Single-cell and spatial transcriptomics inform mechanistic physiology in non-model animals

Abstract

Comparative physiology increasingly requires cellular resolution at the level of cell types, tissue microenvironments, and regulatory programs that associate genotype with function under environmental variation. Bulk transcriptomics and tissue-level assays have revealed pathways associated with stress responses, metabolic shifts, and developmental transitions; however, they average signals across heterogeneous cell populations and frequently obscure which cells contribute to physiological phenotypes. Single-cell RNA sequencing (scRNA-seq), single-nucleus RNA sequencing (snRNA-seq), and spatial transcriptomics (ST) address complementary aspects of this limitation by resolving transcriptional heterogeneity and, for spatial approaches, preserving tissue context. These technologies are extending beyond classical biomedical models to non-model animals, enabling discovery of cell states, comparison of cell-type evolution, and spatially informed hypotheses concerning ionoregulation, respiration, immune defense, endocrine signaling, regeneration, and symbiosis. This review provides a physiology-centered blueprint for applying these approaches to non-model species and critically evaluates dissociation and preservation bias, genome annotation, seasonal and ecological variation, biological replication, pseudoreplication, cross-species integration, and spatial resolution. We distinguish descriptive mapping, replicate-aware association, mechanistic hypothesis generation, and causal validation because expression patterns, colocalization, trajectories, and ligand–receptor predictions do not by themselves demonstrate mechanism. Practical reporting and validation standards are proposed to improve reproducibility and connect cellular maps with independent physiological measurements and perturbation experiments.