Continuous Evaluation of Ligand Protein Predictions: A Weekly Community Challenge for Drug Docking.

Docking calculations can accelerate drug discovery by predicting the bound poses of ligands for a targeted protein. However, it is not clear which docking methods work best. Furthermore, predicting poses requires steps outside the docking algorithm itself, such as preparation of the protein and ligand, and it is not known which components are most in need of improvement. The Continuous Evaluation of Ligand Protein Predictions (CELPP) is a blinded prediction challenge designed to address these issues. Participants create a workflow to predict protein-ligand binding poses, which is then tasked with predicting 10-100 new protein-ligand crystal structures each week. CELPP evaluates the accuracy of each workflow's predictions and posts the scores online. The results can be used to identify the strengths and weaknesses of current approaches, help map docking problems to the algorithms most likely to overcome them, and illuminate areas of unmet need in structure-guided drug design.

Enhancing Virtual Screening Performance of Protein Kinases with Molecular Dynamics Simulations.

In silico virtual screening (VS) is a powerful hit identification technique used in drug discovery projects that aims to effectively distinguish true actives from inactive or decoy molecules. To better capture the dynamic behavior of protein drug targets, compound databases may be screened against an ensemble of protein conformations, which may be experimentally determined or generated computationally, i.e. via molecular dynamics (MD) simulations. Several studies have shown that conformations generated by MD are useful in identifying novel hit compounds, in part because structural rearrangements sampled during MD can provide novel targetable areas. However, it remains difficult to predict a priori when an MD conformation will outperform a VS against the crystal structure alone. Here, we assess whether MD conformations result in improved VS performance for six protein kinases. MD conformations are selected using three different methods, and their VS performances are compared to the corresponding crystal structures. Additionally, these conformations are used to train ensembles, and their VS performance is compared to the individual MD conformations and the corresponding crystal structures using receiver operating characteristic curve (ROC) metrics. We show that performing MD results in at least one conformation that offers better VS performance than the crystal structure, and that, while it is possible to train ensembles to outperform the crystal structure alone, the extent of this enhancement is target dependent. Lastly, we show that the optimal structural selection method is also target dependent and recommend optimizing virtual screens on a kinase-by-kinase basis to improve the likelihood of success.

Simulation-Based Approaches for Determining Membrane Permeability of Small Compounds.

Predicting the rate of nonfacilitated permeation of solutes across lipid bilayers is important to drug design, toxicology, and signaling. These rates can be estimated using molecular dynamics simulations combined with the inhomogeneous solubility-diffusion model, which requires calculation of the potential of mean force and position-dependent diffusivity of the solute along the transmembrane axis. In this paper, we assess the efficiency and accuracy of several methods for the calculation of the permeability of a model DMPC bilayer to urea, benzoic acid, and codeine. We compare umbrella sampling, replica exchange umbrella sampling, adaptive biasing force, and multiple-walker adaptive biasing force for the calculation of the transmembrane PMF. No definitive advantage for any of these methods in their ability to predict the membrane permeability coefficient Pm was found, provided that a sufficiently long equilibration is performed. For diffusivities, a Bayesian inference method was compared to a generalized Langevin method, both being sensitive to chosen parameters and the slow relaxation of membrane defects. Agreement within 1.5 log units of the computed Pm with experiment is found for all permeants and methods. Remaining discrepancies can likely be attributed to limitations of the force field as well as slowly relaxing collective movements within the lipid environment. Numerical calculations based on model profiles show that Pm can be reliably estimated from only a few data points, leading to recommendations for calculating Pm from simulations.

Knowledge-Based Methods To Train and Optimize Virtual Screening Ensembles.

Ensemble docking can be a successful virtual screening technique that addresses the innate conformational heterogeneity of macromolecular drug targets. Yet, lacking a method to identify a subset of conformational states that effectively segregates active and inactive small molecules, ensemble docking may result in the recommendation of a large number of false positives. Here, three knowledge-based methods that construct structural ensembles for virtual screening are presented. Each method selects ensembles by optimizing an objective function calculated using the receiver operating characteristic (ROC) curve: either the area under the ROC curve (AUC) or a ROC enrichment factor (EF). As the number of receptor conformations, N, becomes large, the methods differ in their asymptotic scaling. Given a set of small molecules with known activities and a collection of target conformations, the most resource intense method is guaranteed to find the optimal ensemble but scales as O(2(N)). A recursive approximation to the optimal solution scales as O(N(2)), and a more severe approximation leads to a faster method that scales linearly, O(N). The techniques are generally applicable to any system, and we demonstrate their effectiveness on the androgen nuclear hormone receptor (AR), cyclin-dependent kinase 2 (CDK2), and the peroxisome proliferator-activated receptor δ (PPAR-δ) drug targets. Conformations that consisted of a crystal structure and molecular dynamics simulation cluster centroids were used to form AR and CDK2 ensembles. Multiple available crystal structures were used to form PPAR-δ ensembles. For each target, we show that the three methods perform similarly to one another on both the training and test sets.

