Development of Site-Specific Mg2+-RNA Force Field Parameters: A Dream or Reality? Guidelines from Combined Molecular Dynamics and Quantum Mechanics Simulations

The vital contribution of Mg2+ ions to RNA biology is challenging to dissect at the experimental level. This calls for the integrative support of atomistic simulations, which at the classical level are plagued by limited accuracy. Indeed, force fields intrinsically neglect nontrivial electronic effects that Mg2+ exerts on its surrounding ligands in varying RNA coordination environments. Here, we present a combined computational study based on classical molecular dynamics (MD) and Density Functional Theory (DFT) calculations, aimed at characterizing (i) the performance of five Mg2+ force field (FF) models in RNA systems and (ii) how charge transfer and polarization affect the binding of Mg2+ ions in different coordination motifs. As a result, a total of similar to 2.5 mu s MD simulations (100/200 ns for each run) for two prototypical Mg2+-dependent ribozymes showed remarkable differences in terms of populations of inner-sphere coordination site types. Most importantly, complementary DFT calculations unveiled that differences in charge transfer and polarization among recurrent Mg2+-RNA coordination motifs are surprisingly small. In particular, the charge of the Mg2+ ions substantially remains constant through different coordination sites, suggesting that the common philosophy of developing site-specific Mg2+ ion parameters is not in line with the physical origin of the Mg2+-RNA MD simulations inaccuracies. Overall, this study constitutes a guideline for an adept use of current Mg2+ models and provides novel insights for the rational development of next-generation Mg2+ FFs to be employed for atomistic simulations of RNA.

Integrated computational approaches for spectroscopic studies of molecular systems in the gas phase and in solution: pyrimidine as a test case

Fabio Trani

An integrated computational approach built on quantum mechanical (QM) methods, purposely tailored inter-and intra-molecular force fields and continuum solvent models combined with time-independent and timedependent schemes to account for nuclear motion effects is applied to the spectroscopic investigation of pyrimidine in the gas phase as well as in aqueous and CCl4 solutions. Accurate post-Hartree-Fock methodologies are employed to compute molecular structure, harmonic vibrational frequencies, energies and oscillator strengths for electronic transitions in order to validate the accuracy of approaches rooted into density functional theory with emphasis also on hybrid QM/QM' models. Within the time-independent approaches, IR spectra are computed including anharmonicities through perturbative corrections while UV-vis line-shapes are simulated accounting for the vibrational structure; in both cases, the environmental effects are described by continuum models. The effects of conformational flexibility, including solvent dynamics, are described through time-dependent models based on purposely DFT-tailored force fields applied to molecular dynamics simulations and on QM computations of spectroscopic properties. Such procedures are exploited to simulate IR and UV-vis spectra of pyrimidine in the gas phase and in solutions, leading in all cases to good agreement with experimental observations and allowing to dissect different effects underlying spectral phenomena.

