Coarse-Grained Force Field Development

Coarse-grained modeling is a powerful tool to gain insight into the nanoseconds to (sub)milliseconds dynamics of (bio)molecular systems. Thereby, multiple atoms are grouped together in one interaction site reducing the number of degrees of freedom. Thus, the complexity of the molecular systems is lower which also leads to a faster sampling of the configurational space.

The most popular coarse-grained force field is Martini, for which all major classes of biomolecules – lipids, proteins, sugars, DNA, RNA – are available. In Martini, the chemical properties of the atom groups are primarily taken into account when grouping together multiple atoms to an interaction site. Moreover, functional groups are kept together whenever possible. An acidic group for example does not get split to multiple interaction sites. This strategy guarantees the versatility of the coarse-grained force field Martini. Nowadays, it is not only used for biological systems but it is getting more and more popular in material science as well.

My interests are particularly in the development of force field parameters for lipids and proteins. An improved modeling of structural properties of proteins and lipid membranes as well as an improvement of the flexibility and dynamics of proteins are an important aspect. With respect to the Martini protein model, we are involved in combining structure-based coarse-grained models, such as a Gō-like model, with Martini 3. In addition, we also contributed to the re-parametrization of the Martini 3 model.

Protein-Ligand Interactions

We are also interested in modeling protein-ligand interactions. A detailed understanding of the most important contributions to the energetic stabilization of small molecules (ligands) in protein pockets is decisive to develop new medicinal agents. In this way, the interaction in a specific protein pocket can be strengthened and the specificity of the small molecule towards the target protein can be improved. In addition, a detailed microscopic understanding of protein-lipid interactions is important to unravel cellular processes such as signaling processes.

On the one hand, we work with model systems such as the protein T4 lysozyme which is used as a test system to improve simulation methods. On the other hand, we model pharmacologically important proteins such as G-protein coupled receptors or kinases. Furthermore, we also study proteins which specifically bind one type of lipids. One example is the Tubby protein, which binds the important signaling lipid PI(4,5)P₂. By means of the Tubby protein, we study the behavior of cells after signaling processes during which PI(4,5)P₂ initially is depleted.

To simulate protein-ligand interactions we mostly use coarse-grained molecular dynamics, but also methods at atomistic resolution.

Impact of the Environment on Molecular Reactivity

Molecular reactivity is influenced by the environment. In chemical systems, this is often the solvent; in biological systems lipid membranes or protein pockets are potential environments, too. Besides electrostatics, the flexibility and dynamics of the surrounding plays an important role. The surrounding can e.g. hinder an atomic movement which is necessary for a certain reaction to proceed. Thus, an alternative reaction can take place. Typical biochemical examples are proteins, which orient the reactants in advance so that these can only form one out of two possible enantiomers.

In particular, we studied the dynamical impact of the solvent cage on ultrafast chemical reactions. It turned out that the solvent cage can decisively alter the direction of a bond cleavage so that specific points of the molecular configuration space are reached, so-called conical intersections. They decide the reaction outcome. To model the dynamic solvent effect, we mostly use quantum dynamics which we combine in parts with classical molecular dynamics.

Light-Induced Processes

We are interested in photochemical and photophysical processes as well, which can be initiated by light. The photon energy can be absorbed by molecules and results in an excited electronic and/or vibrational state. This additional energy of the molecule can be used for chemical reactions or it can be released via physical processes.

On the one hand, we worked with light-induced bond cleavages generating reactive molecules. On the other hand, we investigated the control of chemical reactions or of photophysical processes by means of specially shaped light pulses. For example we were able to show that the relaxation of the isolated RNA base uracil after UV excitation is almost optimal and that it is difficult to speed it up. In contrast, a long delay of the relaxation can be achieved using a specially shaped light pulse.

In chemistry, spectroscopic methods are commonly applied to follow the reaction progress. Light-induced processes can be followed for example by means of transient absorption spectroscopy. We also simulated transient absorption spectra on the femto- to nanosecond time scale to better understand experimental spectra. Thus, we could unravel the ultrafast generation of reactive products and their secondary reactions.

In this field we majorly use quantum chemical and quantum dynamical methods as well as optimal control theory.

Photosynthesis

Photosynthesis is a fascinating process during which plants and some bacteria transform light energy into chemical energy. The generated chemical energy storage such as sugars serve countless living organisms as energy resource. By means of multiscale modeling, we aim to improve the current understanding of photosynthesis. In doing so, we are not only aiming at a more clear microscopic picture of biological photosynthesis, but also at contributing to find better technical ways to use the sun as an energy source.

In particular, we are interested in the interplay between protein dynamics and quantum efficiency, the regulation of plant photosynthesis on the level of the thylakoid membrane, as well as in the impact of protein-protein and protein-lipid interactions. We typically apply coarse-grained molecular dynamics and linear response theory to tackle these questions.