Author:

Abstract:

The goal of this PhD project was to use computational methods such as molecular docking and molecular dynamics (MD) simulations to gain a detailed insight into the structural basis of particular protein-ligand complexes. Here these methods were applied on two case studies: a family of Ca2+ transporters called Sarco-/Endoplasmic Reticulum Ca2+-ATPase (SERCA), implicated in heart failure, and biofilms, an aggregate of micro-organisms attached to a surface. Heart failure (HF) is marked by a reduced cardiomyocyte contractility that at least in part is related to a reduced SERCA2a expression. Improving SERCA2a function is therefore a desired therapeutic approach. In the SERCA project we aimed to understand the structural basis of the increased Ca2+ affinity of SERCA2b over SERCA2a, splice variants of the ATP2A2 gene. The presence of an extended C-terminus, the 2b-tail, in SERCA2b accounts for this Ca2+ affinity increasing effect and may serve as a template for the design of molecules that improve the Ca2+ affinity of SERCA2a. Two functional parts are recognized in the 2b-tail: an eleventh transmembrane (TM) helix, TM11, and a luminal extension (LE). LE shows the most promising functional properties and was here used for further functional characterization. Here we identified the presence of a transient D281-K876 salt bridge in SERCA2a and a transient K876-D1038 salt bridge in SERCA2b (i.e. between the pump and LE of the 2b-tail). We investigated their functional importance using MD simulations and in vitro activity and cross-linking experiments. Furthermore the importance of L873 and K876 (in the common parts of the SERCA2a/b pump) and F1040 and W1041 (within LE) for the interaction of the pump with LE was confirmed. Using this information, we developed a pharmacophore model of the LE interaction and we searched for small molecules that can mimic this effect in an initial attempt to search for a therapeutically interesting molecule for the treatment of HF. A total of 10 compounds were tested on both the biochemical Ca2+ ATPase activity of SERCA2a and SERCA1a, the skeletal muscle isoform. 5 molecules displayed interesting allosteric modulatory effects and specificity on SERCA2a and/or SERCA1a. To our knowledge these are the first compounds reported that can affect the SERCA2a activity. In the biofilm project we aimed, as part of a multidisciplinary project, to build in silico drug-target networks with publically available information to discover new anti-biofilm agents. An improved machine-learning algorithm, the 3D neighbourhood kernel (3DNK), will try to learn from this data to predict new protein-ligand interactions allowing us to build drug-target networks for the organism of interest. Through interrogation of the drug-target networks, we aimed to discover new molecules or repurpose existing drugs with potential anti-biofilm properties. In the first two tests we showed that the 3DNK can successfully distinguish positive interactions (ligands with low and intermediate binding affinity values) and negative interactions (ligands with high binding affinity values). Future improvements on the 3DNK will allow us to enhance its predictive power even further.