Advances and applications of binding affinity prediction methods in drug discovery

Nowadays, the improvement of R&D productivity is the primary commitment in pharmaceutical research, both in big pharma and smaller biotech companies. To reduce costs, to speed up the discovery process and to increase the chance of success, advanced methods of rational drug design are very helpful, as demonstrated by several successful applications. Among these, computational methods able to predict the binding affinity of small molecules to specific biological targets are of special interest because they can accelerate the discovery of new hit compounds. Here we provide an overview of the most widely used methods in the field of binding affinity prediction, as well as of our own work in developing BEAR, an innovative methodology specifically devised to overtake some limitations in existing approaches. The BEAR method was successfully validated against different biological targets, and proved its efficacy in retrieving active compounds from virtual screening campaigns. The results obtained so far indicate that BEAR may become a leading tool in the drug discovery pipeline. We primarily discuss advantages and drawbacks of each technique and show relevant examples and applications in drug discovery.     

Prediction of protein-mannose binding sites using random forest

Mannose is an abundant cell surface monosaccharide and has an important role in many biochemical processes. It binds to a great diversity of receptor proteins. In this study we have employed Random Forest for prediction of mannose binding sites. Mannosebinding site is taken to be a sphere around the centroid of the ligand and the sphere is subdivided into different layers and atom wise and residue wise features were extracted for each layer. The method achieves 95.59 % of accuracy using Random Forest with 10 fold cross validation. Prediction of mannose binding site analysis will be quite useful in drug design.     

化合物ADMET性质预测平台的构建

在药物研发早期阶段对化合物成药性和安全性进行评估,对于提高药物研发成功率、降低研发成本具有十分重要的意义。为了能够帮助药物研究工作者快速准确地判断候选化合物的成药性与安全性,开发了一个基于计算机方法的化合物ADMET性质预测平台。首先,通过文本挖掘的方法收集了化合物药代动力学性质和毒性(ADMET)的高质量实验数据。然后,根据原始文献复原了13个预测模型,同时采用支持向量机方法自建了15个具有较高预测能力的计算模型。最后,基于分布式架构,结合高性能计算集群优势,开发了化合物ADMET性质预测平台(http://www.vslead.com/?r=admet/index),用于预测28种重要的化合物ADMET性质。研究者可以使用这一平台快捷方便地对药物研究中比较重要的ADMET性质进行预测,在药物研发早期对候选化合物进行成药性评价和风险评估,有助于提高药物研的成功率,节省研发时间和经费的投入。     

Weak affinity chromatography as a new approach for fragment screening in drug discovery

Fragment-based drug design (FBDD) is currently being implemented in drug discovery, creating a demand for developing efficient techniques for fragment screening. Due to the intrinsic weak or transient binding of fragments (mM–μM in dissociation constant (K_D)) to targets, methods must be sensitive enough to accurately detect and quantify an interaction. This study presents weak affinity chromatography (WAC) as an alternative tool for screening of small fragments. The technology was demonstrated by screening of a selected 23-compound fragment collection of documented binders, mostly amidines, using trypsin and thrombin as model target protease proteins. WAC was proven to be a sensitive, robust, and reproducible technique that also provides information about affinity of a fragment in the range of 1 mM–10 μM. Furthermore, it has potential for high throughput as was evidenced by analyzing mixtures in the range of 10 substances by WAC–MS. The accessibility and flexibility of the technology were shown as fragment screening can be performed on standard HPLC equipment. The technology can further be miniaturized and adapted to the requirements of affinity ranges of the fragment library. All these features of WAC make it a potential method in drug discovery for fragment screening.     

Bioinformatics approaches for new drug discovery: a review

Prolonged antibiotic therapy for the bacterial infections has resulted in high levels of antibiotic resistance. Initially, bacteria are susceptible to the antibiotics, but can gradually develop resistance. Treating such drug-resistant bacteria remains difficult or even impossible. Hence, there is a need to develop effective drugs against bacterial pathogens. The drug discovery process is time-consuming, expensive and laborious. The traditionally available drug discovery process initiates with the identification of target as well as the most promising drug molecule, followed by the optimization of this, in-vitro, in-vivo and in pre-clinical studies to decide whether the compound has the potential to be developed as a drug molecule. Drug discovery, drug development and commercialization are complicated processes. To overcome some of these problems, there are many computational tools available for new drug discovery, which could be cost effective and less time-consuming. In-silico approaches can reduce the number of potential compounds from hundreds of thousands to the tens of thousands which could be studied for drug discovery and this results in savings of time, money and human resources. Our review is on the various computational methods employed in new drug discovery processes.     

