The Identification of New c-FLIP Inhibitors for Restoring Apoptosis in TRAIL-Resistant Cancer Cells.

The catalytically inactive caspase-8-homologous protein, c-FLIP, is a potent antiapoptotic protein highly expressed in various types of cancers. c-FLIP competes with caspase-8 for binding to the adaptor protein FADD (Fas-Associated Death Domain) following death receptors' (DRs) activation via the ligands of the TNF-R family. As a consequence, the extrinsic apoptotic signaling pathway involving DRs is inhibited. The inhibition of c-FLIP activity in tumor cells might enhance DR-mediated apoptosis and overcome immune and anticancer drug resistance. Based on an in silico approach, the aim of this work was to identify new small inhibitory molecules able to bind selectively to c-FLIP and block its anti-apoptotic activity. Using a homology 3D model of c-FLIP, an in silico screening of 1880 compounds from the NCI database (National Cancer Institute) was performed. Nine molecules were selected for in vitro assays, based on their binding affinity to c-FLIP and their high selectivity compared to caspase-8. These molecules selectively bind to the Death Effector Domain 2 (DED2) of c-FLIP. We have tested in vitro the inhibitory effect of these nine molecules using the human lung cancer cell line H1703, overexpressing c-FLIP. Our results showed that six of these newly identified compounds efficiently prevent FADD/c-FLIP interactions in a molecular pull-down assay, as well as in a DISC immunoprecipitation assay. The overexpression of c-FLIP in H1703 prevents TRAIL-mediated apoptosis; however, a combination of TRAIL with these selected molecules significantly restored TRAIL-induced cell death by rescuing caspase cleavage and activation. Altogether, our findings indicate that new inhibitory chemical molecules efficiently prevent c-FLIP recruitment into the DISC complex, thus restoring the caspase-8-dependent apoptotic cascade. These results pave the way to design new c-FLIP inhibitory molecules that may serve as anticancer agents in tumors overexpressing c-FLIP.

Peptide-Conjugated MRI Probe Targeted to Netrin-1, a Novel Metastatic Breast Cancer Biomarker.

Despite significant progress in cancer imaging and treatment over the years, early diagnosis and metastasis detection remain a challenge. Molecular magnetic resonance imaging (MRI), with its high resolution, can be well adapted to fulfill this need, requiring the design of contrast agents which target specific tumor biomarkers. Netrin-1 is an extracellular protein overexpressed in metastatic breast cancer and implicated in tumor progression and the appearance of metastasis. This study focuses on the design and preclinical evaluation of a novel Netrin-1-specific peptide-based MRI probe, GdDOTA-KKTHDAVR (Gd-K), to visualize metastatic breast cancer. The targeting peptide sequence was identified based on the X-ray structure of the complex between Netrin-1 and its transmembrane receptor DCC. Molecular docking simulations support the probe design. In vitro studies evidenced submicromolar affinity of Gd-K for Netrin-1 (KD = 0.29 µM) and good MRI efficacy (proton relaxivity, r1 = 4.75 mM-1 s-1 at 9.4 T, 37 °C). In vivo MRI studies in a murine model of triple-negative metastatic breast cancer revealed successful tumor visualization at earlier stages of tumor development (smaller tumor volume). Excellent signal enhancement, 120% at 2 min and 70% up to 35 min post injection, was achieved (0.2 mmol/kg injected dose), representing a reasonable imaging time window and a superior contrast enhancement in the tumor as compared to Dotarem injection.

Highlighting the Major Role of Cyclin C in Cyclin-Dependent Kinase 8 Activity through Molecular Dynamics Simulations.

