REGISTER NOW

Featuring:

The Ed Carney Predictive Toxicology Award Recipient at the ASCCT 13th Annual Meeting

A brief Q&A session will follow each presentation.

ABSTRACTS

Generating GD211-Aligned Documentation for ToxCast to Support Assay Interpretation and Data Use: The U.S. EPA’s Toxicity Forecaster (ToxCast) program provides a large and impactful public repository of targeted new approach methods (NAMs) for toxicology. Assays employ a variety of technologies to evaluate the effects of chemical exposure on diverse biological targets for almost 10,000 substances. The ToxCast pipeline R software package, tcpl, enables flexible data processing and analysis to process, model, and visualize concentration-response screening data as well as populate a linked MySQL database, invitrodb. Invitrodb contains manually curated and expert-reviewed assay annotation information, describing experimental and biological details that are essential for proper interpretation of screening results. Complementary to any data generation and processing effort, assay documentation using internationally harmonized standards ensures data are transparent, accessible, and interoperable, thereby increasing confidence for the adoption of assay data in next generation chemical assessment. Organisation for Economic Co-operation and Development (OECD) guidance document (GD) 211 suggests components of comprehensive assay documentation describing non-guideline in vitro test methods and their interpretation. The intent of GD 211 is to harmonize non-guideline, in vitro method descriptions to allow assessment of the relevance of the test method for biological response and data quality. Legacy GD 211-based documentation was manually maintained and covered less than 100 assay endpoints. Given major ToxCast software and database enhancements plus a desire to expand assay coverage and automate report generation, a complete overhaul to existing documentation process was undertaken. A compiled report and assay description documents for 809 out of 1507 endpoints in the database were released in September 2024, accompanying the invitrodb v4.2 database release.  For this iteration, assay description documents cover 18 assay sources and a diverse biological space, including, but not limited to, the priority areas of androgen and estrogen receptor (AR/ER), steroidogenesis, thyroid bioactivity, and the developmental neurotoxicity in vitro battery (DNT-IVB). These reports are a work in progress and will be iteratively updated as descriptions improve from user feedback and more information becomes available. This documentation effort represents a novel, semi-automated and large-scale application of the GD211 template with all underlying information also databased. Along with the standardized and documented data processing procedures within tcpl, ToxCast continues to serve as an important publicly available data resource that follows and supports the Findable, Accessible, Interoperable, and Reusable (FAIR) data principles needed for data interoperability in workflows that utilize bioactivity data. This abstract does not necessarily reflect U.S. EPA policy.

Facilitating the regulatory uptake of transcriptional biomarkers for chemical hazard screening: From gene sequence to biology: Regulatory authorities worldwide are pushing for the use of gene expression data to assess chemical toxicity. However, issues with gene expression data reproducibility and interpretation still need to be addressed to take full advantage of this approach. A panel of transcriptional (mRNA) biomarkers can offer valuable insights into molecular response to chemical exposure. We will present a framework and workflow that combines bioinformatics and experimental analyses to characterize gene expression biomarkers by RT-qPCR, an affordable and widely available method. The framework includes the identification of biomarker genes based on conserved evolutionary (orthology) and biological process (ontology). Additionally, the incomplete annotation of the gene transcript diversity resulting from alternative splicing can interfere with proper measurement of transcript levels. As a mitigation strategy, we leveraged human, mouse and rat genomic databases and developed a strategic amplicon sequencing workflow to validate transcript target sequences in orthologous coding exons of biomarker genes from these species. This approach can help to avoid gene regions affected by alternative splicing, and also facilitate the comparison of gene expression data across rodents and humans. Preliminary proof-of-concept work assessing rat primary neuronal cells differentiation and synaptogenesis identified potential biomarkers of developmental neurotoxicity, designed robust gene expression measurement protocols, and generated promising results. Future studies will test additional neurotoxic and non-neurotoxic chemicals to evaluate the assay’s biological relevance and predictivity. Perturbations of these evolutionarily conserved transcriptional biomarkers may be used to monitor Key Events in an Adverse Outcome Pathway approach. 

Development of Mathematical New Approach Methods to Assess Chemical Mixtures: The U.S. EPA’s Toxicity Forecaster (ToxCast) program contains targeted bioactivity screening data for thousands of single chemicals to inform prioritization and hazard prediction. However, environmental contaminants in practice are often encountered as coexposures; thus, it is of high importance to study chemical mixtures. The aim of this work is to leverage the readily available single chemical screening assay data in ToxCast to make predictions of the bioactivity of binary chemical mixtures. Mathematical relationships between mixtures and their single chemical constituents are applied with two different models, concentration addition and independent action. To evaluate the performance of these mixture models, an empirical dataset was collected for 21 chemical mixtures and their single constituents screened in concentration-response using a multidimensional in vitro platform for transcription factor activity. Point of departure estimates were compared among the mixture model predictions and bootstrap resampling was performed to obtain confidence intervals on the models. Out of 1701 mixture-assay endpoint responses, 214 demonstrated a positive response with at least one active single constituent; in these cases, about 90% of the predicted mixture activity concentrations at the cutoff level fell within 0.5 log₁₀-µM of the observed cutoff concentrations and about 80% of the predicted mixture bootstrapped intervals exhibited significant overlap with the observed mixture intervals. As it is resource-prohibitive to screen all combinations of chemicals, the determination of conservative mixture predictions and uncertainties is critical for operationalizing existing ToxCast data for the forecasting of simulated mixtures from real-life coexposures. This abstract does not necessarily reflect U.S. EPA policy.

About the Presenters

Madison Feshuk is a biologist at the US Environmental Protection Agency’s Center for Computational Toxicology and Exposure (CCTE). She supports efforts crucial to promoting the adoption of new approach methods (NAMs) and increasing data interoperability to ultimately achieve reductions in animal testing and improve environmental risk assessment and regulatory decision-making.

Bhaja K. Padhi, PhD, DABT, is a Research Biologist at the Environmental Health Science and Research Bureau of Health Canada. He completed his postdoctoral training at the Ottawa Health Research Institute and earned a Certificate in Bioinformatics accredited by the University of British Columbia, Canada. Additionally, he has served as a Visiting Scientist at Johns Hopkins University in Baltimore, Maryland, USA. Padhi is a Diplomate of the American Board of Toxicology and is a committee member of the Natural Sciences and Engineering Research Council of Canada (NSERC) on the “Genes, Cells, and Molecules” team. To date, he has published more than 40 research papers and authored a book.

Rachel Broughton is an ORISE postdoctoral fellow at the U.S. Environmental Protection Agency within the Center for Computational Toxicology and Exposure. Her research focuses on developing computational techniques to predict the bioactivity behavior of chemical mixtures that are found in common co-exposures. 

Recordings and other materials from this webinar will be posted on the ASCCT webinar archive: https://www.ascctox.org/webinar-archive