Characterization and validation of an in silico toxicology model to predict the mutagenic potential of drug impurities*

Mutagenic and genotoxic in silico QSAR prediction of dimer impurity of gliflozins; canagliflozin, dapaglifozin, and emphagliflozin and in vitro evaluation by Ames and micronucleus test

Canagliflozin, Dapagliflozin, and Empagliflozin, glucagon-like peptide-1 receptor agonists, are indicated for managing type II diabetes. Although the genotoxicity profiles of these drugs are well-explored, limited information exists regarding the genotoxic potential of their impurities. In this investigation, the dimer impurities of Canagliflozin, Dapagliflozin, and Empagliflozin underwent both in silico and in vitro assessments for mutagenic potential. Tester strains of Salmonella typhimurium and Escherichia coli were subjected to the Ames test, utilizing concentrations of up to 1 µg per plate, with and without the presence of metabolic activation. Evaluation of micronucleus induction in TK6 cells was conducted through a micronucleus test, exploring concentrations up to 500 µg/mL, with or without the presence of exogenous metabolic activation. Under the specific test conditions, the dimer impurities of Canagliflozin, Dapagliflozin, and Empagliflozin showed no evidence of mutagenicity or clastrogenicity, establishing their in vitro classification as nonmutagenic. These findings align with negative in silico predictions from quantitative structure-activity relationship (QSAR) analyses for mutagenicity and genotoxicity of the dimer impurities. Collectively, these studies contribute clinically relevant safety information by confirming that the dimer impurities of Canagliflozin, Dapagliflozin, and Empagliflozin are nonmutagenic and nongenotoxic, emphasizing the consistency between in silico and in vitro data.

Revisiting the bacterial mutagenicity assays: Report by a workgroup of the International Workshops on Genotoxicity Testing (IWGT)

The International Workshop on Genotoxicity Testing (IWGT) meets every four years to obtain consensus on unresolved issues associated with genotoxicity testing. At the 2017 IWGT meeting in Tokyo, four sub-groups addressed issues associated with the Organization for Economic Cooperation and Development (OECD) Test Guideline TG471, which describes the use of bacterial reverse-mutation tests. The strains sub-group analyzed test data from >10,000 chemicals, tested additional chemicals, and concluded that some strains listed in TG471 are unnecessary because they detected fewer mutagens than other strains that the guideline describes as equivalent. Thus, they concluded that a smaller panel of strains would suffice to detect most mutagens. The laboratory proficiency sub-group recommended (a) establishing strain cell banks, (b) developing bacterial growth protocols that optimize assay sensitivity, and (c) testing "proficiency compounds" to gain assay experience and establish historical positive and control databases. The sub-group on criteria for assay evaluation recommended that laboratories (a) track positive and negative control data; (b) develop acceptability criteria for positive and negative controls; (c) optimize dose-spacing and the number of analyzable doses when there is evidence of toxicity; (d) use a combination of three criteria to evaluate results: a dose-related increase in revertants, a clear increase in revertants in at least one dose relative to the concurrent negative control, and at least one dose that produced an increase in revertants above control limits established by the laboratory from historical negative controls; and (e) establish experimental designs to resolve unclear results. The in silico sub-group summarized in silico utility as a tool in genotoxicity assessment but made no specific recommendations for TG471. Thus, the workgroup identified issues that could be addressed if TG471 is revised. The companion papers (a) provide evidence-based approaches, (b) recommend priorities, and (c) give examples of clearly defined terms to support revision of TG471.

