A toolkit for covalent docking with GOLD: from automated ligand preparation with KNIME to bound protein-ligand complexes.

Motivation: Current covalent docking tools have limitations that make them difficult to use for performing large-scale structure-based covalent virtual screening (VS). They require time-consuming tasks for the preparation of proteins and compounds (standardization, filtering according to the type of warheads), as well as for setting up covalent reactions. We have developed a toolkit to help accelerate drug discovery projects in the phases of hit identification by VS of ultra-large covalent libraries and hit expansion by exploration of the binding of known covalent compounds. With this application note, we offer the community a toolkit for performing automated covalent docking in a fast and efficient way. Results: The toolkit comprises a KNIME workflow for ligand preparation and a Python program to perform the covalent docking of ligands with the GOLD docking engine running in a parallelized fashion. Availability and implementation: The KNIME workflow entitled 'Evotec_Covalent_Processing_forGOLD.knwf' for the preparation of the ligands is available in the KNIME Hub https://hub.knime.com/emilie_pihan/spaces. Supplementary information: Supplementary data are available at Bioinformatics Advances online.

Fine tuning for success in structure-based virtual screening.

Structure-based virtual screening plays a significant role in drug-discovery. The method virtually docks millions of compounds from corporate or public libraries into a binding site of a disease-related protein structure, allowing for the selection of a small list of potential ligands for experimental testing. Many algorithms are available for docking and assessing the affinity of compounds for a targeted protein site. The performance of affinity estimation calculations is highly dependent on the size and nature of the site, therefore a rationale for selecting the best protocol is required. To address this issue, we have developed an automated calibration process, implemented in a Knime workflow. It consists of four steps: preparation of a protein test set with structures and models of the target, preparation of a compound test set with target-related ligands and decoys, automatic test of 24 scoring/rescoring protocols for each target structure and model, and graphical display of results. The automation of the process combined with execution on high performance computing resources greatly reduces the duration of the calibration phase, and the test of many combinations of algorithms on various target conformations results in a rational and optimal choice of the best protocol. Here, we present this tool and exemplify its application in setting-up an optimal protocol for SBVS against Retinoid X Receptor alpha.

Modeling and Deorphanization of Orphan GPCRs.

Despite tremendous efforts, approximately 120 GPCRs remain orphan. Their physiological functions and their potential roles in diseases are poorly understood. Orphan GPCRs are extremely important because they may provide novel therapeutic targets for unmet medical needs. As a complement to experimental approaches, molecular modeling and virtual screening are efficient techniques to discover synthetic surrogate ligands which can help to elucidate the role of oGPCRs. Constitutively activated mutants and recently published active structures of GPCRs provide stimulating opportunities for building active molecular models for oGPCRs and identifying activators using virtual screening of compound libraries. We describe the molecular modeling and virtual screening process we have applied in the discovery of surrogate ligands, and provide examples for CCKA, a simulated oGPCR, and for two oGPCRs, GPR52 and GPR34.

Plasmodium falciparum CTP:phosphocholine cytidylyltransferase possesses two functional catalytic domains and is inhibited by a CDP-choline analog selected from a virtual screening.

Phosphatidylcholine is the major lipid component of the malaria parasite membranes and is required for parasite multiplication in human erythrocytes. Plasmodium falciparum CTP:phosphocholine cytidylyltransferase (PfCCT) is the rate-limiting enzyme of the phosphatidylcholine biosynthesis pathway and thus considered as a potential antimalarial target. In contrast to its mammalian orthologs, PfCCT contains a duplicated catalytic domain. Here, we show that both domains are catalytically active with similar kinetic parameters. A virtual screening strategy allowed the identification of a drug-size molecule competitively inhibiting the enzyme. This compound also prevented phosphatidylcholine biosynthesis in parasites and exerted an antimalarial effect. This study constitutes the first step towards a rationalized design of future new antimalarial agents targeting PfCCT.

Computational and biophysical approaches to protein-protein interaction inhibition of Plasmodium falciparum AMA1/RON2 complex.

Invasion of the red blood cell by Plasmodium falciparum parasites requires formation of an electron dense circumferential ring called the Moving Junction (MJ). The MJ is anchored by a high affinity complex of two parasite proteins: Apical Membrane Antigen 1 (PfAMA1) displayed on the surface of the parasite and Rhoptry Neck Protein 2 that is discharged from the parasite and imbedded in the membrane of the host cell. Structural studies of PfAMA1 revealed a conserved hydrophobic groove localized to the apical surface that coordinates RON2 and invasion inhibitory peptides. In the present work, we employed computational and biophysical methods to identify competitive P. falciparum AMA1-RON2 inhibitors with the goal of exploring the 'druggability' of this attractive antimalarial target. A virtual screen followed by molecular docking with the PfAMA1 crystal structure was performed using an eight million compound collection that included commercial molecules, the ChEMBL malaria library and approved drugs. The consensus approach resulted in the selection of inhibitor candidates. We also developed a fluorescence anisotropy assay using a modified inhibitory peptide to experimentally validate the ability of the selected compounds to inhibit the AMA1-RON2 interaction. Among those, we identified one compound that displayed significant inhibition. This study offers interesting clues to improve the throughput and reliability of screening for new drug leads.

