Exploring Scoring Function Space: Developing Computational Models for Drug Discovery.

BACKGROUND: The idea of scoring function space established a systems-level approach to address the development of models to predict the affinity of drug molecules by those interested in drug discovery. OBJECTIVE: Our goal here is to review the concept of scoring function space and how to explore it to develop machine learning models to address protein-ligand binding affinity. METHOD: We searched the articles available in PubMed related to the scoring function space. We also utilized crystallographic structures found in the protein data bank (PDB) to represent the protein space. RESULTS: The application of systems-level approaches to address receptor-drug interactions allows us to have a holistic view of the process of drug discovery. The scoring function space adds flexibility to the process since it makes it possible to see drug discovery as a relationship involving mathematical spaces. CONCLUSION: The application of the concept of scoring function space has provided us with an integrated view of drug discovery methods. This concept is useful during drug discovery, where we see the process as a computational search of the scoring function space to find an adequate model to predict receptor-drug binding affinity.

Insect RDL Receptor Models for Virtual Screening: Impact of the Template Conformational State in Pentameric Ligand-Gated Ion Channels.

The RDL receptor is one of the most relevant protein targets for insecticide molecules. It belongs to the pentameric ligand-gated ion channel (pLGIC) family. Given that the experimental structures of pLGICs are difficult to obtain, homology modeling has been extensively used for these proteins, particularly for the RDL receptor. However, no detailed assessments of the usefulness of homology models for virtual screening (VS) have been carried out for pLGICs. The aim of this study was to evaluate which are the determinant factors for a good VS performance using RDL homology models, specially analyzing the impact of the template conformational state. Fifteen RDL homology models were obtained based on different pLGIC templates representing the closed, open, and desensitized states. A retrospective VS process was performed on each model, and their performance in the prioritization of active ligands was assessed. In addition, the three best-performing models among each of the conformations were subjected to molecular dynamics simulations (MDS) in complex with a representative active ligand. The models showed variations in their VS performance parameters that were related to the structural properties of the binding site. VS performance tended to improve in more constricted binding cavities. The best performance was obtained with a model based on a template in the closed conformation. MDS confirmed that the closed model was the one that best represented the interactions with an active ligand. These results imply that different templates should be evaluated and the structural variations between their channel conformational states should be specially examined, providing guidelines for the application of homology modeling for VS in other proteins of the pLGIC family.

Target identification for repurposed drugs active against SARS-CoV-2 via high-throughput inverse docking.

Screening already approved drugs for activity against a novel pathogen can be an important part of global rapid-response strategies in pandemics. Such high-throughput repurposing screens have already identified several existing drugs with potential to combat SARS-CoV-2. However, moving these hits forward for possible development into drugs specifically against this pathogen requires unambiguous identification of their corresponding targets, something the high-throughput screens are not typically designed to reveal. We present here a new computational inverse-docking protocol that uses all-atom protein structures and a combination of docking methods to rank-order targets for each of several existing drugs for which a plurality of recent high-throughput screens detected anti-SARS-CoV-2 activity. We demonstrate validation of this method with known drug-target pairs, including both non-antiviral and antiviral compounds. We subjected 152 distinct drugs potentially suitable for repurposing to the inverse docking procedure. The most common preferential targets were the human enzymes TMPRSS2 and PIKfyve, followed by the viral enzymes Helicase and PLpro. All compounds that selected TMPRSS2 are known serine protease inhibitors, and those that selected PIKfyve are known tyrosine kinase inhibitors. Detailed structural analysis of the docking poses revealed important insights into why these selections arose, and could potentially lead to more rational design of new drugs against these targets.

Effects of gabergic phenols on the dynamic and structure of lipid bilayers: A molecular dynamic simulation approach.