Cross‐docking benchmark for automated pose and ranking prediction of ligand binding

Significant efforts have been devoted in the last decade to improving molecular docking techniques to predict both accurate binding poses and ranking affinities. Some shortcomings in the field are the limited number of standard methods for measuring docking success and the availability of widely accepted standard data sets for use as benchmarks in comparing different docking algorithms throughout the field. In order to address these issues, we have created a Cross‐Docking Benchmark server. The server is a versatile cross‐docking data set containing 4,399 protein‐ligand complexes across 95 protein targets intended to serve as benchmark set and gold standard for state‐of‐the‐art pose and ranking prediction in easy, medium, hard, or very hard docking targets. The benchmark along with a customizable cross‐docking data set generation tool is available at http://disco.csb.pitt.edu . We further demonstrate the potential uses of the server in questions outside of basic benchmarking such as the selection of the ideal docking reference structure.     

Screening for transient biological interactions as applied to albumin ligands: a new concept for drug discovery

The notion that many biological interactions are based on transient binding (dissociation constants (K(d)) in the range of 10-0.01 mM) is familiar, yet the implications for biological sciences have been realized only recently. An important area of biological sciences is drug design, where the traditional "lock and key" view of binding has prevailed and drug candidates are usually selected on their merits as being tight binders. However, the rationale that transient interactions are of importance for drug discovery is slowly gaining acceptance. These interactions may relate not only to the desired target interaction but also to unwanted interactions creating, for example, toxicity problems. Here we demonstrate, in a high-throughput screening format, affinity selection of weak binders to a model target of albumin by zonal retardation chromatography. It is perceived that this approach can define the "transient drug" as a complement to current drug discovery procedures.     

Analysis of binding modes of ligands to multiple conformations of CYP3A4

Cytochromes P450 (CYPs) are extremely versatile enzymes capable of catalyzing a vast number of compounds, and CYP3A4 is no exception metabolizing approximately half of the currently marketed drugs, besides endogenous compounds. To metabolize such a variety of compounds, CYP3A4 has to be extremely flexible, which makes interaction studies difficult. We employ a multi-conformational docking setup where conformations are generated by several molecular dynamics simulations to analyze the binding modes of various ligands, and the docking is considered successful if the ligand site of catalysis (SOC) is within 6.0 Å of the haem Fe. While docking with the X-ray structure proved unsuccessful with all ligands, the multi-conformational docking achieved successful binding of each ligand to at least one protein conformation. Analysis of the docked solutions highlights residues in the active site cavity that may have an important role in access, binding and stabilization of the ligand.     

Molecular modeling and evaluation of binding mode and affinity of artemisinin-quinine hybrid and its congeners with Fe-protoporphyrin-IX as a putative receptor

A recent rational approach to anti-malarial drug design is characterized as "covalent biotherapy" involves linking of two molecules with individual intrinsic activity into a single agent, thus packaging dual activity into a single hybrid molecule. In view of this background and reported anti malaria synergism between artemisinin and quinine; we describe the computer-assisted docking to predict molecular interaction and binding affinity of Artemisinin-Quinine hybrid and its derivatives with the intraparasitic haeme group of human haemoglobin. Starting from a crystallographic structure of Fe-protoporphyrin-IX, binding modes, orientation of peroxide bridge (Fe-O distance), docking score and interaction energy are predicted using the docking molecular mechanics based on generalized Born/surface area (MM-GBSA) solvation model. Seven new ligands were identified with a favourable glide score (XP score) and binding free energy (ΔG) with reference to the experimental structure from a data set of thirty four hybrid derivatives. The result shows the conformational property of the drug-receptor interaction and may lead to rational design and synthesis of improved potent artemisinin based hybrid antimalarial that target haemozoin formation.     