Transitioning to composite bacterial mutagenicity models in ICH M7 (Q)SAR analyses

The International Council on Harmonisation (ICH) M7(R1) guideline describes the use of complementary (quantitative) structure-activity relationship ((Q)SAR) models to assess the mutagenic potential of drug impurities in new and generic drugs. Historically, the CASE Ultra and Leadscope software platforms used two different statistical-based models to predict mutations at G-C (guanine-cytosine) and A-T (adenine-thymine) sites, to comprehensively assess bacterial mutagenesis. In the present study, composite bacterial mutagenicity models covering multiple mutation types were developed. These new models contain more than double the number of chemicals (n = 9,254 and n = 13,514) than the corresponding non-composite models and show better toxicophore coverage. Additionally, the use of a single composite bacterial mutagenicity model simplifies impurity analysis in an ICH M7 (Q)SAR workflow by reducing the number of model outputs requiring review. An external validation set of 388 drug impurities representing proprietary pharmaceutical chemical space showed performance statistics ranging from of 66-82% in sensitivity, 91-95% in negative predictivity and 96% in coverage. This effort represents a major enhancement to these (Q)SAR models and their use under ICH M7(R1), leading to improved patient safety through greater predictive accuracy, applicability, and efficiency when assessing the bacterial mutagenic potential of drug impurities.

Computational prediction of toxicity of small organic molecules: state-of-the-art

The field of computational prediction of various toxicity end-points has evolved over last two decades significantly. Availability of newer modelling techniques, powerful computational resources and good-quality data have made it possible to generate reliable predictions for new chemical entities, impurities, chemicals, natural products and a lot of other substances. The field is still undergoing metamorphosis to take into account molecular complexities underlying toxicity end-points such as teratogenicity, mutagenicity, carcinogenicity, etc. Expansion of the applicability domain of these predictive models into areas other than life sciences, such as environmental and materials sciences have received a great deal of attention from all walks of life, fuelling further development and growth of the field. The present chapter discusses the state-of-the-art computational prediction of toxicity end-points of small organic molecules to balance the trade-off between the molecular complexity and the quality of such predictions, without compromising their immense utility in many fields.

In Silico Approaches in Predictive Genetic Toxicology

Genetic toxicology testing is a weight-of-evidence approach to identify and characterize chemical substances that can cause genetic modifications in somatic and/or germ cells. Prediction of genetic toxicology using computational tools is gaining more attention and preferred by regulatory authorities as an alternate safety assessment for in vivo or in vitro approaches. Due to the cost and time associated with experimental genetic toxicity tests, it is essential to develop more robust in silico methods to predict chemical genetic toxicity. A number of in silico genotoxicity predictive tools/models are developed based on the experimental data gathered over the years. These in silico tools are divided into statistical quantitative structure-activity relationships (QSAR)-based approaches and expert-based systems. This chapter covers the state of the art in silico toxicology approaches and standardized protocols, essential for conducting genetic toxicity predictions of chemicals. This chapter also highlights various parameters for the validation of the prediction results obtained from QSAR models.

Computational approaches to chemical hazard assessment

Computational prediction of toxicity has reached new heights as a result of decades of growth in the magnitude and diversity of biological data. Public packages for statistics and machine learning make model creation faster. New theory in machine learning and cheminformatics enables integration of chemical structure, toxicogenomics, simulated and physical data in the prediction of chemical health hazards, and other toxicological information. Our earlier publications have characterized a toxicological dataset of unprecedented scale resulting from the European REACH legislation (Registration Evaluation Authorisation and Restriction of Chemicals). These publications dove into potential use cases for regulatory data and some models for exploiting this data. This article analyzes the options for the identification and categorization of chemicals, moves on to the derivation of descriptive features for chemicals, discusses different kinds of targets modeled in computational toxicology, and ends with a high-level perspective of the algorithms used to create computational toxicology models.

Applications of Chemoinformatics in Predictive Toxicology for Regulatory Purposes, Especially in the Context of the EU REACH Legislation

Chemoinformatics methodologies such as QSAR/QSPR have been used for decades in drug discovery projects, especially for the finding of new compounds with therapeutic properties and the optimization of ADME properties on chemical series. The application of computational techniques in predictive toxicology is much more recent, and they are experiencing an increasingly interest because of the new legal requirements imposed by national and international regulations. In the pharmaceutical field, the US Food and Drug Administration (FDA) support the use of predictive models for regulatory decision-making when assessing the genotoxic and carcinogenic potential of drug impurities. In Europe, the REACH legislation promotes the use of QSAR in order to reduce the huge amount of animal testing needed to demonstrate the safety of new chemical entities subjected to registration, provided they meet specific conditions to ensure their quality and predictive power. In this review, the authors summarize the state of art of in silico methods for regulatory purposes, with especial emphasis on QSAR models.

