Protein-protein interactions: from specific to global

Direct, non-covalent binding between different biomolecules underlies most of biological activity at the molecular level. Gene expression, signal transduction and cell growth are, for example, just a few fundamental processes, which critically depend on accurate, specific binding interactions between different proteins, nucleic acids, lipids and other molecules. However, our understanding of such processes is still not fully complete. For example, an important open question concerns the connection between the fundamental physico- chemical properties of the basic building blocks of biomolecules, such as amino acids in the case of proteins and nucleobases in the case of nucleic acids, and the specific binding preferences of complete biomolecules. Could one predict the binding preferences of a given protein just by analyzing the properties of its constituent amino acids or their modified versions? Moreover, how does the dynamics of proteins influence their binding to different partners? Our FWF START project Towards a quantitative framework for understanding protein-protein interactions: from specific effects to protein ecology addressed these questions using the techniques of computational biophysics and computational structural biology.As one of the main achievements of the project, we showed that most protein sequences are compositionally complementary to the sequences of messenger RNA (mRNA) molecules that code for them, implying that the two may interact, especially if unstructured. This finding, in turn, also provides support for the hypothesis that the genetic code evolved as a consequence of binding interactions between proteins and their cognate mRNAs. Second, we showed that the physicochemical properties of proteins may easily be encoded in the physicochemical properties of their cognate mRNAs, suggesting a potentially novel mechanism for controlling protein localization. Third, we showed that even a simplified description of protein dynamics, that excluding any correlations between atoms in the protein, may contain enough information to accurately capture an important feature of binding thermodynamics, the conformational entropy. Finally, we have developed a framework for computationally studying protein post-translational modifications (PTMs) and have analyzed their effect on protein properties. In particular, we showed that many PTMs exert a strong effect on proteins interactions with water. Overall, our project has provided a number of conceptual and methodological advances for understanding the behavior of biomolecules in realistic environments, with potential applications in different areas of molecular biology and biomedicine.