Alpha-1-proteinase inhibitor is a heparin binding serpin: molecular interactions with the Lys rich cluster of helix-F domain

Alpha-1-proteinase (alpha-1-PI) inhibitor is the major circulating serine protease inhibitor in humans. The porcine elastase and trypsin inhibitory activity of human and ovine alpha-1-PI is activated several fold in the presence of anti-coagulant heparin. The activation is allosteric and appears to be characterized by two steps of binding; a weak followed by a strong binding. The Kass for ovine and human alpha-1-PI inhibition of porcine pancreatic elastase was increased approximately 45 fold and 38 fold respectively. Using a combinatorial approach of multiple sequence alignment, surface topology, chemical modification and tryptic peptide mapping to identify the sequence of the heparin bound peptide; we demonstrate that heparin binds to the lysyl rich region of the F-helix of alpha-1-PI, which differs from that of heparin-antithrombin (AT) interactions. Molecular docking prediction using the MEDock algorithm approximates the three positively charged lysines (K154, K155, K174) of human alpha-1-PI in this interaction. This heparin alpha-1-PI interaction has been exploited to develop an affinity purification method, which can be used universally to obtain homogenous preparations of mammalian alpha-1-PIs useful for augmentation therapy. Collectively, all these findings imply that alpha-1-PI has a major role in regulating extra cellular protease activity and the physiological activator is heparin.     

Structure activity study of S-trityl-cysteamine dimethylaminopyridine derivatives as SIRT2 inhibitors: Improvement of SIRT2 binding and inhibition

Sirtuin proteins are a highly conserved class of nicotinamide adenine dinucleotide (NAD⁺)-dependent lysine deacylases. The pleiotropic human isoform 2 of Sirtuins (SIRT2) has been engaged in the pathogenesis of cancer in a plethora of reports around the globe. Thus, SIRT2 modulation is deemed as a promising approach for pharmaceutical intervention. Previously, we reported S-Trityl-l-Cysteine (STLC)-ornamented dimethylaminopyridine chemical entity named STC4 with a significant SIRT2 inhibitory capacity; this was separate from the conventional application of STLC scaffold as a kinesin-5 inhibitor. An interactive molecular docking study of SIRT2 and STC4 showed interaction between Asn168 of SIRT2 and the methyl ester of STC4, that appears to hinder STC4 to reach the selective pocket of the protein unlike strong SIRT2 inhibitor SirReal2. To improve its activity, herein, we utilized S-trityl cysteamine pharmacophore lacking the methyl ester. Nine compounds were synthesized and assayed affording three biopertinent SIRT2 inhibitors, and two of them, STCY1 and STCY6 showed higher inhibitory activity than STC4. These compounds have pronounced anti-proliferative activities against different cancer cell lines. A molecular docking study was executed to shed light on the supposed binding mode of the lead compound, STCY1, into the selective pocket of SIRT2 by interaction of the nitrogen of pyridine ring of the compound and Ala135 of the protein. The outcome of the study exposes that the active compounds are effective intermediates to construct more potent biological agents.     

Anchor profiles of HLA-specific peptides: analysis by a novel affinity scoring method and experimental validation

The study of intermolecular interactions is a fundamental research subject in biology. Here we report on the development of a quantitative structure-based affinity scoring method for peptide-protein complexes, named PepScope. The method operates on the basis of a highly specific force field function (CHARMM) that is applied to all-atom structural representations of peptide-receptor complexes. Peptide side-chain contributions to total affinity are scored after detailed rotameric sampling followed by controlled energy refinement. A de novo approach to estimate dehydration energies was developed, based on the simulation of individual amino acids in a solvent box filled with explicit water molecules. Transferability of the method was demonstrated by its application to the hydrophobic HLA-A2 and -A24 receptors, the polar HLA-A1, and the sterically ruled HLA-B7 receptor. A combined theoretical and experimental study on 39 anchor substitutions in FxSKQYMTx/HLA-A2 and -A24 complexes indicated a prediction accuracy of about two thirds of a log-unit in Kd. Analysis of free energy contributions identified a great role of desolvation and conformational strain effects in establishing a given specificity profile. Interestingly, the method rightly predicted that most anchor profiles are less specific than so far assumed. This suggests that many potential T-cell epitopes could be missed with current prediction methods. The results presented in this work may therefore significantly affect T-cell epitope discovery programs applied in the field of peptide vaccine development.     