Prediction models for drug-induced hepatotoxicity by using weighted molecular fingerprints

Background

Drug-induced liver injury (DILI) is a critical issue in drug development because DILI causes failures in clinical trials and the withdrawal of approved drugs from the market. There have been many attempts to predict the risk of DILI based on in vivo and in silico identification of hepatotoxic compounds. In the current study, we propose the in silico prediction model predicting DILI using weighted molecular fingerprints.

Results

In this study, we used 881 bits of molecular fingerprint and used as features describing presence or absence of each substructure of compounds. Then, the Bayesian probability of each substructure was calculated and labeled (positive or negative for DILI), and a weighted fingerprint was determined from the ratio of DILI-positive to DILI-negative probability values. Using weighted fingerprint features, the prediction models were trained and evaluated with the Random Forest (RF) and Support Vector Machine (SVM) algorithms. The constructed models yielded accuracies of 73.8% and 72.6%, AUCs of 0.791 and 0.768 in cross-validation. In independent tests, models achieved accuracies of 60.1% and 61.1% for RF and SVM, respectively. The results validated that weighted features helped increase overall performance of prediction models. The constructed models were further applied to the prediction of natural compounds in herbs to identify DILI potential, and 13,996 unique herbal compounds were predicted as DILI-positive with the SVM model.

Conclusions

The prediction models with weighted features increased the performance compared to non-weighted models. Moreover, we predicted the DILI potential of herbs with the best performed model, and the prediction results suggest that many herbal compounds could have potential to be DILI. We can thus infer that taking natural products without detailed references about the relevant pathways may be dangerous. Considering the frequency of use of compounds in natural herbs and their increased application in drug development, DILI labeling would be very important.

Electronic supplementary material

The online version of this article (doi:10.1186/s12859-017-1638-4) contains supplementary material, which is available to authorized users.

Consequences of New Approach to Chemical Stability Tests to Active Pharmaceutical Ingredients

There is a great need of broaden look on stability tests of active pharmaceutical ingredients (APIs) in comparison with current requirements contained in pharmacopeia. By usage of many modern analytical methods the conception of monitoring the changes of APIs during initial stage of their exposure to harmful factors has been developed. New knowledge must be acquired in terms of identification of each degradation products, especially volatile ones. Further research as toxicology prediction during in silico studies of determined and identified degradation products is necessary. In silico methods are known as computational toxicology or computer-assisted technologies which are used for predicting toxicology of pharmaceutical substances such as impurities or degradation products. This is a specialized software and databases intended to calculate probability of genotoxicity or mutagenicity of these substances through a chemical structure-based screening process and algorithm specific to a given software program. Applying of new analytical approach is proposed as the usage of PAT tools, XRD, HS-SPME GC-MS/MS, LC-MS/MS for stability testing. Described improvements should be taken into account in case of each drug existing already in the market as well as being implemented as new one.

Determination of the impurities in drug products containing montelukast and in silico/in vitro genotoxicological assessments of sulfoxide impurity

