Development of Xanthine Based Inhibitors Targeting
3.2. Materials and methods
3.2.1. Manual designing of xanthine based ligands
Xanthine derivatives are known for their phosphodiesterase inhibition property in
their small size that could be responsible for their not fitting properly in the active site pocket of PDEs. Substitution of xanthine derivatives with fragments according to the active site amino acid composition of the targeted protein can lead to formation of compounds having higher affinity for PDE9A. These derivatives need to be constructed over ‘xanthine’ scaffold. Therefore, modification of xanthine at different positions can give specific compounds targeting PDE9A. It has been reported that substitution at N1
and N3 positions with increasing chain length increases the affinity towards phosphodiesterases and adenosine receptors (Miyamoto et al., 1994; van Galen et al., 1992). Substitution at C8 position of xanthine with aromatic ring (aryl/cycloaryl/
heteroaryl group) also increases the potency towards adenosine receptors (Bandyopadhyay et al., 2012; Chappe et al., 1998). It has also been reported that substitution at N7 position is not favorable while in some cases N9 substitution shows negative effect (Azam et al., 2009). Therefore, in this study, N1, N3, C8 and N9 positions were chosen for the initial phase of modifications of the xanthine scaffold. Here,
“xanthine” was used as a scaffold to construct various derivatives by manual designing approach. Fragments R1, R2, R3 and R4 were chosen for substitution/modification at N1, N3, C8 and N9 positions of the xanthine scaffold, respectively. Ligands with different substituents were manually designed in ChemDraw Ultra 8.0. Figure 3.2 is an illustration of the selected positions where modifications/substitutions at xanthine were carried out in order to construct potent and specific inhibitors for PDE9A.
3.2.2. Molecular docking of manually designed xanthine derivatives with PDE9A
Molecular docking is an in silico approach to predict binding affinity of ligands (substrate or inhibitor) in the active site pocket of enzymes by various interactions such as hydrogen bond interaction, hydrophobic interaction, van der Waals interaction and electrostatic interactions. AutoDock is an important docking tool to identify small molecules which have the possibility to be a future drug by studying their mode of action, interaction pattern and affinity towards a particular enzyme. It shows the interaction between partially flexible macromolecule (i.e. side chains in macromolecules are flexible) and flexible ligands (http://autodock.scripps.edu/). It estimates the free energy of binding by using scoring function based on the AMBER force field (Kumar et al., 2014). In this study, docking was carried out by AutoDock 4.2 using Lamarckian Genetic Algorithm (LGA).
Protein-ligand docking studies were initiated by extracting crystal structure of coding domain of phosphodiesterase proteins - PDE1B, PDE2A, PDE3B, PDE4D, PDE5A, PDE7A, PDE8A, PDE9A and PDE10A from RCSB Protein Data Bank (http://www.rcsb.org; PDB IDs were 1TAZ, 1Z1L, 1SOJ, 1ZKN, 1RKP, 3DBA, 3G3N, 3ECM, 2HD1, 4DFF respectively). Prior to docking, all heteroatoms including ligands and water molecules were removed from the crystal structure using Swiss PDB Viewer.
Two metal ions zinc and magnesium were assigned with charge +2. Macromolecule file for docking was prepared in Auto Dock Tool (ADT) by removing polar hydrogen followed by addition of non-polar hydrogen, computation of gasteiger charges and merging of non-polar hydrogen. Each ligand file was prepared separately by using PRODRG server and ChemDraw Ultra 8.0. Energy minimization of newly designed
extracted from PDB were in the form of 3D co-crystallized structure with inhibitor or substrate, their site of interaction was chosen as the binding site for making grid file.
Parameters for making grid file comprised of 90 points in x, y and z directions with equal spacing of 0.253Å. Each protein was used as rigid model with flexible side chains.
Flexible ligand models were used for docking and further optimization. Manually designed ligands were docked with PDEs by keeping common parameters such as 100 GA run, 300 population sizes, 27000 maximum numbers of generation and 25000000 maximum numbers of evaluations. Clustering of docked complexes was created with 2Å root-mean-square deviation tolerance. Molecular docking was carried out in CentOS Linux system. Docking result was analyzed with the help of AutoDock Tool, PyMol, and Discovery Studio Visualizer which provided information about hydrogen bond interactions and π - π interactions. Hydrophobic interaction, electrostatic interaction and van der Waals interaction were analyzed by pose-view (http://poseview.zbh.uni- hamburg.de/) and Lig Plot (Wallace et al., 1995).
3.2.3. Comparative analysis of pharmaceutical properties of selected compounds
Pharmaceutical properties are among the most important parameters for a compound to be suitable drug candidate. Prediction of drug likeness properties is the final deciding factors for in silico study of inhibitors to qualify for being taken up as future drug candidate. The best compounds obtained from in silico studies, were subjected to study drug-likeness properties. These properties include rule of 5, leadlike rule, CMC like rule, MDDR like rule and BBB permeability. PreADMET software was used to calculate the drug likeness, ADMET and toxicity properties of the selected
3.2.4. Comparative binding study of selected compounds with existing xanthine derivatives
Numerous natural and synthetic xanthine derivatives have been reported with their nonspecific PDE inhibitory activity. Comparative interaction study with existing xanthine based inhibitors was required to ensure the potency of selected compounds towards PDE9A. IBMX is the most common xanthine derivative for most of the phosphodiesterase inhibition. But in case of PDE9A, IBMX does not show any inhibition (Huai et al., 2004). Among xanthine derivatives, only IBMX co-crystal structure with PDE9A has been submitted in Protein Data Bank. This co-crystal structure was extracted from data bank and redocked. The binding affinity of selected compounds for PDE9A from manual designing was compared with the docking result of IBMX-PDE9A complex. Hence, in the present study IBMX was used as a reference molecule (negative control) to analyze the potency of selected molecules towards PDE9A. The docking parameters used for these comparative analyses were 100 GA run, 300 population size, 27000 maximum number of generation and 25000000 maximum number of evaluations.
3.2.5. Comparative inhibition study of selected compounds and known PDE9A inhibitor
In last two decades, extensive work has been carried out in development of potent inhibitors for PDE9A to treat different diseases associated with lowering the level of cGMP (da Silva et al., 2013; Hutson et al., 2011; Nicholas et al., 2011; Singh and Patra, 2014; Verhoest et al., 2012; Wunder et al., 2005). Surprisingly, none of them belong to the xanthine class of compounds. Most of the reported PDE9A inhibitors have been constructed over “pyrazolopyrimidinone” scaffold. Amongst them, BAY73-9961 is an
carried to get the binding pattern inside the active site pocket of PDE9A. A comparative study of the reported inhibitor with selected compounds for PDE9A was carried out by comparative analysis. This was helpful to understand the binding pattern and the role of various active site residues in ligand binding.