Biochemical characterization of Plasmodium falciparum CTP:phosphoethanolamine cytidylyltransferase shows that only one of the two cytidylyltransferase domains is active.

The intra-erythrocytic proliferation of the human malaria parasite Plasmodium falciparum requires massive synthesis of PE (phosphatidylethanolamine) that together with phosphatidylcholine constitute the bulk of the malaria membrane lipids. PE is mainly synthesized de novo by the CDP:ethanolamine-dependent Kennedy pathway. We previously showed that inhibition of PE biosynthesis led to parasite death. In the present study we characterized PfECT [P. falciparum CTP:phosphoethanolamine CT (cytidylyltransferase)], which we identified as the rate-limiting step of the PE metabolic pathway in the parasite. The cellular localization and expression of PfECT along the parasite life cycle were studied using polyclonal antibodies. Biochemical analyses showed that the enzyme activity follows Michaelis-Menten kinetics. PfECT is composed of two CT domains separated by a linker region. Activity assays on recombinant enzymes upon site-directed mutagenesis revealed that the N-terminal CT domain was the only catalytically active domain of PfECT. Concordantly, three-dimensional homology modelling of PfECT showed critical amino acid differences between the substrate-binding sites of the two CT domains. PfECT was predicted to fold as an intramolecular dimer suggesting that the inactive C-terminal domain is important for dimer stabilization. Given the absence of PE synthesis in red blood cells, PfECT represents a potential antimalarial target opening the way for a rational conception of bioactive compounds.

Cytochrome P450-catalyzed degradation of nicotine: fundamental parameters determining hydroxylation by cytochrome P450 2A6 at the 5'-carbon or the n-methyl carbon.

The oxidation of (2'S)-nicotine in the active site of human cytochrome P450 2A6 has been subjected to a detailed analysis by theoretical quantum mechanical/molecular mechanical (QM/MM) calculations linked with a theoretical and experimental study of the associated isotope effects. The study has focused on seeking an explanation as to why oxidation at the 5'-carbon position (A) is favored over oxidation at the methyl carbon (CMe) position (B). It is deduced that the choice of hydrogen for abstraction is not determined by geometric features of the active site, but by the lower energy barrier associated with 5' oxidation. N-Demethylation leading to N-hydroxymethylnornicotine requires ca. 6.5 kcal/mol more energy to transfer a hydrogen atom than is required for oxidation on the carbon 5'. Neither protonation of the pyrrolidine nitrogen (N1') nor inclusion of a water molecule in the reaction process influences the balance between the two oxidation pathways. In both cases, the hydrogen transfer step is rate limiting. An analysis of the calculated kinetic isotope effects indicates that the presence of a (2)H in either the C5' or the CMepositions has a significant effect on the reaction kinetics. However, the experimental values of around 2.2-2.6 are considerably lower than those predicted by theoretical calculations (9.3 and 6.9 for C5' or the CMe positions, respectively, in the LS state of Cpd I), typical of the masking commonly found for CYP450 reactions. The fact that similar values are found for cotinine formation from both substrates, however, may indicate that the measured value is not that for H-abstraction but, rather, is a combined value for (2)H influence on electronic redistribution between iminium states of the pyrrolidine ring. This is the first time that oxidation at the C5' or the CMe positions has been directly compared and that isotope effects have been obtained for this reaction in a human cytochrome P450 reaction.

e-Drug3D: 3D structure collections dedicated to drug repurposing and fragment-based drug design.

MOTIVATION: In the drug discovery field, new uses for old drugs, selective optimization of side activities and fragment-based drug design (FBDD) have proved to be successful alternatives to high-throughput screening. e-Drug3D is a database of 3D chemical structures of drugs that provides several collections of ready-to-screen SD files of drugs and commercial drug fragments. They are natural inputs in studies dedicated to drug repurposing and FBDD. AVAILABILITY: e-Drug3D collections are freely available at http://chemoinfo.ipmc.cnrs.fr/e-drug3d.html either for download or for direct in silico web-based screenings.

Discovery of small molecule inhibitors of protein-protein interactions using combined ligand and target score normalization.

Docking experiments of multiple compounds typically focus on a single protein. However, other targets provide information about relative binding efficiencies that is otherwise lacking. We developed a docking strategy that normalized results in both the ligand and target dimensions. This was applied to dock 287 approved small drugs with 35 peptide-binding proteins, including 15 true positives. The combined docking score was normalized by drug and protein and by incorporating information on contact similarity to the template protein-peptide contacts. The 20 top ranking hits included 6 true positives, and three matches with suggestive evidence in the literature: the cardiac glycoside digitoxin may inhibit WW domain interactions, the 14-3-3 zeta protein may bind negatively charged ligands, and the nuclear receptor coactivator site may bind nuclear receptor agonists. Additionally, the Bcl-2 antiapoptotic protein is predicted to bind pargyline, and the antiapoptic p53 interacting protein MDM2 is suggested to bind clofazimine. These predictions represent starting points for the experimental development of PPI inhibitors based on an existing database of approved drugs and demonstrate that two-dimensional normalization improves docking efficiency.