Dysregulation of cyclin-dependent kinase 8 (CDK8) activity has been associated with many diseases, including colorectal and breast cancer. As usual in the CDK family, the activity of CDK8 is controlled by a regulatory protein called cyclin C (CycC). But, while human CDK family members are generally activated in two steps, that is, the binding of the cyclin to CDK and the phosphorylation of a residue in the CDK activation loop, CDK8 does not require the phosphorylation step to be active. Another peculiarity of CDK8 is its ability to be associated with CycC while adopting an inactive form. These specificities raise the question of the role of CycC in the complex CDK8-CycC, which appears to be more complex than the other members of the CDK family. Through molecular dynamics (MD) simulations and binding free energy calculations, we investigated the effect of CycC on the structure and dynamics of CDK8. In a second step, we particularly focused our investigation on the structural and molecular basis of the protein-protein interaction between the two partners by finely analyzing the energetic contribution of residues and simulating the transition between the active and the inactive form. We found that CycC has a stabilizing effect on CDK8, and we identified specific interaction hotspots within its interaction surface compared to other human CDK/Cyc pairs. Targeting these specific interaction hotspots could be a promising approach in terms of specificity to effectively disrupt the interaction between CDK8. The simulation of the conformational transition from the inactive to the active form of CDK8 suggests that the residue Glu99 of CycC is involved in the orientation of three conserved arginines of CDK8. Thus, this residue may assume the role of the missing phosphorylation step in the activation mechanism of CDK8. In a more general view, these results point to the importance of keeping the CycC in computational studies when studying the human CDK8 protein in both the active and the inactive form.

The Impact of Data on Structure-Based Binding Affinity Predictions Using Deep Neural Networks.

Artificial intelligence (AI) has gained significant traction in the field of drug discovery, with deep learning (DL) algorithms playing a crucial role in predicting protein-ligand binding affinities. Despite advancements in neural network architectures, system representation, and training techniques, the performance of DL affinity prediction has reached a plateau, prompting the question of whether it is truly solved or if the current performance is overly optimistic and reliant on biased, easily predictable data. Like other DL-related problems, this issue seems to stem from the training and test sets used when building the models. In this work, we investigate the impact of several parameters related to the input data on the performance of neural network affinity prediction models. Notably, we identify the size of the binding pocket as a critical factor influencing the performance of our statistical models; furthermore, it is more important to train a model with as much data as possible than to restrict the training to only high-quality datasets. Finally, we also confirm the bias in the typically used current test sets. Therefore, several types of evaluation and benchmarking are required to understand models' decision-making processes and accurately compare the performance of models.

Estimation of Drug-Target Residence Time by Targeted Molecular Dynamics Simulations.

Drug-target residence time has emerged as a key selection factor in drug discovery since the binding duration of a drug molecule to its protein target can significantly impact its in vivo efficacy. The challenge in studying the residence time, in early drug discovery stages, lies in how to cost-effectively determine the residence time for the systematic assessment of compounds. Currently, there is still a lack of computational protocols to quickly estimate such a measure, particularly for large and flexible protein targets and drugs. Here, we report an efficient computational protocol, based on targeted molecular dynamics, to rank drug candidates by their residence time and to obtain insights into ligand-target dissociation mechanisms. The method was assessed on a dataset of 10 arylpyrazole inhibitors of CDK8, a large, flexible, and clinically important target, for which the experimental residence time of the inhibitors ranges from minutes to hours. The compounds were correctly ranked according to their estimated residence time scores compared to their experimental values. The analysis of protein-ligand interactions along the dissociation trajectories highlighted the favorable contribution of hydrophobic contacts to residence time and revealed key residues that strongly affect compound residence time.

Residence Time Prediction of Type 1 and 2 Kinase Inhibitors from Unbinding Simulations.

