University, Diss., 2004--Mainz.

Techn. University, Diss., 2004--Berlin.

Universiẗat, Diss., 2004--Stuttgart.

University, Diss., 2006--Frankfurt (Main). / Zsfassung in engl. und dt. Sprache.

Techn. University, Diss., 2006--Darmstadt.

The aim of this thesis was to develop new automatic enhanced sampling methods by extending the idea of Parrinello’s metadynamics to multistate problems and by introducing new quantum-mechanical electronic collective variables. These methods open up a rich perspective for applications to the photophysical processes in complex molecular systems, which play a major role in many natural processes such as vision and photosynthesis, but also in the development of new materials for organic electronics, whose function depends on specific electronic properties such as biradicalicity. / Das Ziel dieser Arbeit war die Entwicklung neuer automatisierter Methoden für beschleunigtes Sampling molekularer Strukturen durch eine Erweiterung von Parrinellos Metadynamik auf Mehrzustandsprobleme und die Verwendung neuer quantenmechanischer elektronischer kollektiver Variablen. Die entwickelten Methoden bieten einen breiten Anwendungsspielraum im Bereich der photophysikalischen Prozesse komplexer molekularer Systeme, welchen in vielen natürlichen Vorgängen wie beispielsweise dem Sehen und der Photosynthese, aber auch in der Entwicklung neuer Materialien für die organische Elektronik eine Schlüsselrolle zukommt. Die Eigenschaften solcher funktioneller Materialien werden durch spezifische elektronische Eigenschaften wie dem Biradikalcharakter bestimmt.

This paper presents a homogeneous system of magnetic colloidal particles that self-assembles via two structural patterns of different symmetry. Based on a qualitative comparison between a real magnetic particles system, analytical calculations and molecular dynamics simulations, it is shown that bistability can be achieved by a proper tailoring of an anisotropic magnetization distribution inside the particles. The presented bistability opens new possibilities to form two-dimensionally extended and flexible structures where the connectivity between the particles can be changed in vivo. / Dieser Beitrag ist aufgrund einer (DFG-geförderten) Allianz- bzw. Nationallizenz frei zugänglich.

Wassermoleküle spielen oft eine entscheidende Rolle bei der Bindung von Liganden an Proteine. Zum einen ist dies in ihrer Eigenschaft als Wasserstoffbrückendonor und -akzeptor begründet, die es ermöglicht Wechselwirkung zwischen Ligand und Rezeptor zu vermitteln. Zum anderen stellen die Desolvatisierungsenthalpie und -entropie einer Bindetasche während der Ligandbindung einen entscheidenden Anteil der Bindungsaffinität dar. Obwohl man sich dieser Einflüsse seit langem bewusst ist, sind aktuelle Methoden des computerbasierten Wirkstoffdesigns nur in sehr begrenztem Umfang in der Lage, die entsprechenden Effekte zu erfassen und vorherzusagen. Da experimentelle Daten über die Effekte von Wassermolekülen in Protein-Ligand Komplexen von Natur aus schwierig zu erhalten sind, untersucht die vorliegende Arbeit eine Modellbindetasche einer Cytochrom c Peroxidase Mutante (CCP W191G) mit Hilfe von Molecular Modeling Techniken. Diese polare und solvatisierte Kavität ist strukturell sehr gut charakterisiert und bindet kleine, kationische Heterozyklen zusammen mit unterschiedlichen Mengen an Wassermolekülen. Für die Untersuchungen wurden strukturell ähnliche Liganden mit einem unterschiedlichen Wechselwirkungsmuster ausgewählt. Davon ausgehend