Discovery Phase Agrochemical Predictive Safety Assessment Using High Content In Vitro Data to Estimate an In Vivo Toxicity Point of Departure

Journal of Agricultural and Food Chemistry, Год журнала: 2024, Номер 72(30), С. 17099 - 17120

Опубликована: Июль 21, 2024

Utilization of

Язык: Английский

---

The DIKW of Transcriptomics in Ecotoxicology: Extracting Information, Knowledge, and Wisdom From Big Data

Environmental Toxicology and Chemistry, Год журнала: 2024, Номер unknown

Опубликована: Авг. 12, 2024

In the early 2000s, transcriptomics was emerging as a new science with seemingly limitless potential. Originally coined by Charles Auffray (McGettigan, 2013), term refers to measurement of levels all expressed genes across genome (Figure 1). To researchers in our field aughts, promised uncover global molecular response environmental contaminants. We imagined replacing limited view provided individual biomarkers such ethoxyresorufin-O-deethylase (EROD) or vitellogenin, detailed picture changes occurring cell and pathways impacted. The idea that each chemical mixture would have unique signature could be linked organismal-level outcomes prominent. Over 20 years later, many us who are using research still baffled how analyze these big data sets, apply results salient questions ecotoxicology. paradox is we exponentially more information, but capacity interpret what it means. objective this Focus Article explore promise for ecotoxicology break down current realities field. hope convey realistic both challenges opportunities associated scientific approach. Our article geared toward researchers, government regulators, students interested knowledge experience area. organized around Data, Information, Knowledge, Wisdom (DIKW) framework 2), an illustrative case study (Textbox DIKW comes from information science, can used illustrate process moving raw valuable insights. Steps typical experiment. Methodological details provided. involved exposing double-crested cormorant (Nannopterum auritum) embryos ethinylestradiol via egg injection. Detailed methods presented elsewhere (Jeon et al., 2023). One experiment produce hundreds gigabytes (GBs) data. Next generation sequencing allows generate any species. With price dropping precipitously, experiments becoming affordable. leading technology generating RNA (RNA-Seq), cost range $100 USD/sample. New multiplex approaches drive price/sample further still. An like 1) GBs less than $2000 USD. RNA-Seq easily performed basic biology laboratory technicians minimal experience. great advantages (as opposed to, example, microarray technologies) species agnostic; experimental steps same regardless used. come form "sequencing reads," which fragments nucleotide sequences approximately 100 letters (100 base pairs [bp]) long. length gene typically >1000 bp, so overlapping reads need matched they correspond before counted. decreasing, ease increasing. bigger challenge when you figure out extract 100s GB receive laboratory. Extracting not trivial getting easier tools. approach differs without established reference genomes. Statistical uncertainty means there no single "right" analysis: different bioinformatics pipelines results. reads, mapped known then This fairly straightforward publicly available For nonmodel whose genomes yet sequenced (e.g., Nannopterum auritum [Textbox 1]), entire constellation received assembled create de novo transcriptome, place genome. De transcriptome assembly trivial. addition piecing together puzzle, researcher needs identify annotate (i.e., assign function to) within transcriptome. A online tool skip computationally intensive (Ewald 2023; Liu Seq2Fun algorithm (accessible ExpressAnalyst, 2024) aligns functional groups (orthologs) compiled over 600 produces table expression values 12,000 16,000 groups, majority function. Compared annotation, some lost approach, also advantages. one, often dysregulation smaller number rather thousands transcripts, unannotated identified low confidence. Mapping common ortholog database species-agnostic identifiers encourages on helps facilitate cross-species comparisons. seq2Fun practical starting point studying organisms newcomers world bioinformatics. output other will tell were your sample, their relative abundance. Most want compare "counts" conditions. What emerges analysis list differentially (DEGs), is, at control treated samples confirm previously modes action active, suggest hypotheses. Different visualization holistically data, study. Common include principal components (PCA) plots (which similarity between help potential outliers), Venn diagrams up- downregulated under conditions), heat maps highlight patterns samples), volcano p value directionality differences expression). Some illustrated Figures 3 4. differential inherent statistical magnified one working tens genes, high biological variability, small sample sizes. Because toxicogenomics studies expensive secondary goal reducing animal use, tend use lower replicates traditional toxicity studies. Organisation Economic Co-operation Development (OECD) test measuring endpoints related growth, reproduction, survival rats require sizes least 10 animals/experimental group OECD, 1998, 2008), whereas three five experiments. Differential employs t tests differentiate every compared sample. make variety assumptions size only cannot reliably provide set. Bioinformaticians mitigate problem, algorithms return numbers DEGs set even though well considered follow "best practices." end result has power, final produced two scientists analyzing varied depending (Limma vs. EdgeR) fold change cutoffs (Log2FC 0 Log2FC 4). availability evolving impression step never "done" lack confidence