Machine learning methods and docking for predicting human pregnane X receptor activation

Combining graph neural networks and transformers for few-shot nuclear receptor binding activity prediction

Nuclear receptors (NRs) play a crucial role as biological targets in drug discovery. However, determining which compounds can act as endocrine disruptors and modulate the function of NRs with a reduced amount of candidate drugs is a challenging task. Moreover, the computational methods for NR-binding activity prediction mostly focus on a single receptor at a time, which may limit their effectiveness. Hence, the transfer of learned knowledge among multiple NRs can improve the performance of molecular predictors and lead to the development of more effective drugs. In this research, we integrate graph neural networks (GNNs) and Transformers to introduce a few-shot GNN-Transformer, Meta-GTNRP to predict the binding activity of compounds using the combined information of different NRs and identify potential NR-modulators with limited data. The Meta-GTNRP model captures the local information in graph-structured data and preserves the global-semantic structure of molecular graph embeddings for NR-binding activity prediction. Furthermore, a few-shot meta-learning approach is proposed to optimize model parameters for different NR-binding tasks and leverage the complementarity among multiple NR-specific tasks to predict binding activity of compounds for each NR with just a few labeled molecules. Experiments with a compound database containing annotations on the binding activity for 11 NRs shows that Meta-GTNRP outperforms other graph-based approaches. The data and code are available at: https://github.com/ltorres97/Meta-GTNRP .Scientific contributionThe proposed few-shot GNN-Transformer model, Meta-GTNRP captures the local structure of molecular graphs and preserves the global-semantic information of graph embeddings to predict the NR-binding activity of compounds with limited available data; A few-shot meta-learning framework adapts model parameters across NR-specific tasks for different NRs in a joint learning procedure to predict the binding activity of compounds for each NR with just a few labeled molecules in highly imbalanced data scenarios; Meta-GTNRP is a data-efficient approach that combines the strengths of GNNs and Transformers to predict the NR-binding properties of compounds through an optimized meta-learning procedure and deliver robust results valuable to identify potential NR-based drug candidates.

Development and Experimental Validation of Regularized Machine Learning Models Detecting New, Structurally Distinct Activators of PXR

The pregnane X receptor (PXR) regulates the metabolism of many xenobiotic and endobiotic substances. In consequence, PXR decreases the efficacy of many small-molecule drugs and induces drug-drug interactions. The prediction of PXR activators with theoretical approaches such as machine learning (ML) proves challenging due to the ligand promiscuity of PXR, which is related to its large and flexible binding pocket. In this work we demonstrate, by the example of random forest models and support vector machines, that classifiers generated following classical training procedures often fail to predict PXR activity for compounds that are dissimilar from those in the training set. We present a novel regularization technique that penalizes the gap between a model’s training and validation performance. On a challenging test set, this technique led to improvements in Matthew correlation coefficients (MCCs) by up to 0.21. Using these regularized ML models, we selected 31 compounds that are structurally distinct from known PXR ligands for experimental validation. Twelve of them were confirmed as active in the cellular PXR ligand-binding domain assembly assay and more hits were identified during follow-up studies. Comprehensive analysis of key features of PXR biology conducted for three representative hits confirmed their ability to activate the PXR.

Population pharmacokinetic model selection assisted by machine learning

A fit-for-purpose structural and statistical model is the first major requirement in population pharmacometric model development. In this manuscript we discuss how this complex and computationally intensive task could benefit from supervised machine learning algorithms. We compared the classical pharmacometric approach with two machine learning methods, genetic algorithm and neural networks, in different scenarios based on simulated pharmacokinetic data. Genetic algorithm performance was assessed using a fitness function based on log-likelihood, whilst neural networks were trained using mean square error or binary cross-entropy loss. Machine learning provided a selection based only on statistical rules and achieved accurate selection. The minimization process of genetic algorithm was successful at allowing the algorithm to select plausible models. Neural network classification tasks achieved the most accurate results. Neural network regression tasks were less precise than neural network classification and genetic algorithm methods. The computational gain obtained by using machine learning was substantial, especially in the case of neural networks. We demonstrated that machine learning methods can greatly increase the efficiency of pharmacokinetic population model selection in case of large datasets or complex models requiring long run-times. Our results suggest that machine learning approaches can achieve a first fast selection of models which can be followed by more conventional pharmacometric approaches.

Fast screening of covariates in population models empowered by machine learning

One of the objectives of Pharmacometry (PMX) population modeling is the identification of significant and clinically relevant relationships between parameters and covariates. Here, we demonstrate how this complex selection task could benefit from supervised learning algorithms using importance scores. We compare various classical methods with three machine learning (ML) methods applied to NONMEM empirical Bayes estimates: random forest, neural networks (NNs), and support vector regression (SVR). The performance of the ML models is assessed using receiver operating characteristic (ROC) curves. The F1 score, which measures test accuracy, is used to compare ML and PMX approaches. Methods are applied to different scenarios of covariate influence based on simulated pharmacokinetics data. ML achieved similar or better F1 scores than stepwise covariate modeling (SCM) and conditional sampling for stepwise approach based on correlation tests (COSSAC). Correlations between covariates and the number of false covariates does not affect the performance of any method, but effect size has an impact. Methods are not equivalent with respect to computational speed; SCM is 30 and 100-times slower than NN and SVR, respectively. The results are validated in an additional scenario involving 100 covariates. Taken together, the results indicate that ML methods can greatly increase the efficiency of population covariate model building in the case of large datasets or complex models that require long run-times. This can provide fast initial covariate screening, which can be followed by more conventional PMX approaches to assess the clinical relevance of selected covariates and build the final model.

Designing Out PXR Activity on Drug Discovery Projects: A Review of Structure-Based Methods, Empirical and Computational Approaches

This perspective discusses the role of pregnane xenobiotic receptor (PXR) in drug discovery and the impact of its activation on CYP3A4 induction. The use of structural biology to reduce PXR activity on drug discovery projects has become more common in recent years. Analysis of this work highlights several important molecular interactions, and the resultant structural modifications to reduce PXR activity are summarized. The computational approaches undertaken to support the design of new drugs devoid of PXR activation potential are also discussed. Finally, the SAR of empirical design strategies to reduce PXR activity is reviewed, and the key SAR transformations are discussed and summarized. In conclusion, this perspective demonstrates that PXR activity can be greatly diminished or negated on active drug discovery projects with the knowledge now available. This perspective should be useful to anyone who seeks to reduce PXR activity on a drug discovery project.

