Data-Driven Characterization of Genetic Variability in Disease Pathways and Pesticide-Induced Nervous System Disease in the United States Population.

BACKGROUND: Genetic susceptibility to chemicals is incompletely characterized. However, nervous system disease development following pesticide exposure can vary in a population, implying some individuals may have higher genetic susceptibility to pesticide-induced nervous system disease. OBJECTIVES: We aimed to build a computational approach to characterize single-nucleotide polymorphisms (SNPs) implicated in chemically induced adverse outcomes and used this framework to assess the link between differential population susceptibility to pesticides and human nervous system disease. METHODS: We integrated publicly available datasets of Chemical-Gene, Gene-Pathway, and SNP-Disease associations to build Chemical-Pathway-Gene-SNP-Disease linkages for humans. As a case study, we integrated these linkages with spatialized pesticide application data for the US from 1992 to 2018 and spatialized nervous system disease rates for 2018. Through this, we characterized SNPs that may be important in states with high disease occurrence based on the pesticides used there. RESULTS: We found that the number of SNP hits per pesticide in US states positively correlated with disease incidence and prevalence for Alzheimer's disease, Parkinson disease, and multiple sclerosis. We performed frequent itemset mining to differentiate pesticides used over time in states with high and low disease occurrence and found that only 19% of pesticide sets overlapped between 10 states with high disease occurrence and 10 states with low disease occurrence rates, and more SNPs were implicated in pathways in high disease occurrence states. Through a cross-validation of subsets of five high and low disease occurrence states, we characterized SNPs, genes, pathways, and pesticides more frequently implicated in high disease occurrence states. DISCUSSION: Our findings support that pesticides contribute to nervous system disease, and we developed priority lists of SNPs, pesticides, and pathways for further study. This data-driven approach can be adapted to other chemicals, diseases, and locations to characterize differential population susceptibility to chemical exposures. https://doi.org/10.1289/EHP14108.

Potential for Machine Learning to Address Data Gaps in Human Toxicity and Ecotoxicity Characterization.

Machine Learning (ML) is increasingly applied to fill data gaps in assessments to quantify impacts associated with chemical emissions and chemicals in products. However, the systematic application of ML-based approaches to fill chemical data gaps is still limited, and their potential for addressing a wide range of chemicals is unknown. We prioritized chemical-related parameters for chemical toxicity characterization to inform ML model development based on two criteria: (1) each parameter's relevance to robustly characterize chemical toxicity described by the uncertainty in characterization results attributable to each parameter and (2) the potential for ML-based approaches to predict parameter values for a wide range of chemicals described by the availability of chemicals with measured parameter data. We prioritized 13 out of 38 parameters for developing ML-based approaches, while flagging another nine with critical data gaps. For all prioritized parameters, we performed a chemical space analysis to assess further the potential for ML-based approaches to predict data for diverse chemicals considering the structural diversity of available measured data, showing that ML-based approaches can potentially predict 8-46% of marketed chemicals based on 1-10% with available measured data. Our results can systematically inform future ML model development efforts to address data gaps in chemical toxicity characterization.

To Split or Not to Split: Characterizing Chemical Pollution Impacts in Aquatic Ecosystems with Species Sensitivity Distributions for Specific Taxonomic Groups.

Bridging applied ecology and ecotoxicology is key to protect ecosystems. These disciplines show a mismatch, especially when evaluating pressures. Contrasting to applied ecology, ecotoxicological impacts are often characterized for whole species assemblages based on Species Sensitivity Distributions (SSDs). SSDs are statistical models describing per chemical across-species sensitivity variation based on laboratory toxicity tests. To assist in the aligning of the disciplines and improve decision-support uses of SSDs, we investigate taxonomic-group-specific SSDs for algae/cyanobacteria/aquatic plants, invertebrates, and vertebrates for 180 chemicals with sufficient test data. We show that splitting improves pollution impact assessments for chemicals with a specific mode of action and, surprisingly, for narcotic chemicals. We provide a framework for splitting SSDs that can be applied to serve in environmental protection, life cycle assessment, and management of freshwater ecosystems. We illustrate that using split SSDs has potentially large implications for the decision-support of SSD-based outputs around the globe.

