Search

The term glycan refers to a broad category of molecules composed of monosaccharide units linked to each other in a variety of ways, whose structural diversity is related to different functions in living organisms. Among others, glycans are recognized by proteins with the aim of carrying information and for signaling purposes. Determining the three-dimensional structures of protein-glycan complexes is essential both for the understanding of the mechanisms glycans are involved in and for applications such as drug design. In this context, molecular docking approaches are of undoubted importance as complementary approaches to experiments. In this study, we show how high ambiguity-driven DOCKing (HADDOCK) can be efficiently used for the prediction of protein-glycan complexes. Using a benchmark of 89 complexes, starting from their bound or unbound forms, and assuming some knowledge of the binding site on the protein, our protocol reaches a 70% and 40% top 5 success rate on bound and unbound data sets, respectively. We show that the main limiting factor is related to the complexity of the glycan to be modeled and the associated conformational flexibility.

This PhD thesis focuses on the employment of computational approaches for the study of the interactions between biomolecules, a broad term that accounts for different molecular species ranging from proteins to small ligands. Understanding how biomolecules recognize each other, thus giving rise to complexes or assemblies, is a key point for the comprehension of biological mechanisms in living organisms and for application purposes in a variety of fields, among which drug design. These difficult and multidisciplinary issues strongly exploit in silico approaches, which, in the last decades, have become increasingly efficient and essential for supporting and guiding the experiments. The research activity carried out during my PhD work has mainly dealt with two projects. The first one revolves around protein-protein interactions and concerns a specific use-case, namely the necessity to predict how two affitins bind the human epidermal growth factor receptor 2 (HER2). This project has been carried out in collaboration with Dr. Alessandro Maiocchi (Bracco S.p.A – owner of two patents that cover the use of the two affitins as molecular probes targeting HER2), and Dr. Elisabetta Moroni (SCITEC, Italian National Research Council). The second project has been conducted at the Computational Structural Biology group (Bijvoet Centre for Biomolecular Research, Universiteit Utrecht) under the supervision of Prof. Alexandre Bonvin and Dr. Marco Giulini. It aims at building a reliable protocol, based on the software HADDOCK3, which is developed at the CSB group, for the prediction of protein-glycan complexes., This PhD thesis focuses on the employment of computational approaches for the study of the interactions between biomolecules, a broad term that accounts for different molecular species ranging from proteins to small ligands. Understanding how biomolecules recognize each other, thus giving rise to complexes or assemblies, is a key point for the comprehension of biological mechanisms in living organisms and for application purposes in a variety of fields, among which drug design. These difficult and multidisciplinary issues strongly exploit in silico approaches, which, in the last decades, have become increasingly efficient and essential for supporting and guiding the experiments. The research activity carried out during my PhD work has mainly dealt with two projects. The first one revolves around protein-protein interactions and concerns a specific use-case, namely the necessity to predict how two affitins bind the human epidermal growth factor receptor 2 (HER2). This project has been carried out in collaboration with Dr. Alessandro Maiocchi (Bracco S.p.A – owner of two patents that cover the use of the two affitins as molecular probes targeting HER2), and Dr. Elisabetta Moroni (SCITEC, Italian National Research Council). The second project has been conducted at the Computational Structural Biology group (Bijvoet Centre for Biomolecular Research, Universiteit Utrecht) under the supervision of Prof. Alexandre Bonvin and Dr. Marco Giulini. It aims at building a reliable protocol, based on the software HADDOCK3, which is developed at the CSB group, for the prediction of protein-glycan complexes.