Impurities affecting safety, efficacy, and quality of pharmaceuticals are of increasing concern for regulatory agencies and pharmaceutical industries, since genotoxic impurities are understood to play important role in carcinogenesis. The study aimed to analyse impurities of montelukast chronically used in asthma theraphy and perform genotoxicological assessment considering regulatory approaches. Impurities (sulfoxide, cis-isomer, Michael adducts-I&II, methylketone, methylstyrene) were quantified using RP-HPLC analysis on commercial products available in Turkish market. For sulfoxide impurity, having no toxicity data and found to be above the qualification limit, in silico mutagenicity prediction analysis, miniaturized bacterial gene mutation test, mitotic index determination and in vitro chromosomal aberration test w/wo metabolic activation system were conducted. In the analysis of different batches of 20 commercial drug products from 11 companies, only sulfoxide impurity exceeded qualification limit in pediatric tablets from 2 companies and in adult tablets from 7 companies. Leadscope and ToxTree programs predicted sulfoxide impurity as nonmutagenic. It was also found to be nonmutagenic in Ames MPF Penta I assay. Sulfoxide impurity was dose-dependent cytotoxic in human peripheral lymphocytes, however, it was found to be nongenotoxic. It was concluded that sulfoxide impurity should be considered as nonmutagenic and can be classified as ordinary impurity according to guidelines.

An evaluation of in-house and off-the-shelf in silico models: implications on guidance for mutagenicity assessment

The evaluation of impurities for genotoxicity using in silico models are commonplace and have become accepted by regulatory agencies. Recently, the ICH M7 Step 4 guidance was published and requires two complementary models for genotoxicity assessments. Over the last ten years, many companies have developed their own internal genotoxicity models built using both public and in-house chemical structures and bacterial mutagenicity data. However, the proprietary nature of internal structures prevents sharing of data and the full OECD compliance of such models. This analysis investigated whether using in-house internal compounds for training models is needed and substantially impacts the results of in silico genotoxicity assessments, or whether using commercial-off-the-shelf (COTS) packages such as Derek Nexus or Leadscope provide adequate performance. We demonstrated that supplementation of COTS packages with a Support Vector Machine (SVM) QSAR model trained on combined in-house and public data does, in fact, improve coverage and accuracy, and reduces the number of compounds needing experimental assessment, i.e., the liability load. This result indicates that there is added value in models trained on both internal and public structures and incorporating such models as part of a consensus approach improves the overall evaluation. Lastly, we optimized an in silico consensus decision-making approach utilizing two COTS models and an internal (SVM) model to minimize false negatives.

Computational toxicology, friend or foe?

Computational approaches offer the attraction of being both fast and cheap to run being able to process thousands of chemical structures in a few minutes. As with all new technology, there is a tendency for these approaches to be hyped up and claims of reliability and performance may be exaggerated. So just how good are these computational methods?

Nonclinical Safety and Toxicology

Nonclinical safety pharmacology and toxicology testing of drug candidates assess the potential adverse effects caused by the drug in relation to its intended use in humans. Hazards related to a drug have to be identified and the potential risks at the intended exposure have to be evaluated in comparison to the potential benefit of the drug. Preclinical safety is thus an integral part of drug discovery and drug development. It still causes significant attrition during drug development.Therefore, there is a need for smart selection of drug candidates in drug discovery including screening of important safety endpoints. In the recent years,there was significant progress in computational and in vitro technology allowing in silico assessment as well as high-throughput screening of some endpoints at very early stages of discovery. Despite all this progress, in vivo evaluation of drug candidates is still an important part to safety testing. The chapter provides an overview on the most important areas of nonclinical safety screening during drug discovery of small molecules.

Evidence-based evaluation of potential benefits and safety of beta-alanine supplementation for military personnel

This Department of Defense-sponsored evidence-based review evaluates the safety and putative outcomes of enhancement of athletic performance or improved recovery from exhaustion in studies involving beta-alanine alone or in combination with other ingredients. Beta-alanine intervention studies and review articles were collected from 13 databases, and safety information was collected from adverse event reporting portals. Due to the lack of systematic studies involving military populations, all the available literature was assessed with a subgroup analysis of studies on athletes to determine if beta-alanine would be suitable for the military. Available literature provided only limited evidence concerning the benefits of beta-alanine use, and a majority of the studies were not designed to address safety. Overall, the strength of evidence in terms of the potential for risk of bias in the quality of the available literature, consistency, directness, and precision did not support the use of beta-alanine by military personnel. The strength of evidence for a causal relation between beta-alanine and paresthesia was moderate.

