PREDICTOR: A Non-Enzymatic Catalytic Cascade Tool for in Situ Visualization of Small Extracellular Vesicle Surface glycoRNAs

Glycosylated RNAs (glycoRNAs) are membrane-displayed RNA-glycan conjugates, but quantitative in situ analysis of low-abundance glycoRNAs on small extracellular vesicles (sEVs) remains challenging. Here, we develop PREDICTOR (proximity-encoded non-linear hybridization chain reaction circuit), an enzyme-free catalytic DNA cascade for imaging and quantifying surface glycoRNAs on intact sEVs. PREDICTOR uses a sialic acid aptamer and an RNA-sequence probe as split recognition modules, whose proximity on a single glycoRNA reconstitutes an initiator that drives dendritic DNA self-assembly and non-linear fluorescence amplification on the sEV surface. Across serial sEV inputs (5×10³-5×10⁹ particles·mL^{- 1}), PREDICTOR exhibited a steeper concentration-response than representative amplification assays, with higher regression slopes in bulk fluorescence (0.372, R² = 0.9846) compared with ARPLA (0.301, R² = 0.9748) and HieCo 2 (0.230, R² = 0.9663), and similarly improved performance in bead-based imaging (4.105, R² = 0.9816 vs. 3.495 and 2.711). Applying PREDICTOR to a breast Cancer transformation model (MCF-10A → MCF-7 → MDA-MB-231) showed that multiple surface glycoRNA candidates decrease with malignancy, accompanied by progressive softening of sEV membranes (Young's modulus reduced by 25.08% and 54.72%). Functionally, enzymatic depletion of surface RNA and disruption of sialylated N-glycan features attenuated macrophage uptake and inflammatory activation, supporting a contribution of surface glycoRNAs to sEV recognition. Collectively, PREDICTOR provides a rapid, enzyme-free platform for quantitative and spatial profiling of sEV surface glycoRNAs and links their abundance to vesicle mechanics and immune-cell interactions.