Cholesterol affinity recognition for extracellular vesicle isolation and metabolomics analysis in liver disease

Extracellular vesicles (EVs), abundant in bodily fluids such as blood, urine, and saliva, have emerged as promising biomarker sources for disease diagnosis due to their rich content of metabolites, proteins, and nucleic acids. Efficient isolation of EVs remains a significant challenge. Currently, commonly used EV enrichment methods include ultracentrifugation (UC), size exclusion chromatography (SEC), antibody-based affinity methods, and polymer precipitation. These methods face limitations such as high costs, complex operations, and contamination risks. Consequently, there is an urgent need for the development of rapid and efficient methods for EV isolation.

In this study, we developed a strategy for synthesizing cholesterol-imprinted polymers via a one-pot reverse microemulsion approach, enabling rapid and high-yield enrichment of urinary EVs. Notably, this approach requires only a minimal amount of sample (4 mg) to achieve rapid separation (50 min) with high recovery rate (79 %), demonstrating significant improvements in both efficiency and practicality compared to conventional techniques. Ultra-performance liquid chromatography-ion mobility-mass spectrometry, we achieved high-resolution separation and detection of EV metabolites. The analytical system demonstrated excellent sensitivity, with a signal-to-noise ratio >2000:1 for 1 pg of reserpine, and high mass accuracy (resolution ≥80,000 FWHM). Metabolomic analysis of EVs from clinical samples (9 Cirrhosis, 12 Liver cancer, and 10 Healthy controls) successfully differentiated the groups. Receiver operating characteristic (ROC) analysis of four representative metabolites yielded area under the curve (AUC) values greater than 0.7, indicating promising diagnostic potential.

In conclusion, the cholesterol-imprinted polymer developed in this study enables high-yield enrichment and isolation of EVs from 2 mL of urine within 50 min. Metabolomic analysis of the enriched urinary EVs from clinical samples successfully classified and predicted different pathological conditions. This approach provides an efficient strategy for urinary EV enrichment and offers a promising platform for broader liquid biopsy applications.