Probabilistic Points of Departure and Reference Doses for Characterizing Human Noncancer and Developmental/Reproductive Effects for 10,145 Chemicals.

BACKGROUND: Regulatory toxicity values used to assess and manage chemical risks rely on the determination of the point of departure (POD) for a critical effect, which results from a comprehensive and systematic assessment of available toxicity studies. However, regulatory assessments are only available for a small fraction of chemicals. OBJECTIVES: Using in vivo experimental animal data from the U.S. Environmental Protection Agency's Toxicity Value Database, we developed a semiautomated approach to determine surrogate oral route PODs, and corresponding toxicity values where regulatory assessments are unavailable. METHODS: We developed a curated data set restricted to effect levels, exposure routes, study designs, and species relevant for deriving toxicity values. Effect levels were adjusted to chronic human equivalent benchmark doses (BMDh). We hypothesized that a quantile of the BMDh distribution could serve as a surrogate POD and determined the appropriate quantile by calibration to regulatory PODs. Finally, we characterized uncertainties around the surrogate PODs from intra- and interstudy variability and derived probabilistic toxicity values using a standardized workflow. RESULTS: The BMDh distribution for each chemical was adequately fit by a lognormal distribution, and the 25th percentile best predicted the available regulatory PODs [R2≥0.78, residual standard error (RSE)≤0.53 log10 units]. We derived surrogate PODs for 10,145 chemicals from the curated data set, differentiating between general noncancer and reproductive/developmental effects, with typical uncertainties (at 95% confidence) of a factor of 10 and 12, respectively. From these PODs, probabilistic reference doses (1% incidence at 95% confidence), as well as human population effect doses (10% incidence), were derived. DISCUSSION: In providing surrogate PODs calibrated to regulatory values and deriving corresponding toxicity values, we have substantially expanded the coverage of chemicals from 744 to 8,023 for general noncancer effects, and from 41 to 6,697 for reproductive/developmental effects. These results can be used across various risk assessment and risk management contexts, from hazardous site and life cycle impact assessments to chemical prioritization and substitution. https://doi.org/10.1289/EHP11524.

High-content analysis shows synergistic effects of low perfluorooctanoic acid (PFOS) and perfluorooctane sulfonic acid (PFOA) mixture concentrations on human breast epithelial cell carcinogenesis.

Perfluoroalkyl substances (PFAS) have been associated with cancer, but the potential underlying mechanisms need to be further elucidated and include studies of PFAS mixtures. This mechanistic study revealed that very low concentrations (500 pM) of the binary PFOS and PFOA mixture induced synergistic effects on human epithelial breast cell (MCF-10A) proliferation. The cell proliferation was mediated by pregnane X receptor (PXR) activation, an increase in cyclin D1 and CDK6/4 levels, decrease in p21 and p53 levels, and by regulation of phosphor-Akt and ß-catenin. The PFAS mixture also altered histone modifications, epigenetic mechanisms implicated in tumorigenesis, and promoted cell migration and invasion by reducing the levels of occludin. High-content screening using the cell painting assay, revealed that hundreds of cell features were affected by the PFAS mixture even at the lowest concentration tested (100 pM). The detailed phenotype profiling further demonstrated that the PFAS mixture altered cell morphology, mostly in parameters related to intensity and texture associated with mitochondria, endoplasmic reticulum, and nucleoli. Exposure to higher concentrations (≥50 µM) of the PFOS and PFOA mixture caused cell death through synergistic interactions that induced oxidative stress, DNA/RNA damage, and lipid peroxidation, illustrating the complexity of mixture toxicology. Increased knowledge about mixture-induced effects is important for better understanding of PFAS' possible role in cancer etiology, and may impact the risk assessment of these and other compounds. This study shows the potential of image-based multiplexed fluorescence assays and high-content screening for development of new approach methodologies in toxicology.