A molecular thermodynamic view of DNA-drug interactions: a case study of 25 minor-groove binders

Developing a molecular view of the thermodynamics of DNA recognition is essential to the design of ligands for regulating gene expression. In a first comprehensive attempt at sketching an atlas of DNA-drug energetics, we present here a detailed thermodynamic view of minor-groove recognition by small molecules via a computational study on 25 DNA-drug complexes. The studies are configured in the MMGBSA (Molecular Mechanics-Generalized Born-Solvent Accessibility) framework at the current state of the art and facilitate a structure-energy component correlation. Analyses were conducted on both energy minimized structures of DNA-drug complexes and molecular dynamics trajectories developed for the purpose of this study. While highlighting the favorable role of packing, shape complementarity, and van der Waals and hydrophobic interactions of the drugs in the minor groove in conformity with experiment, the studies reveal an interesting annihilation of favorable electrostatics by desolvation. Structural modifications attempted on the ligands point to the requisite physico-chemical factors for obtaining improved binding energies. Hydrogen bonds predicted to be important for specificity based on structural considerations do not always turn out to be significant to binding in post facto analyses of molecular dynamics trajectories, which treat thermal averaging, solvent, and counterion effects rigorously. The strength of the hydrogen bonds retained between the DNA and drug during the molecular dynamics simulations is approximately 1kcal/mol. Overall, the study reveals the compensatory nature of the diverse binding free energy components, possible threshold limits for some of these properties, and the availability of a computationally viable free energy methodology which could be of value in drug-design endeavors.     

Coumarin-suberoylanilide hydroxamic acid as a fluorescent probe for determining binding affinities and off-rates of histone deacetylase inhibitors

Histone deacetylases (HDACs) are intimately involved in epigenetic regulation and, thus, are one of the key therapeutic targets for cancer, and two HDAC inhibitors, namely suberoylanilide hydroxamic acid (SAHA) and romidepsin, have been recently approved for cancer treatment. Because the screening and detailed characterization of HDAC inhibitors has been time-consuming, we synthesized coumarin-SAHA (c-SAHA) as a fluorescent probe for determining the binding affinities (K_d) and the dissociation off-rates (k_off) of the enzyme–inhibitor complexes. The determination of the above parameters relies on the changes in the fluorescence emission intensity (λ_ex = 325 nm, λ_em = 400 nm) of c-SAHA due to its competitive binding against other HDAC inhibitors, and such determination neither requires employment of polarization accessories nor is dependent on the fluorescence energy transfer from the enzyme’s tryptophan residues to the probe. Our highly sensitive and robust analytical protocol presented here is applicable to most of the HDAC isozymes, and it can be easily adopted in a high-throughput mode for screening the HDAC inhibitors as well as for quantitatively determining their K_d and k_off values.     

Towards understanding the unbound state of drug compounds: Implications for the intramolecular reorganization energy upon binding

There has been an explosion of structural information for pharmaceutical compounds bound to biological targets, but the conformations and dynamics of compounds free in solution are poorly characterized, if at all. Yet, knowledge of the unbound state is essential to understand the fundamentals of molecular recognition, including the much debated conformational intramolecular reorganization energy of a compound upon binding (ΔE_Reorg). Also, dependable observation of the unbound compounds is important for ligand-based drug discovery, e.g. with pharmacophore modelling. Here, these questions are addressed with long (⩾0.5 μs) state-of-the-art molecular dynamics (MD) simulations of 26 compounds (including 7 approved drugs) unbound in explicit solvent. These compounds were selected to be chemically diverse, with a range of flexibility, and good quality bioactive X-ray structures. The MD-simulated free compounds are compared to their bioactive structure and conformers generated with ad hoc sampling in vacuo or with implicit generalized Born (GB) aqueous solvation models. The GB conformational models clearly depart from those obtained in explicit solvent, and suffer from conformational collapse almost as severe as in vacuo. Thus, the global energy minima in vacuo or with GB are not suitable representations of the unbound state, which can instead be extensively sampled by MD simulations. Many, but not all, MD-simulated compounds displayed some structural similarity to their bioactive structure, supporting the notion of conformational pre-organization for binding. The ligand–protein complexes were also simulated in explicit solvent, to estimate ΔE_Reorg as an enthalpic difference ΔH_Reorg between the intramolecular energies in the bound and unbound states. This fresh approach yielded ΔH_Reorg values ⩽ 6 kcal/mol for 18 out of 26 compounds. For three particularly polar compounds 15 ⩽ ΔH_Reorg ⩽ 20 kcal/mol, supporting the notion that ΔH_Reorg can be substantial. Those large ΔH_Reorg values correspond to a redistribution of electrostatic interactions upon binding. Overall, the study illustrates how MD simulations offer a promising avenue to characterize the unbound state of medicinal compounds.     