γ-Aminobutyric acid (GABA) is the major inhibitory neurotransmitter in the vertebrate and invertebrate nervous system. GABAA receptors are activated by GABA and their agonists, and modulated by a wide variety of recognized drugs, including barbiturates, anesthetics, and benzodiazepines. The phenols propofol, thymol, chlorothymol, carvacrol and eugenol act as positive allosteric modulators on GABAA-R receptor. These GABAergic phenols interact with the lipid membrane, therefore, their anesthetic activity could be the combined result of their specific activity (with receptor proteins) as well as nonspecific interactions (with surrounding lipid molecules) modulating the supramolecular organization of the receptor environment. Therefore, we aimed to contribute to a description of the molecular events that occur at the membrane level as part of the mechanism of general anesthesia, using a molecular dynamic simulation approach. Equilibrium molecular dynamics simulations indicate that the presence of GABAergic phenols in a DPPC bilayer orders lipid acyl chains for carbons near the interface and their effect is not significant at the bilayer center. Phenols interacts with the polar interface of phospholipid bilayer, particularly forming hydrogen bonds with the glycerol and phosphate group. Also, potential of mean force calculations using umbrella sampling show that propofol partition is mainly enthalpic driven at the polar region and entropic driven at the hydrocarbon chains. Finally, potential of mean force indicates that propofol partition into a gel DPPC phase is not favorable. Our in silico results were positively contrasted with previous experimental data.

Relevance of the protein macrodipole in the membrane-binding process. Interactions of fatty-acid binding proteins with cationic lipid membranes.

The fatty acid-binding proteins L-BABP and Rep1-NCXSQ bind to anionic lipid membranes by electrostatic interactions. According to Molecular Dynamics (MD) simulations, the interaction of the protein macrodipole with the membrane electric field is a driving force for protein binding and orientation in the interface. To further explore this hypothesis, we studied the interactions of these proteins with cationic lipid membranes. As in the case of anionic lipid membranes, we found that both proteins, carrying a negative as well as a positive net charge, were bound to the positively charged membrane. Their major axis, those connecting the bottom of the ß-barrel with the α-helix portal domain, were rotated about 180 degrees as compared with their orientations in the anionic lipid membranes. Fourier transform infrared (FTIR) spectroscopy of the proteins showed that the positively charged membranes were also able to induce conformational changes with a reduction of the ß-strand proportion and an increase in α-helix secondary structure. Fatty acid-binding proteins (FABPs) are involved in several cell processes, such as maintaining lipid homeostasis in cells. They transport hydrophobic molecules in aqueous medium and deliver them into lipid membranes. Therefore, the interfacial orientation and conformation, both shown herein to be electrostatically determined, have a strong correlation with the specific mechanism by which each particular FABP exerts its biological function.

Coalescence of Nanoclusters Analyzed by Well-Tempered Metadynamics. Comparison with Straightforward Molecular Dynamics.

The coalescence process of two nanoparticles to yield a core-shell structure is analyzed by a well-tempered metadynamics procedure. This methodology has been shown to be useful in understanding the present phenomenon in terms of two collective variables: the distance between the center of mass of the coalescing particles and the gyration radius of the resulting core element. The free-energy contour plots clearly show that the coalescence process involves the deformation of the core material, which is manifested in the residence of the system in regions with a larger gyration radius. Results from molecular dynamics for the same system were found helpful to reach the definition of this second collective variable. The advantages and limitations of the latter approach are discussed.

Improved prediction of bilayer and monolayer properties using a refined BMW-MARTINI force field.

Coarse-grained (CG) models allow enlarging the size and time scales that are reachable by atomistic molecular dynamics simulations. A CG force field (FF) for lipids and amino acids that possesses a polarizable water model has been developed following the MARTINI parametrization strategy, the BMW-MARTINI [1]. We tested the BMW-MARTINI FF capability to describe some structural and thermodynamical properties of lipid monolayers and bilayers. We found that, since the surface tension values of oil/water interfaces calculated with the model are not correct, compression isotherms of lipid monolayers present artifacts. Also, this FF predicts DPPC and DAPC bilayers to remain in the Lα phase at temperatures as low as 283K, contrary to the expected from their experimental Tm values. Finally, simulations at constant temperature of bilayers of saturated lipids belonging to PC homologous, showed an increase in the mean molecular area (Mma) upon increasing the chain length, inversely to the experimental observation. We refined BMW-MARTINI FF by modifying as few parameters as possible in order to bring simulated and experimental measurements closer. We have also modified structural parameters of the lipid geometry that do not have direct influence in global properties of the bilayer membranes or monolayers, but serve to approach the obtained CG geometry to atomistic reference values. The refined FF is able to better reproduce phase transition temperatures and Mma for saturated PC bilayers than BMW-MARTINI and MARTINI FF. Finally, the simulated surface pressure-Mma isotherms of PC monolayers resemble the experimental ones and eliminate serious artifacts of previous models.

