Structure, function and evolution of protein-protein interactions

In this thesis we are concerned with structural, functional and evolutionary aspects of protein interactions. Chapter 1 introduces background information including important concepts in protein structural classification and gene regulation. It provides an overview of methods and databases that are pertinent to this thesis. Most proteins in genomes are the product of recombination of two or more proteins. Previous studies of domain combinations have shown that homologous combinations of domains tend to strongly conserve their N-to-C terminal order of domains as well as their 3-dimensional orientation. In Chapter 2, in order to investigate the extent of this conservation and the nature of domain-domain interactions that result in divergent domain geometry, homologous proteins with 128 different combinations are examined. We demonstrate that about two thirds of combinations conserve their domain geometry and divergent geometry results form five different types of domain interfaces. Chapter 3 extends concepts applied in the analysis of domain geometry to discuss the implications of domain-domain interactions to the folding of multi-domain proteins. By analysing all known multi-domain proteins that have been subjected to experimental folding studies, a correlation between the nature of domain interfaces and folding interdependency of domains is described. In Chapter 4 a set of gene regulatory protein-protein interactions formed by over one thousand mammalian proteins is analysed. Here, the concept of protein families defined by the ordered domain content of proteins used in previous chapters is employed in order to extract families of homologous interactions. In addition a functional classification of gene regulatory proteins is outlined. Chapter 5 summarises the novel results presented in this thesis and discusses their relevance in the context of other studies.