DS-1150b

Selecting a suitable non-mouse animal model is a structured decision based on the research question, required anatomy or physiology, disease mechanism, endpoint feasibility, translational relevance, and ethical justification. Non-mouse models are preferred when mice cannot reproduce key human-relevant features, such as organ size, surgical anatomy, cardiovascular physiology, neuroanatomy, immune features, pharmacology, toxicology, or long-term clinical procedures. Candidate species may include rats, rabbits, guinea pigs, ferrets, zebrafish, pigs, sheep, goats, dogs, cats, horses, and non-human primates, but each species must be justified by its specific scientific advantage rather than convenience or tradition. Unresolved questions include how to quantify translational superiority across species, how to balance increased biological relevance against higher ethical burden, and when human-derived systems or new approach methodologies should replace animal use.

AMP-activated protein kinase, AMPK, is a conserved cellular energy sensor that responds to reduced cellular energy status and coordinates metabolism by increasing ATP-generating catabolic pathways while suppressing ATP-consuming anabolic processes. In metabolic disease research, the AMPK pathway is experimentally relevant because it regulates hepatic lipid synthesis, fatty acid oxidation, glucose production, skeletal-muscle glucose disposal, mTORC1-linked biosynthesis, autophagy, mitochondrial homeostasis, and whole-body energy balance. The central pathway logic is that energy stress, metformin, exercise-like stimulation, or direct AMPK activators increase AMPKα Thr172 phosphorylation and downstream substrate phosphorylation, including ACC and RAPTOR. Phosphorylation of ACC suppresses lipogenesis and supports fatty acid oxidation, whereas phosphorylation of RAPTOR suppresses mTORC1 signaling and links cellular energy status to growth and protein synthesis control. The pathway is linked

Choosing the right model animal is a validity-driven decision in which the species, strain, sex, age, genetic background, disease-induction method, outcome measures, and welfare burden must match the scientific question rather than laboratory tradition or convenience. A model should be selected by judging face validity, construct validity, and predictive validity: whether it resembles the human phenotype, whether it reproduces relevant mechanisms, and whether results are likely to predict human biology or treatment response. Animal studies often fail to translate because of species differences, weak disease resemblance, poor experimental design, inadequate reporting, publication bias, and underuse of randomization, blinding, and sample-size justification. Unresolved questions include how to rank competing models objectively, how much human-disease complexity must be reproduced for a given objective, and when non-animal systems such as organoids, ex vivo tissue, or computational models