Vinardo: A Scoring Function Based on Autodock Vina Improves Scoring, Docking, and Virtual Screening.

Autodock Vina is a very popular, and highly cited, open source docking program. Here we present a scoring function which we call Vinardo (Vina RaDii Optimized). Vinardo is based on Vina, and was trained through a novel approach, on state of the art datasets. We show that the traditional approach to train empirical scoring functions, using linear regression to optimize the correlation of predicted and experimental binding affinities, does not result in a function with optimal docking capabilities. On the other hand, a combination of scoring, minimization, and re-docking on carefully curated training datasets allowed us to develop a simplified scoring function with optimum docking performance. This article provides an overview of the development of the Vinardo scoring function, highlights its differences with Vina, and compares the performance of the two scoring functions in scoring, docking and virtual screening applications. Vinardo outperforms Vina in all tests performed, for all datasets analyzed. The Vinardo scoring function is available as an option within Smina, a fork of Vina, which is freely available under the GNU Public License v2.0 from http://smina.sf.net. Precompiled binaries, source code, documentation and a tutorial for using Smina to run the Vinardo scoring function are available at the same address.

Reversing the peptide sequence impacts on molecular surface behaviour.

The protein's primary structure has all the information for specific protein/peptide folding and, in many cases, can define specific amphiphilic regions along molecules that are important for interaction with membranes. In order to shed light on how peptide sequence is important for the surface properties of amphiphilic peptides, we designed three pairs of peptides with the following characteristics: (1) all molecules have the same hydrophobic residues; (2) the couples differ from each other in their hydrophilic amino acids: positively, negatively and non-charged; (3) each pair has the same residues (same global molecular hydrophobicity) but the primary structure is reversed in comparison to its partner (retro-isomer), giving a molecule with a hydrophilic N or C-terminus and a hydrophobic C or N-terminus. Using the Langmuir monolayer approach, we observed that sequence reversal has a central role in the lateral stability of peptide monolayers, in the ability of the molecules to partition into the air-water interface and in the rheological properties of peptide films, whereas the peptide's secondary structure, determined by ATR-FTIR, was the same for all peptides. Reversing the sequence also gives a differential way of peptide/lipid interaction when peptides are in the presence of POPC lipid bilayers. Our results show how sequence inversion confers a distinctive peptide surface behaviour and lipid interaction for molecules with a similar structure.

Interactions of the fatty acid-binding protein ReP1-NCXSQ with lipid membranes. Influence of the membrane electric field on binding and orientation.

The regulatory protein of the squid nerve sodium calcium exchanger (ReP1-NCXSQ) is a 15kDa soluble, intracellular protein that regulates the activity of the Na(+)/Ca(2+) exchanger in the squid axon. It is a member of the cellular retinoic acid-binding proteins family and the fatty acid-binding proteins superfamily. It is composed of ten beta strands defining an inner cavity and a domain of two short alpha helix segments. In this work, we studied the binding and orientation of ReP1-NCXSQ in anionic and zwitterionic lipid membranes using molecular dynamics (MD) simulations. Binding to lipid membranes was also measured by filtration binding assay. ReP1-NCXSQ acquired an orientation in the anionic membranes with the positive end of the macrodipole pointing to the lipid membrane. Potential of mean force calculations, in agreement with experimental measurements, showed that the binding to the anionic interfaces in low ionic strength was stronger than the binding to anionic interfaces in high ionic strength or to zwitterionic membranes. The results of MD showed that the electrostatic binding can be mediated not only by defined patches or domains of basic residues but also by a global asymmetric distribution of charges. A combination of dipole-electric field interaction and local interactions determined the orientation of ReP1-NCXSQ in the interface.

Analysis of the interaction interfaces of the N-terminal domain from Pseudomonas aeruginosa MutL.

