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.

Building Biological Relevance Into Integrative Modelling of Macromolecular Assemblies.

Recent advances in structural biophysics and integrative modelling methods now allow us to decipher the structures of large macromolecular assemblies. Understanding the dynamics and mechanisms involved in their biological function requires rigorous integration of all available data. We have developed a complete modelling pipeline that includes analyses to extract biologically significant information by consistently combining automated and interactive human-guided steps. We illustrate this idea with two examples. First, we describe the ryanodine receptor, an ion channel that controls ion flux across the cell membrane through transitions between open and closed states. The conformational changes associated with the transitions are small compared to the considerable system size of the receptor; it is challenging to consistently track these states with the available cryo-EM structures. The second example involves homologous recombination, in which long filaments of a recombinase protein and DNA catalyse the exchange of homologous DNA strands to reliably repair DNA double-strand breaks. The nucleoprotein filament reaction intermediates in this process are short-lived and heterogeneous, making their structures particularly elusive. The pipeline we describe, which incorporates experimental and theoretical knowledge combined with state-of-the-art interactive and immersive modelling tools, can help overcome these challenges. In both examples, we point to new insights into biological processes that arise from such interdisciplinary approaches.

Deep generative models for ligand-based de novo design applied to multi-parametric optimization.

Multi-parameter optimization (MPO) is a major challenge in new chemical entity (NCE) drug discovery. Recently, promising results were reported for deep learning generative models applied to de novo molecular design, but, to our knowledge, until now no report was made of the value of this new technology for addressing MPO in an actual drug discovery project. In this study, we demonstrate the benefit of applying AI technology in a real drug discovery project. We evaluate the potential of a ligand-based de novo design technology using deep learning generative models to accelerate the obtention of lead compounds meeting 11 different biological activity objectives simultaneously. Using the initial dataset of the project, we built QSAR models for all the 11 objectives, with moderate to high performance (precision between 0.67 and 1.0 on an independent test set). Our DL-based AI de novo design algorithm, combined with the QSAR models, generated 150 virtual compounds predicted as active on all objectives. Eleven were synthetized and tested. The AI-designed compounds met 9.5 objectives on average (i.e., 86% success rate) versus 6.4 (i.e., 58% success rate) for the initial molecules measured on all objectives. One of the AI-designed molecules was active on all 11 measured objectives, and two were active on 10 objectives while being in the error margin of the assay for the last one. The AI algorithm designed compounds with functional groups, which, although being rare or absent in the initial dataset, turned out to be highly beneficial for the MPO.

Development of a chemogenomics library for phenotypic screening.

With the development of advanced technologies in cell-based phenotypic screening, phenotypic drug discovery (PDD) strategies have re-emerged as promising approaches in the identification and development of novel and safe drugs. However, phenotypic screening does not rely on knowledge of specific drug targets and needs to be combined with chemical biology approaches to identify therapeutic targets and mechanisms of actions induced by drugs and associated with an observable phenotype. In this study, we developed a system pharmacology network integrating drug-target-pathway-disease relationships as well as morphological profile from an existing high content imaging-based high-throughput phenotypic profiling assay known as "Cell Painting". Furthermore, from this network, a chemogenomic library of 5000 small molecules that represent a large and diverse panel of drug targets involved in diverse biological effects and diseases has been developed. Such a platform and a chemogenomic library could assist in the target identification and mechanism deconvolution of some phenotypic assays. The usefulness of the platform is illustrated through examples.

Shape-Based Descriptors for Efficient Structure-Based Fragment Growing.

Structure-based fragment growing is one of the key techniques in fragment-based drug design. Fragment growing is commonly practiced based on structural and biophysical data. Computational workflows are employed to predict which fragment elaborations could lead to high-affinity binders. Several such workflows exist but many are designed to be long running noninteractive systems. Shape-based descriptors have been proven to be fast and perform well at virtual-screening tasks. They could, therefore, be applied to the fragment-growing problem to enable an interactive fragment-growing workflow. In this work, we describe and analyze the use of specific shape-based directional descriptors for the task of fragment growing. The performance of these descriptors that we call ray volume matrices (RVMs) is evaluated on two data sets containing protein-ligand complexes. While the first set focuses on self-growing, the second measures practical performance in a cross-growing scenario. The runtime of screenings using RVMs as well as their robustness to three dimensional perturbations is also investigated. Overall, it can be shown that RVMs are useful to prefilter fragment candidates. For up to 84% of the 3299 generated self-growing cases and for up to 66% of the 326 generated cross-growing cases, RVMs could create poses with less than 2 Å root-mean-square deviation to the crystal structure with average query speeds of around 30,000 conformations per second. This opens the door for fast explorative screenings of fragment libraries.

