Dear Editor, Serological assays such as enzyme-linked immunosorbent assays (ELISA) using SARS-CoV-2 Spike (S) proteins are practical methods to determine the extent of COVID-19 immunity in a population upon SARS-CoV-2 vaccination 1. However, the robustness of these assays, which depends on the sensitivity and specificity of S proteins in detecting anti-SARS-CoV-2 antibodies, is not well characterized. Here, we report that the S1 subunit of the SARS-CoV-2 S protein has superiority over the receptorbinding domain (RBD) and the native state S trimer in terms of sensitivity and specificity, respectively, in measuring anti-SARS-CoV-2 antibodies from COVID-19 convalescent patients. S1 and S trimer are more sensitive than RBD antigen because they are able to capture non-RBD, as well as RBD-binding COVID-19 antibodies. However, the full-length S trimer, which harbors S2 subunit in addition to S1, cross-reacted with antibodies elicited by circulating coronavirus (CoV), such as HCoV-OC43 and HCoV-HKU1, making it less specific than S1 in detecting COVID-19 antibodies. Our results show that the S1 subunit protein of SARS-CoV-2 is both sensitive and specific in distinguishing COVID-19 antibodies, suggesting its potential implication in large-scale serological testing, especially in determining the efficacy of SARS-CoV-2 vaccine.

CoV infection has been known to occur in humans. Some human CoV strains that are widely circulating, such as HCoV-HKU1, HCoV-OC43, HCoV-NL63, and HCoV-229E caused mild respiratory disease, while other endemic strains, such as SARS-CoV and MERS-CoV caused a more severe disease 2. The recent outbreak of SARS-CoV-2, a virus from the Betacoronavirus family that has high homology to SARS-CoV, caused an unprecedented CoV-related pandemic, and posed major threat to global health and stability 3, 4. SARS-CoV-2 infection, like other viral infections, triggers adaptive immune response, including generation of neutralizing antibodies 5. In response to the pandemic, there is a pressing need for extensive serological testing to estimate the levels of COVID-19 antibodies in high-risk communities and to evaluate the effectiveness of SARS-CoV-2 vaccine in generating neutralizing antibodies. Hence, identification of a robust antigen and serological method is important to facilitate large-scale screening of COVID-19 antibodies. The SARS-CoV-2 S protein is a good candidate for antigen as many COVID-19 antibodies have been found to target the S protein 6–10. In this regard, neutralizing antibodies against SARS-CoV-2 or the closely related SARS-CoV mainly target the RBD on the S1 subunit of the S protein (Fig. 1 a), a region that engages the receptor, ACE2, on the host cell 11. The full-length S protein, which consists of the S1 and S2 subunits, normally exists in a trimeric state with one of the three RBDs being in an accessible conformation 12, 13. Our aim is to determine which components or forms of the S protein is more sensitive and specific in ELISA assays to detect and differentiate anti-SARS-CoV-2-specific antibodies from antibodies elicited by the widely circulating CoVs.