Molecular docking to flexible targets.

It is widely accepted that protein receptors exist as an ensemble of conformations in solution. How best to incorporate receptor flexibility into virtual screening protocols used for drug discovery remains a significant challenge. Here, stepwise methodologies are described to generate and select relevant protein conformations for virtual screening in the context of the relaxed complex scheme (RCS), to design small molecule libraries for docking, and to perform statistical analyses on the virtual screening results. Methods include equidistant spacing, RMSD-based clustering, and QR factorization protocols for ensemble generation and ROC analysis for ensemble selection.

Computation-guided discovery of influenza endonuclease inhibitors.

Influenza is a global human health threat, and there is an immediate need for new antiviral therapies to circumvent the limitations of vaccination and current small molecule therapies. During viral transcription, influenza incorporates the 5'-end of the host cell's mRNA in a process that requires the influenza endonuclease. Based on recently published endonuclease crystalized structures, a three-dimensional pharmacophore was developed and used to virtually screen 450,000 compounds for influenza endonuclease inhibitors. Of 264 compounds tested in a FRET-based endonuclease-inhibition assay, 16 inhibitors (IC50 <50 µM) that span 5 molecular classes novel to this endonuclease were found (6.1% hit rate). To determine cytotoxicity and antiviral activity, subsequent cellular assays were performed. Two compounds suppress viral replication with negligible cell toxicity.

Progress towards automated Kepler scientific workflows for computer-aided drug discovery and molecular simulations.

We describe the development of automated workflows that support computed-aided drug discovery (CADD) and molecular dynamics (MD) simulations and are included as part of the National Biomedical Computational Resource (NBCR). The main workflow components include: file-management tasks, ligand force field parameterization, receptor-ligand molecular dynamics (MD) simulations, job submission and monitoring on relevant high-performance computing (HPC) resources, receptor structural clustering, virtual screening (VS), and statistical analyses of the VS results. The workflows aim to standardize simulation and analysis and promote best practices within the molecular simulation and CADD communities. Each component is developed as a stand-alone workflow, which allows easy integration into larger frameworks built to suit user needs, while remaining intuitive and easy to extend.

Back to the future: can physical models of passive membrane permeability help reduce drug candidate attrition and move us beyond QSPR?

It is widely recognized that adsorption, distribution, metabolism, excretion, and toxicology liabilities kill the majority of drug candidates that progress to clinical trials. The development of computational models to predict small molecule membrane permeability is therefore of considerable scientific and public health interest. Empirical qualitative structure permeability relationship models of permeability have been a mainstay in industrial applications, but lack a deep understanding of the underlying biologic physics. Others and we have shown that implicit solvent models to predict passive permeability for small molecules exhibit mediocre predictive performance when validated across experimental test sets. Given the vast increase in computer power, more efficient parallelization schemes, and extension of current atomistic simulation codes to general use graphical processing units, the development and application of physical models based on all-atom simulations may now be feasible. Preliminary results from rigorous free energy calculations using all-atom simulations indicate that performance relative to implicit solvent models may be improved, but many outstanding questions remain. Here, we review the current state-of-the-art physical models for passive membrane permeability prediction and present a prospective look at promising new directions for all-atom approaches.

Magnesium-induced nucleophile activation in the guanylyltransferase mRNA capping enzyme.

