Making the Case for Quantum Mechanics in Predictive Toxicology─Nearly 100 Years Too Late?

The use of quantum mechanics (QM) has long been the norm to study covalent-binding phenomena in chemistry and biochemistry. The pharmaceutical industry leverages QM models explicitly in covalent drug discovery and implicitly to characterize short-range interactions in noncovalent binding. Predictive toxicology has resisted widespread adoption of QM, including in the pharmaceutical industry, despite its obvious relevance to the metabolic processes in the upstream of adverse outcome pathways and advances in both QM methods and computational resources, which support fit-for-purpose applications in reasonable timeframes. Here, we make the case for embracing QM as an indispensable part of a toxicologist's toolkit. We argue that QM provides the necessary orthogonality to alert-based expert systems and traditional QSARs, consistent with calls for animal-free integrated testing strategies for safety assessments of commercial chemicals. We outline existing roadblocks to this transition, including the need to train model developers in QM and the shift toward service-based toxicity models that utilize high-performance computing clusters. Lastly, we describe recent examples of successful implementations of QM in hazard assessments and propose how in silico toxicology can be further advanced by integrating QM with artificial intelligence.

Quantum-Mechanical Approach to Predicting the Carcinogenic Potency of N-Nitroso Impurities in Pharmaceuticals.

N-Nitroso contaminants in medicinal products are of concern due to their high carcinogenic potency; however, not all these compounds are created equal, and some are relatively benign chemicals. Understanding the structure-activity relationships (SARs) that drive hazards in one molecule versus another is key to both protecting human health and alleviating costly and sometimes inaccurate animal testing. Here, we report on an extension of the CADRE (computer-aided discovery and REdesign) platform, which is used broadly by the pharmaceutical and personal care industries to assess environmental and human health endpoints, to predict the carcinogenic potency of N-nitroso compounds. The model distinguishes compounds in three potency categories with 77% accuracy in external testing, which surpasses the reproducibility of rodent cancer bioassays and constraints imposed by limited (high-quality) data. The robustness of predictions for more complex pharmaceuticals is maximized by capturing key SARs using quantum mechanics, that is, by hinging the model on the underlying chemistry versus chemicals in the training set. To this end, the present approach can be leveraged in a quantitative hazard assessment and to offer qualitative guidance using electronic structure comparisons between well-studied analogues and unknown contaminants.

Computer-Aided Discovery and Redesign for Respiratory Sensitization: A Tiered Mechanistic Model to Deliver Robust Performance Across a Diverse Chemical Space.

Asthma is among the most common occupational diseases with considerable public health and economic costs. Chemicals that induce hypersensitivity in the airways can cause respiratory distress and comorbidities with respiratory infections such as COVID. Robust predictive models for this end point are still elusive due to the lack of an experimental benchmark and the over-reliance of existing in silico tools on structural alerts and structural (vs chemical) similarities. The Computer-Aided Discovery and REdesign (CADRE) platform is a proven strategy for providing robust computational predictions for hazard end points using a tiered hybrid system of expert rules, molecular simulations, and quantum mechanics calculations. The recently developed CADRE model for respiratory sensitization is based on a highly curated data set of structurally diverse chemicals with high-fidelity biological data. The model evaluates absorption kinetics in lung mucosa using Monte Carlo simulations, assigns reactive centers in a molecule and possible biotransformations via expert rules, and determines subsequent reactivity with cell proteins via quantum-mechanics calculations using a multi-tiered regression. The model affords an accuracy above 0.90, with a series of external validations based on literature data in the range of 0.88-0.95. The model is applicable to all low-molecular-weight organics and can inform not only chemical substitution but also chemical redesign to advance development of safer alternatives.

O data, where art thou? Revolutionizing data sharing to advance our sustainability goals through smart chemical innovation.

Antiquated and inefficient data-sharing practices represent one of the key obstacles to advancing sustainability goals through green chemistry. To this end, we need to robustly link data on chemical impacts with new chemical design strategies, which requires the development of next-generation data-sharing platforms to harmonize both data and efforts. These decentralized and interactive programs should be structured as live ecosystems for data generation and exchange, inviting conversations about the reliability and relevance of information used to make decisions regarding chemical performance and safety.