Comparative selective pressure potential of antibiotics in the environment.

To guide both environmental and public health policy, it is important to assess the degree of antibiotic resistance selection pressure under measured environmental concentrations (MECs), and to compare the efficacy of different mitigation strategies to minimize the spread of resistance. To this end, the resistance selection and enrichment potential due to antibiotic emissions into the environment must be analysed from a life cycle perspective, for a wide range of antibiotics, and considering variations in the underlying fitness costs between different resistance mutations and genes. The aim of this study is to consistently derive fitness cost-dependent minimum selective concentrations (MSCs) from readily available bacterial inhibition data and to build MSC-based species sensitivity distributions (SSDs). These are then used to determine antibiotic-specific resistance selection concentrations predicted to promote resistance in 5% of exposed bacterial species (RSC5). Using a previously developed competition model, we provide estimated MSC10 endpoints for 2,984 antibiotic and bacterial species combinations; the largest set of modelled MSCs available to date. Based on constructed SSDs, we derive RSC5 for 128 antibiotics with four orders of magnitude difference in their 'selective pressure potential' in the environment. By comparing our RSC5 to MECs, we highlight specific environmental compartments (e.g. hospital and wastewater effluents, lakes and rivers), as well as several antibiotics (e.g. ciprofloxacin, norfloxacin, enrofloxacin, and tetracycline), to be scrutinized for their potential role in resistance selection and dissemination. In addition to enabling comparative risk screening of the selective pressure potential of multiple antibiotics, our SSD-derived RSC5 provide the point of departure for calculating new life cycle-based characterization factors for antibiotics to compare mitigation strategies, thereby contributing towards a 'One-Health' approach to tackling the global antibiotic resistance crisis.

Advancing exposure data analytics and repositories as part of the European Exposure Science Strategy 2020-2030.

High-quality and comprehensive exposure-related data are critical for different decision contexts, including environmental and human health monitoring, and chemicals risk assessment and management. However, exposure-related data are currently scattered, frequently of unclear quality and structure, not readily accessible, and stored in various-partly overlapping-data repositories, leading to inefficient and ineffective data usage in Europe and globally. We propose strategic guidance for an integrated European exposure data production and management framework for use in science and policy, building on current and future data analysis and digitalization trends. We map the existing exposure data landscape to requirements for data analytics and repositories across European policies and regulations. We further identify needs and ways forward for improving data generation, sharing, and usage, and translate identified needs into an operational action plan for European and global advancement of exposure data for policies and regulations. Identified key areas of action are to develop consistent exposure data standards and terminology for data production and reporting, increase data transparency and availability, enhance data storage and related infrastructure, boost automation in data management, increase data integration, and advance tools for innovative data analysis. Improving and streamlining exposure data generation and uptake into science and policy is crucial for the European Chemicals Strategy for Sustainability and European Digital Strategy, in line with EU Data policies on data management and interoperability.

Toward Assessing Absolute Environmental Sustainability of Chemical Pollution.

Chemicals are widely used in modern society, which can lead to negative impacts on ecosystems. Despite the urgent relevance for global policy setting, there are no established methods to assess the absolute sustainability of chemical pressure at relevant spatiotemporal scales. We propose an absolute environmental sustainability framework (AESA) for chemical pollution where (1) the chemical pressure on ecosystems is quantified, (2) the ability for ecosystems to withstand chemical pressure (i.e., their carrying capacity) is determined, and (3) the "safe space" is derived, wherein chemical pressure is within the carrying capacity and hence does not lead to irreversible adverse ecological effects. This space is then allocated to entities contributing to the chemical pressure. We discuss examples involving pesticide use in Europe to explore the associated challenges in implementing this framework (e.g., identifying relevant chemicals, conducting analyses at appropriate spatiotemporal scales) and ways forward (e.g., chemical prioritization approaches, data integration). The proposed framework is the first step toward understanding where and how much chemical pressure exceeds related ecological limits and which sources and actors are contributing to the chemical pressure. This can inform sustainable levels of chemical use and help policy makers establish relevant and science-based protection goals from regional to global scale.