The Suzuki-Miyaura reaction between the aryl halide (1) and the phenyl boronic acid (2), in the presence of the palladium(0) complex (3) as catalyst, gives the cross-coupling product (4) in quantitative yield when performed in basic aqueous solution of the nonionic surfactant Kolliphor-EL (K-EL). The partition between the aqueous and micellar environments of the species of this reaction has been investigated by means of Molecular Dynamics (MD) simulations. Starting from the K-EL molecules dispersed in water, a micelle model has been generated by MD simulations, adopting the 2016H66 force field. Reagent and product species have been described with the same force field, once the reliability of this force field has been tested comparing the n-octanol/water partition free energies calculated from the MD and Free Energy Perturbation (FEP) method with those obtained from the quantum-mechanical SMD method. The potential of mean force for the transfer process between water and the micellar phase of the different species has been calculated by the MD simulations and the Umbrella Sampling (US) method. The overall picture that emerges from these results confirms that the molecular species involved in this reaction prefers the micellar environment and concentrates in different but close zones of the micelle. This supports the experimental evidence that the use of suitable surfactant agents promotes reactivity, allowing micelles to behave as nanoreactors in which reactive species are solubilized and enhance their local concentration.

The photogeneration of electricity and solar fuels by solar irradiation in photoelectrochemical cells is one of the sectors with the highest growth potential in the decarbonised society. However, the use of different components, in particular photosensitizers and catalysts, can present problems of charge transfer efficiency at the interface, leading to lower final efficiencies. In this work we present novel integrated photosensitizer-catalyst dyads based on robust and, at the same time, flexible host-guest non-covalent interactions through the use of calix[4]arene cavities. Current photogeneration in photoelectrochemical cells showed twice greater efficiency in the integrated calixarene-based host-guest dyads compared to the traditional architecture based on the separate photosensitizer-catalyst pair. Molecular dynamics studies have shown that the enhanced performance originates from an optimization of the distances between the centres of the photosensitizer, catalyst and semiconductor involved in the charge transfer processes, thus allowing a higher final efficiency of the charge photogeneration process.

Affitins constitute a class of small proteins belonging to Sul7d family, which, in microorganisms such as Sulfolobus acidocaldarius, bind DNA preventing its denaturation. Thanks to their stability and small size (60–66 residues in length) they have been considered as ideal candidates for engineering and have been used for more than 10 years now, for different applications. The individuation of a mutant able to recognize a specific target does not imply the knowledge of the binding geometry between the two proteins. However, its identification is of undoubted importance but not always experimentally accessible. For this reason, computational approaches such as protein-protein docking can be helpful for an initial structural characterization of the complex. This method, which produces tens of putative binding geometries ordered according to a binding score, needs to be followed by a further reranking procedure for finding the most plausible one. In the present paper, we use the server ClusPro for generating docking models of affitins with different protein partners whose experimental structures are available in the Protein Data Bank. Then, we apply two protocols for reranking the docking models. The first one investigates their stability by means of Molecular Dynamics simulations; the second one, instead, compares the docking models with the interacting residues predicted by the Matrix of Local Coupling Energies method. Results show that the more efficient way to deal with the reranking problem is to consider the information given by the two protocols together, i.e. employing a consensus approach.

Intrinsically Disordered Peptides and Proteins (IDPs) in solution can span a broad range of conformations that often are hard to characterize by both experimental and computational methods. However, obtaining a significant representation of the conformational space is important to understand mechanisms underlying protein functions such as partner recognition. In this work, we investigated the behavior of the Sic1 Kinase-Inhibitor Domain (KID) in solution by Molecular Dynamics (MD) simulations. Our results point out that application of common descriptors of molecular shape such as Solvent Accessible Surface (SAS) area can lead to misleading outcomes. Instead, more appropriate molecular descriptors can be used to define 3D structures. In particular, we exploited Weighted Holistic Invariant Molecular (WHIM) descriptors to get a coarse-grained but accurate definition of the variegated Sic1 KID conformational ensemble. We found that Sic1 is able to form a variable amount of folded structures even in absence of partners. Among them, there were some conformations very close to the structure that Sic1 is supposed to assume in the binding with its physiological complexes. Therefore, our results support the hypothesis that this protein relies on the conformational selection mechanism to recognize the correct molecular partners.