In the early stage of a drug discovery process, the selection and optimization of a ligand is mainly based on equilibrium thermodynamic constants such as KD or IC50 values, which are representatives of the affinity of the compound for its target. However, these criteria are not able to correctly evaluate the efficacy of compounds in vivo and result in many failures of compound development during phase II of clinical trials. Residence time (RT) is an important parameter associated to an in vivo drug's safety and efficacy. The determination or modulation of kinetic rates correlated to RT may be performed to identify the best drug candidates in the early stages of a drug design project. For this purpose, a number of experimental methodologies were developed but remain costly in both time and money. Herein, we developed a novel protocol based on biased molecular dynamics simulations and transition-state theory in order to predict relative ligand kinetic rates and relative RTs of a series of compounds. First, we have repeatedly simulated the unbinding process of the ligand from its binding site to the outside of the target. Next, we sample the conformational space along the determined unbinding paths to allow relevant statistical distributions of the system. The free energy profiles associated to these distributions are then computed and used to predict the kinetics parameters. The studied set was composed of eight ligands with fast, intermediate, and slow dissociation rates and binding to the active and inactive states of p38α protein kinase. The proposed method provides an excellent correlation between the predicted values and the experimentally measured kinetic rates, in addition to a detailed characterization of the kinetic paths at the atomic level.

Comparative Assessment of Protein Kinase Inhibitors in Public Databases and in PKIDB.

Since the first approval of a protein kinase inhibitor (PKI) by the Food and Drug Administration (FDA) in 2001, 55 new PKIs have reached the market, and many inhibitors are currently being evaluated in clinical trials. This is a clear indication that protein kinases still represent major drug targets for the pharmaceutical industry. In a previous work, we have introduced PKIDB, a publicly available database, gathering PKIs that have already been approved (Phase 4), as well as those currently in clinical trials (Phases 0 to 3). This database is updated frequently, and an analysis of the new data is presented here. In addition, we compared the set of PKIs present in PKIDB with the PKIs in early preclinical studies found in ChEMBL, the largest publicly available chemical database. For each dataset, the distribution of physicochemical descriptors related to drug-likeness is presented. From these results, updated guidelines to prioritize compounds for targeting protein kinases are proposed. The results of a principal component analysis (PCA) show that the PKIDB dataset is fully encompassed within all PKIs found in the public database. This observation is reinforced by a principal moments of inertia (PMI) analysis of all molecules. Interestingly, we notice that PKIs in clinical trials tend to explore new 3D chemical space. While a great majority of PKIs is located on the area of "flatland", we find few compounds exploring the 3D structural space. Finally, a scaffold diversity analysis of the two datasets, based on frequency counts was performed. The results give insight into the chemical space of PKIs, and can guide researchers to reach out new unexplored areas. PKIDB is freely accessible from the following website: http://www.icoa.fr/pkidb.

Specific Targeting of Plant and Apicomplexa Parasite Tubulin through Differential Screening Using In Silico and Assay-Based Approaches.

Dinitroanilines are chemical compounds with high selectivity for plant cell α-tubulin in which they promote microtubule depolymerization. They target α-tubulin regions that have diverged over evolution and show no effect on non-photosynthetic eukaryotes. Hence, they have been used as herbicides over decades. Interestingly, dinitroanilines proved active on microtubules of eukaryotes deriving from photosynthetic ancestors such as Toxoplasma gondii and Plasmodium falciparum, which are responsible for toxoplasmosis and malaria, respectively. By combining differential in silico screening of virtual chemical libraries on Arabidopsis thaliana and mammal tubulin structural models together with cell-based screening of chemical libraries, we have identified dinitroaniline related and non-related compounds. They inhibit plant, but not mammalian tubulin assembly in vitro, and accordingly arrest A. thaliana development. In addition, these compounds exhibit a moderate cytotoxic activity towards T. gondii and P. falciparum. These results highlight the potential of novel herbicidal scaffolds in the design of urgently needed anti-parasitic drugs.

Virtual screening of natural products to enhance melanogenosis.

Natural products have long been an important source of inspiration for medicinal chemistry and drug discovery. In the cosmetic field, they remain the major elements of the composition and serve as marketing asset. Recent research showed the implication of salt-inducible kinases on the melanin production in skin via MITF regulation. Finding new potent modulators on such target could open the way to several cosmetic applications to attenuate visible signs of photoaging and improve the tan without sun. Since virtual screening can be a powerful tool for detecting hit compounds in the early stages of a drug discovery process, we applied this method on salt-inducible kinase 2 to discover potential interesting compounds. Here, we present the different steps from the construction of a database of natural products, to the validation of a docking protocol and the results of the virtual screening. Hits from the screening were tested in vitro to confirm their efficiency and results are discussed.