wurde die Möglichkeit zweier Docking-Programme, den Grad der Wasserverdrängung durch den Liganden zusammen mit dem Bindungsmodus vorherzusagen, untersucht. Die dynamischen Eigenschaften der Bindetaschenwassermoleküle wurden mittels Molekulardynamiksimulationen studiert. Schließlich wurden diese rein strukturellen Betrachtungen durch eine energetische/thermodynamische Analyse komplettiert. Die Anwendung dieser unterschiedlichen Verfahren liefert einige neue Erkenntnisse über die untersuchte Modellbindetasche. Trotz der relativen Einfachheit der kleinen Kavität der CCP W191G Mutante war die vollständige Charakterisierung und eine korrekte (retrospektive) Vorhersage des Wasser-Wechselwirkungsmuster der Ligand-Komplexe nicht trivial. Zusammenfassend kann man festhalten, dass insgesamt eine gute Übereinstimmung zwischen den durch Computersimulationen erhaltenen Ergebnissen und den kristallographischen Daten erzielt wurde. Unerwartete Befunde, die auf den ersten Blick mit den kristallographischen Beobachtungen nicht übereinstimmen, können ebenso durch Limitationen in den Kristallstrukturen bedingt sein. Darüber hinaus gaben die Ergebnisse auch eine Hilfestellung, welches Verfahren zur Beantwortung einer Fragestellung im Rahmen von Wassermolekülen im Wirkstoffdesign geeignet sind. Schließlich wurden ebenso die Begrenzungen der jeweiligen Methoden aufgezeigt. / Water molecules play an important role for the binding of small molecule ligands to proteins. One of the reasons for this is their ability to act as a hydrogen bond donor and acceptor at the same time. Additionally, the enthalpy and entropy of desolvation of the pocket is one large contribution to the overall binding affinity. Although this is long known, prediction of these effects by current methods of computer-aided drug design is rather limited. Since experimental information about water effects in protein-ligand complexes are inherently difficult to obtain, in the present work a well-suited model binding site of a mutant of the cytochrome c peroxidase (CCP W191G) is studied using molecular modeling techniques. This polar and solvated cavity is structurally very well characterized and several small, cationic heterocycles bind together with a different amount of water molecules. For this study structurally similar ligands which have a different interaction pattern where chosen. First, the ability of two docking programs to predict cavity desolvation upon ligand binding was investigated. The dynamic properties of the binding site water molecules where studied by means of molecular dynamic simulations. Ultimately, the pure structural considerations addressed in this work were complemented by an energetic/thermodynamic analysis. The application of the different methods offered some new insights into the studied model binding site. Despite the relative simplicity of the small cavity of the CCP W191G mutant, a complete characterization and a correct (retrospective) prediction of the water interaction network in ligand complexes of this model binding site is not trivial. In summary, an overall good agreement between computational results and crystallographic data is obtained. Unexpected findings, which at first sight disagree with crystallographic observations, may also be due to limitations of the crystal structures. In addition, the results help to decide which method is appropriate to address a certain question in the context of water molecules in drug design. Also, the limitations of the respective methods are exposed.

