The Bench

Classical toxicology is hyperfocused on characterizing toxins: identifying harm, quantifying dose-response, and setting safe reference doses.[1] Success is measured in proving to regulators that reducing exposure justifies policy implementation: lead bans in gasoline,[2] PCBs restricted under TSCA,[3] and air pollution mitigation via the Clean Air Act.[4]

The EPA's biotechnology initiatives developed high-throughput screening to assess thousands of chemicals at scale,[5] aiming to convert toxicology from observation to prediction across targets, pathways, and mechanisms. Yet mechanistic work remained supporting evidence for regulatory decisions,[6][7] and considering this data as starting material for therapeutic interventions was far outside the EPA's mandate.

Innovation now in toxicology moves toward advanced computational and multimodal methodologies,[8] addressing concerns of multiple, cumulative exposures while migrating away from animal testing toward new alternative methods (NAMs). However, next-generation toxicology could also apply findings to biopharma by identifying addressable molecular targets that indicate biological damage and increased disease risk.

Toxicology has an underappreciated role in molecular biology. The canonical case: 1960s–70s studies of carcinogens (PAHs, coal tar metabolites) binding to DNA, RNA, and proteins[9] led to discovery that antioxidants in food and natural products were safe detoxification candidates.[10]

In 1976, dioxins were found to bind the aryl hydrocarbon receptor (AhR).[11] Decades of characterization revealed AhR as a master integrator of environmental signals and immune homeostasis.[12] AhR is now a validated drug target, and Tapinarof became the first FDA-approved AhR drug for psoriasis,[13] with 20+ clinical trials exploring ligands and synthetic molecules for chronic conditions and immunotherapies.[14]

A landmark 2000 study used the pesticide rotenone to develop an accurate Parkinson's disease animal model,[15] later confirmed in human epidemiological studies.[16] The rotenone model was mechanistically informative because it implicated mitochondria and downstream oxidative stress as drivers of neurodegeneration,[17] pointing directly to cellular machinery that subsequent PD therapeutic efforts have targeted.[18] Tools like biosensors targeting mitochondrial compartments[19] are now being deployed by labs like the Looger Lab at UCSD to study pesticide-driven cell death and develop better treatments. Our goal is to support this shift at the lab bench.

Next-generation toxicology can enhance biopharma pipelines by leveraging molecular features of pollutant exposures. Endocrine disruptors, heavy metals, and microplastics all stimulate similar immune and inflammatory pathways that could potentially be reversed or mitigated.[20] State-of-the-art approaches to detect and test this damage, as surveyed in the Deep Science Ventures Toxicity Report,[21] include organ-specific blood biomarkers (Teal Omics), high-throughput proteomics (SomaLogic), and epigenetic aging clocks based on methylation profiles (MyDNAge). Target selection should begin at intersections with the highest joint probability of vulnerability and treatment likelihood.

The exposome framework, which measures the totality of noninheritable environmental contributions to disease,[22] enables quantification via exposome risk scores by leaders like Chirag Patel[23] and Gary Miller,[24] with implications for drug development. We recommend a prioritization schema layered on network analysis that evaluates damage to genetic material, proteins, and the microbiome most likely to yield viable intervention points.

Foundation models like Tahoe-1x (pretrained on 266M single-cell profiles from the Tahoe-100M perturbation compendium) from Tahoe Therapeutics, Arc Institute, and CZ Biohub aim to produce predictive virtual cell models.[25] Screening pollutants would build intuition on cell state shifts leading to disease while improving model generalization. This complements EvE Bio's "pharmome" mapping, a platform of druggable proteins that could screen pollutants for hits warranting intervention.[26] Additionally, pollutants disrupting epigenetic processes make epigenetic reprogramming a viable countermeasure. Startups like Altos Labs, NewLimit, and Life Biosciences are developing therapies that reset methylation and chromatin marks.[27]