Orthogonal DNA barcoding enables subpopulation-resolved extracellular vesicle miRNA profiling

Extracellular vesicle-associated microRNAs (EV-miRNAs) hold promise for liquid biopsy, but their clinical utility is hindered by EV heterogeneity. Here, we report DEEPER, a DNA-encoded orthogonal recognition platform for selective barcoding of tumor-derived EV subpopulations and in situ miRNA profiling. DEEPER uses four aptamers targeting Cluster of Differentiation 44 (CD44), Epidermal Growth Factor Receptor‌ (EGFR), Human Epidermal Growth Factor Receptor 2‌ (HER2), and ‌Programmed Death-Ligand 1(PD-L1) to generate orthogonal barcodes that distinguish 15 EV subpopulations. Barcoded EVs then undergo targeted membrane fusion with liposome probes, enabling sensitive in situ analysis of subpopulation-specific miRNAs. In a 60-sample cohort of patients with gastric cancer (GC) and healthy donors (HDs), DEEPER identified a high-performing signature, GC-EV miRNA Prime, comprising EGFR⁺, EGFR⁺HER2⁺, and EGFR⁺HER2⁺CD44⁺ EVs, which achieved 98.3% diagnostic accuracy. The same profiling framework also assessed tumor invasiveness with 95.5% accuracy. Mechanistic studies further showed that these EV-miRNAs contribute to tumor progression. In summary, DEEPER not only confirmed the clinical application value of EV-miRNA-based liquid biopsy but also provided unique insights through the EV subpopulations into the mechanisms driving cancer development.