Improved Label-Free Identification of Individual Exosome-like Vesicles with Au@Ag Nanoparticles as SERS Substrate

摘要

Exosome-like vesicles (ELVs) are nanovectors released by cells that are endowed with a variety of molecules, including proteins, nucleic acids, and chemicals that reflect the molecular signature of the producing cell. Given their presence in many biofluids, they form an easily accessible biomarker for early disease detection. Previously we demonstrated the possibility of identifying individual ELVs by analyzing their molecular signatures with surface-enhanced Raman scattering (SERS) after functionalization of ELVs with 4-(dimethylamino)pyridine (DMAP)-stabilized gold nanoparticles (AuNP). Although this strategy was capable of distinguishing ELVs from different cellular origins, the quality of the SERS spectra was suboptimal due to high background coming from the DMAP stabilizing molecules at the AuNP surface. In this study we demonstrate that it is possible to eliminate interfering SERS signals from stabilizing molecules at the AuNP surface by overgrowing in situ the ELV-attached AuNPs with a silver layer so as to form a core–shell nanoparticle ([email?protected]) directly at the ELV surface. As such it represents the first known strategy to generate clear SERS spectral fingerprints of delicate biological structures without interference of linker molecules that are needed to ensure colloidal stability of the plasmonic NP and to allow them to associate to the ELV surface. This new strategy using core–shell plasmonic NPs as SERS substrate showed higher near-field enhancements than previous approaches, which resulted in SERS spectra with improved signal-to-noise ratio. This allowed us to discriminate individual vesicles derived from B16F10 melanoma cells and red blood cells (RBC) with an unprecedented sensitivity and specificity >90%. Importantly, thanks to the higher near field enhancement the acquisition time could be reduced by 20-fold in comparison to previously reported strategies, paving the way toward high-throughput label-free single ELV identification.