MMIP-assisted selective molecular depletion of icaritin-type aglycones for causality-guided metabolite identification in fermented Epimedium

A magnetic icaritin-imprinted polymer was developed for selective aglycone depletion. • Time-resolved LC–MS/MS profiling revealed formation of icaritin-type aglycones during fermentation. • The imprinted polymer enabled controlled removal of icaritin from fermented matrices. • Selective aglycone depletion attenuated cartilage-promoting activity in zebrafish. • An analytical depletion strategy was established for metabolite–activity causality analysis. Establishing causal relationships between individual metabolites and biological activity in complex natural product systems remains a significant analytical challenge, as conventional chromatographic profiling is largely limited to correlation-based interpretation. In this study, a chromatography-guided analytical intervention strategy was developed based on a magnetic molecularly imprinted polymer (MMIP) to enable selective and quantitative manipulation of target metabolites in fermented Epimedium flavonoids. Time-resolved UPLC–MS/MS profiling revealed rapid microbial deglycosylation of prenylated flavonoid glycosides, accompanied by the accumulation of icaritin-type aglycones, particularly icaritin and desmethylicaritin. An icaritin-imprinted MMIP was synthesized and systematically characterized, exhibiting high binding capacity (58.14 mg g⁻¹), favorable adsorption kinetics, and selective recognition toward target aglycones. Under optimized conditions, the MMIP enabled controlled depletion of icaritin from fermented matrices, and residual levels were accurately quantified by a validated HPLC–UV method, achieving a removal efficiency of 84.5%. Selective removal of aglycones resulted in a pronounced attenuation of cartilage-promoting effects in zebrafish models, and the magnitude of functional loss was quantitatively correlated with the degree of depletion, providing direct evidence of metabolite–activity causality. Compared with conventional fractionation approaches, the proposed strategy enables selective chromatographic modulation of target analytes with minimal disturbance to the surrounding chemical matrix, extending chromatographic analysis from passive profiling to active molecular intervention and providing a generalizable framework for causality-oriented metabolite identification in complex systems.