A comparison of four serological assays for detecting anti–SARS-CoV-2 antibodies in human serum samples from different populations

It is of paramount importance to evaluate the prevalence of both asymptomatic and symptomatic cases of SARS-CoV-2 infection and their differing antibody response profiles. Here, we performed a pilot study of four serological assays to assess the amounts of anti–SARS-CoV-2 antibodies in serum samples obtained from 491 healthy individuals before the SARS-CoV-2 pandemic, 51 individuals hospitalized with COVID-19, 209 suspected cases of COVID-19 with mild symptoms, and 200 healthy blood donors. We used two ELISA assays that recognized the full-length nucleoprotein (N) or trimeric spike (S) protein ectodomain of SARS-CoV-2. In addition, we developed the S-Flow assay that recognized the S protein expressed at the cell surface using flow cytometry, and the luciferase immunoprecipitation system (LIPS) assay that recognized diverse SARS-CoV-2 antigens including the S1 domain and the carboxyl-terminal domain of N by immunoprecipitation. We obtained similar results with the four serological assays. Differences in sensitivity were attributed to the technique and the antigen used. High anti–SARS-CoV-2 antibody titers were associated with neutralization activity, which was assessed using infectious SARS-CoV-2 or lentiviral-S pseudotype virus. In hospitalized patients with COVID-19, seroconversion and virus neutralization occurred between 5 and 14 days after symptom onset, confirming previous studies. Seropositivity was detected in 32% of mildly symptomatic individuals within 15 days of symptom onset and in 3% of healthy blood donors. The four antibody assays that we used enabled a broad evaluation of SARS-CoV-2 seroprevalence and antibody profiling in different subpopulations within one region.