Development and validation of a cross-species breath metabolomics platform for translational VOC analysis

Abstract

The lack of universal cross-species sampling and quality control methods has limited the potential of breath metabolomics to advance from clinical discovery to mechanistic validation. To address this challenge, this study developed and systematically validated an integrated cross-species breath analysis platform, with its core components comprising a high-sensitivity mouse breath sampling system (FaunaScope) and a quality control (QC) strategy incorporating behavioral monitoring. By identifying ethyl acetate and dimethyl sulfide as characteristic interference markers, the platform exhibited high detection capability and good analytical reproducibility in the analysis of 33 core volatile organic compounds (VOCs) in mouse breath, with detection rates exceeding 88.2% for 30 VOCs and coefficients of variation below 30% for more than 70% of the compounds. Using inflammatory bowel disease (IBD) as a demonstration model, the platform enabled a full-chain study from validation of breath fingerprints in clinical patients to longitudinal monitoring in mouse models, successfully capturing dynamic metabolic changes in short-chain fatty acids (SCFAs) and responses to exclusive enteral nutrition (EEN) intervention, while revealing metabolic feature divergence between humans and mice attributable to differences in pathological mechanisms. Overall, this platform exhibits high sensitivity and strong resistance to interference, providing an effective translational medicine tool for linking clinical findings with fundamental mechanistic research.