Hammerhead: fast,fully automated docking of flexible ligands to protein binding sites

Background: Molecular docking seeks to predict the geometry and affinity of the binding of a small molecule to a given protein of known structure. Rigid docking has long been used to screen databases of small molecules, because docking techniques that account for ligand flexibility have either been too slow or have required significant human intervention. Here we describe a docking algorithm, Hammerhead, which is a fast, automated tool to screen for the binding of flexible molecules to protein binding sites.Results: We used Hammerhead to successfully dock a variety of positive control ligands into their cognate proteins. The empirically tuned scoring function of the algorithm predicted binding affinities within 1.3 log units of the known affinities for these ligands. Conformations and alignments close to those determined crystallographically received the highest scores. We screened 80 000 compounds for binding to streptavidin, and biotin was predicted as the top-scoring ligand, with other known ligands included among the highest-scoring dockings. The screen ran in a few days on commonly available hardware.Conclusions: Hammerhead is suitable for screening large databases of flexible molecules for binding to a protein of known structure. It correctly docks a variety of known flexible ligands, and it spends an average of only a few seconds on each compound during a screen. The approach is completely automated, from the elucidation of protein binding sites, through the docking of molecules, to the final selection of compounds for assay.     

New approach to molecular docking and its application to virtual screening of chemical databases

This paper describes the validation of a molecular docking method and its application to virtual database screening. The code flexibly docks ligand molecules into rigid receptor structures using a tabu search methodology driven by an empirically derived function for estimating the binding affinity of a protein-ligand complex. The docking method has been tested on 70 ligand-receptor complexes for which the experimental binding affinity and binding geometry are known. The lowest energy geometry produced by the docking protocol is within 2.0 A root mean square of the experimental binding mode for 79% of the complexes. The method has been applied to the problem of virtual database screening to identify known ligands for thrombin, factor Xa, and the estrogen receptor. A database of 10,000 randomly chosen "druglike" molecules has been docked into the three receptor structures. In each case known receptor ligands were included in the study. The results showed good separation between the predicted binding affinities of the known ligand set and the database subset.     

Solid-phase synthesis for the identification of high-affinity bivalent lectin ligands

The development of carbohydrate-based therapeutics has been frustrated by the low affinities that characterize protein-carbohydrate complexation. Because of the oligomeric nature of most lectins, the use of multivalency may offer a successful strategy for the creation of high-affinity ligands. The solid-phase evaluation of libraries of peptide-linked multivalent ligands facilitates rapid examination of a large fraction of linker structure space. If such solid-phase assays are to replicate solution binding behavior, the potential for intermolecular bivalent binding on bead surfaces must be eliminated. Here we report the solid-phase synthesis and analysis of peptide-linked, spatially segregated mono- and bivalent ligands for the legume lectin concanavalin A. Bead shaving protocols were used for the creation of beads displaying spatially segregated binding sequences on the surface of Tentagel resins. The same ligands were also synthesized on PEGA resin to determine the effect of ligand presentation on solid-phase binding. While we set out to determine the lower limit of assay sensitivity, the unexpected observation that intermolecular bivalent ligand binding is enhanced for bivalent ligands relative to monovalent ligands allowed direct observation of the level of surface blocking required to prevent intermolecular bivalent ligand binding. For a protein with binding sites separated by 65 A, approximately 99.9% of Tentagel(1) surface sites and 99.99% of the total sites on a PEGA bead must be blocked to prevent intermolecular bivalent binding. We also report agglutination and calorimetric solution-phase binding studies of mono- and bivalent peptide-linked ligands.     

Influencing receptor-ligand binding mechanisms with multivalent ligand architecture

Multivalent ligands can function as inhibitors or effectors of biological processes. Potent inhibitory activity can arise from the high functional affinities of multivalent ligand-receptor interactions. Effector functions, however, are influenced not only by apparent affinities but also by alternate factors, including the ability of a ligand to cluster receptors. Little is known about the molecular features of a multivalent ligand that determine whether it will function as an inhibitor or effector. We envisioned that, by altering multivalent ligand architecture, ligands with preferences for different binding mechanisms would be generated. To this end, a series of 28 ligands possessing structural diversity was synthesized. This series provides the means to explore the effects of ligand architecture on the inhibition and clustering of a model protein, the lectin concanavalin A (Con A). The structural parameters that were varied include scaffold shape, size, valency, and density of binding elements. We found that ligands with certain architectures are effective inhibitors, but others mediate receptor clustering. Specifically, high molecular weight, polydisperse polyvalent ligands are effective inhibitors of Con A binding, whereas linear oligomeric ligands generated by the ring-opening metathesis polymerization have structural properties that favor clustering. The shape of a multivalent ligand also influences specific aspects of receptor clustering. These include the rate at which the receptor is clustered, the number of receptors in the clusters, and the average interreceptor distance. Our results indicate that the architecture of a multivalent ligand is a key parameter in determining its activity as an inhibitor or effector. Diversity-oriented syntheses of multivalent ligands coupled with effective assays that can be used to compare the contributions of different binding parameters may afford ligands that function by specific mechanisms.     