An Evaluation of the Occupational Health Hazards of Peptide Couplers.

Peptide couplers (also known as amide bond-forming reagents or coupling reagents) are broadly used in organic chemical syntheses, especially in the pharmaceutical industry. Yet, occupational health hazards associated with this chemical class are largely unexplored, which is disconcerting given the intrinsic reactivity of these compounds. Several case studies involving occupational exposures reported adverse respiratory and dermal health effects, providing initial evidence of chemical sensitization. To address the paucity of toxicological data, a pharmaceutical cross-industry task force was formed to evaluate and assess the potential of these compounds to cause eye and dermal irritation as well as corrosivity and dermal sensitization. The goal of our work was to inform health and safety professionals as well as pharmaceutical and organic chemists of the occupational health hazards associated with this chemical class. To that end, 25 of the most commonly used peptide couplers and five hydrolysis products were selected for in vivo, in vitro, and in silico testing. Our findings confirmed that dermal sensitization is a concern for this chemical class with 21/25 peptide couplers testing positive for dermal sensitization and 15 of these being strong/extreme sensitizers. We also found that dermal corrosion and irritation (8/25) as well as eye irritation (9/25) were health hazards associated with peptide couplers and their hydrolysis products (4/5 were dermal irritants or corrosive and 4/5 were eye irritants). Resulting outcomes were synthesized to inform decision making in peptide coupler selection and enable data-driven hazard communication to workers. The latter includes harmonized hazard classifications, appropriate handling recommendations, and accurate safety data sheets, which support the industrial hygiene hierarchy of control strategies and risk assessment. Our study demonstrates the merits of an integrated, in vivo -in silico analysis, applied here to the skin sensitization endpoint using the Computer-Aided Discovery and REdesign (CADRE) and Derek Nexus programs. We show that experimental data can improve predictive models by filling existing data gaps while, concurrently, providing computational insights into key initiating events and elucidating the chemical structural features contributing to adverse health effects. This interactive, interdisciplinary approach is consistent with Green Chemistry principles that seek to improve the selection and design of less hazardous reagents in industrial processes and applications.

Structure-to-process design framework for developing safer pesticides.

Rational design of pesticides with tunable degradation properties and minimal ecotoxicity is among the grand challenges of green chemistry. While computational approaches have gained traction in predictive toxicology, current methods lack the necessary multifaceted approach and design-vectoring tools needed for system-based chemical development. Here, we report a tiered computational framework, which integrates kinetics and thermodynamics of indirect photodegradation with predictions of ecotoxicity and performance, based on cutoff values in mechanistically derived physicochemical properties and electronic parameters. Extensively validated against experimental data and applied to 700 pesticides on the U.S. Environmental Protection Agency's registry, our simple yet powerful approach can be used to screen existing molecules to identify application-ready candidates with desirable characteristics. By linking structural attributes to process-based outcomes and by quantifying trade-offs in safety, depletion, and performance, our method offers a user-friendly roadmap to rational design of novel pesticides.

Structure-Energetics-Property Relationships Support Computational Design of Photodegradable Pesticides.

Development of high-performing pesticides with tunable degradation properties is vital to increasing the safety and effectiveness of tomorrow's analogs. Chromophoric dissolved organic matter in the excited triple state (3CDOM*) is known to play a key role in the removal of pesticides via indirect photodegradation. However, the potential of these transformations to guide the design of safer chemicals has not yet been fully realized. Here, we report a two-tier computational framework developed to probe and predict both kinetics and thermodynamics of 3CDOM*-pesticide interactions. In the first tier, robust in silico models were constructed by fitting free energies obtained from density functional theory (DFT) calculations to cell potentials and second-order rate constants for the 3CDOM*-pesticide electron transfer. In the second tier, Gibbs free energies and corresponding free energy barriers, determined in solution using the Marcus theory, were applied to develop a quick yet accurate screening approach based on the frontier molecular orbital (FMO) Theory. Being highly mechanistic and spanning ca. 1500 unique 3CDOM*-pesticide interactions, our approach is both robust and broadly applicable. To that end, the outcomes of our computational models were integrated into an easy-to-use decision framework that can guide structure-based design of less persistent pesticide analogs.

