A revolutionary approach to chemical safety that uses computational methods to protect ecosystems without endless animal testing
Imagine you're tasked with protecting every animal in a vast rainforest from thousands of human-made chemicals. You'd need to test each chemical on every species—from jaguars and toucans to tree frogs and ants—an utterly impossible endeavor requiring millions of tests and endless resources 1 . This isn't just a rainforest dilemma; it's the monumental challenge facing toxicologists worldwide who work to safeguard both human health and ecosystems from potentially harmful chemicals.
For decades, chemical safety assessment has relied heavily on animal testing, using a handful of model species like lab rats to represent humans, and a few select fish and invertebrates to represent ecological health.
These two streams of research have largely remained separate, with little collaboration between them. But now, a revolutionary international initiative is breaking down these silos and pioneering sophisticated computational methods to protect all species more efficiently and humanely 3 .
Ecosystems are typically populated by countless species, each with potentially different sensitivities to the numerous chemical compounds present in their environment. Traditionally, regulatory toxicology has approached this complexity by testing chemicals on a limited number of species and extrapolating the results to others. Mammalian data drives human health assessments, while data from a few non-mammalian species informs environmental protection, with surprisingly little collaboration between these two approaches 3 .
The fundamental problem is straightforward: experimentally testing all possible species-chemical combinations is physically, financially, and ethically impossible 1 . Additionally, global regulatory trends are moving toward reducing animal testing. Europe has banned animal testing for cosmetics, and the U.S. Environmental Protection Agency has committed to eliminating mammalian studies by 2035 3 . These developments have created an urgent need for innovative, efficient, and humane alternatives.
In response to these challenges, the International Consortium to Advance Cross-Species Extrapolation (ICACSER) was established. This global, cross-sector consortium brings together researchers, regulators, and advocates to integrate bioinformatics approaches that can revolutionize how we assess chemical safety 4 .
ICACSER aligns with the One Health approach—a collaborative model that recognizes the interconnectedness between people, animals, plants, and their shared environment. This perspective acknowledges that you cannot protect human health without simultaneously protecting the ecosystems we're embedded within 3 .
The consortium builds upon the vision articulated by the National Research Council in 2007, which advocated for a shift away from traditional animal testing toward more efficient cell-based and computational approaches for evaluating chemical safety in the 21st century 3 .
By bringing together experts from academia, industry, regulatory agencies, and advocacy groups worldwide, ICACSER fosters the interdisciplinary collaboration needed to tackle the complex challenge of cross-species chemical extrapolation.
This field has developed specific terminology that's essential to understanding the revolution in toxicology:
A conceptual framework that maps the pathway from a chemical's initial interaction with a biomolecule (Molecular Initiating Event) through subsequent biological changes to an adverse outcome relevant to risk assessment 3 .
Defines how broadly across species a biological pathway is applicable based on conservation of structure and function 3 .
An umbrella term for approaches that reduce animal use, including in silico (computational), in chemico, and in vitro assays, along with omics technologies 3 .
The Sequence Alignment to Predict Across Species Susceptibility tool, which computationally predicts protein conservation and potential chemical susceptibility across species 6 .
The AOP framework serves as the backbone of modern cross-species extrapolation efforts. Think of it as creating a detailed roadmap of how a chemical causes harm, starting at the molecular level and progressing through cellular, tissue, and organism-level effects 3 .
The power of AOPs lies in their ability to identify conserved biological pathways across species. If a Molecular Initiating Event (such as a chemical binding to a specific protein) is conserved between humans and fish, then data from one can inform understanding of effects on the other. This framework removes silos between human and ecological toxicology by focusing on the fundamental biology shared across species 3 .
To understand how cross-species extrapolation works in practice, let's examine a crucial experiment that utilized the SeqAPASS tool to determine the applicability of high-throughput screening data across species.
The U.S. EPA's ToxCast™ program has screened thousands of chemicals in hundreds of mammalian-based high-throughput screening (HTS) assays. These rapid, automated tests prioritize which chemicals warrant further testing. Initially designed for human health assessment, scientists realized this data might also protect wildlife—if they could determine its relevance across species 5 .
Researchers used SeqAPASS to assess conservation of 484 protein targets represented in ToxCastâ„¢ assays. The tool employs a tiered approach:
Compares full primary amino acid sequences of proteins across species
Focuses comparison on specific functional domains within proteins
Examines individual amino acid residues known to be critical for chemical-protein interactions 5
The underlying premise is simple: the greater the similarity between a chemical's protein target in a known sensitive species and the corresponding protein in another species, the greater the likelihood that the chemical will interact similarly in both species 5 .
To demonstrate the method, researchers applied SeqAPASS to proteins targeted by endocrine-disrupting chemicals. The endocrine system—which governs growth, development, and reproduction through hormones—is particularly vulnerable to chemical disruption in both humans and wildlife 5 .