Identification of novel pregnane X receptor (PXR) agonists by In silico and biological activity analyses and reversal of cigarette smoke-induced PXR downregulation

Cigarette smoke (CS) contains many toxins that collectively harm nearly every organ in the body, and smoking is a key risk factor for many chronic diseases. Aside from its toxic actions, CS may alter expression of the drug- and steroid-binding pregnane X receptor (PXR), which when activated upregulates expression of cytochrome P450 (CYP) enzymes, glutathione transferases (GSTs), and multidrug resistance protein 1 (MDR1), an adaptive metabolic array that mediates clearance of CS component toxins. We sought to identify new PXR agonists that may be useful for restoring PXR activity in conditions wherein it is suppressed, and their mechanisms of PXR binding and activation. PXR has a uniquely larger, hydrophobic, and highly flexible ligand-binding domain (LBD) vs. other nuclear receptors, enabling it to interact with structurally diverse molecules. We tested certain calcium channel blockers (CCBs) as a pharmacological subset of potential PXR ligands, analyzing by molecular docking methods, and identified a putative active site in the PXR LBD, along with the relevant bonds and bonding energies. We analyzed felodipine binding and agonist activity in detail, as it showed the lowest binding energy among CCBs tested. We found felodipine was a potent PXR agonist as measured by luciferase reporter assay, whereas CCBs with higher binding energies were less potent (amlodipine) or nearly inactive (manidipine), and it induced CYP3A4 expression in HepG2 cells, a known target of PXR agonism. Felodipine also both induced PXR mRNA in HepG2 hepatocytes and reduced CS extract-induced diminution of PXR levels, indicating it modulates PXR expression. The results illuminate mechanisms of ligand-induced PXR activation and identify felodipine as a novel PXR agonist.

NURA: A curated dataset of nuclear receptor modulators

Nuclear receptors (NRs) are key regulators of human health and constitute a relevant target for medicinal chemistry applications as well as for toxicological risk assessment. Several open databases dedicated to small molecules that modulate NRs exist; however, depending on their final aim (i.e., adverse effect assessment or drug design), these databases contain a different amount and type of annotated molecules, along with a different distribution of experimental bioactivity values. Stemming from these considerations, in this work we aim to provide a unified dataset, NURA (NUclear Receptor Activity) dataset, collecting curated information on small molecules that modulate NRs, to be intended for both pharmacological and toxicological applications. NURA contains bioactivity annotations for 15,247 molecules and 11 selected NRs, and it was obtained by integrating and curating data from toxicological and pharmacological databases (i.e., Tox21, ChEMBL, NR-DBIND and BindingDB). Our results show that NURA dataset is a useful tool to bridge the gap between toxicology- and medicinal-chemistry-related databases, as it is enriched in terms of number of molecules, structural diversity and covered atomic scaffolds compared to the single sources. To the best of our knowledge, NURA dataset is the most exhaustive collection of small molecules annotated for their modulation of the chosen nuclear receptors. NURA dataset is intended to support decision-making in pharmacology and toxicology, as well as to contribute to data-driven applications, such as machine learning. The dataset and the data curation pipeline can be downloaded free of charge on Zenodo at the following DOI: https://doi.org/10.5281/zenodo.3991561.

Computational prediction of cytochrome P450 inhibition and induction

Cytochrome P450 (CYP) enzymes play an important role in the phase I metabolism of many xenobiotics. Most drug-drug interactions (DDIs) associated with CYP are caused by either CYP inhibition or induction. The early detection of potential DDIs is highly desirable in the pharmaceutical industry because DDIs can cause serious adverse events, which can lead to poor patient health and drug development failures. Recently, many computational studies predicting CYP inhibition and induction have been reported. The current computational modeling approaches for CYP metabolism are classified as ligand- and structure-based; various techniques, such as quantitative structure-activity relationships, machine learning, docking, and molecular dynamic simulation, are involved in both the approaches. Recently, combining these two approaches have resulted in improvements in the prediction accuracy of DDIs. In this review, we present important, recent developments in the computational prediction of the inhibition of four clinically crucial CYP isoforms (CYP1A2, 2C9, 2D6, and 3A4) and three nuclear receptors (aryl hydrocarbon receptor, constitutive androstane receptor, and pregnane X receptor) involved in the induction of CYP1A2, 2B6, and 3A4, respectively.

Back to the future: Advances in development of broad-spectrum capsid-binding inhibitors of enteroviruses

The hydrophobic pocket within viral capsid protein 1 is a target to combat the rhino- and enteroviruses (RV and EV) using small molecules. The highly conserved amino acids lining this pocket enable the development of antivirals with broad-spectrum of activity against numerous RVs and EVs. Inhibitor binding blocks: the attachment of the virion to the host cell membrane, viral uncoating, and/or production of infectious virus particles. Syntheses and biological studies of the most well-known antipicornaviral capsid binders have been reviewed and we propose next steps in this research.

Prediction of Drug-Drug Interactions Related to Inhibition or Induction of Drug-Metabolizing Enzymes

Drug-drug interaction (DDI) is the phenomenon of alteration of the pharmacological activity of a drug(s) when another drug(s) is co-administered in cases of so-called polypharmacy. There are three types of DDIs: pharmacokinetic (PK), pharmacodynamic, and pharmaceutical. PK is the most frequent type of DDI, which often appears as a result of the inhibition or induction of drug-metabolising enzymes (DME). In this review, we summarise in silico methods that may be applied for the prediction of the inhibition or induction of DMEs and describe appropriate computational methods for DDI prediction, showing the current situation and perspectives of these approaches in medicinal and pharmaceutical chemistry. We review sources of information on DDI, which can be used in pharmaceutical investigations and medicinal practice and/or for the creation of computational models. The problem of the inaccuracy and redundancy of these data are discussed. We provide information on the state-of-the-art physiologically- based pharmacokinetic modelling (PBPK) approaches and DME-based in silico methods. In the section on ligand-based methods, we describe pharmacophore models, molecular field analysis, quantitative structure-activity relationships (QSAR), and similarity analysis applied to the prediction of DDI related to the inhibition or induction of DME. In conclusion, we discuss the problems of DDI severity assessment, mention factors that influence severity, and highlight the issues, perspectives and practical using of in silico methods.

