Abstract
Rhodiola species are traditionally used to mitigate hypoxia-related symptoms, but comparative evidence on their chemical bases and active constituents is limited. We implemented an integrated "qualitative analysis-pharmacological exploration-correlation analysis-molecular validation" workflow to compare Rhodiola crenulata, R. kirilowii, and R. rosea. Ultra-high-performance liquid chromatography-Q Exactive mass spectrometry (UPLC-QE-MS) profiling identified 175 metabolites across the three species, of which 161 were shared; multivariate analyses (principal component analysis, PCA; partial least squares-discriminant analysis, PLS-DA) revealed 30 differential compounds. In a normobaric hypoxia mouse model using herbal powder solutions, all three species significantly increased survival time versus control (p < 0.05), with mean survival times of 48.16 min (RR), 47.07 min (RC), and 44.82 min (RK) compared with 44.34 min for the positive control. Chemometric correlation (partial least squares regression, PLSR) combined with grey relational analysis (GRA) prioritized 14 compounds consistently associated with anti-hypoxia efficacy; six representative metabolites-epicatechin, 3-O-galloylquinic acid, salidroside, p-coumaric acid-4-O-glucoside, citric acid, and geraniol-were selected for in silico assessment. Molecular docking against hypoxia-inducible factor-1α (HIF-1α) yielded favorable binding poses (docking scores < -4.0), providing preliminary molecular-level plausibility without claiming mechanistic proof. This multi-level approach clarifies chemical-pharmacological relationships among Rhodiola species and provides prioritized candidate compounds for targeted isolation and mechanistic validation.