3D structural analysis of RAS effector binding and investigation of context-specific networks

RAS is a signalling switch important in tissue homeostasis and cancer. In its active form, numerous effector proteins with a RAS binding domain are able to compete for binding. While some of these effectors such as RAF are well characterized, a systems understanding of the RAS effector system in the context of effector competition and signalling, oncogenic mutations, and co-stimulatory contexts is lacking. In this thesis, the RAS effector system was investigated from a structural and an experimental perspective. For the structural perspective, 54 putative effector proteins in complex with RAS were modelled. Based on the structural models, binding affinities were estimated which were incorporated into a mathematical model of Ras effector complex formation in 29 different tissues. Interface mutations interfering with the complex structures were energetically evaluated in silico and then integrated in the complex formation modelling. For the experimental perspective, KRAS interactomes from Caco-2 cells overexpressing different KRAS plasmids (WT, G12C, G12D, and G12V KRAS) and treated with different stimulations/inhibitors (untreated, DMOG, TNF-α, IL-6, PGE2, and EGF) were investigated for functional differences. Then, using random walks on a network of the proteins found in the interactomes, we predicted the contribution of individual effectors to functional processes. Our work numerically describes the Ras effector system in a methodologically coherent way and lays the foundation to experimentally engineer the Ras effector system with interface mutations to selectively enhance or abolish (“rewire”) some effectors. In addition, we describe how different combinations of mutations and treatments shape functional annotations of the KRAS interactome.