if yield opinion, "final" can, ways, justified. Researchers must shift understanding encompass implicit look large overall definitive answer. It helpful think prioritizing features investigation instead producing set-in-stone Alternatively, stringent multiple packages publish consensus DEGs. Regardless used, unlocking full deriving requires analysis. Generating complementary genome, pathways, organismal level response. transcriptomic dose–response (TDRA) hold making comparisons responses. model system, design, underlying question. initial scan through allow dysregulated Clustering profiles organize perturbed into similar samples. Prior about lead inferences mechanism chemical. study, classical estrogen exposure, vitellogenin apovitellenin 1, apparent, expected individuals exposed 3B). type gene-by-gene build results, tends focus attention already familiar away Unlocking broader transcriptional may involved. Pathway analysis, next pipeline, attempts do this. As ever-evolving pathway disparate packages. Two currently popular enrichment (GSEA) overrepresentation (ORA). ecotoxicology, particularly challenging. databases Kyoto Encyclopedia Genes Genomes, Gene Ontology) largely based biomedical models humans, rodents) therefore missing appropriate context nonmammalian illustrates problem highlights Parkinson's, Alzheimer's, Huntington's diseases seem relevant 5). However, identifying human disease nonhuman should necessarily interpreted nonsensical. Parkinson's mitochondrial respiration nicotinamide adenine dinucleotide:ubiquinone oxidoreductase cytochrome C oxidase isoforms), differently birds. cases, key very missing. classic biomarker mammals. well-conserved oviparous animals missed relies built rodent More needed area develop tailored ecotoxicological models. Another TDRA. rapidly gaining because provides way responses apical regulatory concern (Johnson 2022). Specifically, TDRA fits curves calculates gene-level benchmark dose significant curve fits. departure calculated distribution doses estimates chronic (Pagé-Larivière al, 2019). Transcriptomic advantage being able capture requiring belong especially attractive involving These promising, building wisdom sound decisions remains elusive goal. Transcriptomics contribute required informed chemicals environment. Global efforts reduce stimulating interest (NAMs). done develop, standardize, validate incorporated frameworks. Adopter-centric user experience, fit purpose, implementation critical success efforts. field, defined real-world impact contamination integration sources fields Insights derived piece puzzle. context, decision-making. Regulators consider assessments, traditionally been endpoints, effects standardized Although types direct link concern, detect subtle nonetheless indicative long-term harm, elucidate mechanisms action. Additionally, decision-making widely understood inefficient, prohibitively expensive, ethically concerning (Pain 2020). Given shortcomings status quo, organizations US Environmental Protection Agency (USEPA) Health Canada now promoting NAMs reduce, refine, replace use. vitro embryo-based assays integrate high-throughput screen prioritize purposes commercial frameworks easy. approaches, standardization, validation, evidence overprotective. Social sciences reveals design toxicogenomic simplicity compatibility routines workflows adopter-centric view) versus focusing novel functionality innovation-centric view; Pain Toward end, introduction USEPA Assessment Product 2024 transcriptomic-based represents important forward. stimulates exploration innovation. materials hazardous, meeting green chemistry principle "benign-by-design." Knowledge construct pay careful technical validity implementation. critically given desire favor approaches. years, ecotoxicogenomics met excitement skepticism ecologically meaningful outcomes. ability harness power improves, optimism grows. reflection leaves several concluding thoughts path technologies, sometimes perceived over-promise under-deliver. regard, interpretation communication benefit honest reckoning where framework. experiments, those species, stop stage hypotheses lists struggle perform first affected accomplished well-annotated Finally, driven adopters bring closer desired harnessing support knowledge-based regarding safety chemicals. hierarchy, strengths limitations technologies aware bioinformatic pipeline acceptable options step. spending effort transparent reproducible, sharing adhering Findability, Accesibility, Interoperability, Reusability [FAIR] principles) beneficial agonizing specific parameters standardization exemplified Reporting Framework initiative OECD's Extended Advisory Group Molecular Screening Toxicogenomics (Harrill 2021). discussed expect conclusive either context. omics lies simultaneously, few predetermined ones mortality response), does requirements adequate size, consideration toxicokinetics, replicated regulations restricting waiting evaluation, days "kill 'em count 'em" testing receding past. up trainees thinking modernize strategies. handling its efficiently wisely protect all. adapted plenary presentation delivered J. H. Canadian Ecotoxicology Workshop (CEW) annual October 2021 (Halifax, NS). authors thank CEW organizers B. Jourdan invitation. acknowledge Genome Quebec (via following programs: 2016 Large Scale Applied Research Program; 2018 Bioinformatics Computational Biology 2023 Genomics Applications Partnership Program) helped role Jessica A. Head: Conceptualization; Data curation; Formal analysis; Visualization; Writing—original draft. D. Ewald: Writing—review & editing. Niladri Basu: Expression Omnibus National Center Biotechnology Information accession GSE214620.

