein interactions are very important for the spatial organisation of metabolic protein networks. This thesis has also inferred the likely time of origin of the human protein-protein interactions that were examined, and suggests that most of the protein interactions in human are likely to be conserved in chimpanzee, macaque, mouse, rat, horse, dog and cow. This suggests they date from at least the point in evolution when placental mammals radiated out. Analyses of the function of proteins analyses in the thesis also suggests that the protein interactions have been conserved because they perform fundamental cellular functions. Finally, the duplicability of protein self-interactions (interactions between two proteins encoded by the same gene i.e., homodimers) are investigated. These genes have higher duplicability in comparison to other protein encoding genes, and also contain more interactions in general, which can perhaps be explained by the greater evolutionary age of these genes.