Computational research of superoxide dismutase 1 and its modulators by CADD approaches to reveal potential drug candidates for amyotrophic lateral sclerosis treatment
Provider:
IT4Innovations
Investigator:
Rafael Dolezal, Michaela Melikova, Jan Trejbal, Ondrej Krejcar
Abstract:
Amyotrophic lateral sclerosis (ALS) is a severe neurodegenerative disease that leads to destruction of motoric neurons and finally to profound muscular paralysis. According to recent reports, this sporadic illness is linked with mutations of superoxide dismutase 1 (SOD1) enzyme that exhibits, besides its normal catalytic functions, also considerable side toxic effects. Experiments with SOD1 knock-out mice have shown no ALS symptoms, while transgenic mice with incorporated malfunctioning human gene G93A of SOD1 have developed the symptoms [1]. The latest findings have also proved abundant proteinaceous aggregates of SOD1 in post-mortem brains of ALS patients. Although extensive biochemical experiments have confirmed a clear connection between malfunctioning SOD1 and ALS, especially the familial type, no drug has been discovered so far to treat ALS by modulation of SOD1. In the project implementing advanced computer-aided drug design (CADD) methods, we will investigate the structure of available SOD1 3D X-ray models by bioinformatics tools to reveal the extent of the mutations, further the structure of SOD1 will be deeply analyzed to find potential binding sites, then structure-based virtual screening (SBVS) by high-throughput flexible molecular docking will be performed to reveal potential drug candidates for modulation of SOD1. Finally, the interaction stability of top-scoring candidates in SOD1 will be evaluated by molecular dynamics. All these innovative computational approaches represent extremely demanding tasks, which need parallelized computer performance to be accomplished in reasonable time span. In particular, the most exhaustive calculations within SBVS will be carried out by self-developed applications utilizing “pleasing” parallelization principles and message passing interface (MPI) based techniques. Remaining tasks such as molecular dynamics will be performed with the use of CPU/GPU hybridized computational protocols.
Project duration:
23-May-2017–18-February-2018
Discovery of first small-molecule orexin receptor type 2 (OX2R) agonists as orphan drugs to combat narcolepsy by molecular dynamics and virtual high-throughput screening
Provider:
IT4Innovations
Investigator:
Eugenie Nepovimová, Rafael Doležal, Jan Korábečný, Tomáš Kučera, Kamil Kuča
Abstract:
Narcolepsy is a chronic neurologic disease characterized by excessive daytime sleepiness. Etiology of this rare disease with prevalence about 0.05% is still not fully elucidated. The year 1998 meant a milestone in narcolepsy research with the discovery of the neuropeptide – orexin, the absence of which is now believed to be responsible for most of the symptoms of narcolepsy. Currently, patients with narcolepsy are treated symptomatically. An alternative to the symptomatic treatment of narcolepsy with cataplexy would be a direct orexigenic system-targeted therapy. Experiments on orexin/ataxin-3 narcolepsy mouse model have shown that
intracerebroventricular (ICV) injections of orexin A can induce activation of orexin receptors and thus compensate for orexin cell loss and reverse most of the narcolepsy symptoms, including cataplexy. As ICV injections are not feasible in humans, for this reason, IV application methods have also been investigated. However, the size of both orexin neuropeptides has suggested that they would be unlikely to penetrate the blood brain barrier (BBB), this assumption was confirmed. Therefore, development of small-molecule orexin receptors agonists that are able to pass across the BBB represents a hopeful way. Searching for orexin receptor type 2 (OX2R) agonists will be performed by molecular dynamics (MD) study and subsequent virtual high-throughput screening (vHTS) employing GPU accelerated, MPI enhanced and pleasingly parallel computing approaches. MD study is supposed to elucidate critical conditions for inducing agonistic conformational
response of OX2R, whilst vHTS will serve to discover potential OX2R agonists with suitable chemical structure from virtual ligand databases.
Project duration:
30-March-2016 - 31-December-2016
Rational drug design strategy for novel selective cathepsin B inhibitors to combat cancer and Alzheimer’s disease
Provider:
IT4Innovations
Investigator:
Rafael Doležal, Ondrej Soukup, Jana Janočková, Eva Novotná, Markéta Pasdiorová, David Maliňák, Šárka Salajková
Abstract:
In healthy organisms, cathepsin lysosomal proteases play a vital role in degradation and turnover of proteins and polypeptides. Among a dozen of substrate specific cathepsins occurring in human cells, cathepsin B has been shown to be embroiled in an array of tangled pathologies and oncogenic processes involving brain, lung, prostate, breast, and colorectal cancer. Besides the oncogenic role, cathepsin B has been suspected to increase the production of Amyloid-beta peptides responsible for dissemination of neural plaque in Alzheimer’s disease brains. Thus, targeting over-expressed cathepsin B is currently deemed as a salient approach to combat cancer and Alzheimer’s disease. However, selective reducing the proteolytic activity of cathepsin B requires a specifically acting inhibitor which is able to penetrate into the cells, to recognize the right enzyme on the molecular level and deactivate it, while avoiding inhibitory inference towards the other enzyme variants as much as possible. In order to accomplish finding of such molecule in a rational manner worthy of that name, computer aided drug design approaches seems to be of invaluable assistance. In the current project, custom virtual ligand databases will be screened by high throughput flexible molecular docking utilizing multi-threading and massage passing interface for parallelization to reveal potential cathepsin B inhibitors. The top scoring candidates will be finally evaluated for in silico binding affinities towards cathepsin A, S and L to identify and
eliminate non selective inhibitors. The last phase of the project will focus on in silico scaffold derivation of the candidate structures to optimize their binding properties.
