Rebecca Goodrum, Huiyan Li
Abstract Extracellular vesicles (EVs) are membrane bound nanoscale particles released by cells that contain molecular cargo reflective of their parental cell and can be found in many biofluids. The overexpression of EVs and EV-related protein markers has been linked to various diseased states, making them a promising tool for liquid biopsy-based disease diagnostics. Many complex diseases, like cancer, impact multiple markers simultaneously, and during early stages, are present at low concentrations. Current EV analysis technology is limited in sensitivity, multiplexing, and ease of use. We have developed a silver nanoparticle embedded membrane (sNEM) platform that utilizes the 3D structure of nitrocellulose membrane, metal-enhanced fluorescence (MEF)-based detection and a novel wax-based compartmentalization technique for highly sensitive multiplex EV protein detection from minimal sample volume. We compared various nanoparticle shapes, sizes, and metal types with fluorophores of different wavelengths to determine which provided optimal MEF-based detection with high sensitivity. Fluorescence intensity from FITC was much lower than that from Cy5 and was found to pronounce the effects of autofluorescence by 2 times. After selecting 30 nm silver nanoparticles at a concentration of 10 9 particles ml −1 and the Cy5 fluorophore based on greatest fluorescence enhancement, we then demonstrated its application for multiplexed detection of surface and intravesicular proteins directly from lysed EVs in both buffer and human plasma. In PBS, detection limits of 2–3 orders of magnitude lower than traditional ELISA were achieved. Directly from human plasma, detection limits of 1.97 × 10 5 EVs ml −1 , 1.94 × 10 6 EVs ml −1 , and 2.17 × 10 4 EVs ml −1 for TGF- β 1, AKT1, and TSG101 were achieved. These results demonstrate the suitability of sNEM for highly sensitive, multiplexed detection of EV markers from complex biofluids for early diagnostics while offering advantages such as low reagent/sample consumption, scalability, reduced sample preparation, and ease of use.