Diese Arbeit charakterisiert die Solvatation des organischen Farbstoffes Coumarin 153 in Gemischen aus jeweils einem Alkan und einem Alkohol. Dabei werden Methoden der statischen und zeitaufgeloesten optischen Spektroskopie sowie klassische molekulardynamische Simulationen fuer die Untersuchungen angewendet. Die experimentellen Ergebnisse zeigen, dass der Farbstoff im Gemisch selektiv durch den Alkohol solvatisiert ist. Die Staerke dieser Solvatation ist für den elektronischen Grund- und Anregungszustand des Farbstoffes unterschiedlich. Aus diesem Grund wird die Solvatationsdynamik in Alkohol/Alkan Gemischen durch einen Translationsdiffusionsprozess bestimmt. Die molekulardynamischen Simulationen veranschaulichen die selektive Solvatation des Coumarin 153 in einem Methanol/Hexan Gemisch. Die Solvathuelle enhaelt im Anregungszustand des Farbstoffes bis zu dreimal mehr Molekuele als im Grundzustand. Im Unterschied zur Solvatation in reinem Methanol spielen spezifische Bindungen wie Wasserstoffbrueckenbindungen in einem Methanol/Hexan Gemisch eine wesentliche Rolle.

Glycosaminoglycans (GAGs) are linear polysaccharides, built of periodically occurring disaccharide units. GAGs are ubiquitous in the extracellular matrix (ECM), where they exhibit multifarious biological activities. This diversity arises from - among others - their ability to interact with and regulate a large number of proteins, such as cytokines, chemokines, and growth factors. As of the huge variety in their chemical configuration, GAGs are further sub-classified into different types (heparin, for instance, is one of these sub-classes). Hence, GAGs are a diverse class of molecules, which surely contributes to the broadness of their spectrum of biological functions. Through varying arrangements of sulfate groups and different types of saccharide units, individual GAG molecules can establish specific atomic contacts to proteins. One of the best-studied examples is antithrombin-heparin, whose biologically relevant interaction requires a specific pentasaccharide sequence. It is valid to assume, however, that various proteins are yet to be discovered whose biological functions are in some way affected by GAGs. In other cases, and this is true for the cytokine interleukin-10 (IL-10), there are already experimental indications for a biologically relevant protein-GAG interaction, but the details are still obscure and the fundamental molecular interaction mechanism has still not been clarified. IL-10 has been shown to bind GAGs. So far, however, no structural detail about IL-10-GAG interaction is known. Function-wise, IL-10 is mainly considered to be immunosuppressive and therefore anti-inflammatory, but it in fact has the pleiotropic ability to influence the immune system in both directions, i.e. it constitutes a complex regulation system on its own. Therefore, the role of GAGs in this system is potentially substantial, but is yet to be clarified. In vitro experiments have yielded indications for GAGs being able to modulate IL-10\'s biological function, and obviously IL-10 and GAGs are simultaneously present in the ECM. This gives rise to the assumption that IL-10-GAG interaction is of biological significance, and that understanding the impact of GAGs on IL-10 biology is important - from the basic research point of view, but also for the development of therapies, potentially involving artificially designed ECMs. A promising approach for obtaining knowledge about the nature of IL-10-GAG interaction is its investigation on the structural level, i.e. the identification and characterization of the molecular interaction mechanisms that govern the IL-10-GAG system. In this PhD project it was my goal to reveal structural and molecular details about IL-10-GAG interaction with theoretical and computational means, and with the help of experiments performed by collaborators in the framework of the Collaborative Research Centre DFG Transregio 67. For achieving this, I developed three methods for the in silico investigation of protein-GAG systems in general and subsequently applied them to the IL-10-GAG system. Parts of that work have been published in scientific journals, as outlined further below. I proposed and validated a systematic approach for predicting GAG binding regions on a given protein, based on the numerical simulation and analysis of its Coulomb potential. One advantage of this method is its intrinsic ability to provide clues about the reliability of the resulting prediction. Application of this approach to IL-10 lead to the observation that its Coulomb attraction for GAGs is significantly weaker than in case of exemplary protein-GAG systems (such as FGF2-heparin). Still, a distinct IL-10-GAG binding region centered on the residues R102, R104, R106, R107 of the human IL-10 sequence was identified. This region can be assumed to play a major role in IL-10-GAG interaction, as described in chapter 3. Molecular docking methods are used to generate binding mode predictions for a given receptor-ligand system. In chapter 4, I clarify the importance of data clustering as an essential step for post-processing docking results and present a clustering methodology optimized for GAG molecules. It allows for a reproducible analysis, enabling systematic comparisons among different docking studies. The approach has become standard procedure in our research group. It has been applied in a variety of studies, and served as an essential tool for studying IL-10-GAG interaction, as described in chapter 3. Motivated by the shortcomings of classical docking approaches, especially with respect to protein-GAG systems, I worked on the development of a molecular dynamics-based docking method with less radical approximations than usually applied in classical docking. The goal was to make the computational model properly account for the special physical properties of GAGs, and to include the effects of receptor flexibility and solvation. The methodology was named Dynamic Molecular Docking (DMD) and published in the Journal of Chemical Information and Modeling-together with a validation study. The subsequent application of DMD in a variety of studies required enormous amounts of computational resources. For tackling this challenge, I established a graphics processing unit-based high-performance computing environment in our research group and developed a software framework for reliably performing DMD studies on this hardware, as well as on other computing resources of the TU Dresden. The investigation of the IL-10-GAG system via DMD was focused on the IL-10-GAG binding region predicted earlier, and made heavy usage of the optimized clustering approach named above. An important result of this endeavor is that IL-10's amino acid residue R107 significantly stands out compared to all other residues and supposedly plays a particularly important role in IL-10-GAG recognition. The collaboration with the NMR laboratory of Prof. Daniel Huster at the Universität Leipzig was fruitful: I post-processed nuclear Overhauser effect data and obtained heparin structure models, which revealed that IL-10-heparin interaction has a measurable impact on the backbone structure of the heparin molecule. These results were published in Glycobiology. In chapter 8, I propose two different scenarios about how GAG-binding to IL-10 might affect its biological function, based on the findings made in this thesis project. In conclusion, a set of methods has been developed, all of which are generically applicable for the investigation of protein-GAG systems. Regarding the IL-10-GAG system, valuable structural insights for increasing the understanding about its molecular mechanisms were derived. These observations pave the way towards unraveling GAG-mediated bioactivity of IL-10, which may then be specifically exploited, for instance in artificial ECMs for improved wound healing.