Tetrahydropyridine LIMK inhibitors: Structure activity studies and biological characterization.

LIM Kinases, LIMK1 and LIMK2, have become promising targets for the development of inhibitors with potential application for the treatment of several major diseases. LIMKs play crucial roles in cytoskeleton remodeling as downstream effectors of small G proteins of the Rho-GTPase family, and as major regulators of cofilin, an actin depolymerizing factor. In this article we describe the conception, synthesis, and biological evaluation of novel tetrahydropyridine pyrrolopyrimidine LIMK inhibitors. Homology models were first constructed to better understand the binding mode of our preliminary compounds and to explain differences in biological activity. A library of over 60 products was generated and in vitro enzymatic activities were measured in the mid to low nanomolar range. The most promising derivatives were then evaluated in cell on cofilin phosphorylation inhibition which led to the identification of 52 which showed excellent selectivity for LIMKs in a kinase selectivity panel. We also demonstrated that 52 affected the cell cytoskeleton by disturbing actin filaments. Cell migration studies with this derivative using three different cell lines displayed a significant effect on cell motility. Finally, the crystal structure of the kinase domain of LIMK2 complexed with 52 was solved, greatly improving our understanding of the interaction between 52 and LIMK2 active site. The reported data represent a basis for the development of more efficient LIMK inhibitors for future in vivo preclinical validation.

Synthesis and properties of 2'-O-neopentyl modified oligonucleotides.

2'-O-Neopentyldeoxyuridine (Un) was synthesized and incorporated into a series of oligodeoxyribonucleotides. Single and triple incorporations in various arrangements were performed. The Watson and Crick pairing properties with complementary DNA and RNA were investigated by UV melting curves, CD spectroscopy, and molecular dynamic simulations. The results were compared to those obtained with DNA-DNA and DNA-RNA duplexes involving dU at the same positions. Oligonucleotides containing Un clearly demonstrated their ability to form duplexes with both complementary DNA and RNA but with higher stabilities for the DNA-RNA duplexes similar to the one of the parent DNA-RNA duplex. Investigations into the thermodynamic properties of these 17-base-pair duplexes revealed ΔG values (37 °C) that are in line with the measured T(m) values for both the DNA-DNA and DNA-RNA duplexes. CD spectroscopic structural investigations indicated that the conformations of the DNA-DNA and DNA-RNA duplexes involving Un are similar to those of the dT-rA and dU-rA containing duplexes. Only small changes in intensities and weak blue shifts were observed when three Uns were incorporated into the duplexes. The results of the molecular dynamic simulations showed, for the six duplexes involving the modified nucleoside Un, calculated curvatures similar to those of the corresponding unmodified duplexes without base-pair disruption. The neopentyl group is able to be accommodated in the minor grooves of both the DNA-DNA and RNA-DNA duplexes. However, molecular dynamic simulations indicated that the Uns adopt a C2'-exo sugar pucker conformation close to an A-helix type without perturbing the C2'-endo sugar pucker conformations of their 2'-deoxynucleoside neighbours. These results confirm the potential of 2'-O-neopentyldeoxyuridine as a nucleoside surrogate for oligonucleotide based therapeutic strategies.

A perspective on the sharing of docking data.

Computational approaches are nowadays largely applied in drug discovery projects. Among these, molecular docking is the most used for hit identification against a drug target protein. However, many scientists in the field shed light on the lack of availability and reproducibility of the data obtained from such studies to the whole community. Consequently, sustaining and developing the efforts toward a large and fully transparent sharing of those data could be beneficial for all researchers in drug discovery. The purpose of this article is first to propose guidelines and recommendations on the appropriate way to conduct virtual screening experiments and second to depict the current state of sharing molecular docking data. In conclusion, we have explored and proposed several prospects to enhance data sharing from docking experiment that could be developed in the foreseeable future.

