Into the structure of human full-length Smad proteins and the impact of cancer mutations

Abstract

Signal transduction pathways are essential mechanisms that cells use to perform a myriad of biological processes that when aberrant, can cause serious diseases. Understanding the molecular basis of these pathways in detail can help us interpret these signaling cascades. The TGFβ (Transforming Growth Factor β) signaling pathway is one of the best-studied ones in metazoans. The main effectors of this pathway are the Smad transcription factors, protein family. The work described in this thesis postulated mechanisms correlating Smad protein structure with function using structural biology.

This work has been devoted to the study of Smad full-length proteins, with an emphasis on Smad4 and Smad2. The project started with the production of full- length human Smad proteins, which required an optimized protocol to obtain homogeneous samples in milligram quantities. Once this step was achieved, we have proceeded to the acquisition of structural data combining several techniques (NMR, SAXS and Ion Mobility Mass Spectrometry), together with software development for data integration. The analysis of this information allowed us to obtain a description of the full-length ensemble of Smad2 and Smad4 proteins. The main findings stressed that the inter-domain linkers of Smad2 and Smad4 behave as intrinsically disordered proteins. Smad4 full- length is a monomeric and flexible protein in an equilibrium between elongated and more compact conformations. In solution, Smad4 is not predominantly in an auto- inhibited conformation and the MH1 and MH2 are independently functioning domains. Smad2 is an oligomeric protein populating a monomer- dimer-trimer equilibrium shaped by phosphorylation. Phosphorylation and MH1 deletion shift the equilibrium towards trimer formation, deletion of the C-terminal phosphosites abolishes trimer formation and the inter-domain association is concentration- dependent. We also analyzed how cancer mutations could affect protein structure and found that: Smad4 MH1 mutations mainly affect charged and hydrophobic residues. Cancer mutations seem to affect protein stability while maintaining DNA binding functionality. By comparing melting temperature analysis and molecular dynamics simulations, we proposed a mutational landscape mechanism for the Smad4 MH1 domain that exerts its effects by disrupting salt-bridge networks. Overall, we have established a framework for describing Smad protein structure from an intra- and inter-molecular perspective