Toward a Universal Acute Fish Threshold of Toxicological Concern.

Threshold of toxicological concern (TTC) is a concept that has been around for decades in human health sciences. Ecotoxicology recently adopted a variant of this concept as eco-TTC. Adoption of the concept of TTC considerably reduces the amount of animal testing required for regulatory purposes. We provide an application of a universal TTC for the entirety of acute fish toxicity data (i.e., establishment of an exposure level below which there would be minimal probability of acute fish toxicity for any chemical, without consideration of mechanism of action). We calculated TTC values for a number of subgroups using various approaches. These approaches were evaluated using data from a cohort of 69 999 acute fish toxicological assays. This database was normalized/curated for units, exposure duration, quality assurance/control, and duplicates, which reduced it to 47 694 assays. Data were not normally but log-normally distributed, making geometric means the most appropriate statistical parameter. Thus, we developed descriptive statistics using geometric means with 95, 99, and 99.9% confidence intervals. Various assessment factors (akin to predicted-no-effect concentration derivation) were applied to the geometric means to derive TTCs. Other approaches employed were the calculation of y = 0 intercepts as well as development of 95 and 99.75% cutoffs of cumulative data as well as modular uncertainty scoring tool (MUST) analysis. All of the methodologies derived highly congruent TTCs ranging from to 2 to 8 µg/L except for the 99.75th percentile cutoff of 0.3 µg/L. The data would be most useful in making a binary testing/no testing required decision. For acute fish toxicity, a TTC value of 2 µg/L was most appropriate, based on the 95th percentile of data distribution without any assessment factor. Environ Toxicol Chem 2021;40:1740-1749. © 2021 SETAC.

Quantifying Uncertainty in Ecotoxicological Risk Assessment: MUST, a Modular Uncertainty Scoring Tool.

Whether conducting a risk, hazard, or alternatives assessment, one invariably struggles with the task of reconciling multiple available values of toxicological thresholds into a single outcome. When combining multiple pieces of evidence from many different sources, it is important to consider the role of data uncertainty. Uncertainty is inherent to all scientific data. However, in toxicological assessments, controversies and uncertainties are typically understated; they lack methodological transparency; or they poorly integrate qualitative and quantitative sources of information. Similarly, in model development, data curation is rarely performed with sufficient rigor, particularly when applying big data statistics. To overcome the hurdles of a decision process that must reconcile divergent data, we developed an uncertainty scoring tool that can be trained to reproduce specific decision-making paradigms and ensure consistency in the practitioner's judgment across complex scenarios. While designed to aid with ecotoxicological assessments and predictive model development, the tool's applicability extends to any decision-making process that calls for synthesis of incongruent data. Here, we highlight the development process, as well as demonstrate the method's utility in several prototypical ecotoxicological case studies.

Going All In: A Strategic Investment in In Silico Toxicology.

As vast numbers of new chemicals are introduced to market annually, we are faced with the grand challenge of protecting humans and the environment while minimizing economically and ethically costly animal testing. In silico models promise to be the solution we seek, but we find ourselves at crossroads of future development efforts that would ensure standalone applicability and reliability of these tools. A conscientious effort that prioritizes experimental testing to support the needs of in silico models (versus regulatory needs) is called for to achieve this goal. Using economic analogy in the title of this work, we argue that a prudent investment is to go all-in to support in silico model development, rather than gamble our future by keeping the status quo of a "balanced portfolio" of testing approaches. We discuss two paths to future in silico toxicology-one based on big-data statistics ("broadsword"), and the other based on direct modeling of molecular interactions ("scalpel")-and offer rationale that the latter approach is more transparent, is better aligned with our quest for fundamental knowledge, and has a greater potential to succeed if we are willing to transform our toxicity-testing paradigm.

Toxicology Advances for 21st Century Chemical Pollution.