Unraveling Plant Natural Chemical Diversity for Drug Discovery Purposes.

The screening and testing of extracts against a variety of pharmacological targets in order to benefit from the immense natural chemical diversity is a concern in many laboratories worldwide. And several successes have been recorded in finding new actives in natural products, some of which have become new drugs or new sources of inspiration for drugs. But in view of the vast amount of research on the subject, it is surprising that not more drug candidates were found. In our view, it is fundamental to reflect upon the approaches of such drug discovery programs and the technical processes that are used, along with their inherent difficulties and biases. Based on an extensive survey of recent publications, we discuss the origin and the variety of natural chemical diversity as well as the strategies to having the potential to embrace this diversity. It seemed to us that some of the difficulties of the area could be related with the technical approaches that are used, so the present review begins with synthetizing some of the more used discovery strategies, exemplifying some key points, in order to address some of their limitations. It appears that one of the challenges of natural product-based drug discovery programs should be an easier access to renewable sources of plant-derived products. Maximizing the use of the data together with the exploration of chemical diversity while working on reasonable supply of natural product-based entities could be a way to answer this challenge. We suggested alternative ways to access and explore part of this chemical diversity with in vitro cultures. We also reinforced how important it was organizing and making available this worldwide knowledge in an "inventory" of natural products and their sources. And finally, we focused on strategies based on synthetic biology and syntheses that allow reaching industrial scale supply. Approaches based on the opportunities lying in untapped natural plant chemical diversity are also considered.

A compact review of molecular property prediction with graph neural networks.

As graph neural networks are becoming more and more powerful and useful in the field of drug discovery, many pharmaceutical companies are getting interested in utilizing these methods for their own in-house frameworks. This is especially compelling for tasks such as the prediction of molecular properties which is often one of the most crucial tasks in computer-aided drug discovery workflows. The immense hype surrounding these kinds of algorithms has led to the development of many different types of promising architectures and in this review we try to structure this highly dynamic field of AI-research by collecting and classifying 80 GNNs that have been used to predict more than 20 molecular properties using 48 different datasets.

Fluorescent analogues of BeKm-1 with high and specific activity against the hERG channel.

Peptidic toxins that target specifically mammalian channels and receptors can be found in the venom of animals. These toxins are rarely used directly as tools for biochemical experiments, and need to be modified via the attachment of chemical groups (e.g., radioactive or fluorescent moieties). Ideally, such modifications should maintain the toxin specificity and affinity for its target. With the goal of obtaining fluorescent derivatives of BeKm-1, a toxin from the scorpion species Buthus eupeus that selectively inhibits the voltage-gated potassium ion channel hERG, we produced four active analogues using a model of BeKm-1 docking to the outer mouth of the channel. In these BeKm-1 analogues, the natural peptide was linked to the fluorescent cyanine 5 (Cy5) probe via four different linkers at Arg1 or Arg/Lys27. All analogues retained their specificity towards the hERG channel in electrophysiological experiments but displayed a lesser affinity. These results validate our strategy for designing toxin analogues and demonstrate that different chemical groups can be attached to different residues of BeKm-1.

S29434, a Quinone Reductase 2 Inhibitor: Main Biochemical and Cellular Characterization.