Mismatch Repair System corrects mutations arising from DNA replication that escape from DNA polymerase proofreading activity. This system consists of three main proteins, MutS-L-H, responsible for lesion recognition and repair. MutL is a member of GHKL ATPase family and its ATPase cycle has been proposed to modulate MutL activity during the repair process. Pseudomonas aeruginosa MutL (PaMutL) contains an N-terminal (NTD) ATPase domain connected by a linker to a C-terminal (CTD) dimerization domain that possesses metal ion-dependent endonuclease activity. With the aim to identify characteristics that allow the PaMutL NTD allosteric control of CTD endonuclease activity, we used an in silico and experimental approach to determine the interaction surfaces of P. aeruginosa NTD (PaNTD), and compared it with the well characterized Escherichia coli MutL NTD (EcNTD). Molecular dynamics simulations of PaNTD and EcNTD bound to or free of adenosine nucleotides showed that a significant difference exists between the behavior of the EcNTD and PaNTD dimerization interface, particularly in the ATP lid. Structure based simulations of MutL homologues with endonuclease activity were performed that allowed an insight of the dimerization interface behavior in this family of proteins. Our experimental results show that, unlike EcNTD, PaNTD is dimeric in presence of ADP. Simulations in mixed solvent allowed us to identify the PaNTD putative DNA binding patch and a putative interaction patch located opposite to the dimerization face. Structure based simulations of PaNTD dimer in presence of ADP or ATP suggest that nucleotide binding could differentially modulate PaNTD protein-protein interactions. Far western assays performed in presence of ADP or ATP are in agreement with our in silico analysis.

Thermodynamic and structural analysis of homodimeric proteins: model of ß-lactoglobulin.

The energetics of protein homo-oligomerization was analyzed in detail with the application of a general thermodynamic model. We have studied the thermodynamic aspects of protein-protein interaction employing ß-lactoglobulin A from bovine milk at pH=6.7 where the protein is mainly in its dimeric form. We performed differential calorimetric scans at different total protein concentration and the resulting thermograms were analyzed with the thermodynamic model for oligomeric proteins previously developed. The thermodynamic model employed, allowed the prediction of the sign of the enthalpy of dimerization, the analysis of complex calorimetric profiles without transitions baselines subtraction and the obtainment of the thermodynamic parameters from the unfolding and the association processes and the compared with association parameters obtained with Isothermal Titration Calorimetry performed at different temperatures. The dissociation and unfolding reactions were also monitored by Fourier-transform infrared spectroscopy and the results indicated that the dimer of ß-lactoglobulin (N(2)) reversibly dissociates into monomeric units (N) which are structurally distinguishable by changes in their infrared absorbance spectra upon heating. Hence, it is proposed that ß-lactoglobulin follows the conformational path induced by temperature:N(2)⇌2N⇌2D. The general model was validated with these results indicating that it can be employed in the study of the thermodynamics of other homo-oligomeric protein systems.

A Straightforward Approach for the Determination of the Maximum Time Step for the Simulation of Nanometric Metallic Systems.

In the present work, we report on a systematic analysis to determine the maximum time step allowed in molecular dynamics simulations applied to study metal systems of current interest in nanoscience. Using the velocity Verlet integration scheme, we have found that it is possible to use a 20 fs time step for the simulation of gold nanosystems. This is roughly an order of magnitude greater than the usually employed integration step (2 to 5 fs). We also propose a general criterion to select this maximum time step for other metallic nanosystems, even in the case of bimetallic nanosystems.

Catalytic and glycan-binding abilities of ppGalNAc-T2 are regulated by acetylation.

Post-translational acetylation is an important molecular regulatory mechanism affecting the biological activity of proteins. Polypeptide GalNAc transferases (ppGalNAc-Ts) are a family of enzymes that catalyze initiation of mucin-type O-glycosylation. All ppGalNAc-Ts in mammals are type II transmembrane proteins having a Golgi lumenal region that contains a catalytic domain with glycosyltransferase activity, and a C-terminal R-type ("ricin-like") lectin domain. We investigated the effect of acetylation on catalytic activity of glycosyltransferase, and on fine carbohydrate-binding specificity of the R-type lectin domain of ppGalNAc-T2. Acetylation effect on ppGalNAc-T2 biological activity in vitro was studied using a purified human recombinant ppGalNAc-T2. Mass spectrometric analysis of acetylated ppGalNAc-T2 revealed seven acetylated amino acids (K103, S109, K111, K363, S373, K521, and S529); the first five are located in the catalytic domain. Specific glycosyltransferase activity of ppGalNAc-T2 was reduced 95% by acetylation. The last two amino acids, K521 and S529, are located in the lectin domain, and their acetylation results in alteration of the carbohydrate-binding ability of ppGalNAc-T2. Direct binding assays showed that acetylation of ppGalNAc-T2 enhances the recognition to αGalNAc residue of MUC1αGalNAc, while competitive assays showed that acetylation modifies the fine GalNAc-binding form of the lectin domain. Taken together, these findings clearly indicate that biological activity (catalytic capacity and glycan-binding ability) of ppGalNAc-T2 is regulated by acetylation.