基于雌激素β受体结构雌激素类化合物的三维定量结构-活性关系与分子对接研究

雌激素类化合物由于其对人和野生动物健康的负面影响而受到广泛关注.雌激素受体存在α和β两种亚型,由于雌激素β受体（ERβ）与α受体（ERα）两者结合腔中的氨基酸序列存在明显差异,因此配体化合物在与雌激素β受体和α受体的结合活性和模式上也可能存在较大差别.本文以50个与雌激素β受体结合的化合物为研究对象,应用比较分子相似性指数分析（COMSIA）的三维定量结构-活性关系（3D-QSAR）分析方法研究化合物结构与活性之间的关系,比较了原子契合和基于受体结构两种分子叠合方式对模型质量的影响,建立了相关性显著、预测能力强的定量活性预测模型（R^2=0.961,qLOO^2=0.671,R^2Pred=0.722）,并结合分子对接方法揭示了影响化合物活性的分子结构特征和分子机理.     

Optimization of tether length in nonglycosidically linked bivalent ligands that target sites 2 and 1 of a Shiga-like toxin

A series of bivalent ligands for a Shiga-like toxin have been synthesized, their experimentally determined inhibitory activities were compared with a simplified thermodynamic model, and computer simulations were used to predict the optimal tether length in bivalent ligands. The design of the inhibitors exploits the proximity of the C-2' hydroxyl groups of two P(k)-trisaccharides when bound to two different, neighboring carbohydrate recognizing binding sites located on the surface of Shiga-like toxin. NMR studies of the complex between the toxin and bivalent ligands show that site 2 and site 1 of a single B subunit are simultaneously occupied by a tethered P(k)-trisaccharide dimer. A simplified thermodynamic treatment provides the intrinsic affinities and binding energies for the intermolecular and intramolecular association events and permits the deconvolution of the contributions to the relative binding energies for the set of bivalent ligands. Conformational analysis based on MD simulations for bivalent galabioside dimers containing different tethers demonstrated that the calculated local concentrations of the pendant ligand at the second binding site correlate with the experimentally determined relative affinity values of the respective bivalent ligands, thereby providing a predictive method to optimize tether length.     

<Emphasis Type="Italic">In-silico</Emphasis> Screening using Flexible Ligand Binding Pockets: A Molecular Dynamics-based Approach

Summary In-silico screening of flexible ligands against flexible ligand binding pockets (LBP) is an emerging approach in structure-based drug discovery. Here, we describe a molecular dynamics (MD) based docking approach to investigate the influence on the high-throughput in-silico screening of small molecules against flexible ligand binding pockets. In our approach, an ensemble of 51 energetically favorable structures of the LBP of human estrogen receptor α (hERα) were collected from 3 ns MD simulations. In-silico screening of 3500 endocrine disrupting compounds against these flexible ligand binding pockets resulted in thousands of ER–ligand complexes of which 582 compounds were unique. Detailed analysis of MD generated structures showed that only 17 of the LBP residues significantly contribute to the overall binding pocket flexibility. Using the flexible LBP conformations generated, we have identified 32 compounds that bind better to the flexible ligand-binding pockets compared to the crystal structure. These compounds, though chemically divergent, are structurally similar to the natural hormone. Our MD-based approach in conjunction with grid–based distributed computing could be applied routinely for in-silico screening of large databases against any given target.     