Machine Learning Methods in Computational Toxicology

Various methods of machine learning, supervised and unsupervised, linear and nonlinear, classification and regression, in combination with various types of molecular descriptors, both "handcrafted" and "data-driven," are considered in the context of their use in computational toxicology. The use of multiple linear regression, variants of naïve Bayes classifier, k-nearest neighbors, support vector machine, decision trees, ensemble learning, random forest, several types of neural networks, and deep learning is the focus of attention of this review. The role of fragment descriptors, graph mining, and graph kernels is highlighted. The application of unsupervised methods, such as Kohonen's self-organizing maps and related approaches, which allow for combining predictions with data analysis and visualization, is also considered. The necessity of applying a wide range of machine learning methods in computational toxicology is underlined.

Predictive models for identifying the binding activity of structurally diverse chemicals to human pregnane X receptor

Toxic chemicals entered into human body would undergo a series of metabolism, transport and excretion, and the key roles played in there processes were metabolizing enzymes, which was regulated by the pregnane X receptor (PXR). However, some chemicals in environment could activate or antagonize human pregnane X receptor, thereby leading to a disturbance of normal physiological systems. In this study, based on a larger number of 2724 structurally diverse chemicals, we developed qualitative classification models by the k-nearest neighbor method. Moreover, the logarithm of 20 and 50% effective concentrations (log EC ₂₀ and log EC ₅₀) was used to establish quantitative structure-activity relationship (QSAR) models. With the classification model, two descriptors were enough to establish acceptable models, with the sensitivity, specificity, and accuracy being larger than 0.7, highlighting a high classification performance of the models. With two QSAR models, the statistics parameters with the correlation coefficient (R ²) of 0.702-0.749 and the cross-validation and external validation coefficient (Q ²) of 0.643-0.712, this indicated that the models complied with the criteria proposed in previous studies, i.e., R ² > 0.6, Q ² > 0.5. The small root mean square error (RMSE) of 0.254-0.414 and the good consistency between observed and predicted values proved satisfactory goodness of fit, robustness, and predictive ability of the developed QSAR models. Additionally, the applicability domains were characterized by the Euclidean distance-based approach and Williams plot, and results indicated that the current models had a wide applicability domain, which especially included a few classes of environmental contaminant, those that were not included in the previous models.

Establishment of In Silico Prediction Models for CYP3A4 and CYP2B6 Induction in Human Hepatocytes by Multiple Regression Analysis Using Azole Compounds

Drug-drug interactions (DDIs) via cytochrome P450 (P450) induction are one clinical problem leading to increased risk of adverse effects and the need for dosage adjustments and additional therapeutic monitoring. In silico models for predicting P450 induction are useful for avoiding DDI risk. In this study, we have established regression models for CYP3A4 and CYP2B6 induction in human hepatocytes using several physicochemical parameters for a set of azole compounds with different P450 induction as characteristics as model compounds. To obtain a well-correlated regression model, the compounds for CYP3A4 or CYP2B6 induction were independently selected from the tested azole compounds using principal component analysis with fold-induction data. Both of the multiple linear regression models obtained for CYP3A4 and CYP2B6 induction are represented by different sets of physicochemical parameters. The adjusted coefficients of determination for these models were of 0.8 and 0.9, respectively. The fold-induction of the validation compounds, another set of 12 azole-containing compounds, were predicted within twofold limits for both CYP3A4 and CYP2B6. The concordance for the prediction of CYP3A4 induction was 87% with another validation set, 23 marketed drugs. However, the prediction of CYP2B6 induction tended to be overestimated for these marketed drugs. The regression models show that lipophilicity mostly contributes to CYP3A4 induction, whereas not only the lipophilicity but also the molecular polarity is important for CYP2B6 induction. Our regression models, especially that for CYP3A4 induction, might provide useful methods to avoid potent CYP3A4 or CYP2B6 inducers during the lead optimization stage without performing induction assays in human hepatocytes.

Screening Ingredients from Herbs against Pregnane X Receptor in the Study of Inductive Herb-Drug Interactions: Combining Pharmacophore and Docking-Based Rank Aggregation

The issue of herb-drug interactions has been widely reported. Herbal ingredients can activate nuclear receptors and further induce the gene expression alteration of drug-metabolizing enzyme and/or transporter. Therefore, the herb-drug interaction will happen when the herbs and drugs are coadministered. This kind of interaction is called inductive herb-drug interactions. Pregnane X Receptor (PXR) and drug-metabolizing target genes are involved in most of inductive herb-drug interactions. To predict this kind of herb-drug interaction, the protocol could be simplified to only screen agonists of PXR from herbs because the relations of drugs with their metabolizing enzymes are well studied. Here, a combinational in silico strategy of pharmacophore modelling and docking-based rank aggregation (DRA) was employed to identify PXR's agonists. Firstly, 305 ingredients were screened out from 820 ingredients as candidate agonists of PXR with our pharmacophore model. Secondly, DRA was used to rerank the result of pharmacophore filtering. To validate our prediction, a curated herb-drug interaction database was built, which recorded 380 herb-drug interactions. Finally, among the top 10 herb ingredients from the ranking list, 6 ingredients were reported to involve in herb-drug interactions. The accuracy of our method is higher than other traditional methods. The strategy could be extended to studies on other inductive herb-drug interactions.