Dye-ligand membranes as selective adsorbents for rapid purification of enzymes: a case study

A new adsorbent for the selective binding of enzymes, in the form of microporous membranes carrying triazine dyes as pseudo-affinity ligand, has been implemented in the recovery of glucose-6-phosphate dehydrogenase from yeast. A detailed investigation of the process parameters has been performed. In the adsorption step, the contact time for binding G6PDH could be reduced down to 0.25 s without significant decrease of the capture efficiency. Hence, fast filtration allowed to isolate G6PDH from a dilute extract (1.6 mug G6PDH . mL(-1)), where the enzyme accounted for 1% of the proteins. The yield of the selective elution step using NADP was only 70% at best. It could be improved to near 100% by supplementing the eluent with ethylene glycol, without loss of selectivity. A Scale-up of the cross-section of the membrane by a factor of 40 allowed to purify 1140 U from 0.6 L extract from 1% to 57% purity with 82% yield, within 10 minutes. The case study presented here demonstrates the applicability of general-purpose membrane adsorbents for the purification of enzymes.     

Interactions of urocanic acid with bovine serum albumin and the influence of pH on binding affinities: a docking simulation study

In the present study, the interactions of urocanic acid (UA) with bovine serum albumins (BSA) at pH 5.0 and 7.4 were investigated by means of docking simulations. The binding modes of trans- and cis-UA to BSA at pH 5.0 and 7.4 were analysed. In addition, the theoretically predicted binding abilities of zwitterion and anion of UA with BSA are in good agreement with the experimental results. Through comparison with the binding patterns, we revealed that the stronger interactions of UA anion with BSA relative to the zwitterion primarily result from: (1) the increased number of hydrogen bonds between UA anion and BSA; (2) the attractive electrostatic interaction between the deprotoned carboxyl group in UA anion and Arg433 in comparison with the repulsion between the imidazole moiety in zwitterion and the same residue in BSA. This provides a rational explanation for the experimental finding that the binding of UA to BSA at pH 7.4 is much stronger than at pH 5.0.     

1,1-Difluoroethyl-substituted triazolothienopyrimidines as inhibitors of a human urea transport protein (UT-B): New analogs and binding model

The kidney urea transport protein UT-B is an attractive target for the development of small-molecule inhibitors with a novel diuretic (‘urearetic’) action. Previously, two compounds in the triazolothienopyrimidine scaffold (1a and 1c) were reported as UT-B inhibitors. Compound 1c incorporates a 1,1-difluoroethyl group, which affords improved microsomal stability when compared to the corresponding ethyl-substituted compound 1a. Here, a small focused library (4a–4f) was developed around lead inhibitor 1c to investigate the requirement of an amidine-linked thiophene in the inhibitor scaffold. Two compounds (4a and 4b) with nanomolar inhibitory potency (IC₅₀ ≈ 40 nM) were synthesized. Computational docking of lead structure 1c and 4a–4f into a homology model of the UT-B cytoplasmic surface suggested binding with the core heterocycle buried deep into the hydrophobic pore region of the protein.     

Alkyloxy carbonyl modified hexapeptides as a high affinity compounds for Wnt5A protein in the treatment of psoriasis

Psoriasis is one of the most prevalent chronic inflammatory diseases of the skin. The Wnt pathways have been documented to play essential role in stem cell self-renewal and keratinocyte differentiation in the skin. Antagonizing the Wnt5a protein would emerge as a novel therapeutics in psoriasis treatment. In this view, we have developed and characterized series of compounds by attaching varied tertiary alkyloxy carbonyl groups at the N-terminal end of the hexapeptide (Met-Asp-Gly-Cys-Glu-Leu) bestowed to inhibit Wnt/Ca2+ signaling in psoriasis. Hexapeptide compound with 1,1-diphenylethoxy carbonyl group attached to N-terminal end of hexapeptide demonstrated highest binding affinity amongst all the evaluated compounds. The compound identified in the study can be subjected further for in vitro and in vivo studies for ADMET properties.