Pollution represents a leading threat to global health and ecosystems. Systems-based initiatives, including Planetary Health, EcoHealth, and One Health, require theoretical and translational platforms to address chemical pollution. Comparative and predictive toxicology are providing integrative approaches for identifying problematic contaminants, designing less hazardous alternatives, and reducing the impacts of chemical pollution.

Decarbonylative Olefination of Aldehydes to Alkenes.

New atom-economical alternatives to Wittig chemistry are needed to construct olefins from carbonyl compounds, but none have been developed to-date. Here we report an atom-economical olefination of carbonyls via aldol-decarbonylative coupling of aldehydes using robust and recyclable supported Pd catalysts, producing only CO and H2O as waste. The reaction affords homocoupling of aliphatic aldehydes, as well as heterocoupling of aliphatic and aromatic ones. Computations provide insight into the selectivity and thermodynamics of the reaction. The tandem aldol-decarbonylation reaction opens the door to exploration of new carbonyl reactivity to construct olefins.

CRISPR-Generated Nrf2a Loss- and Gain-of-Function Mutants Facilitate Mechanistic Analysis of Chemical Oxidative Stress-Mediated Toxicity in Zebrafish.

The transcription factor Nrf2a induces a cellular antioxidant response and provides protection against chemical-induced oxidative stress, as well as playing a critical role in development and disease. Zebrafish are a powerful model to study the role of Nrf2a in these processes but have been limited by reliance on transient gene knockdown techniques or mutants with only partial functional alteration. We developed several lines of zebrafish carrying different null (loss of function, LOF) or hyperactive (gain of function, GOF) mutations to facilitate our understanding of the Nrf2a pathway in protecting against oxidative stress. The mutants confirmed Nrf2a dependence for induction of the antioxidant genes gclc, gstp, prdx1, and gpx1a and identified a role for Nrf2a in the baseline expression of these genes, as well as for sod1. Specifically, the 4-fold induction of gstp by tert-butyl hydroperoxide (tBHP) in wild type fish was abolished in LOF mutants. In addition, baseline gstp expression in GOF mutants increased by 12.6-fold and in LOF mutants was 0.8-fold relative to wild type. Nrf2a LOF mutants showed increased sensitivity to the acute toxicity of cumene hydroperoxide (CHP) and tBHP throughout the first 4 days of development. Conversely, GOF mutants were less sensitive to CHP toxicity during the first 4 days of development and were protected against the toxicity of both hydroperoxides after 4 dpf. Neither gain nor loss of Nrf2a modulated the toxicity of R-(-)-carvone (CAR), despite the ability of this compound to potently induce Nrf2a-dependent antioxidant genes. Similar to other species, GOF zebrafish mutants exhibited significant growth and survival defects. In summary, these new genetic tools can be used to facilitate the identification of downstream gene targets of Nrf2a, better define the role of Nrf2a in the toxicity of environmental chemicals, and further the study of diseases involving altered Nrf2a function.

Toward Less Hazardous Industrial Compounds: Coupling Quantum Mechanical Computations, Biomarker Responses, and Behavioral Profiles To Identify Bioactivity of SN2 Electrophiles in Alternative Vertebrate Models.

Sustainable molecular design of less hazardous chemicals promises to reduce risks to public health and the environment. Computational chemistry modeling coupled with alternative toxicology models (e.g., larval fish) present unique high-throughput opportunities to understand structural characteristics eliciting adverse outcomes. Numerous environmental contaminants with reactive properties can elicit oxidative stress, an important toxicological response associated with diverse adverse outcomes (i.e., cancer, diabetes, neurodegenerative disorders, etc.). We examined a common chemical mechanism (bimolecular nucleophilic substitution (SN2)) associated with oxidative stress using property-based computational modeling coupled with acute (mortality) and sublethal (glutathione, photomotor behavior) responses in the zebrafish (Danio rerio) and the fathead minnow (Pimephales promelas) models to identify whether relationships exist among biological responses and molecular attributes of industrial chemicals. Following standardized methods, embryonic zebrafish and larval fathead minnows were exposed separately to eight different SN2 compounds for 96 h. Acute and sublethal responses were compared to computationally derived in silico chemical descriptors. Specifically, frontier molecular orbital energies were significantly related to acute LC50 values and photomotor response (PMR) no observed effect concentrations (NOECs) in both fathead minnow and zebrafish. This reactivity index, LC50 values, and PMR NOECs were also significantly related to whole body glutathione (GSH) levels, suggesting that acute and chronic toxicity results from protein adduct formation for SN2 electrophiles. Shared refractory locomotor response patterns among study compounds and two alternative vertebrate models appear informative of electrophilic properties associated with oxidative stress for SN2 chemicals. Electrophilic parameters derived from frontier molecular orbitals were predictive of experimental in vivo acute and sublethal toxicity. These observations provide important implications for identifying and designing less hazardous industrial chemicals with reduced potential to elicit oxidative stress through bimolecular nucleophilic substitution.