Interfacial properties of the M1 segment of the nicotinic acetylcholine receptor.

We have studied the thermodynamic, surface, and structural properties of alphaM1 transmembrane sequence of the nicotinic acetylcholine receptor (nAChR) by using Langmuir monolayer, FT-IR spectroscopy and molecular dynamics simulation techniques in membrane-mimicking environments. M1 spontaneously incorporates into a lipid-free air-water interface, showing a favourable adsorption free energy of -7.2 kcal/mol. A cross-sectional molecular area of 210 A(2)/molecule, a surface potential of 4.2 fV/molecule and a high stability of the film were deducted from pure M1 monolayers. FT-IR experiments and molecular dynamics simulations in membrane-mimicking environments (sodium-dodecyl-sulfate and CCl(4), respectively) indicate coexistence between helical and non-helical structures. Furthermore, mixed peptide-lipid monolayers and monolayer penetration experiments were performed in order to study the peptide-lipid interaction. Mixed with condensed lipids (dipalmitoyl-phosphocholine, and dipalmitoyl-phosphoglycerol), M1 shows immiscible/miscible behaviour at low/high peptide concentration, respectively. Conversely, a complete miscible peptide-lipid interface is observed with liquid-expanded lipids (palmitoyl-oleoyl-phosphocholine, and palmitoyl-oleoyl-phosphoglycerol). Peptide penetration experiments demonstrate that the M1 peptide preferentially interacts with zwitterionic phosphocholine interfaces.

Fine carbohydrate recognition of Euphorbia milii lectin.

Glycans are key structures involved in biological processes such as cell attachment, migration, and invasion. Information coded on cell-surface glycans is frequently deciphered by proteins, as lectins, that recognize specific carbohydrate topology. Here, we describe the fine carbohydrate specificity of Euphorbia milii lectin (EML). Competitive assays using various sugars showed that GalNAc was the strongest inhibitor, and that the hydroxyl axial position of C4 and acetamido on C2 of GalNAc are critical points of EML recognition. A hydrophobic locus adjacent to GalNAc is also an important region for EML binding. Direct binding assays of EML revealed a stereochemical requirement for a structure adjacent to terminal GalNAc, showing that GalNAc residue is a necessary but not sufficient condition for EML interaction. The capacity of EML to bind epithelial tumor cells makes it a potentially useful tool for study of some over-expressed GalNAc glycoconjugates.

Vanadate inhibits the ATPase activity and DNA binding capability of bacterial MutS. A structural model for the vanadate-MutS interaction at the Walker A motif.

MutS, a member of the ABC ATPases superfamily, is a mismatch DNA-binding protein constituent of the DNA post-replicative mismatch repair system (MMRS). In this work, it is shown that the ATPase activity of Pseudomonas aeruginosa and Escherichia coli MutS is inhibited by ortho- and decavanadate. Structural comparison of the region involved in the ATP binding of E.coli MutS with the corresponding region of other ABC ATPases inhibited by vanadate, including the myosin- orthovanadate-Mg complex, showed that they are highly similar. From these results it is proposed that the orthovanadate inhibition of MutS ATPase can take place by a similar mechanism to that described for other ATPases. Docking of decavanadate on the ATP-binding region of MutS showed that the energetically more favorable interaction of this compound would take place with the complex MutS- ADP-Mg, suggesting that the inhibitory effect could be produced by a steric impediment of the protein ATP/ADP exchange. Besides the effect observed on the ATPase activity, vanadate also affects the DNA-binding capability of the protein, and partially inhibits the oligomerization of MutS and the temperature-induced inactivation of the protein. From the results obtained, and considering that vanadate is an intracellular trace component, this compound could be considered as a new modulator of the MMRS.