In silico methods for predicting drug-drug interactions with cytochrome P-450s, transporters and beyond

Drug-drug interactions (DDIs) are associated with severe adverse effects that may lead to the patient requiring alternative therapeutics and could ultimately lead to drug withdrawal from the market if they are severe. To prevent the occurrence of DDI in the clinic, experimental systems to evaluate drug interaction have been integrated into the various stages of the drug discovery and development process. A large body of knowledge about DDI has also accumulated through these studies and pharmacovigillence systems. Much of this work to date has focused on the drug metabolizing enzymes such as cytochrome P-450s as well as drug transporters, ion channels and occasionally other proteins. This combined knowledge provides a foundation for a hypothesis-driven in silico approach, using either cheminformatics or physiologically based pharmacokinetics (PK) modeling methods to assess DDI potential. Here we review recent advances in these approaches with emphasis on hypothesis-driven mechanistic models for important protein targets involved in PK-based DDI. Recent efforts with other informatics approaches to detect DDI are highlighted. Besides DDI, we also briefly introduce drug interactions with other substances, such as Traditional Chinese Medicines to illustrate how in silico modeling can be useful in this domain. We also summarize valuable data sources and web-based tools that are available for DDI prediction. We finally explore the challenges we see faced by in silico approaches for predicting DDI and propose future directions to make these computational models more reliable, accurate, and publically accessible.

Open Source Bayesian Models. 1. Application to ADME/Tox and Drug Discovery Datasets

On the order of hundreds of absorption, distribution, metabolism, excretion, and toxicity (ADME/Tox) models have been described in the literature in the past decade which are more often than not inaccessible to anyone but their authors. Public accessibility is also an issue with computational models for bioactivity, and the ability to share such models still remains a major challenge limiting drug discovery. We describe the creation of a reference implementation of a Bayesian model-building software module, which we have released as an open source component that is now included in the Chemistry Development Kit (CDK) project, as well as implemented in the CDD Vault and in several mobile apps. We use this implementation to build an array of Bayesian models for ADME/Tox, in vitro and in vivo bioactivity, and other physicochemical properties. We show that these models possess cross-validation receiver operator curve values comparable to those generated previously in prior publications using alternative tools. We have now described how the implementation of Bayesian models with FCFP6 descriptors generated in the CDD Vault enables the rapid production of robust machine learning models from public data or the user’s own datasets. The current study sets the stage for generating models in proprietary software (such as CDD) and exporting these models in a format that could be run in open source software using CDK components. This work also demonstrates that we can enable biocomputation across distributed private or public datasets to enhance drug discovery.

Toward predicting drug-induced liver injury: parallel computational approaches to identify multidrug resistance protein 4 and bile salt export pump inhibitors

Drug-induced liver injury (DILI) is an important cause of drug toxicity. Inhibition of multidrug resistance protein 4 (MRP4), in addition to bile salt export pump (BSEP), might be a risk factor for the development of cholestatic DILI. Recently, we demonstrated that inhibition of MRP4, in addition to BSEP, may be a risk factor for the development of cholestatic DILI. Here, we aimed to develop computational models to delineate molecular features underlying MRP4 and BSEP inhibition. Models were developed using 257 BSEP and 86 MRP4 inhibitors and noninhibitors in the training set. Models were externally validated and used to predict the affinity of compounds toward BSEP and MRP4 in the DrugBank database. Compounds with a score above the median fingerprint threshold were considered to have significant inhibitory effects on MRP4 and BSEP. Common feature pharmacophore models were developed for MRP4 and BSEP with LigandScout software using a training set of nine well characterized MRP4 inhibitors and nine potent BSEP inhibitors. Bayesian models for BSEP and MRP4 inhibition/noninhibition were developed with cross-validated receiver operator curve values greater than 0.8 for the test sets, indicating robust models with acceptable false positive and false negative prediction rates. Both MRP4 and BSEP inhibitor pharmacophore models were characterized by hydrophobic and hydrogen-bond acceptor features, albeit in distinct spatial arrangements. Similar molecular features between MRP4 and BSEP inhibitors may partially explain why various drugs have affinity for both transporters. The Bayesian (BSEP, MRP4) and pharmacophore (MRP4, BSEP) models demonstrated significant classification accuracy and predictability.

Thiazide-like diuretic drug metolazone activates human pregnane X receptor to induce cytochrome 3A4 and multidrug-resistance protein 1

Human pregnane X receptor (hPXR) regulates the expression of drug-metabolizing enzyme cytochrome P450 3A4 (CYP3A4) and drug transporters such as multidrug-resistance protein 1 (MDR1). PXR can be modulated by small molecules, including Federal Drug Administration (FDA)-approved drugs, thus altering drug metabolism and causing drug-drug interactions. To determine the role of FDA-approved drugs in PXR-mediated regulation of drug metabolism and clearance, we screened 1481 FDA-approved small-molecule drugs by using a luciferase reporter assay in HEK293T cells and identified the diuretic drug metolazone as an activator of hPXR. Our data showed that metolazone activated hPXR-mediated expression of CYP3A4 and MDR1 in human hepatocytes and intestine cells and increased CYP3A4 promoter activity in various cell lines. Mammalian two-hybrid assays showed that hPXR recruits its co-activator SRC-1 upon metolazone binding in HepG2 cells, explaining the mechanism of hPXR activation. To understand the role of other commonly-used diuretics in hPXR activation and the structure-activity relationship of metolazone, thiazide and non-thiazide diuretics drugs were also tested but only metolazone activates hPXR. To understand the molecular mechanism, docking studies and mutational analysis were carried out and showed that metolazone binds in the ligand-binding pocket and interacts with mostly hydrophobic amino acid residues. This is the first report showing that metolazone activates hPXR. Because activation of hPXR might cause drug-drug interactions, metolazone should be used with caution for drug treatment in patients undergoing combination therapy.

Binary and ternary combinations of anti-HIV protease inhibitors: effect on gene expression and functional activity of CYP3A4 and efflux transporters

BACKGROUND

The purpose of this study is to identify the effect of binary and ternary combinations of anti-HIV protease inhibitors (PIs) on the expression of metabolizing enzyme (CYP3A4) and efflux transporters [multidrug resistance-associated protein 2 (MRP2), P-glycoprotein (P-gp) and breast cancer resistant protein (BCRP)] in a model intestinal cell line (LS-180).

METHODS

LS-180 cells were treated with various combinations of PIs (amprenavir, indinavir, saquinavir and lopinavir), and the mRNA expression levels of metabolizing enzyme and efflux transporters were measured using quantitative reverse transcription polymerase chain reaction. The alteration of gene expression was further correlated to the expression of nuclear hormone receptor PXR. Uptake of fluorescent and radioactive substrates was carried out to study the functional activity of these proteins. Cytotoxicity and adenosine triphosphate (ATP) assays were carried out to measure stress responses.