Comprehension of drug toxicity: software and databases

Quantitative structure-property/activity relationships (QSPRs/QSARs) are a tool (in silico) to rapidly predict various endpoints in general, and drug toxicity in particular. However, this dynamic evolution of experimental data (expansion of existing experimental data on drugs toxicity) leads to the problem of critical estimation of the data. The carcinogenicity, mutagenicity, liver effects and cardiac toxicity should be evaluated as the most important aspects of the drug toxicity. The toxicity is a multidimensional phenomenon. It is apparent that the main reasons for the increase in applications of in silico prediction of toxicity include the following: (i) the need to reduce animal testing; (ii) computational models provide reliable toxicity prediction; (iii) development of legislation that is related to use of new substances; (iv) filling data gaps; (v) reduction of cost and time; (vi) designing of new compounds; (vii) advancement of understanding of biology and chemistry. This mini-review provides analysis of existing databases and software which are necessary for use of robust computational assessments and robust prediction of potential drug toxicities by means of in silico methods.

A novel QSAR model of Salmonella mutagenicity and its application in the safety assessment of drug impurities

As indicated in ICH M7 draft guidance, in silico predictive tools including statistically-based QSARs and expert analysis may be used as a computational assessment for bacterial mutagenicity for the qualification of impurities in pharmaceuticals. To address this need, we developed and validated a QSAR model to predict Salmonella t. mutagenicity (Ames assay outcome) of pharmaceutical impurities using Prous Institute's Symmetry(SM), a new in silico solution for drug discovery and toxicity screening, and the Mold2 molecular descriptor package (FDA/NCTR). Data was sourced from public benchmark databases with known Ames assay mutagenicity outcomes for 7300 chemicals (57% mutagens). Of these data, 90% was used to train the model and the remaining 10% was set aside as a holdout set for validation. The model's applicability to drug impurities was tested using a FDA/CDER database of 951 structures, of which 94% were found within the model's applicability domain. The predictive performance of the model is acceptable for supporting regulatory decision-making with 84±1% sensitivity, 81±1% specificity, 83±1% concordance and 79±1% negative predictivity based on internal cross-validation, while the holdout dataset yielded 83% sensitivity, 77% specificity, 80% concordance and 78% negative predictivity. Given the importance of having confidence in negative predictions, an additional external validation of the model was also carried out, using marketed drugs known to be Ames-negative, and obtained 98% coverage and 81% specificity. Additionally, Ames mutagenicity data from FDA/CFSAN was used to create another data set of 1535 chemicals for external validation of the model, yielding 98% coverage, 73% sensitivity, 86% specificity, 81% concordance and 84% negative predictivity.

In silico methods to predict drug toxicity

This review describes in silico methods to characterize the toxicity of pharmaceuticals, including tools which predict toxicity endpoints such as genotoxicity or organ-specific models, tools addressing ADME processes, and methods focusing on protein-ligand docking binding. These in silico tools are rapidly evolving. Nowadays, the interest has shifted from classical studies to support toxicity screening of candidates, toward the use of in silico methods to support the expert. These methods, previously considered useful only to provide a rough, initial estimation, currently have attracted interest as they can assist the expert in investigating toxic potential. They provide the expert with safety perspectives and insights within a weight-of-evidence strategy. This represents a shift of the general philosophy of in silico methodology, and it is likely to further evolve especially exploiting links with system biology.

Predictive computational toxicology to support drug safety assessment

Use of predictive technologies is an important aspect of many efforts in today's research, development, and regulatory landscapes. Computational methods as predictive tools for supporting drug safety assessments is of widespread interest as the field of in silico assessments rapidly changes with emerging technologies and the large amount of existing data available for modeling. There are challenges associated with application of in silico analyses for drug toxicity predictions and need for strategies and harmonization to enable an acceptable in silico evaluation for prediction of specific toxicity assay outcomes. This chapter will provide an overview focused on computational tools using structure-activity relationships and will highlight initiatives for use of computational assessments and realistic applications for predictive modeling in evaluating potential toxicities of drug-related molecules.