Quinone reductase 2 (QR2, E.C. 1.10.5.1) is an enzyme with a feature that has attracted attention for several decades: in standard conditions, instead of recognizing NAD(P)H as an electron donor, it recognizes putative metabolites of NADH, such as N-methyl- and N-ribosyl-dihydronicotinamide. QR2 has been particularly associated with reactive oxygen species and memory, strongly suggesting a link among QR2 (as a possible key element in pro-oxidation), autophagy, and neurodegeneration. In molecular and cellular pharmacology, understanding physiopathological associations can be difficult because of a lack of specific and powerful tools. Here, we present a thorough description of the potent, nanomolar inhibitor [2-(2-methoxy-5H-1,4b,9-triaza(indeno[2,1-a]inden-10-yl)ethyl]-2-furamide (S29434 or NMDPEF; IC50 = 5-16 nM) of QR2 at different organizational levels. We provide full detailed syntheses, describe its cocrystallization with and behavior at QR2 on a millisecond timeline, show that it penetrates cell membranes and inhibits QR2-mediated reactive oxygen species (ROS) production within the 100 nM range, and describe its actions in several in vivo models and lack of actions in various ROS-producing systems. The inhibitor is fairly stable in vivo, penetrates cells, specifically inhibits QR2, and shows activities that suggest a key role for this enzyme in different pathologic conditions, including neurodegenerative diseases.

Integrated Electromechanical Transduction Schemes for Polymer MEMS Sensors.

Polymer Micro ElectroMechanical Systems (MEMS) have the potential to constitute a powerful alternative to silicon-based MEMS devices for sensing applications. Although the use of commercial photoresists as structural material in polymer MEMS has been widely reported, the integration of functional polymer materials as electromechanical transducers has not yet received the same amount of interest. In this context, we report on the design and fabrication of different electromechanical schemes based on polymeric materials ensuring different transduction functions. Piezoresistive transduction made of carbon nanotube-based nanocomposites with a gauge factor of 200 was embedded within U-shaped polymeric cantilevers operating either in static or dynamic modes. Flexible resonators with integrated piezoelectric transduction were also realized and used as efficient viscosity sensors. Finally, piezoelectric-based organic field effect transistor (OFET) electromechanical transduction exhibiting a record sensitivity of over 600 was integrated into polymer cantilevers and used as highly sensitive strain and humidity sensors. Such advances in integrated electromechanical transduction schemes should favor the development of novel all-polymer MEMS devices for flexible and wearable applications in the future.

Molecule-based microelectromechanical sensors.

Incorporating functional molecules into sensor devices is an emerging area in molecular electronics that aims at exploiting the sensitivity of different molecules to their environment and turning it into an electrical signal. Among the emergent and integrated sensors, microelectromechanical systems (MEMS) are promising for their extreme sensitivity to mechanical events. However, to bring new functions to these devices, the functionalization of their surface with molecules is required. Herein, we present original electronic devices made of an organic microelectromechanical resonator functionalized with switchable magnetic molecules. The change of their mechanical properties and geometry induced by the switching of their magnetic state at a molecular level alters the device's dynamical behavior, resulting in a change of the resonance frequency. We demonstrate that these devices can be operated to sense light or thermal excitation. Moreover, thanks to the collective interaction of the switchable molecules, the device behaves as a non-volatile memory. Our results open up broad prospects of new flexible photo- and thermo-active hybrid devices for molecule-based data storage and sensors.

Network-based Approaches in Pharmacology.

In drug discovery, network-based approaches are expected to spotlight our understanding of drug action across multiple layers of information. On one hand, network pharmacology considers the drug response in the context of a cellular or phenotypic network. On the other hand, a chemical-based network is a promising alternative for characterizing the chemical space. Both can provide complementary support for the development of rational drug design and better knowledge of the mechanisms underlying the multiple actions of drugs. Recent progress in both concepts is discussed here. In addition, a network-based approach using drug-target-therapy data is introduced as an example.

Binding mode prediction and MD/MMPBSA-based free energy ranking for agonists of REV-ERBα/NCoR.

The knowledge of the free energy of binding of small molecules to a macromolecular target is crucial in drug design as is the ability to predict the functional consequences of binding. We highlight how a molecular dynamics (MD)-based approach can be used to predict the free energy of small molecules, and to provide priorities for the synthesis and the validation via in vitro tests. Here, we study the dynamics and energetics of the nuclear receptor REV-ERBα with its co-repressor NCoR and 35 novel agonists. Our in silico approach combines molecular docking, molecular dynamics (MD), solvent-accessible surface area (SASA) and molecular mechanics poisson boltzmann surface area (MMPBSA) calculations. While docking yielded initial hints on the binding modes, their stability was assessed by MD. The SASA calculations revealed that the presence of the ligand led to a higher exposure of hydrophobic REV-ERB residues for NCoR recruitment. MMPBSA was very successful in ranking ligands by potency in a retrospective and prospective manner. Particularly, the prospective MMPBSA ranking-based validations for four compounds, three predicted to be active and one weakly active, were confirmed experimentally.