RESULTS

Binary and ternary combinations of PIs appeared to modulate the expression of CYP3A4, MRP2, P-gp and BCRP in a considerable manner. Unlike the individual PIs, their binary combinations showed much greater induction of metabolizing enzyme and efflux proteins. However, such pronounced induction was not observed in the presence of ternary combinations. The observed trend of altered mRNA expression was found to correlate well with the change in expression levels of PXR. The gene expression was found to correlate with activity assays. Lack of cytotoxicity and ATP activity was observed in the treatment samples, suggesting that these alterations in expression levels were probably not stress responses.

CONCLUSIONS

In the present study, we demonstrated that combinations of drugs can have serious consequences toward the treatment of HIV infection by altering their bioavailability and disposition.

Progress in computational toxicology

INTRODUCTION

Computational methods have been widely applied to toxicology across pharmaceutical, consumer product and environmental fields over the past decade. Progress in computational toxicology is now reviewed.

METHODS

A literature review was performed on computational models for hepatotoxicity (e.g. for drug-induced liver injury (DILI)), cardiotoxicity, renal toxicity and genotoxicity. In addition various publications have been highlighted that use machine learning methods. Several computational toxicology model datasets from past publications were used to compare Bayesian and Support Vector Machine (SVM) learning methods.

RESULTS

The increasing amounts of data for defined toxicology endpoints have enabled machine learning models that have been increasingly used for predictions. It is shown that across many different models Bayesian and SVM perform similarly based on cross validation data.

DISCUSSION

Considerable progress has been made in computational toxicology in a decade in both model development and availability of larger scale or 'big data' models. The future efforts in toxicology data generation will likely provide us with hundreds of thousands of compounds that are readily accessible for machine learning models. These models will cover relevant chemistry space for pharmaceutical, consumer product and environmental applications.

An updated review on drug-induced cholestasis: mechanisms and investigation of physicochemical properties and pharmacokinetic parameters

Drug-induced cholestasis is an important form of acquired liver disease and is associated with significant morbidity and mortality. Bile acids are key signaling molecules, but they can exert toxic responses when they accumulate in hepatocytes. This review focuses on the physiological mechanisms of drug-induced cholestasis associated with altered bile acid homeostasis due to direct (e.g., bile acid transporter inhibition) or indirect (e.g., activation of nuclear receptors, altered function/expression of bile acid transporters) processes. Mechanistic information about the effects of a drug on bile acid homeostasis is important when evaluating the cholestatic potential of a compound, but experimental data often are not available. The relationship between physicochemical properties, pharmacokinetic parameters, and inhibition of the bile salt export pump among 77 cholestatic drugs with different pathophysiological mechanisms of cholestasis (i.e., impaired formation of bile vs. physical obstruction of bile flow) was investigated. The utility of in silico models to obtain mechanistic information about the impact of compounds on bile acid homeostasis to aid in predicting the cholestatic potential of drugs is highlighted.

Induction of P-glycoprotein by antiretroviral drugs in human brain microvessel endothelial cells

The membrane-associated drug transporter P-glycoprotein (P-gp) plays an essential role in drug efflux from the brain. Induction of this protein at the blood-brain barrier (BBB) could further affect the ability of a drug to enter the brain. At present, P-gp induction mediated by antiretroviral drugs at the BBB has not been fully investigated. Since P-gp expression is regulated by ligand-activated nuclear receptors, i.e., human pregnane X receptor (hPXR) and human constitutive androstane receptor (hCAR), these receptors could represent potential pathways involved in P-gp induction by antiretroviral drugs. The aims of this study were (i) to determine whether antiretroviral drugs currently used in HIV pharmacotherapy are ligands for hPXR or hCAR and (ii) to examine P-gp function and expression in human brain microvessel endothelial cells treated with antiretroviral drugs identified as ligands of hPXR and/or hCAR. Luciferase reporter gene assays were performed to examine the activation of hPXR and hCAR by antiretroviral drugs. The hCMEC/D3 cell line, which is known to display several morphological and biochemical properties of the BBB in humans, was used to examine P-gp induction following 72 h of exposure to these agents. Amprenavir, atazanavir, darunavir, efavirenz, ritonavir, and lopinavir were found to activate hPXR, whereas abacavir, efavirenz, and nevirapine were found to activate hCAR. P-gp expression and function were significantly induced in hCMEC/D3 cells treated with these drugs at clinical concentrations in plasma. Together, our data suggest that P-gp induction could occur at the BBB during chronic treatment with antiretroviral drugs identified as ligands of hPXR and/or hCAR.

LASSO-ing Potential Nuclear Receptor Agonists and Antagonists: A New Computational Method for Database Screening

Nuclear receptors (NRs) are important biological macromolecular transcription factors that are implicated in multiple biological pathways and may interact with other xenobiotics that are endocrine disruptors present in the environment. Examples of important NRs include the androgen receptor (AR), estrogen receptors (ER), and the pregnane X receptor (PXR). In this study we have utilized the Ligand Activity by Surface Similarity Order (LASSO) method, a ligand-based virtual screening strategy to derive structural (surface/shape) molecular features used to generate predictive models of biomolecular activity for AR, ER, and PXR. For PXR, twenty-five models were built using between 8 to 128 agonists and tested using 3000, 8000, and 24,000 drug-like decoys including PXR inactive compounds (N=228). Preliminary studies with AR and ER using LASSO suggested the utility of this approach with 2-fold enrichment factors at 20%. We found that models with 64–128 PXR actives provided enrichment factors of 10-fold (10% actives in the top 1% of compounds screened). The LASSO models for AR and ER have been deployed and are freely available online, and they represent a ligand-based prediction method for putative NR activity of compounds in this database.