Kinetics of Glutathione Depletion and Antioxidant Gene Expression as Indicators of Chemical Modes of Action Assessed in Vitro in Mouse Hepatocytes with Enhanced Glutathione Synthesis.

Here we report a vertically integrated in vitro - in silico study that aims to elucidate the molecular initiating events involved in the induction of oxidative stress (OS) by seven diverse chemicals (cumene hydroperoxide, t-butyl hydroperoxide, hydroquinone, t-butyl hydroquinone, bisphenol A, Dinoseb, and perfluorooctanoic acid). To that end, we probe the relationship between chemical properties, cell viability, glutathione (GSH) depletion, and antioxidant gene expression. Concentration-dependent effects on cell viability were assessed by MTT assay in two Hepa-1 derived mouse liver cell lines: a control plasmid vector transfected cell line (Hepa-V), and a cell line with increased glutamate-cysteine ligase (GCL) activity and GSH content (CR17). Changes to intracellular GSH content and mRNA expression levels for the Nrf2-driven antioxidant genes Gclc, Gclm, heme oxygenase-1 ( Hmox1), and NADPH quinone oxidoreductase-1 ( Nqo1) were monitored after sublethal exposure to the chemicals. In silico models of covalent and redox reactivity were used to rationalize differences in activity of quinones and peroxides. Our findings show CR17 cells were generally more resistant to chemical toxicity and showed markedly attenuated induction of OS biomarkers; however, differences in viability effects between the two cell lines were not the same for all chemicals. The results highlight the vital role of GSH in protecting against oxidative stress-inducing chemicals as well as the importance of probing molecular initiating events in order to identify chemicals with lower potential to cause oxidative stress.

Comparative behavioral toxicology with two common larval fish models: Exploring relationships among modes of action and locomotor responses.

Behavioral responses inform toxicology studies by rapidly and sensitively detecting molecular initiation events that propagate to physiological changes in individuals. These behavioral responses can be unique to chemical specific mechanisms and modes of action (MOA) and thus present diagnostic utility. In an initial effort to explore the use of larval fish behavioral response patterns in screening environmental contaminants for toxicity and to identify behavioral responses associated with common chemical specific MOAs, we employed the two most common fish models, the zebrafish and the fathead minnow, to define toxicant induced swimming activity alterations during interchanging photoperiods. Though the fathead minnow (Pimephales promelas) is a common model for aquatic toxicology research and regulatory toxicology practice, this model has received little attention in behavioral studies compared to the zebrafish, a common biomedical model. We specifically compared behavioral responses among 7 different chemicals (1-heptanol, phenol, R-(-)-carvone, citalopram, diazinon, pentylenetetrazole (PTZ), and xylazine) that were selected and classified based on anticipated MOA (nonpolar narcosis, polar narcosis, electrophile, specific mechanism) according to traditional approaches to examine whether these comparative responses differ among chemicals with various structure-based predicted toxicity. Following standardized experimental guidelines, zebrafish embryos and fathead minnow larvae were exposed for 96 h to each compound then were observed using digital behavioral analysis. Behavioral observations included photomotor responses, distance traveled, and stimulatory, refractory and cruising locomotor activity. Though fathead minnow larvae displayed greater behavioral sensitivity to 1-heptanol, phenol and citalopram, zebrafish were more sensitive to diazinon and R-(-)-carvone. Both fish models were equally sensitive to xylazine and PTZ. Further, the pharmaceuticals citalopram and xylazine significantly affected behavior at therapeutic hazard values, and each of the seven chemicals elicited unique behavioral response profiles. Larval fish behaviors appear useful as early tier diagnostics to identify mechanisms and pathways associated with diverse biological activities for chemicals lacking mechanistic data.