Distinct genetic regions are associated with differential population susceptibility to chemical exposures.

Interactions between environmental factors and genetics underlie the majority of chronic human diseases. Chemical exposures are likely an underestimated contributor, yet gene-environment (GxE) interaction studies rarely assess their modifying effects. Here, we describe a novel method to profile the human genome and identify regions associated with differential population susceptibility to chemical exposures. Single nucleotide polymorphisms (SNPs) implicated in enriched chemical-disease intersections were identified and validated for three chemical classes with expected GxE interaction potential (neuroactive, hepatoactive, and cardioactive compounds). The same approach was then used to characterize consumer product classes with unknown risk for GxE interactions (washing products, cosmetics, and adhesives). Additionally, high-risk variant sets that may confer differential population susceptibility were identified for these consumer product groups through frequent itemset mining and pathway analysis. A dataset of 2454 consumer product chemical-disease linkages, with risk values, SNPs, and pathways for each association was developed, describing the interplay between environmental factors and genetics in human disease progression. We found that genetic hotspots implicated in GxE interactions differ across chemical classes (e.g., washing products had high-risk SNPs implicated in nervous system disease) and illustrate how this approach can discover new associations (e.g., washing product n-butoxyethanol implicated SNPs in the PI3K-Akt signaling pathway for Alzheimer's disease). Hence, our approach can predict high-risk genetic regions for differential population susceptibility to chemical exposures and characterize chemical modifying factors in specific diseases. These methods show promise for describing how chemical exposures can lead to varied health outcomes in a population and for incorporating inter-individual variability into chemical risk assessment.

Concentration-response evaluation of ToxCast compounds for multivariate activity patterns of neural network function.

The US Environmental Protection Agency's ToxCast program has generated toxicity data for thousands of chemicals but does not adequately assess potential neurotoxicity. Networks of neurons grown on microelectrode arrays (MEAs) offer an efficient approach to screen compounds for neuroactivity and distinguish between compound effects on firing, bursting, and connectivity patterns. Previously, single concentrations of the ToxCast Phase II library were screened for effects on mean firing rate (MFR) in rat primary cortical networks. Here, we expand this approach by retesting 384 of those compounds (including 222 active in the previous screen) in concentration-response across 43 network activity parameters to evaluate neural network function. Using hierarchical clustering and machine learning methods on the full suite of chemical-parameter response data, we identified 15 network activity parameters crucial in characterizing activity of 237 compounds that were response actives ("hits"). Recognized neurotoxic compounds in this network function assay were often more potent compared to other ToxCast assays. Of these chemical-parameter responses, we identified three k-means clusters of chemical-parameter activity (i.e., multivariate MEA response patterns). Next, we evaluated the MEA clusters for enrichment of chemical features using a subset of ToxPrint chemotypes, revealing chemical structural features that distinguished the MEA clusters. Finally, we assessed distribution of neurotoxicants with known pharmacology within the clusters and found that compounds segregated differentially. Collectively, these results demonstrate that multivariate MEA activity patterns can efficiently screen for diverse chemical activities relevant to neurotoxicity, and that response patterns may have predictive value related to chemical structural features.

Integration of curated and high-throughput screening data to elucidate environmental influences on disease pathways.