Identification of novel breast cancer resistance protein (BCRP) inhibitors by virtual screening

Breast cancer resistance protein (BCRP; ABCG2) is an efflux transporter that plays an important role in multidrug resistance to antineoplastic drugs. The identification of drugs as BCRP inhibitors could aid in designing better therapeutic strategies for cancer treatment and will be critical for identifying potential drug-drug interactions. In the present study, we applied ligand-based virtual screening combined with experimental testing for the identification of novel drugs that can possibly interact with BCRP. Bayesian and pharmacophore models generated with known BCRP inhibitors were validated with an external test set. The resulting models were applied to predict new potential drug candidates from a database with more than 2000 FDA-approved drugs. Thirty-three drugs were tested in vitro for their inhibitory effects on BCRP-mediated transport of [(3)H]-mitoxantrone in MCF-7/AdrVp cells. Nineteen drugs were identified with significant inhibitory effect on BCRP transport function. The combined strategy of computational and experimental approaches in this paper has suggested potential drug candidates and thus represents an effective tool for rational identification of modulators of other proteins.

PXR ligand classification model with SFED-weighted WHIM and CoMMA descriptors

Understanding which type of endogenous and exogenous compounds serve as agonists for the nuclear pregnane X receptor (PXR) would be valuable for drug discovery and development, because PXR regulates a large number of genes related to xenobiotic metabolism. Although several models have been proposed to classify human PXR activators and non-activators, models with better predictability are necessary for practical purposes in drug discovery. Grid-weighted holistic invariant molecular (G-WHIM) and comparative molecular moment analysis (G-CoMMA) type 3D descriptors that contain information about the solvation free energy of target molecules were developed. With these descriptors, prediction models built using decision tree (DT)-, support vector machine (SVM)-, and Kohonen neural network (KNN)-based models exhibited better predictability than previously proposed models. Solvation free energy density-weighted G-WHIM and G-CoMMA descriptors reveal new insights into PXR ligand classification, and incorporation with machine learning methods (DT, SVM, KNN) exhibits promising results, especially SVM and KNN. SVM- and KNN-based models exhibit accuracy around 0.90, and DT-based models exhibit accuracy around 0.8 for both the training and test sets.

Development of in silico filters to predict activation of the pregnane X receptor (PXR) by structurally diverse drug-like molecules

The pregnane X receptor (PXR), a member of the nuclear hormone superfamily, regulates the expression of several enzymes and transporters involved in metabolically relevant processes. The significant induction of CYP450 enzymes by PXR, in particular CYP3A4, might significantly alter the metabolism of prescribed drugs. In order to early identify molecules in drug discovery with a potential to activate PXR as antitarget, we developed fast and reliable in silico filters by ligand-based QSAR techniques. Two classification models were established on a diverse dataset of 434 drug-like molecules. A second augmented set allowed focusing on interesting regions in chemical space. These classifiers are based on decision trees combined with a genetic algorithm based variable selection to arrive at predictive models. The classifier for the first dataset on 29 descriptors showed good performance on a test set with a correct classification of both 100% for PXR activators and non-activators plus 87% for activators and 83% for non-activators in an external dataset. The second classifier then correctly predicts 97% activators and 91% non-activators in a test set and 94% for activators and 64% non-activators in an external set of 50 molecules, which still qualifies for application as a filter focusing on PXR activators. Finally a quantitative model for PXR activation for a subset of these molecules was derived using a regression-tree approach combined with GA variable selection. This final model shows a predictive r(2) of 0.774 for the test set and 0.452 for an external set of 33 molecules. Thus, the combination of these filters consistently provide guidelines for lowering PXR activation in novel candidate molecules.

From Molecular Docking to 3D-Quantitative Structure-Activity Relationships (3D-QSAR): Insights into the Binding Mode of 5-Lipoxygenase Inhibitors

Pharmacological intervention with 5-Lipoxygenase (5-LO) is a promising strategy for treatment of inflammatory and allergic ailments, including asthma. With the aim of developing predictive models of 5-LO affinity and gaining insights into the molecular basis of ligand-target interaction, we herein describe QSAR studies of 59 diverse nonredox-competitive 5-LO inhibitors based on the use of molecular shape descriptors and docking experiments. These studies have successfully yielded a predictive model able to explain much of the variance in the activity of the training set compounds while predicting satisfactorily the 5-LO inhibitory activity of an external test set of compounds. The inspection of the selected variables in the QSAR equation unveils the importance of specific interactions which are observed from docking experiments. Collectively, these results may be used to design novel potent and selective nonredox 5-LO inhibitors.

Ligand- and structure-based pregnane X receptor models

The human pregnane X receptor (PXR) is a ligand dependent transcription factor that can be activated by structurally diverse agonists including steroid hormones, bile acids, herbal drugs, and prescription medications. PXR regulates the transcription of several genes involved in xenobiotic detoxification and apoptosis. Activation of PXR has the potential to initiate adverse effects by altering drug pharmacokinetics or perturbing physiological processes. Hence, more reliable prediction of PXR activators would be valuable for pharmaceutical drug discovery to avoid potential toxic effects. Ligand- and protein structure-based computational models for PXR activation have been developed in several studies. There has been limited success with structure-based modeling approaches to predict human PXR activators, which can be attributed to the large and promiscuous site of this protein. Slightly better success has been achieved with ligand-based modeling methods including quantitative structure-activity relationship (QSAR) analysis, pharmacophore modeling and machine learning that use appropriate descriptors to account for the diversity of the ligand classes that bind to PXR. These combined computational approaches using molecular shape information may assist scientists to more confidently identify PXR activators. This chapter reviews the various ligand and structure based methods undertaken to date and their results.

Identification and validation of novel human pregnane X receptor activators among prescribed drugs via ligand-based virtual screening

Human pregnane X receptor (hPXR) plays a key role in regulating metabolism and clearance of endogenous and exogenous substances. Identification of novel hPXR activators among commercial drugs may aid in avoiding drug-drug interactions during coadministration. We applied ligand-based computational approaches for virtual screening of a commonly prescribed drug database (SCUT). Bayesian classification models were generated with a training set comprising 177 compounds using Fingerprints and 117 structural descriptors. A cell-based luciferase reporter assay was used for evaluation of chemical-mediated hPXR activation in HepG2 cells. All compounds were tested at 10 μM concentration with rifampicin and dimethyl sulfoxide as positive and negative controls, respectively. The Bayesian models showed specificity and overall prediction accuracy up to 0.92 and 0.69 for test set compounds. Screening the SCUT database with this model retrieved 105 hits and 17 compounds from the top 25 hits were chosen for in vitro testing. The reporter assay confirmed that nine drugs, i.e., fluticasone, nimodipine, nisoldipine, beclomethasone, finasteride, flunisolide, megestrol, secobarbital, and aminoglutethimide, were previously unidentified hPXR activators. Thus, the present study demonstrates that novel hPXR activators can be efficiently identified among U.S. Food and Drug Administration-approved and commonly prescribed drugs, which should lead to detection and prevention of potential drug-drug interactions.