Язык: Английский

---

High-Throughput Transcriptomics Toxicity Assessment of Eleven Data-Poor Bisphenol A Alternatives

Environmental Pollution, Год журнала: 2024, Номер 361, С. 124827 - 124827

Опубликована: Авг. 27, 2024

Bisphenol A (BPA), a widely used chemical in the production of plastics and epoxy resins, has garnered significant attention due to its association with adverse health effects, particularly endocrine-disrupting properties. Regulatory measures aimed at reducing human exposure BPA have led proliferation alternative chemicals various consumer industrial products. While these alternatives serve reduce exposure, concerns arisen regarding their safety potential toxicity as regrettable substitutes. Previous efforts demonstrated that vitro high-throughput transcriptomics (HTTr) studies can be assess alternatives, this strategy produces transcriptomic points-of-departure (tPODs) are protective when compared PODs from traditional rodent studies. In study, we HTTr for eleven data-poor legacy sharing structural similarities BPA. Human breast cancer MCF-7 cells were exposed 11 concentrations ranging 0.1 25 μM toxicity. Analysis global changes previously characterized estrogen receptor alpha (ERα) biomarker signature revealed 9 altered gene expression relative controls. One (2,4'-Bisphenol A) activated ERα same concentration (i.e., 4,4'-BPA) but was deemed more potent it induced lower concentrations. These results address data gaps support ongoing screening assessments identify hazard help candidates may safer alternatives.

Язык: Английский

---

Bioinformatic Workflows for Deriving Transcriptomic Points of Departure: Current status, Data Gaps, and Research Priorities

Toxicological Sciences, Год журнала: 2024, Номер unknown

Опубликована: Ноя. 5, 2024

Abstract There is a pressing need to increase the efficiency and reliability of toxicological safety assessment for protecting human health environment. Although conventional toxicology tests rely on measuring apical changes in vertebrate models, there increasing interest use molecular information from animal vitro studies inform assessment. One promising pragmatic application involves derivation transcriptomic points departure (tPODs). Transcriptomic analyses provide snapshot global that reflect cellular responses stressors progression toward disease. A tPOD identifies dose level below which concerted change gene expression not expected biological system response chemical. common approach derive such consists modeling dose–response behavior each independently then aggregating gene-level data into single tPOD. different implementations this are possible, as discussed manuscript, research strongly supports overall idea reference doses produced using tPODs protective. An advantage can be generated shorter term (e.g. days) compared with endpoints 90-d subchronic rodent tests). Moreover, Given potential regulatory testing, rigorous reproducible wet dry laboratory methodologies their required. This review summarizes current state science regarding study design bioinformatics workflows derivation. We identify standards practice sources variability generation, gaps, areas uncertainty. recommendations address barriers promote adoption decision making.

Язык: Английский

---

Comparison of in vitro transcriptomic points of departure with fish acute and chronic toxicity values: A case study of rainbow trout cells exposed to pesticides

Environmental Toxicology and Chemistry, Год журнала: 2025, Номер unknown

Опубликована: Янв. 16, 2025

There is growing interest in transcriptomic points of departure (tPOD) values from vitro experiments as an alternative to animal test method. The study objective was calculate tPODs rainbow trout gill cells (RTgill-W1 following OECD 249) exposed pesticides, and evaluate how these compare fish acute chronic toxicity data. Cells were one fungicide (chlorothalonil), ten herbicides (atrazine, glyphosate, imazethapyr, metolachlor, diquat, s-metolachlor, AMPA, dicamba, dimethenamid-P, metribuzin), eight insecticides (chlorpyrifos, diazinon, permethrin, carbaryl, clothianidin, imidacloprid, thiamethoxam, chlorantraniliprole), 249 positive control 3,4-dichloroaniline. Pesticide concentrations wells modeled with IV-MBM EQP v2.1. Sequencing libraries prepared UPXome, calculated ExpressAnalyst. method identified 14,449 unique genes, 1,115 genes having >5 counts the 576 samples sequenced. For all chemicals, derived tPOD mode ranged 0.0004 125µM average 36µM. significant correlations between (x-value) EC50s RTgill-W1 (y = 0.92x+1.2, R2=0.9, p < 0.00001; n 11), LC50s 0.81x+0.8, R2=0.63, 0.0001; 20), sub-lethal effect 0.53x-0.2, R2=0.4, 0.009; 16) lethal 0.64x-0.023, R2=0.59, 0.0013; 14). Bland-Altman plot analyses comparisons also showed good agreement. Pathway-level benchmark doses when statistical requirements met, only possible for four pesticides. These findings support notion that short-term studies may be comparable concentration data vivo durations.

Язык: Английский

---