Conformational Transition Pathways in Signaling and Enzyme Catalysis Explored by Computational Methods

Biomolecules are dynamic in nature and visit a number of states while performing their biological function. However, understanding how they interconvert between functional substates is a challenging task. In this thesis, we employ enhanced computational strategies to reveal in atomistic resolution transition states and molecular mechanism along conformational pathways of the signaling protein Nitrogen Regulatory Protein C (NtrC) and the enzyme Adenylate Kinase (Adk).;Targeted Molecular Dynamics (TMD) simulations and NMR experiments have previously found the active/inactive interconversion of NtrC is stabilized by non-native transient contacts. To find where along the conformational pathway they lie and probe the existence of multiple intermediates, a beyond 8mus-extensive mapping of the conformational landscape was performed by a multitude of straightforward MD simulations relaxed from the biased TMD pathway. A number of metastable states stabilized by local interactions was found to underline the conformational pathway of NtrC. Two spontaneous transitions of the last stage of the active-to-inactive conversion were identified and used in path sampling procedures to generate an ensemble of truly dynamic reactive pathways. The transition state ensemble (TSE) and mechanistic descriptors of this transition were revealed in atomic detail and verified by committor analysis.;By analyzing how pressure affects the dynamics and function of two homologous Adk proteins - the P.Profundum Adk surviving at 700atm pressure in the deep sea, and the E. coli Adk that lives at ambient pressures - we indirectly obtained atomic information about the TSE of the large-amplitude rate-limiting conformational opening of the Adk lids. Guided by NMR experiments showing significantly decreased activation volumes of the piezophile compared to its mesophilic counterpart, TMD simulations revealed the formation of an extended hydrogen-bonded water network in the transition state of the piezophile that can explain the experimentally measured activation volume differences. The transition state of the conformational change was proposed to lie close to the closed state. Additionally, a number of descriptors were used to characterize the free energy landscape of the mesophile. It was found that the features of landscape are highly sensitive to the binding of different ligands, their protonation states and the presence of magnesium.