Identification of clinically used drugs that activate pregnane X receptors

The pregnane X receptor (PXR) binds xenobiotics and regulates the expression of several drug-metabolizing enzymes and transporters. Human PXR (hPXR) activation and CYP3A4 induction can be involved in drug-drug interactions, resulting in reduced efficacy or increased toxicity. However, there are known species-specific differences with regard to PXR activation that should be taken into account when animal PXR data are extrapolated to humans. We profiled 2816 clinically used drugs from the National Institutes of Health Chemical Genomics Center Pharmaceutical Collection for their ability to activate hPXR and rat PXR (rPXR) at the cellular level, induce human CYP3A4 at the cellular level, and bind human PXR at the protein level. From 6 to 11% of drugs were identified as active across the four assays, which included assay-specific and pan-active compounds. The lowest concordance was observed between the hPXR and rPXR assays, and many compounds active in both assays nonetheless demonstrated significant potency differences between species. Analysis based on clustering potency values demonstrated the greatest activity correlation between the hPXR activation and CYP3A4 induction assays. Structure-activity relationship analysis identified chemical scaffolds that were pan-active (e.g., dihydropyridine calcium channel blockers) and others that were uniquely active in individual assays (e.g., steroids and fatty acids). These results provide important information on PXR activation by clinically used drugs, highlight the species specificity of PXR activation by xenobiotics, and provide a means of prioritizing compounds for follow-up studies and optimization efforts.

Regulation of human pregnane X receptor and its target gene cytochrome P450 3A4 by Chinese herbal compounds and a molecular docking study

The pregnane X receptor (PXR) plays a critical role in the regulation of human cytochrome P450 3A4 (CYP3A4) gene. In this study, we investigated the effect of an array of compounds isolated from Chinese herbal medicines on the activity of PXR using a luciferase reporter gene assay in transiently transfected HepG2 and Huh7 cells and on the expression of PXR and CYP3A4 in LS174T cells. Furthermore, molecular docking was performed to investigate the binding modes of herbal compounds with PXR. Praeruptorin A and C, salvianolic acid B, sodium danshensu, protocatechuic aldehyde, cryptotanshinone, emodin, morin, and tanshinone IIA significantly transactivated the CYP3A4 reporter gene construct in either HepG2 or Huh7 cells. The PXR mRNA expression in LS174T cells was significantly induced by physcion, protocatechuic aldehyde, salvianolic acid B, and sodium danshensu. However, epifriedelanol, morin, praeruptorin D, mulberroside A, tanshinone I, and tanshinone IIA significantly down-regulated the expression of PXR mRNA in LS174T cells. All the herbal compounds tested can be readily docked into the ligand-binding cavity of PXR mainly through hydrogen bond and aromatic interactions with Ser247, Gln285, His407, and Arg401. These findings suggest that herbal medicines can significantly regulate PXR and CYP3A4 and this has important implication in herb-drug interactions.

Evaluation of computational docking to identify pregnane X receptor agonists in the ToxCast database

BACKGROUND

The pregnane X receptor (PXR) is a key transcriptional regulator of many genes [e.g., cytochrome P450s (CYP2C9, CYP3A4, CYP2B6), MDR1] involved in xenobiotic metabolism and excretion.

OBJECTIVES

As part of an evaluation of different approaches to predict compound affinity for nuclear hormone receptors, we used the molecular docking program GOLD and a hybrid scoring scheme based on similarity weighted GoldScores to predict potential PXR agonists in the ToxCast database of pesticides and other industrial chemicals. We present some of the limitations of different in vitro systems, as well as docking and ligand-based computational models.

METHODS

Each ToxCast compound was docked into the five published crystallographic structures of human PXR (hPXR), and 15 compounds were selected based on their consensus docking scores for testing. In addition, we used a Bayesian model to classify the ToxCast compounds into PXR agonists and nonagonists. hPXR activation was determined by luciferase-based reporter assays in the HepG2 and DPX-2 human liver cell lines.

RESULTS

We tested 11 compounds, of which 6 were strong agonists and 2 had weak agonist activity. Docking results of additional compounds were compared with data reported in the literature. The prediction sensitivity of PXR agonists in our sample ToxCast data set (n = 28) using docking and the GoldScore was higher than with the hybrid score at 66.7%. The prediction sensitivity for PXR agonists using GoldScore for the entire ToxCast data set (n = 308) compared with data from the NIH (National Institutes of Health) Chemical Genomics Center data was 73.8%.

CONCLUSIONS

Docking and the GoldScore may be useful for prioritizing large data sets prior to in vitro testing with good sensitivity across the sample and entire ToxCast data set for hPXR agonists.

In silico prediction and screening of gamma-secretase inhibitors by molecular descriptors and machine learning methods

Gamma-secretase inhibitors have been explored for the prevention and treatment of Alzheimer's disease (AD). Methods for prediction and screening of gamma-secretase inhibitors are highly desired for facilitating the design of novel therapeutic agents against AD, especially when incomplete knowledge about the mechanism and three-dimensional structure of gamma-secretase. We explored two machine learning methods, support vector machine (SVM) and random forest (RF), to develop models for predicting gamma-secretase inhibitors of diverse structures. Quantitative analysis of the receiver operating characteristic (ROC) curve was performed to further examine and optimize the models. Especially, the Youden index (YI) was initially introduced into the ROC curve of RF so as to obtain an optimal threshold of probability for prediction. The developed models were validated by an external testing set with the prediction accuracies of SVM and RF 96.48 and 98.83% for gamma-secretase inhibitors and 98.18 and 99.27% for noninhibitors, respectively. The different feature selection methods were used to extract the physicochemical features most relevant to gamma-secretase inhibition. To the best of our knowledge, the RF model developed in this work is the first model with a broad applicability domain, based on which the virtual screening of gamma-secretase inhibitors against the ZINC database was performed, resulting in 368 potential hit candidates.