Addressing the complex relationship between public health and environmental exposure requires multiple types and sources of data. An important source of chemical data derives from high-throughput screening (HTS) efforts, such as the Tox21/ToxCast program, which aim to identify chemical hazard using primarily in vitro assays to probe toxicity. While most of these assays target specific genes, assessing the disease-relevance of these assays remains challenging. Integration with additional data sets may help to resolve these questions by providing broader context for individual assay results. The Comparative Toxicogenomics Database (CTD), a publicly available database that builds networks of chemical, gene, and disease information from manually curated literature sources, offers a promising solution for contextual integration with HTS data. Here, we tested the value of integrating data across Tox21/ToxCast and CTD by linking elements common to both databases (i.e., assays, genes, and chemicals). Using polymarcine and Parkinson's disease as a case study, we found that their union significantly increased chemical-gene associations and disease-pathway coverage. Integration also enabled new disease associations to be made with HTS assays, expanding coverage of chemical-gene data associated with diseases. We demonstrate how integration enables development of predictive adverse outcome pathways using 4-nonylphenol, branched as an example. Thus, we demonstrate enhancements to each data source through database integration, including scenarios where HTS data can efficiently probe chemical space that may be understudied in the literature, as well as how CTD can add biological context to those results.

Determination of chemical-disease risk values to prioritize connections between environmental factors, genetic variants, and human diseases.

Traditional methods for chemical risk assessment are too time-consuming and resource-intensive to characterize either the diversity of chemicals to which humans are exposed or how that diversity may manifest in population susceptibility differences. The advent of novel toxicological data sources and their integration with bioinformatic databases affords opportunities for modern approaches that consider gene-environment (GxE) interactions in population risk assessment. Here, we present an approach that systematically links multiple data sources to relate chemical risk values to diseases and gene-disease variants. These data sources include high-throughput screening (HTS) results from Tox21/ToxCast, chemical-disease relationships from the Comparative Toxicogenomics Database (CTD), hazard data from resources like the Integrated Risk Information System, exposure data from the ExpoCast initiative, and gene-variant-disease information from the DisGeNET database. We use these integrated data to identify variants implicated in chemical-disease enrichments and develop a new value that estimates the risk of these associations toward differential population responses. Finally, we use this value to prioritize chemical-disease associations by exploring the genomic distribution of variants implicated in high-risk diseases. We offer this modular approach, termed DisQGOS (Disease Quotient Genetic Overview Score), for relating overall chemical-disease risk to potential for population variable responses, as a complement to methods aiming to modernize aspects of risk assessment.

Associations between access to healthcare, environmental quality, and end-stage renal disease survival time: Proportional-hazards models of over 1,000,000 people over 14 years.

Prevalence of end-stage renal disease (ESRD) in the US increased by 74% from 2000 to 2013. To investigate the role of the broader environment on ESRD survival time, we evaluated average distance to the nearest hospital by county (as a surrogate for access to healthcare) and the Environmental Quality Index (EQI), an aggregate measure of ambient environmental quality composed of five domains (air, water, land, built, and sociodemographic), at the county level across the US. Associations between average hospital distance, EQI, and survival time for 1,092,281 people diagnosed with ESRD between 2000 and 2013 (age 18+, without changes in county residence) from the US Renal Data System were evaluated using proportional-hazards models adjusting for gender, race, age at first ESRD service date, BMI, alcohol and tobacco use, and rurality. The models compared the average distance to the nearest hospital (<10, 10-20, >20 miles) and overall EQI percentiles [0-5), [5-20), [20-40), [40-60), [60-80), [80-95), and [95-100], where lower percentiles are interpreted as better EQI. In the full, non-stratified model with both distance and EQI, there was increased survival for patients over 20 miles from a hospital compared to those under 10 miles from a hospital (hazard ratio = 1.14, 95% confidence interval = 1.12-1.15) and no consistent direction of association across EQI strata. In the full model stratified by average hospital distance, under 10 miles from a hospital had increased survival in the worst EQI strata (median survival 3.0 vs. 3.5 years for best vs. worst EQI, respectively), however for people over 20 miles from a hospital, median survival was higher in the best (4.2 years) vs worst (3.4 years) EQI. This association held across different rural/urban categories and age groups. These results demonstrate the importance of considering multiple factors when studying ESRD survival and future efforts should consider additional components of the broader environment.