The successful development of effective viral vaccines depends on well-known correlates of protection, high immunogenicity, acceptable safety criteria, low reactogenicity, and well-designed immune monitoring and serology. Virus-neutralizing antibodies are often a good correlate of protective immunity, and their serum concentration is a key parameter during the pre-clinical and clinical testing of vaccine candidates. Viruses are inherently infectious and potentially harmful, but we and others developed replication-defective SARS-CoV-2 virus-like-particles (VLPs) as surrogates for infection to quantitate neutralizing antibodies with appropriate target cells using a split enzyme-based approach. Here, we show that SARS-CoV-2 and Epstein-Barr virus (EBV)-derived VLPs associate and fuse with extracellular vesicles in a highly specific manner, mediated by the respective viral fusion proteins and their corresponding host receptors. We highlight the capacity of virus-neutralizing antibodies to interfere with this interaction and demonstrate a potent application using this technology. To overcome the common limitations of most virus neutralization tests, we developed a quick in vitro diagnostic assay based on the fusion of SARS-CoV-2 VLPs with susceptible vesicles to quantitate neutralizing antibodies without the need for infectious viruses or living cells. We validated this method by testing a set of COVID-19 patient serum samples, correlated the results with those of a conventional test, and found good sensitivity and specificity. Furthermore, we demonstrate that this serological assay can be adapted to a human herpesvirus, EBV, and possibly other enveloped viruses.
GrantsBundesministerium fuer Bildung und Forschung Deutsche Krebshilfe NIH Deutsche Forschungsgemeinschaft Initiative and Networking Fund of the Helmholtz Association of German Research Centres (HGF) under the CORAERO project Free State of Bavaria under the Excellence Strategy of the Federal Government Federal Ministry of Education and Research (BMBF) LMUexcellent