Molecular Dynamics Simulations and Kinetic Measurements to Estimate and Predict Protein-Ligand Residence Times.

Ligand-target residence time is emerging as a key drug discovery parameter because it can reliably predict drug efficacy in vivo. Experimental approaches to binding and unbinding kinetics are nowadays available, but we still lack reliable computational tools for predicting kinetics and residence time. Most attempts have been based on brute-force molecular dynamics (MD) simulations, which are CPU-demanding and not yet particularly accurate. We recently reported a new scaled-MD-based protocol, which showed potential for residence time prediction in drug discovery. Here, we further challenged our procedure's predictive ability by applying our methodology to a series of glucokinase activators that could be useful for treating type 2 diabetes mellitus. We combined scaled MD with experimental kinetics measurements and X-ray crystallography, promptly checking the protocol's reliability by directly comparing computational predictions and experimental measures. The good agreement highlights the potential of our scaled-MD-based approach as an innovative method for computationally estimating and predicting drug residence times.

Optimization Of PVDF-TrFE Processing Conditions For The Fabrication Of Organic MEMS Resonators.

This paper reports a systematic optimization of processing conditions of PVDF-TrFE piezoelectric thin films, used as integrated transducers in organic MEMS resonators. Indeed, despite data on electromechanical properties of PVDF found in the literature, optimized processing conditions that lead to these properties remain only partially described. In this work, a rigorous optimization of parameters enabling state-of-the-art piezoelectric properties of PVDF-TrFE thin films has been performed via the evaluation of the actuation performance of MEMS resonators. Conditions such as annealing duration, poling field and poling duration have been optimized and repeatability of the process has been demonstrated.

Alternative Radioligands for Investigating the Molecular Pharmacology of Melatonin Receptors.

Melatonin exerts a variety of physiologic activities that are mainly relayed through the melatonin receptors MT1 and MT2 Low expressions of these receptors in tissues have led to widespread experimental use of the agonist 2-[(125)I]-iodomelatonin as a substitute for melatonin. We describe three iodinated ligands: 2-(2-[(2-iodo-4,5-dimethoxyphenyl)methyl]-4,5-dimethoxy phenyl) (DIV880) and (2-iodo-N-2-[5-methoxy-2-(naphthalen-1-yl)-1H-pyrrolo[3,2-b]pyridine-3-yl])acetamide (S70254), which are specific ligands at MT2 receptors, and N-[2-(5-methoxy-1H-indol-3-yl)ethyl]iodoacetamide (SD6), an analog of 2-[(125)I]-iodomelatonin with slightly different characteristics. Here, we further characterized these new ligands with regards to their molecular pharmacology. We performed binding experiments, saturation assays, association/dissociation rate measurements, and autoradiography using sheep and rat tissues and recombinant cell lines. Our results showed that [(125)I]-S70254 is receptor, and can be used with both cells and tissue. This radioligand can be used in autoradiography. Similarly, DIV880, a partial agonist [43% of melatonin on guanosine 5'-3-O-(thio)triphosphate binding assay], selective for MT2, can be used as a tool to selectively describe the pharmacology of this receptor in tissue samples. The molecular pharmacology of both human melatonin receptors MT1 and MT2, using a series of 24 ligands at these receptors and the new radioligands, did not lead to noticeable variations in the profiles. For the first time, we described radiolabeled tools that are specific for one of the melatonin receptors (MT2). These tools are amenable to binding experiments and to autoradiography using sheep or rat tissues. These specific tools will permit better understanding of the role and implication in physiopathologic processes of the melatonin receptors.

Dynamics of hERG closure allow novel insights into hERG blocking by small molecules.