The mRNA guanylyltransferase, or mRNA capping enzyme, cotranscriptionally caps the 5'-end of nascent mRNA with GMP during the second reaction in a set of three enzymatic reactions that result in the formation of an N7-methylguanosine cap during mRNA maturation. The mRNA capping enzyme is characterized, in part, by a conserved lysine nucleophile that attacks the α-phosphorus atom of GTP, forming a lysine-GMP intermediate. Experiments have firmly established that magnesium is required for efficient intermediate formation but have provided little insight into the requirement's molecular origins. Using empirical and thermodynamic integration pK(a) estimates, along with conventional molecular dynamics simulations, we show that magnesium binding likely activates the lysine nucleophile by increasing its acidity and by biasing the deprotonated nucleophile into conformations conducive to intermediate formation. These results provide additional functional understanding of an important enzyme in the mRNA transcript life cycle and allow functional analogies to be drawn that affect our understanding of the metal dependence of related superfamily members.

Rational prediction with molecular dynamics for hit identification.

Although the motions of proteins are fundamental for their function, for pragmatic reasons, the consideration of protein elasticity has traditionally been neglected in drug discovery and design. This review details protein motion, its relevance to biomolecular interactions and how it can be sampled using molecular dynamics simulations. Within this context, two major areas of research in structure-based prediction that can benefit from considering protein flexibility, binding site detection and molecular docking, are discussed. Basic classification metrics and statistical analysis techniques, which can facilitate performance analysis, are also reviewed. With hardware and software advances, molecular dynamics in combination with traditional structure-based prediction methods can potentially reduce the time and costs involved in the hit identification pipeline.

Variable ligand- and receptor-binding hot spots in key strains of influenza neuraminidase.

Influenza A continues to be a major public health concern due to its ability to cause epidemic and pandemic disease outbreaks in humans. Computational investigations of structural dynamics of the major influenza glycoproteins, especially the neuraminidase (NA) enzyme, are able to provide key insights beyond what is currently accessible with standard experimental techniques. In particular, all-atom molecular dynamics simulations reveal the varying degrees of flexibility for such enzymes. Here we present an analysis of the relative flexibility of the ligand- and receptor-binding area of three key strains of influenza A: highly pathogenic H5N1, the 2009 pandemic H1N1, and a human N2 strain. Through computational solvent mapping, we investigate the various ligand- and receptor-binding "hot spots" that exist on the surface of NA which interacts with both sialic acid receptors on the host cells and antiviral drugs. This analysis suggests that the variable cavities found in the different strains and their corresponding capacities to bind ligand functional groups may play an important role in the ability of NA to form competent reaction encounter complexes with other species of interest, including antiviral drugs, sialic acid receptors on the host cell surface, and the hemagglutinin protein. Such considerations may be especially useful for the prediction of how such complexes form and with what binding capacity.

Modeling the pharmacodynamics of passive membrane permeability.

Small molecule permeability through cellular membranes is critical to a better understanding of pharmacodynamics and the drug discovery endeavor. Such permeability may be estimated as a function of the free energy change of barrier crossing by invoking the barrier domain model, which posits that permeation is limited by passage through a single "barrier domain" and assumes diffusivity differences among compounds of similar structure are negligible. Inspired by the work of Rezai and co-workers (JACS 128:14073-14080, 2006), we estimate this free energy change as the difference in implicit solvation free energies in chloroform and water, but extend their model to include solute conformational affects. Using a set of eleven structurally diverse FDA approved compounds and a set of thirteen congeneric molecules, we show that the solvation free energies are dominated by the global minima, which allows solute conformational distributions to be effectively neglected. For the set of tested compounds, the best correlation with experiment is obtained when the implicit chloroform global minimum is used to evaluate the solvation free energy difference.

Biochemical, structural and molecular dynamics analyses of the potential virulence factor RipA from Yersinia pestis.

Human diseases are attributed in part to the ability of pathogens to evade the eukaryotic immune systems. A subset of these pathogens has developed mechanisms to survive in human macrophages. Yersinia pestis, the causative agent of the bubonic plague, is a predominately extracellular pathogen with the ability to survive and replicate intracellularly. A previous study has shown that a novel rip (required for intracellular proliferation) operon (ripA, ripB and ripC) is essential for replication and survival of Y. pestis in postactivated macrophages, by playing a role in lowering macrophage-produced nitric oxide (NO) levels. A bioinformatics analysis indicates that the rip operon is conserved among a distally related subset of macrophage-residing pathogens, including Burkholderia and Salmonella species, and suggests that this previously uncharacterized pathway is also required for intracellular survival of these pathogens. The focus of this study is ripA, which encodes for a protein highly homologous to 4-hydroxybutyrate-CoA transferase; however, biochemical analysis suggests that RipA functions as a butyryl-CoA transferase. The 1.9 Å X-ray crystal structure reveals that RipA belongs to the class of Family I CoA transferases and exhibits a unique tetrameric state. Molecular dynamics simulations are consistent with RipA tetramer formation and suggest a possible gating mechanism for CoA binding mediated by Val227. Together, our structural characterization and molecular dynamic simulations offer insights into acyl-CoA specificity within the active site binding pocket, and support biochemical results that RipA is a butyryl-CoA transferase. We hypothesize that the end product of the rip operon is butyrate, a known anti-inflammatory, which has been shown to lower NO levels in macrophages. Thus, the results of this molecular study of Y. pestis RipA provide a structural platform for rational inhibitor design, which may lead to a greater understanding of the role of RipA in this unique virulence pathway.