Estrogen-derived steroidal metal complexes: agents for cellular delivery of metal centers to estrogen receptor-positive cells

Targeted cellular delivery of drugs to specific tissues is an important goal in biomedical chemistry. Achieving this requires harnessing and applying molecular-level recognition events prevalent in (or specific to) the desired tissue type. Tissues rich in estrogen receptors (ERs), which include many types of breast cancer, accumulate molecules that have high binding affinities for these receptors. Therefore, molecules that (i) bind to the ER, (ii) have favorable cellular transport properties, and (iii) contain a second functionality (such as a center that may be used for diagnostic imaging or medical therapy) are exciting synthetic targets in the field of drug delivery. To this end, we have prepared a range of metallo-estrogens based on 17alpha-ethynylestradiol and examined their binding to the ER both as isolated receptor and in whole cell assays (ER positive MCF-7 cells). Estrogens functionalized with metal binding units are prepared by palladium-catalyzed cross-coupling reactions and a wide range of metal centers introduced readily. All the compounds prepared and tested exhibit effective binding to the estrogen receptor and are delivered across the cell membrane into MCF-7 cells. In the whole cell assays, despite their monocationic nature, the palladium and platinum complexes prepared exhibit similar (and even enhanced) receptor binding affinities compared to their corresponding neutral free ligands. It is unprecedented for a higher ER binding affinity to be observed for a cationic complex than for its metal-free ligand.     

Molecular mechanics methods for predicting protein-ligand binding

Ligand binding affinity prediction is one of the most important applications of computational chemistry. However, accurately ranking compounds with respect to their estimated binding affinities to a biomolecular target remains highly challenging. We provide an overview of recent work using molecular mechanics energy functions to address this challenge. We briefly review methods that use molecular dynamics and Monte Carlo simulations to predict absolute and relative ligand binding free energies, as well as our own work in which we have developed a physics-based scoring method that can be applied to hundreds of thousands of compounds by invoking a number of simplifying approximations. In our previous studies, we have demonstrated that our scoring method is a promising approach for improving the discrimination between ligands that are known to bind and those that are presumed not to, in virtual screening of large compound databases. In new results presented here, we explore several improvements to our computational method including modifying the dielectric constant used for the protein and ligand interiors, and empirically scaling energy terms to compensate for deficiencies in the energy model. Future directions for further improving our physics-based scoring method are also discussed.     

Anti-estrogenic activity of mansorins and mansonones from the heartwood of Mansonia gagei DRUMM

Through an anti-estrogenic bioassay-guided fractionation of the methanol extract of Mansonia gagei, three new coumarins, called mansorins I (1), II (2) and III (3) and a new naphthoquinone, mansonone I (4), were isolated. Their structures were determined based on their NMR data and CD spectroscopy. The anti-estrogenic activity of the fractions and the isolated compounds were investigated using a yeast two-hybrid assay method expressing estrogen receptors alpha (ERalpha) and beta (ERbeta). In addition, an ERalpha competitor screening system (ligand binding screen) was used to verify the binding affinities of the isolated compounds to the estrogen receptor. 1,2-Naphthoquinones (mansonones) showed more binding affinities to ER in both assay systems. All the tested compounds showed higher binding affinities to ERbeta than to ERalpha in the yeast two-hybrid assay. Mansonones F and S showed the most potent estrogen binding and estrogen antagonistic effects.     

Supervised scoring models with docked ligand conformations for structure-based virtual screening

Protein-ligand docking programs have been used to efficiently discover novel ligands for target proteins from large-scale compound databases. However, better scoring methods are needed. Generally, scoring functions are optimized by means of various techniques that affect their fitness for reproducing X-ray structures and protein-ligand binding affinities. However, these scoring functions do not always work well for all target proteins. A scoring function should be optimized for a target protein to enhance enrichment for structure-based virtual screening. To address this problem, we propose the supervised scoring model (SSM), which takes into account the protein-ligand binding process using docked ligand conformations with supervised learning for optimizing scoring functions against a target protein. SSM employs a rough linear correlation between binding free energy and the root mean square deviation of a native ligand for predicting binding energy. We applied SSM to the FlexX scoring function, that is, F-Score, with five different target proteins: thymidine kinase (TK), estrogen receptor (ER), acetylcholine esterase (AChE), phosphodiesterase 5 (PDE5), and peroxisome proliferator-activated receptor gamma (PPARgamma). For these five proteins, SSM always enhanced enrichment better than F-Score, exhibiting superior performance that was particularly remarkable for TK, AChE, and PPARgamma. We also demonstrated that SSM is especially good at enhancing enrichments of the top ranks of screened compounds, which is useful in practical drug screening.     