2012

Computational Investigation of Intracellular Signalling Cascades

Siri Camee van Keulen

In order to adjust to changes in the environment, cells can communicate via signalling cascades. A common pathway for intercellular communication is the G-protein-coupled- receptor (GPCR) signal transduction cascade which is triggered by a (chemical or physical) extracellular signal, such as light or small molecules. This external interaction induces conformational changes that can lead either to GPCR activation or deactivation. In the active form, GPCRs can induce the activation of trimeric G proteins, which results in the dissociation of the Galpha from the beta and gamma subunits. Subsequently, active Galpha can interact with other cytosolic proteins, such as adenylyl cyclase (AC), which converts adenosine triphosphate to cyclic adenosine monophosphate (cAMP) upon stimulation. AC activation can be induced via stimulatory Galpha subunit (Galpha_s) interactions, while inhibition is related to inhibitory Galpha (Galpha_i) association. X-ray crystallography is crucial for the computational study of the GPCR signal transduction pathway. However, sometimes important molecules or moieties can become lost or cannot be incorporated during crystallisation procedures, which can have a significant effect on a protein’s conformation and function. In this thesis, three steps in the GPCR signal transduction cascade have been studied, rhodopsin activation (i), the conformation of active Galpha_i (ii) and the interaction between Galpha_i and AC (iii). In all three cases some groups or molecules that are absent in the X-ray structures play a vital role in the function of the proteins. In the first project, rhodopsin’s early intermediates after photon exposure are investigated together with a deprotonation reaction that takes place in the later stages of the activation pathway. The classical molecular dynamics (MD) and mixed quantum mechanics/molecular mechanics MD results suggest that photon exposure induces active site rearrangements which increase the stability of rhodopsin’s deprotonated Schiff base state. A bridging water molecule in the active site that is absent in the X-ray structure appears to play an important role during the deprotonation mechanism, which demonstrates the dependence of the protein’s function on its environment. The second study investigates the conformation of the active soluble form of Galpha_i, which has not yet been fully resolved via experimental studies due to crystallisation difficulties of the lipidated N-terminus. To investigate the effect of lipidation via classical MD simulations, a model of soluble lipid-bound Galpha_i was constructed. The simulations show that Galpha_i can form a hydrophobic pocket for the lipid on the protein surface. Other regions of the protein, important for protein-protein interaction, are adjusted as well. Hence, the post-translational modification appears to have a significant impact on Galpha_i’s active conformation and function, which was not captured via the unlipidated X-ray structures. The last project is devoted to a complex of active soluble Galpha_i and AC (Galpha_i :AC). While the interaction site for Galpha_s on AC is known, the interaction site for Galpha_i has not yet been identified. Classical MD simulations were performed on a docked Galpha_i:AC complex to understand the inhibition mechanisms as well as the differentiation between AC stimulation/inhibition. The results show that Galpha_i binding not only leads to AC inhibition via impeding ATP binding but also by preventing re-activation via Galpha_s through closure of the subunit’s interaction site on AC. Keywords: GPCR signal transduction pathway, G-protein-coupled recepter, rhodopsin, G protein, complex formation, cis-trans isomerisation, adenylyl cyclase.

EPFL2017

Computational studies of the structural and optical properties of organic-inorganic lead halide perovskites.

Ariadni Boziki

The rapid rise of organic-inorganic lead halide perovskites has amazed the photovoltaic community. As the efficiency race continues for this revolutionary class of light-harvesting materials, many questions on the structural, electronic and optical properties of perovskite solar cells have still to be addressed. Computational chemistry as well as computational materials' science and engineering have developed into interdisciplinary and powerful scientific fields. They enclose a portfolio of theories and methods developed in mathematics, physics and chemistry, to help scrutinize and tackle problems that can often not be solved in real experiments. In this dissertation, a compendium of available computational methods that treat matter in the quantum mechanical sense and appropriately link its microscopic behavior with its macroscopic properties are employed. Density Functional Theory (DFT) both in the form of static calculations at zero temperature and in the form of first-principles Molecular Dynamics (MD) simulations have been performed in order to study the structural characteristics, the relative stability and the electronic and optical properties of a variety of lead halide perovskites. Many of the computational studies presented here have been performed in close collaboration with experimental groups. More specifically, in order to tackle the challenge of limited phase stability, the atomistic origins of the preferential stabilization or destabilization of the perovskite phase at room temperature is addressed in the case of mixed cations/mixed halide lead perovskites. Their electronic properties, in turn are investigated in order to examine how mixing affects the properties of the pure compounds. Through these studies, it was possible to formulate new design principles for the synthesis of stable mixed cation/mixed halide perovskites with enhanced optical properties. Furthermore, some fundamental open issues related to the nature of the excited states of lead halide perovskites are addressed by simulating the transient absorption spectra and by studying the possible formation of polarons at room temperature for cesium lead bromide. An understanding of the correlations between the nature of the excited states and the atomistic characteristics of these materials is paving the way for the design of lead halide perovskites with enhanced optical properties. Finally, the anomalous low-temperature behavior of the photoluminescence spectra of cesium lead bromide is rationalized in terms of the structural characteristics of the material at low temperatures. The exploration of such phenomena allowed to gain more insights about the properties of organic-inorganic lead halide perovskites that may make them more suitable for applications within and beyond photovoltaics.