Project duration:
30-March-2016 - 31-December-2016
Discovery of novel ATR inhibitors as potential anticancer drugs by homology modeling and virtual high-throughput screening
Provider:
IT4Innovations
Investigator:
Jan Korabecny, Rafael Dolezal, Martin Andrs, Kamil Musilek, Lukas Hroch, Srka Salajkova
Abstract:
Cancer belongs to the most severe diseases of the modern civilization. It carries global health, economic and social burden. Therefore it is not surprising, that looking for new drugs and methods for cancer treatment and prevention is one of the main focuses of modern medicine. Thanks to achievements in this field research, the average 5-year survival rate of all cancer diagnosed in 2004-2010 reached to 68 %. Nevertheless, cancer still accounts for about 1 in every 7 deaths worldwide[1]. The treatment of this multiform condition is very difficult, expensive and unfortunately in many cases insufficient. A great interest is nowadays given to targeted therapy, which can specifically target cancer cells. Targeted drugs have two main forms: monoclonal antibodies and small-molecule drugs, usually protein kinase inhibitors[1].Research in protein kinase inhibitors is the domain of the last decade and majority of these drugs have been permitted for clinic use in last three years. One of the very promising kinases is ataxia telangiectasia and Rad3-related (ATR)[2].Although ATR is an interesting target, to this date there are only a few potent and selective inhibitors known, two of them entered the first stage clinical trials in 2014 (AZD6738 and VE-822)[3,4]. Development of a novel and selective ATR kinase inhibitor could be very difficult by using conventional synthetic approach. Normally, large libraries of compounds have to be synthesized and evaluated for identification of new scaffold structure for further development. Virtual high-throughput screening (VHTS) methods could provide a great advantage in drug discovery, because it can go through large compound libraries in relatively short time and identify new potential inhibitors.
Project duration:
1-Jun-2015 - 29-January-2016
Computerized drug design of novel glycogen synthase kinase 3 beta inhibitors for Alzheimer’s disease treatment enhanced by structure-based virtual screening and QM/MM calculations
Provider:
IT4Innovations
Investigator:
Rafael Dolezal, Kamil Kuca, Jan Korabecny, David Malinak, Ondrej Benek, Marketa Pasdiorova
Abstract:
Many neurodegenerative diseases are associated with pathological deposition of specific proteins in the brain. In the case of Alzheimer’s disease (AD), emergency of amyloid beta (Aβ) and hyperphosphorylated tau proteins (Tau) denotes that the abnormal biological processes in neurons have reached a fatal level. Currently, most attention in anti-AD drug discovery is devoted to prevent Aβ and Tau from spreading prion-like seeds over cerebral tissue in the initial phase of AD. Strong experimental evidence supports the hypothesis that hyperphosphorylated Tau is able to propagate neurofibrillary lesions throughout different brain regions. Amongst a number of proteins involved in the Tau hyperphosphorylation, glycogen synthase kinase 3 beta (GSK-3β) enzyme was selected as a biological target in this project. Employing structure-based virtual screening (SBVS), novel GSK-3β inhibitors as potential anti-AD drugs will be searched. Herein, we will perform a general computer analysis of possible binding sites of GSK-3β to design suitable geometrical parameters for scanning spatial ligand-enzyme interactions. By ligand-based virtual screening (LBVS), hundreds of thousands of chemical structures will be filtered in order to reduce the starting chemical space to structures bearing common features with known GSK-3β inhibitors. The most demanding calculations will be carried out within SBVS to indicate the promising candidates of GSK-3β inhibitors. The precision of the calculated binding energies resulting from SBVS will be further investigated by advanced QM/MM calculations. The majority of intended calculations will utilize parallelized and distributed algorithms to properly exploit the power of Salomon supercomputer.
Project duration:
1-Jun-2015 - 29-January-2016
An in silico approach for development of novel vaccine adjuvants – ligands of Toll-like receptor 4
Provider:
National Supercomputing Center - IT4Innovations
Investigator:
Rafael Dolezal, Jan Honegr, Kamil Kuca, Roman Prymula
Abstract:
The main objective of the project is to carry out an extensive molecular docking study of ~10000 drug-like compounds by distributed and parallelized computer calculations, with prospect to discover novel compounds which can enhance immunity response of organism to vaccines through TLR4.
Molecular docking involves usage of 3D structure (see picture) of the target receptors (TLR4). Each of the database compounds is docked into the receptor binding site and the best electrostatic fit is predicted. Best candidates and their derivatives, will be than synthesized and evaluated in-vitro and subsequently in-vivo. With granted computational time of 500 000 corehours we will be able to perform this task in a timespan of several weeks instead of years.
Project duration:
1-December-2013 - 31-July-2014