Intervals and the deduction of drug binding site models

In the search for new drugs, it often occurs that the binding affinities of several compounds to a common receptor macromolecule are known experimentally, but the structure of the receptor is not known. This article describes an extraordinarily objective computer algorithm for deducing the important geometric and energetic features of the common binding site, starting only from the chemical structures of the ligands and their observed binding. The user does not have to propose a pharmacophore, guess the bioactive conformations of the ligands, or suggest ways to superimpose the active compounds. The method takes into account conformational flexibility of the ligands, stereospecific binding, diverse or unrelated chemical structures, inaccurate or qualitative binding data, and the possibility that chemically similar ligands may or may not bind to the receptor in similar orientations. The resulting model can be viewed graphically and interpreted in terms of one or more binding regions of the receptor, each preferring to be occupied by various sorts of chemical groups. The model always fits the given data completely and can predict the binding of any other ligand, regardless of chemical structure. The method is an outgrowth of distance geometry and Voronoi polyhedra site modeling but incorporates several novel features. The geometry of the ligand molecules and the site is described in terms of intervals of internal distances. Determining the site model consists of reducing the uncertainty in the interregion distance intervals, and this uncertainty is described as intervals of intervals. Similarly, the given binding affinities and their experimental uncertainties are treated as intervals in the affinity scale. The final site model specifies an entire region of interaction energy parameters that satisfy the training set rather than a single set of parameters. Predicted binding for test compounds results in an interval which, when compared to the experimental interval, may be correct, incorrect, or vague. There is a pervasive ternary logic involved in the assessment of predictions, in the search for a satisfactory model, and in judging whether a given molecule may bind in a particular orientation: true, false, or maybe. The approach is illustrated on an extremely simple artificial example and on a real data set of cocaine analogues binding to a nerve membrane receptor in vitro. © 1995 by John Wiley & Sons, Inc.     

Synthesis and Langmuir studies of bivalent and monovalent alpha-D-mannopyranosides with lectin Con A

Highly avid interaction between carbohydrate ligands and lectin receptors nominally requires the ligand presentation in a clustered form. We present herein an approach involving Langmuir monolayer formation of the sugar ligands and the assessment of their lectin binding at the air-water interface. Bivalent alpha-D-mannopyranoside containing the glycolipid ligand was used to study its binding profiles with lectin Con A, in comparison to the corresponding monovalent glycolipid. In addition to the bivalent and monovalent nature of the glycolipid ligands at the molecular level, the ligand densities at the monolayer level were varied with the aid of a nonsugar lipid molecule so as to obtain mixed monolayers with various sugar-nonsugar ratios. Lectin binding of bivalent and monovalent ligands at different ratios was monitored by differential changes in the surface area per molecule of the mixed monolayer, with and without the lectin. The present study shows that maximal binding of the lectin to the bivalent ligand occurs at lower sugar densities at the interface ( approximately 10% sugar in the mixed monolayer) than for that of the monovalent ligand ( approximately 20% sugar in the mixed monolayer). It is observed that complete coverage of the monolayer with only the sugar ligands does not allow all of the sugars to be functionally active.     

Experimental versus predicted affinities for ligand binding to estrogen receptor: iterative selection and rescoring of docked poses systematically improves the correlation

The computational determination of binding modes for a ligand into a protein receptor is much more successful than the prediction of relative binding affinities (RBAs) for a set of ligands. Here we consider the binding of a set of 26 synthetic A-CD ligands into the estrogen receptor ERα. We show that the MOE default scoring function (London dG) used to rank the docked poses leads to a negligible correlation with experimental RBAs. However, switching to an energy-based scoring function, using a multiple linear regression to fit experimental RBAs, selecting top-ranked poses and then iteratively repeating this process leads to exponential convergence in 4–7 iterations and a very strong correlation. The method is robust, as shown by various validation tests. This approach may be of general use in improving the quality of predicted binding affinities.     

Assessment of programs for ligand binding affinity prediction

The prediction of the binding free energy between a ligand and a protein is an important component in the virtual screening and lead optimization of ligands for drug discovery. To determine the quality of current binding free energy estimation programs, we examined FlexX, X-Score, AutoDock, and BLEEP for their performance in binding free energy prediction in various situations including cocrystallized complex structures, cross docking of ligands to their non-cocrystallized receptors, docking of thermally unfolded receptor decoys to their ligands, and complex structures with "randomized" ligand decoys. In no case was there a satisfactory correlation between the experimental and estimated binding free energies over all the datasets tested. Meanwhile, a strong correlation between ligand molecular weight-binding affinity correlation and experimental predicted binding affinity correlation was found. Sometimes the programs also correctly ranked ligands' binding affinities even though native interactions between the ligands and their receptors were essentially lost because of receptor deformation or ligand randomization, and the programs could not decisively discriminate randomized ligand decoys from their native ligands; this suggested that the tested programs miss important components for the accurate capture of specific ligand binding interactions.     

