Vesicle-Surface Recognition Coupled with Graphdiyne-zyme Arrays for Marker-Independent Colorimetric Fingerprinting of Bacterial Extracellular Vesicles

Abstract Bacterial extracellular vesicles (bEVs) that encode disease-relevant molecular information are released into body fluids at early stages of infection, rendering them promising biomarkers for early diagnosis. However, the lack of universal surface markers causes limitation of conventional antibody-based strategies, severely restricting the use of bEVs for early diagnosis. Here, we report a marker-independent recognition strategy that translates the interfacial heterogeneity of bEV membranes into amplified catalytic signatures. Graphdiyne-zyme sensor arrays are designed and constructed, in which the graphdiyne spatially fixes and organizes three kinds of metal oxides to form multiple nanozymes with confined, multivalent catalytic interfaces. By coupling vesicle-surface recognition with graphdiyne-zyme arrays, the interfaces efficiently transform subtle membrane variations into distinct colorimetric fingerprints. The graphdiyne-zyme array generated distinguishable colorimetric fingerprints for bEVs from six UTI-associated bacterial species and, when integrated with an XGBoost classifier, achieved an accuracy of 96.9% in a clinical test set comprising 32 samples. This study establishes an analytical framework for the marker-independent recognition of bEVs, delivering an alternative approach to replace conventional affinity-based biosensing strategies.