The Molecular Design Research Network.

Herein, we provide an overview of a research network that is aimed at fostering interdisciplinary collaboration between chemists and toxicologists with the goal of rationally designing safer commercial chemicals. The collaborative is the Molecular Design Research Network (MoDRN) that was created in 2013 with funding from the EPA-National Science Foundation Networks for Sustainable Molecular Design and Synthesis (NSMDS) program. MoDRN is led by 4 universities, Baylor University, University of Washington, The George Washington University, and Yale University. The overarching goal of the network is to enable and empower the design of safer chemicals based on the fourth Principle of Green Chemistry that states, "chemical products should be designed to preserve efficacy of function while minimizing toxicity."

Advancing alternatives analysis: The role of predictive toxicology in selecting safer chemical products and processes.

Alternatives analysis (AA) is a method used in regulation and product design to identify, assess, and evaluate the safety and viability of potential substitutes for hazardous chemicals. It requires toxicological data for the existing chemical and potential alternatives. Predictive toxicology uses in silico and in vitro approaches, computational models, and other tools to expedite toxicological data generation in a more cost-effective manner than traditional approaches. The present article briefly reviews the challenges associated with using predictive toxicology in regulatory AA, then presents 4 recommendations for its advancement. It recommends using case studies to advance the integration of predictive toxicology into AA, adopting a stepwise process to employing predictive toxicology in AA beginning with prioritization of chemicals of concern, leveraging existing resources to advance the integration of predictive toxicology into the practice of AA, and supporting transdisciplinary efforts. The further incorporation of predictive toxicology into AA would advance the ability of companies and regulators to select alternatives to harmful ingredients, and potentially increase the use of predictive toxicology in regulation more broadly. Integr Environ Assess Manag 2017;13:915-925. © 2017 SETAC.

Toward the Design of Less Hazardous Chemicals: Exploring Comparative Oxidative Stress in Two Common Animal Models.

Sustainable molecular design of less hazardous chemicals presents a potentially transformative approach to protect public health and the environment. Relationships between molecular descriptors and toxicity thresholds previously identified the octanol-water distribution coefficient, log D, and the HOMO-LUMO energy gap, ΔE, as two useful properties in the identification of reduced aquatic toxicity. To determine whether these two property-based guidelines are applicable to sublethal oxidative stress (OS) responses, two common aquatic in vivo models, the fathead minnow (Pimephales promelas) and zebrafish (Danio rerio), were employed to examine traditional biochemical biomarkers (lipid peroxidation, DNA damage, and total glutathione) and antioxidant gene activation following exposure to eight structurally diverse industrial chemicals (bisphenol A, cumene hydroperoxide, dinoseb, hydroquinone, indene, perfluorooctanoic acid, R-(-)-carvone, and tert-butyl hydroperoxide). Bisphenol A, cumene hydroperoxide, dinoseb, and hydroquinone were consistent inducers of OS. Glutathione was the most consistently affected biomarker, suggesting its utility as a sensitivity response to support the design of less hazardous chemicals. Antioxidant gene expression (changes in nrf2, gclc, gst, and sod) was most significantly (p < 0.05) altered by R-(-)-carvone, cumene hydroperoxide, and bisphenol A. Results from the present study indicate that metabolism of parent chemicals and the role of their metabolites in molecular initiating events should be considered during the design of less hazardous chemicals. Current empirical and computational findings identify the need for future derivation of sustainable molecular design guidelines for electrophilic reactive chemicals (e.g., SN2 nucleophilic substitution and Michael addition reactivity) to reduce OS related adverse outcomes in vivo.