Design and synthesis of a bivalent ligand to explore the putative heterodimerization of the mu opioid receptor and the chemokine receptor CCR5

The bivalent ligand approach has been utilized not only to study the underlying mechanism of G protein-coupled receptors dimerization and/or oligomerization, but also to enhance ligand affinity and/or selectivity for potential treatment of a variety of diseases by targeting this process. Substance abuse and addiction have made both the prevention and the treatment of human immunodeficiency virus (HIV) infection more difficult to tackle. Morphine, a mu opioid receptor (MOR) agonist, can accelerate HIV infection through up-regulating the expression of the chemokine receptor CCR5, a well-known co-receptor for HIV invasion to the host cells and this has been extensively studied. Meanwhile, two research groups have described the putative MOR-CCR5 heterodimers in their independent studies. The purpose of this paper is to report the design and synthesis of a bivalent ligand to explore the biological and pharmacological process of the putative MOR-CCR5 dimerization phenomenon. The developed bivalent ligand thus contains two distinct pharmacophores linked through a spacer; ideally one of which will interact with the MOR and the other with the CCR5. Naltrexone and Maraviroc were selected as the pharmacophores to generate such a bivalent probe. The overall reaction route to prepare this bivalent ligand was convergent and efficient, and involved sixteen steps with moderate to good yields. The preliminary biological characterization showed that the bivalent compound 1 retained the pharmacological characteristics of both pharmacophores towards the MOR and the CCR5 respectively with relatively lower binding affinity, which tentatively validated our original molecular design.     

Nonspanning bivalent ligands as improved surface receptor binding inhibitors of the cholera toxin B pentamer

A series of bivalent ligands of varying length were synthesized to inhibit the receptor-binding process of cholera toxin. Competitive surface receptor binding assays showed that significant potency gains relative to the constituent monovalent ligands were achieved independently from the ability of the extended bivalent ligands to span binding sites within the toxin pentamer. Several models that could account for the unexpected improvement in IC(50) values are examined, taking into account crystallographic analysis of each ligand in complex with the toxin pentamer. Evidence is presented that steric blocking at the receptor binding surface may play a role. The results of our study suggest that the use of relatively short, "nonspanning" bivalent ligands, or monovalent ligands of similar topology and bulk may be an effective way of blocking the interaction of multimeric proteins with their cell surface receptors.     

Ab initio quantum mechanical study of the binding energies of human estrogen receptor alpha with its ligands: an application of fragment molecular orbital method

We have theoretically examined the relative binding affinities (RBA) of typical ligands, 17beta-estradiol (EST), 17alpha-estradiol (ESTA), genistein (GEN), raloxifene (RAL), 4-hydroxytamoxifen (OHT), tamoxifen (TAM), clomifene (CLO), 4-hydroxyclomifene (OHC), diethylstilbestrol (DES), bisphenol A (BISA), and bisphenol F (BISF), to the alpha-subtype of the human estrogen receptor ligand-binding domain (hERalpha LBD), by calculating their binding energies. The ab initio fragment molecular orbital (FMO) method, which we have recently proposed for the calculations of macromolecules such as proteins, was applied at the HF/STO-3G level. The receptor protein was primarily modeled by 50 amino acid residues surrounding the ligand. The number of atoms in these model complexes is about 850, including hydrogen atoms. For the complexes with EST, RAL, OHT, and DES, the binding energies were calculated again with the entire ERalphaLBD consisting of 241 residues or about 4000 atoms. No significant difference was found in the calculated binding energies between the model and the real protein complexes. This indicates that the binding between the protein and its ligands is well characterized by the model protein with the 50 residues. The calculated binding energies relative to EST were very well correlated with the experimental RBA (the correlation coefficient r=0.837) for the ligands studied in this work. We also found that the charge transfer between ER and ligands is significant on ER-ligand binding. To our knowledge, this is the first achievement of ab initio quantum mechanical calculations of large molecules such as the entire ERalphaLBD protein.