The structural basis of pregnane X receptor binding promiscuity

The steroid and xenobiotic-responsive human pregnane X receptor (PXR) binds a broad range of structurally diverse compounds. The structures of the apo and ligand-bound forms of PXR are very similar, in contrast to most promiscuous proteins that generally adapt their shape to different ligands. We investigated the structural origins of PXR's recognition promiscuity using computational solvent mapping, a technique developed for the identification and characterization of hot spots, i.e., regions of the protein surface that are major contributors to the binding free energy. Results reveal that the smooth and nearly spherical binding site of PXR has a well-defined hot spot structure, with four hot spots located on four different sides of the pocket and a fifth close to its center. Three of these hot spots are already present in the ligand-free protein. The most important hot spot is defined by three structurally and sequentially conserved residues, W299, F288, and Y306. This largely hydrophobic site is not very specific and interacts with all known PXR ligands. Depending on their sizes and shapes, individual PXR ligands extend into two, three, or four more hot spot regions. The large number of potential arrangements within the binding site explains why PXR is able to accommodate a large variety of compounds. All five hot spots include at least one important residue, which is conserved in all mammalian PXRs, suggesting that the hot spot locations have remained largely invariant during mammalian evolution. The same side chains also show a high level of structural conservation across hPXR structures. However, each of the hPXR hot spots also includes residues with moveable side chains, further increasing the size variation in ligands that PXR can bind. Results also suggest a unique signal transduction mechanism between the PXR homodimerization interface and its coactivator binding site.

Challenges predicting ligand-receptor interactions of promiscuous proteins: the nuclear receptor PXR

Transcriptional regulation of some genes involved in xenobiotic detoxification and apoptosis is performed via the human pregnane X receptor (PXR) which in turn is activated by structurally diverse agonists including steroid hormones. Activation of PXR has the potential to initiate adverse effects, altering drug pharmacokinetics or perturbing physiological processes. Reliable computational prediction of PXR agonists would be valuable for pharmaceutical and toxicological research. There has been limited success with structure-based modeling approaches to predict human PXR activators. Slightly better success has been achieved with ligand-based modeling methods including quantitative structure-activity relationship (QSAR) analysis, pharmacophore modeling and machine learning. In this study, we present a comprehensive analysis focused on prediction of 115 steroids for ligand binding activity towards human PXR. Six crystal structures were used as templates for docking and ligand-based modeling approaches (two-, three-, four- and five-dimensional analyses). The best success at external prediction was achieved with 5D-QSAR. Bayesian models with FCFP_6 descriptors were validated after leaving a large percentage of the dataset out and using an external test set. Docking of ligands to the PXR structure co-crystallized with hyperforin had the best statistics for this method. Sulfated steroids (which are activators) were consistently predicted as non-activators while, poorly predicted steroids were docked in a reverse mode compared to 5α-androstan-3β-ol. Modeling of human PXR represents a complex challenge by virtue of the large, flexible ligand-binding cavity. This study emphasizes this aspect, illustrating modest success using the largest quantitative data set to date and multiple modeling approaches.

Promiscuous proteins generally bind a large array of diverse ligand structures. This may be facilitated by a very large binding site, multiple binding sites, or a flexible binding site that can adjust to the size of the ligand. These aspects also increase the complexity of predicting whether a molecule will bind or not to such proteins which frequently function as exogenous compound sensors to respond to toxic stress. For example, transporters may prevent absorption of some molecules, and enzymes may convert them to more readily excretable compounds (or alternatively activate them prior to further clearance by other detoxification enzymes). Nuclear hormone receptors may respond to ligands and then affect downstream gene expression to upregulate both enzymes and transporters to increase the clearance for the same or different molecules. We have assessed the ability of many different ligand-based and structure-based computational approaches to model and predict the activation of human PXR by steroidal compounds. We find the most effective computational approach to identify potential steroidal PXR agonists which are clinically relevant due to their widespread use in clinical medicine and the presence of mimics in the environment.

Polymer-drug interactions in tyrosine-derived triblock copolymer nanospheres: a computational modeling approach

A combination of molecular dynamics (MD) simulations and docking calculations was employed to model and predict polymer-drug interactions in self-assembled nanoparticles consisting of ABA-type triblock copolymers, where A-blocks are poly(ethylene glycol) units and B-blocks are low molecular weight tyrosine-derived polyarylates. This new computational approach was tested on three representative model compounds: nutraceutical curcumin, anticancer drug paclitaxel and prehormone vitamin D3. Based on this methodology, the calculated binding energies of polymer-drug complexes can be correlated with maximum drug loading determined experimentally. Furthermore, the modeling results provide an enhanced understanding of polymer-drug interactions, revealing subtle structural features that can significantly affect the effectiveness of drug loading (as demonstrated for a fourth tested compound, anticancer drug camptothecin). The present study suggests that computational calculations of polymer-drug pairs hold the potential of becoming a powerful prescreening tool in the process of discovery, development and optimization of new drug delivery systems, reducing both the time and the cost of the process.

Elucidating the 'Jekyll and Hyde' nature of PXR: the case for discovering antagonists or allosteric antagonists

The pregnane X receptor belongs to the nuclear hormone receptor superfamily and is involved in the transcriptional control of numerous genes. It was originally thought that it was a xenobiotic sensor controlling detoxification pathways. Recent studies have shown an increasingly important role in inflammation and cancer, supporting its function in abrogating tissue damage. PXR orthologs and PXR-like pathways have been identified in several non-mammalian species which corroborate a conserved role for PXR in cellular detoxification. In summary, PXR has a multiplicity of roles in vivo and is being revealed as behaving like a "Jekyll and Hyde" nuclear hormone receptor. The importance of this review is to elucidate the need for discovery of antagonists of PXR to further probe its biology and therapeutic applications. Although several PXR agonists are already reported, virtually nothing is known about PXR antagonists. Here, we propose the development of PXR antagonists through chemical, genetic and molecular modeling approaches. Based on this review it will be clear that antagonists of PXR and PXR-like pathways will have widespread utility in PXR biology and therapeutics.