Mechanism of 150-cavity formation in influenza neuraminidase.

The recently discovered 150-cavity in the active site of group-1 influenza A neuraminidase (NA) proteins provides a target for rational structure-based drug development to counter the increasing frequency of antiviral resistance in influenza. Surprisingly, the 2009 H1N1 pandemic virus (09N1) neuraminidase was crystalized without the 150-cavity characteristic of group-1 NAs. Here we demonstrate, through a total sum of 1.6 µs of biophysical simulations, that 09N1 NA exists in solution preferentially with an open 150-cavity. Comparison with simulations using avian N1, human N2 and 09N1 with a I149V mutation and an extensive bioinformatics analysis suggests that the conservation of a key salt bridge is crucial in the stabilization of the 150-cavity across both subtypes. This result provides an atomic-level structural understanding of the recent finding that antiviral compounds designed to take advantage of contacts in the 150-cavity can inactivate both 2009 H1N1 pandemic and avian H5N1 viruses.

Novel naphthalene-based inhibitors of Trypanosoma brucei RNA editing ligase 1.

BACKGROUND: Neglected tropical diseases, including diseases caused by trypanosomatid parasites such as Trypanosoma brucei, cost tens of millions of disability-adjusted life-years annually. As the current treatments for African trypanosomiasis and other similar infections are limited, new therapeutics are urgently needed. RNA Editing Ligase 1 (REL1), a protein unique to trypanosomes and other kinetoplastids, was identified recently as a potential drug target. METHODOLOGY/PRINCIPAL FINDINGS: Motivated by the urgent need for novel trypanocidal therapeutics, we use an ensemble-based virtual-screening approach to discover new naphthalene-based TbREL1 inhibitors. The predicted binding modes of the active compounds are evaluated within the context of the flexible receptor model and combined with computational fragment mapping to determine the most likely binding mechanisms. Ultimately, four new low-micromolar inhibitors are presented. Three of the four compounds may bind to a newly revealed cleft that represents a putative druggable site not evident in any crystal structure. CONCLUSIONS/SIGNIFICANCE: Pending additional optimization, the compounds presented here may serve as precursors for future novel therapies useful in the fight against several trypanosomatid pathogens, including human African trypanosomiasis, a devastating disease that afflicts the vulnerable patient populations of sub-Saharan Africa.

Using multistate free energy techniques to improve the efficiency of replica exchange accelerated molecular dynamics.

Replica exchange accelerated molecular dynamics (REXAMD) is a method that enhances conformational sampling while retaining at least one replica on the original potential, thus avoiding the statistical problems of exponential reweighting. In this article, we study three methods that can combine the data from the accelerated replicas to enhance the estimate of properties on the original potential: weighted histogram analysis method (WHAM), pairwise multistate Bennett acceptance ratio (PBAR), and multistate Bennett acceptance ratio (MBAR). We show that the method that makes the most efficient use of equilibrium data from REXAMD simulations is the MBAR method. This observation holds for both alchemical free energy and structural observable prediction. The combination of REXAMD and MBAR should allow for more efficient scaling of the REXAMD method to larger biopolymer systems.

Substrate induced population shifts and stochastic gating in the PBCV-1 mRNA capping enzyme.