Design, synthesis and characterisation of a novel type II B-RAF paradox breaker inhibitor.

The mutation V600E in B-Raf leads to mitogen activated protein kinase (MAPK) pathway activation, uncontrolled cell proliferation, and tumorigenesis. ATP competitive type I B-Raf inhibitors, such as vemurafenib (1) and PLX4720 (4) efficiently block the MAPK pathways in B-Raf mutant cells, however these inhibitors induce conformational changes in the wild type B-Raf (wtB-Raf) kinase domain leading to heterodimerization with C-Raf, causing paradoxical hyperactivation of the MAPK pathway. This unwanted activation may be avoided by another class of inhibitors (type II) which bind the kinase in the DFG-out conformation, such as AZ628 (3) preventing heterodimerization. Here we present a new B-Raf kinase domain inhibitor, based on a phenyl(1H-pyrrolo [2,3-b]pyridin-3-yl)methanone template, that represents a hybrid between 4 and 3. This novel inhibitor borrows the hinge binding region from 4 and the back pocket binding moiety from 3. We determined its binding mode, performed activity/selectivity studies, and molecular dynamics simulations in order to study the conformational effects induced by this inhibitor on wt and V600E mutant B-Raf kinase. We discovered that the inhibitor was active and selective for B-Raf, binds in a DFG-out/αC-helix-in conformation, and did not induce the aforementioned paradoxical hyperactivation in the MAPK pathway. We propose that this merging approach can be used to design a novel class of B-Raf inhibitors for translational studies.

A triple spin-labeling strategy coupled with DEER analysis to detect DNA modifications and enzymatic repair.

In a spin: Spin-labeled oligonucleotides produced by click chemistry can be studied by EPR, by using a DEER sequence. This was used to test a complex triple-labeling strategy with damaged DNA. Extensive and accurate analysis of DNA structure and enzymatic repair processes were performed after digestion by EndoIV. Modified DNA structures and DNA-protein interactions can now be readily studied.

Lesion-induced DNA weak structural changes detected by pulsed EPR spectroscopy combined with site-directed spin labelling.

Double electron-electron resonance (DEER) was applied to determine nanometre spin-spin distances on DNA duplexes that contain selected structural alterations. The present approach to evaluate the structural features of DNA damages is thus related to the interspin distance changes, as well as to the flexibility of the overall structure deduced from the distance distribution. A set of site-directed nitroxide-labelled double-stranded DNA fragments containing defined lesions, namely an 8-oxoguanine, an abasic site or abasic site analogues, a nick, a gap and a bulge structure were prepared and then analysed by the DEER spectroscopic technique. New insights into the application of 4-pulse DEER sequence are also provided, in particular with respect to the spin probes' positions and the rigidity of selected systems. The lesion-induced conformational changes observed, which were supported by molecular dynamics studies, confirm the results obtained by other, more conventional, spectroscopic techniques. Thus, the experimental approaches described herein provide an efficient method for probing lesion-induced structural changes of nucleic acids.

Comparing native and irradiated E. coli lactose repressor-operator complex by molecular dynamics simulation.

The function of the E. coli lactose operon requires the binding of the tetrameric repressor protein to the operator DNA. We have previously shown that gamma-irradiation destabilises the repressor-operator complex because the repressor gradually loses its DNA-binding ability (Radiat Res 170:604-612, 2008). It was suggested that the observed oxidation of tyrosine residues and the concomitant structural changes of irradiated headpieces (DNA-binding domains of repressor monomers) could be responsible for the inactivation. To unravel the mechanisms that lead to repressor-operator complex destabilisation when tyrosine oxidation occurs, we have compared by molecular dynamic simulations two complexes: (1) the native complex formed by two headpieces and the operator DNA, and (2) the damaged complex, in which all tyrosines are replaced by their oxidation product 3,4-dihydroxyphenylalanine (DOPA). On a 20 ns time scale, MD results show effects consistent with complex destabilisation: increased flexibility, increased DNA bending, modification of the hydrogen bond network, and decrease of the positive electrostatic potential at the protein surface and of the global energy of DNA-protein interactions.