Today, drug discovery routinely uses experimental assays to determine very early if a lead compound can yield certain types of off-target activity. Among such off targets is hERG. The ion channel plays a primordial role in membrane repolarization and altering its activity can cause severe heart arrhythmia and sudden death. Despite routine tests for hERG activity, rather little information is available for helping medicinal chemists and molecular modelers to rationally circumvent hERG activity. In this article novel insights into the dynamics of hERG channel closure are described. Notably, helical pairwise closure movements have been observed. Implications and relations to hERG inactivation are presented. Based on these dynamics novel insights on hERG blocker placement are presented, compared to literature, and discussed. Last, new evidence for horizontal ligand positioning is shown in light of former studies on hERG blockers.

2P2I HUNTER: a tool for filtering orthosteric protein-protein interaction modulators via a dedicated support vector machine.

Over the last 10 years, protein-protein interactions (PPIs) have shown increasing potential as new therapeutic targets. As a consequence, PPIs are today the most screened target class in high-throughput screening (HTS). The development of broad chemical libraries dedicated to these particular targets is essential; however, the chemical space associated with this 'high-hanging fruit' is still under debate. Here, we analyse the properties of 40 non-redundant small molecules present in the 2P2I database (http://2p2idb.cnrs-mrs.fr/) to define a general profile of orthosteric inhibitors and propose an original protocol to filter general screening libraries using a support vector machine (SVM) with 11 standard Dragon molecular descriptors. The filtering protocol has been validated using external datasets from PubChem BioAssay and results from in-house screening campaigns. This external blind validation demonstrated the ability of the SVM model to reduce the size of the filtered chemical library by eliminating up to 96% of the compounds as well as enhancing the proportion of active compounds by up to a factor of 8. We believe that the resulting chemical space identified in this paper will provide the scientific community with a concrete support to search for PPI inhibitors during HTS campaigns.

Encoding protein-ligand interaction patterns in fingerprints and graphs.

We herewith present a novel and universal method to convert protein-ligand coordinates into a simple fingerprint of 210 integers registering the corresponding molecular interaction pattern. Each interaction (hydrophobic, aromatic, hydrogen bond, ionic bond, metal complexation) is detected on the fly and physically described by a pseudoatom centered either on the interacting ligand atom, the interacting protein atom, or the geometric center of both interacting atoms. Counting all possible triplets of interaction pseudoatoms within six distance ranges, and pruning the full integer vector to keep the most frequent triplets enables the definition of a simple (210 integers) and coordinate frame-invariant interaction pattern descriptor (TIFP) that can be applied to compare any pair of protein-ligand complexes. TIFP fingerprints have been calculated for ca. 10,000 druggable protein-ligand complexes therefore enabling a wide comparison of relationships between interaction pattern similarity and ligand or binding site pairwise similarity. We notably show that interaction pattern similarity strongly depends on binding site similarity. In addition to the TIFP fingerprint which registers intermolecular interactions between a ligand and its target protein, we developed two tools (Ishape, Grim) to align protein-ligand complexes from their interaction patterns. Ishape is based on the overlap of interaction pseudoatoms using a smooth Gaussian function, whereas Grim utilizes a standard clique detection algorithm to match interaction pattern graphs. Both tools are complementary and enable protein-ligand complex alignments capitalizing on both global and local pattern similarities. The new fingerprint and companion alignment tools have been successfully used in three scenarios: (i) interaction-biased alignment of protein-ligand complexes, (ii) postprocessing docking poses according to known interaction patterns for a particular target, and (iii) virtual screening for bioisosteric scaffolds sharing similar interaction patterns.

Use of reduced graphs to encode bioisosterism for similarity-based virtual screening.

This paper describes a project to include explicit information about bioisosteric equivalences between pairs of fragment substructures in a system for similarity-based virtual screening. Data from the BIOSTER database show that reduced graphs provide a simple way of encoding known bioisosteric equivalences in a manner that can be used during similarity searching. Scaffold-hopping experiments with the WOMBAT database show that including such information enables similarities to be identified between the reference structures and active structures from the database that contain different, but equivalent, fragment substructures. However, such equivalences also contribute to the similarities between the reference structures and inactives, and the latter equivalences can swamp those involving the actives. This presents serious problems for the routine use of information about bioisosteric fragments in similarity-based virtual screening.