Enthalpic and entropic contributions to biomolecular recognition

A fundamental understanding of biomolecular recognition is crucial to grasp complex processes in cells, e.g., signaling pathways, cell development, or even the effects of drugs. Water and especially water replacement can be one of the main driving forces for biomolecular recognition, but its role in many biological processes is still elusive. On one hand, water can help to bridge enthalpic interactions, hence, contributing to the enthalpic part of binding. On the other hand, also the replacement of entropically unfavored, bound water molecules can lead to entropic binding contributions. To elucidate the role of water in biomolecular recognition, we employed physics-based computational chemistry methods to intensively investigate the thermodynamic quantities of water molecules in close proximity to amino acids (AAs), proteins, and protein ligand complexes. As a central part of this thesis, I present a new, broadly applicable AA hydrophobicity scale based on molecular dynamics (MD) simulations analyzed with grid inhomogeneous solvation theory (GIST). Therefore, the solvation thermodynamics of single, capped AAs was analyzed. The scale we developed from these results shows two main advantages in contrast to previously published models: Firstly, the entropic contributions are calculated directly from the phase space water molecules can occupy; secondly, a spatial resolution of thermodynamic quantities, e.g., enthalpy or entropy, is provided. As a result, we gain information on the local differences in hydrophobicity in the environment of an AA. Applying an analogous methodology, we investigated the distinct binding modes of highly similar ligand fragments in the kinase domain of TGFBR1. With the help of GIST and additional pKA calculations, we were able to explain the highly interesting binding pose flips of these fragments. Whereas one binding pose is rationalized by enthalpic interactions between the ligand and the protein due to formation of a salt-bridge, the second binding pose is explained through the replacement of entropically unfavorable water molecules. In further studies, the ice recognition ability of antifreeze proteins (AFPs) was investigated with the combined methods of MD and GIST. We were able to show that the entropic and enthalpic contributions to hydration for active ice-binding sites show very unique properties especially the ratio between these thermodynamic values is key for their ability to bind ice. We used the obtained insight to further improve the currently proposed mechanism of ice-binding, adding an additional piece to the puzzle of molecular recognition involving ice: weak enthalpic interactions between a protein and its hydrating water allow this hydration layer to adapt an ice-like conformation, which is beneficial for ice-binding. Imitating biomolecular recognition with artificial materials is not only of academic interest, understanding, these processes is crucial also for industrial use. We investigated the theoretical performance limitations of molecularly imprinted polymers (MIPs) a prominent representative of such artificial materials. It was possible to show that the heterogeneity of binding sites restricts their ability to differentiate between similar molecules, not allowing these MIPs to show selectivities similar to proteins. Besides the proof and examination of GIST as a reliable tool to investigate protein and protein ligand hydration, further contributions to the field of proteases were made. In the field of protease science, the convergence behavior and reliability of a knowledge-based protease specificity metric, the cleavage entropy, was examined. This metric intrinsically depends on the known number of substrates: we were able to demonstrate that at least 30 substrates are necessary to adequately describe the specificity of a protease. Furthermore, we used the substrate data as a training set for a virtual screening approach for the identification of active small molecules. The developed approach shows excellent performance for the four tested proteases with the used data set. By correlating protease substrate specificity with conformational binding enthalpy contributions, we could show that considering the crystallographically obtained structure alone does not suffice. In order to correctly explain specificity, an ensemble of conformations has to be considered. Furthermore, a fruitful collaboration with our Institutes division of inorganic chemistry was established in the course of my PhD studies. Density functional theory methods were used to assist fundamental research efforts in solid state chemistry. Theoretical calculations allowed to interpret experimentally obtained vibrational spectra, and to predict band structures and band gaps of novel materials.A fundamental understanding of biomolecular recognition is crucial to grasp complex processes in cells, e.g., signaling pathways, cell development, or even the effects of drugs. Water and especially water replacement can be one of the main driving forces for biomolecular recognition, but its role in many biological processes is still elusive. On one hand, water can help to bridge enthalpic interactions, hence, contributing to the enthalpic part of binding. On the other hand, also the replacement of entropically unfavored, bound water molecules can lead to entropic binding contributions. To elucidate the role of water in biomolecular recognition, we employed physics-based computational chemistry methods to intensively investigate the thermodynamic quantities of water molecules in close proximity to amino acids (AAs), proteins, and protein ligand complexes. As a central part of this thesis, I present a new, broadly applicable AA hydrophobicity scale based on molecular dynamics (MD) simulations analyzed with grid inhomogeneous solvation theory (GIST). Therefore, the solvation thermodynamics of single, capped AAs was analyzed. The scale we developed from these results shows two main advantages in contrast to previously published models: Firstly, the entropic contributions are calculated directly from the phase space water molecules can occupy; secondly, a spatial resolution of thermodynamic quantities, e.g., enthalpy or entropy, is provided. As a result, we gain information on the local differences in hydrophobicity in the environment of an AA. Applying an analogous methodology, we investigated the distinct binding modes of highly similar ligand fragments in the kinase domain of TGFBR1. With the help of GIST and additional pKA calculations, we were able to explain the highly interesting binding pose flips ofby Michael SchauperlAbweichender Titel laut Übersetzung der Verfasserin/des VerfassersKumulative Dissertation aus 14 ArtikelnUniversity Innsbruck, Dissertation, 2017OeBB(VLID)188137