Critical for male sexual development and function
Involved in producing sex hormones
Essential for metabolism, growth, and brain development 5
For each protein, researchers input the known protein sequence from model organisms into SeqAPASS, which then compared these sequences against the National Center for Biotechnology Information database containing over 153 million proteins from more than 95,000 organisms 6 .
The SeqAPASS analysis revealed remarkable conservation of endocrine targets across vertebrate species. The data indicated that the androgen receptor, steroidogenesis enzymes, and thyroid pathway proteins are widely conserved across mammals, birds, reptiles, amphibians, and fish 5 .
| Protein Target | Mammals | Birds | Reptiles | Amphibians | Fish |
|---|---|---|---|---|---|
| Androgen Receptor | Conserved | Conserved | Conserved | Conserved | Mostly Conserved |
| Aromatase | Conserved | Conserved | Conserved | Conserved | Conserved |
| Thyroid Peroxidase | Conserved | Conserved | Conserved | Conserved | Mostly Conserved |
| Transthyretin | Conserved | Conserved | Conserved | Conserved | Conserved |
These findings were scientifically important because they provided initial evidence that HTS data generated primarily from mammalian systems could reasonably predict potential endocrine disruption in diverse wildlife species. This significantly expands the utility of existing data and helps prioritize limited testing resources on chemicals of greatest concern 5 .
| Analysis Level | Number of Species Predicted Susceptible |
|---|---|
| Level 1 (Primary Sequence) | 952 species |
| Level 2 (Functional Domains) | 976 species |
| Level 3 (Specific Residues) | 750 species |
The tables above illustrate how SeqAPASS generates quantitative predictions of potential chemical susceptibility across species. The slight variations between analysis levels reflect increasing specificity—Level 3 incorporates knowledge of exact amino acids involved in chemical binding, providing the highest resolution predictions .
The field of cross-species extrapolation relies on a sophisticated suite of computational tools and databases. Here are the key resources enabling this research:
| Tool/Resource | Function | Application Example |
|---|---|---|
| SeqAPASS | Compares protein sequences across species to predict chemical susceptibility | Determining if a chemical that binds human estrogen receptor likely affects fish |
| ExpressAnalyst | Cross-species platform for RNAseq annotation, quantification, and visualization | Identifying conserved gene expression patterns across species |
| ECOTOX Knowledgebase | Curated database of ecotoxicology literature | Accessing tested toxicity values for species with existing data |
| Molecular Docking & Dynamics | Simulates chemical-protein interactions at atomic level | Understanding binding strength and mechanism across species |
| NCBI Protein Database | Central repository for protein sequence data from numerous organisms | Source data for cross-species protein comparisons |
These tools represent a powerful shift from traditional toxicology testing toward computational predictive toxicology. By combining multiple lines of evidence from different tools, researchers can build compelling cases for chemical safety decisions without extensive animal testing 4 6 .
Recent advances are making cross-species extrapolation increasingly sophisticated. In a 2025 study, researchers combined SeqAPASS with molecular dynamics simulations to examine the interaction between perfluorooctanoic acid (PFOA) and transthyretin (a thyroid hormone transport protein) across species. This approach provided quantitative metrics on binding strength and confirmed that Lysine-15 is a key residue for this interaction across numerous vertebrate species .
The regulatory landscape is evolving rapidly to embrace these new methods. As one EPA scientist noted, "The data landscape is changing. There is now greater focus on the generation of mechanistic, cell-based, and computationally derived information for consideration as alternatives to animal testing" 3 .
ICACSER continues to drive innovation by bringing together tool developers, regulators, and researchers to define needs, create standardized approaches, and demonstrate real-world applications. Their work ensures that these sophisticated methods will transition from research labs to regulatory decision-making, creating a more efficient and humane chemical safety assessment system 3 .
The effort to advance cross-species extrapolation represents more than just technical innovation—it signifies a fundamental shift in how we conceptualize chemical safety. By recognizing the biological connections between species and developing tools to explore those connections computationally, toxicology is transforming from a science that tests what we can (typically rodents and a few standard test species) to a science that predicts what we need to protect (all species, including humans).
This approach aligns with the profound truth captured in the One Health paradigm: the health of people, animals, and ecosystems are inextricably linked. As ICACSER and similar initiatives continue to break down disciplinary silos, we move closer to comprehensive chemical protection that honors this interconnectedness—ensuring safety for all inhabitants of our shared planet through smarter science, not just more testing.
As this field advances, we may finally solve the impossible task of protecting thousands of species from thousands of chemicals—not through endless testing, but through the intelligent application of cutting-edge computational science that recognizes the fundamental unity of life on Earth.