The 317 residue PBCV-1 mRNA capping enzyme catalyzes the second enzymatic reaction in the formation of an N-7-methyl-GMP cap on the 5'-end of the nascent mRNA. It is composed of two globular domains bound by a short flexible peptide linker, which have been shown to undergo opening and closing events. The small size and experimentally demonstrated domain mobility make the PBCV-1 capping enzyme an ideally suited model system to explore domain mobility in context of substrate binding. Here, we specifically address the following four questions: (1) How does substrate binding affect relative domain mobility: is the system better described by an induced fit or population shift mechanism? (2) What are the gross characteristics of a conformation capable of binding substrate? (3) Does "domain gating" of the active site affect the rate of substrate binding? (4) Does the magnitude of receptor conformational fluctuations confer substrate specificity by sterically occluding molecules of a particular size or geometry? We answer these questions using a combination of theory, Brownian dynamics, and molecular dynamics. Our results show that binding efficiency is a function of conformation but that isomerization between efficient and inefficient binding conformations does not impact the substrate association rate. Additionally, we show that conformational flexibility alone is insufficient to explain single stranded mRNA specificity. While our results are specific to the PBCV-1 mRNA capping enzyme, they provide a useful context within which the substrate binding behavior of similarly structured enzymes or proteins may be considered.

Toward understanding the conformational dynamics of RNA ligation.

Members of the genus Trypanosoma, which include the pathogenic species Trypanosoma brucei and Trypanosoma cruzi, edit their post-transcriptional mitochondrial RNA via a multiprotein complex called the editosome. In T. brucei, the RNA is nicked prior to uridylate insertion and deletion. Following editing, nicked RNA is religated by one of two RNA-editing ligases (TbREL). This study describes a recent 70 ns molecular dynamics simulation of TbREL1, an ATP-dependent RNA-editing ligase of the nucleotidyltransferase superfamily that is required for the survival of T. brucei insect and bloodstream forms. In this work, a model of TbREL1 in complex with its full double-stranded RNA (dsRNA) substrate is created on the basis of the homologous relation between TbREL1 and T4 Rnl2. The simulation captures TbREL1 dynamics in the state immediately preceding RNA ligation, providing insights into the functional dynamics and catalytic mechanism of the kinetoplastid ligation reaction. Important features of RNA binding and specificity are revealed for kinetoplastid ligases and the broader nucleotidyltransferase superfamily.

Discovery and design of DNA and RNA ligase inhibitors in infectious microorganisms.

BACKGROUND: Members of the nucleotidyltransferase superfamily known as DNA and RNA ligases carry out the enzymatic process of polynucleotide ligation. These guardians of genomic integrity share a three-step ligation mechanism, as well as common core structural elements. Both DNA and RNA ligases have experienced a surge of recent interest as chemotherapeutic targets for the treatment of a range of diseases, including bacterial infection, cancer, and the diseases caused by the protozoan parasites known as trypanosomes. OBJECTIVE: In this review, we will focus on efforts targeting pathogenic microorganisms; specifically, bacterial NAD(+)-dependent DNA ligases, which are promising broad-spectrum antibiotic targets, and ATP-dependent RNA editing ligases from Trypanosoma brucei, the species responsible for the devastating neurodegenerative disease, African sleeping sickness. CONCLUSION: High quality crystal structures of both NAD(+)-dependent DNA ligase and the Trypanosoma brucei RNA editing ligase have facilitated the development of a number of promising leads. For both targets, further progress will require surmounting permeability issues and improving selectivity and affinity.

Catalytically requisite conformational dynamics in the mRNA-capping enzyme probed by targeted molecular dynamics.

The addition of a N7-methyl guanosine cap to the 5' end of nascent mRNA is carried out by the mRNA-capping enzyme, a two-domain protein that is a member of the nucleotidyltransferase superfamily. The mRNA-capping enzyme is composed of a catalytic nucleotidyltransferase domain and a noncatalytic oligonucleotide/oligosaccharide binding (OB) domain. Large-scale domain motion triggered by substrate binding mediates catalytically requisite conformational rearrangement of the GTP substrate prior to the chemical step. In this study, we employ targeted molecular dynamics (TMD) on the PBCV-1 capping enzyme to probe the global domain dynamics and internal dynamics of conserved residues during the conformational transformation from the open to the closed state. Analysis of the resulting trajectories along with structural and sequence homology to other members of the superfamily allows us to suggest a conserved mechanism of conformational rearrangements spanning all mRNA-capping enzymes and all ATP-dependent DNA ligases. Our results suggest that the OB domain moves quasi-statically toward the nucleotidyltransferase domain, pivoting about a short linker region. The approach of the OB domain brings a conserved RxDK sequence, an element of conserved motif VI, within proximity of the triphosphate of GTP, destabilizing the unreactive conformation and thereby allowing thermal fluctuations to partition the substrate toward the catalytically competent state.