Computational Analysis of Crystallization Additives for the Identification of New Allosteric Sites.

Allosteric effect can modulate the biological activity of a protein. Thus, the discovery of new allosteric sites is very attractive for designing new modulators or inhibitors. Here, we propose an innovative way to identify allosteric sites, based on crystallization additives (CA), used to stabilize proteins during the crystallization process. Density and clustering analyses of CA, applied on protein kinase and nuclear receptor families, revealed that CA are not randomly distributed around protein structures, but they tend to aggregate near common sites. All orthosteric and allosteric cavities described in the literature are retrieved from the analysis of CA distribution. In addition, new sites were identified, which could be associated to putative allosteric sites. We proposed an efficient and easy way to use the structural information of CA to identify allosteric sites. This method could assist medicinal chemists for the design of new allosteric compounds targeting cavities of new drug targets.

VSPrep: A KNIME Workflow for the Preparation of Molecular Databases for Virtual Screening.

Drug discovery is a challenging and expensive field. Hence, novel in silico tools have been developed in early discovery stage to identify and prioritize novel molecules with suitable physicochemical properties. In many in silico drug design projects, molecular databases are screened by virtual screening tools to search for potential bioactive molecules. The preparation of the molecules is therefore a key step in the success of well-established techniques such as docking, similarity or pharmacophore searching. We review here the lists of several toolkits used in different steps during the cleaning of molecular databases, integrated within a KNIME workflow. During the first step of the automatic workflow, salts are removed, and mixtures are split to get one compound per entry. Then compounds with unwanted features are filtered. Duplicated entries are then deleted while considering stereochemistry. As a compromise between exhaustiveness and computational time, most distributed tautomers at physiological pH are computed. Additionally, various flags are applied to molecules by using either classical molecular descriptors, similarity search to known libraries or substructure search rules. Moreover, stereoisomers are enumerated depending on the unassigned chiral centers. Then, three-dimensional coordinates, and optionally conformers, are generated. This workflow has been already applied to several drug design projects and can be used for molecular database preparation upon request.

Enhanced Molecular Dynamics Methods Applied to Drug Design Projects.

Nobel Laureate Richard P. Feynman stated: "[…] everything that living things do can be understood in terms of jiggling and wiggling of atoms […]." The importance of computer simulations of macromolecules, which use classical mechanics principles to describe atom behavior, is widely acknowledged and nowadays, they are applied in many fields such as material sciences and drug discovery. With the increase of computing power, molecular dynamics simulations can be applied to understand biological mechanisms at realistic timescales. In this chapter, we share our computational experience providing a global view of two of the widely used enhanced molecular dynamics methods to study protein structure and dynamics through the description of their characteristics, limits and we provide some examples of their applications in drug design. We also discuss the appropriate choice of software and hardware. In a detailed practical procedure, we describe how to set up, run, and analyze two main molecular dynamics methods, the umbrella sampling (US) and the accelerated molecular dynamics (aMD) methods.

VSPrep: A General KNIME Workflow for the Preparation of Molecules for Virtual Screening.

Over the past decades, virtual screening has proved itself to be a valuable asset to identify new bioactive compounds. The vast majority of commonly used techniques can be described in three steps: pre-processing the dataset i. e. small (ligands) and eventually larger (receptors) molecules, execute the method and finally analyse the results. Hence, the preparation of ligands is a critical step for success of commonly used virtual screening approaches such as protein-ligand docking, similarity or pharmacophore search. We present here a new workflow, VSPrep, for the pre-processing of small molecules; it is based on freely accessible tools for academics and is integrated within the KNIME platform. It can be used to perform several chemoinformatics tasks such as molecular database cleaning, tautomer and stereoisomer enumeration, focused library design and conformer generation. Additionally, graphical reports of the results are provided to the user as a convenient analysis tool.