key: cord-0789250-q5tswda1
authors: Huey, Leah; Andersen, Gillian; Merkel, Patricia A.; Morrison, Thomas E.; McCarthy, Mary; DomBourian, Melkon G.; Annen, Kyle; Dawson, Erica D.; Rowlen, Kathy L.; Knight, Vijaya
title: Evaluation of a multiplexed coronavirus antigen array for detection of SARS-CoV-2 specific IgG in COVID-19 convalescent plasma
date: 2021-07-22
journal: J Immunol Methods
DOI: 10.1016/j.jim.2021.113104
sha: 90e159b0ca711878b30a055656e8517412a383a7
doc_id: 789250
cord_uid: q5tswda1

Mitigation of the COVID-19 pandemic requires an understanding of the antibody response to SARS-CoV-2. However, throughout the development of SARS-CoV-2 IgG antibody assays during the past year, cross-reactivity to other coronaviruses remained a question. To address these issues, we evaluated IgG in COVID-19 convalescent plasma samples for reactivity against three SARS-CoV-2 antigens including full-length spike, receptor binding domain, and the proximal extracellular fusion domain, and spike antigens from other coronaviruses (SARS-CoV, MERS-CoV, hCoV-HKU1, hCoV-OC43, hCoV-229E and hCoV-NL63) using the VaxArray Coronavirus SeroAssay which is a multiplexed antigen assay developed by InDevR Inc. These results were compared to two commercial SARS-CoV-2 IgG ELISAs targeting either the SARS-CoV-2 nucleocapsid or spike antigens and a live virus focus reduction neutralizing antibody test (FRNT). The VaxArray platform showed high specificity for detection of SARS-CoV-2 IgG, evident from lack of reactivity to SARS-CoV-2 antigens despite significant reactivity to endemic coronavirus antigens in pre-pandemic samples. SARS-CoV-2 IgG positive samples reacted weakly to SARS-CoV spike but not to MERS-CoV. While the VaxArray platform had overall comparable results to the spike and nucleocapsid IgG ELISAs, results were more similar to the spike antigen ELISA and the platform displayed a higher sensitivity and specificity than both ELISAs. Samples with FRNT titers below 1/23 reported negative on VaxArray, while positive samples on VaxArray had significantly higher neutralizing antibody titers. These results suggest that the VaxArray Coronavirus SeroAssay performs with high sensitivity and specificity for the detection of SARS-CoV-2 IgG, and positive results on the platform indicate SARS-CoV-2 neutralizing activity.

A plethora of antibody assays have been developed in response to the novel Coronavirus Disease-19 (COVID-19) pandemic that continues to spread across the globe. Analysis of SARS-CoV-2 antibodies has utility for understanding the prevalence of infection within communities [1] [2] [3] [4] , providing evidence of past infection or recent infection when viral detection by PCR is negative 5, 6 , and for the analysis of convalescent plasma that has been used therapeutically to treat COVID-19 7,8 .

Additionally, with the implementation of COVID-19 vaccination, analysis of the antibody response to these vaccines is necessary in order to understand both the short-term immune response and the longevity of the immune response.

Initially, the development of antibody assays was fraught with concerns of crossreactivity to pre-existing antibodies to the non-SARS-CoV-2 human coronaviruses (hCoVs) hCoV-HKU1, hCoV-OC43, hCoV-229E and hCoV-NL63 and therefore, lower specificity of antibody detection. Over the course of the past several months, antibody assays with high specificity and sensitivity (depending on the timing of sample collection post infection) have been developed and implemented in clinical laboratories 9, 10 .

However, few have incorporated antigens from related hCoVs to address the question of antibody specificity.

The VaxArray Coronavirus SeroAssay incorporates the S1 domain from the spike (S) protein of SARS-CoV, MERS-COV, and either the S1 domain or the full-length S protein of four human endemic CoVs, hCoV-HKU1, hCoV-OC43, hCoV-229E and hCoV-NL63 along with the full-length S protein, the receptor binding domain (RBD), and the proximal extracellular fusion domain (S2) of the SARS-CoV-2 S antigen in a single assay 11 . Each of these antigens is spotted, in nine replicates, onto a functionalized glass slide in a 9x9 grid. When incubated with human serum followed by anti-human Table 1 ). The manufacturer's published characteristics for the assay include 98.5% sensitivity, 100% specificity and an average precision of 11% CV 11 . Plasma and serum samples, including human pooled AB serum as a negative control, were diluted 1:100 in kit-provided protein blocking buffer and added to the microarray slide. The slide was then incubated for an hour in a humidity chamber at room temperature on an orbital shaker set to 80 rpm. Following incubation, the slide was washed with kit specific wash buffer and incubated under the same conditions as previously stated for five minutes. After removal of the wash buffer, slides were incubated for thirty minutes with detection solution containing fluorescently labeled IgG antibody under the same conditions as previously stated. Slides were then washed with kit specific wash buffer, followed by 70% ethanol and deionized water. Once dry, slides were analyzed with the VaxArray Imaging System (InDevR). For each CoV antigen the median relative fluorescent unit of the nine replicates was divided by the median background signal to give a signal to background ratio that was used for the analysis of the SeroAssay. were reported as borderline.

signal to background ratio for nCoV(i), nCoV(ii), and nCoV(iii) was greater than 6.18. If either criterion was not met, then the sample was considered negative.

The FRNT 50 titer was calculated relative to a virus only control (no serum) set at 100%, using GraphPad Prism 9 (San Diego, CA, USA) default nonlinear curve fit constrained between 0 and 100%. (Figure 2A and 2E) . Weak reactivity to SARS-CoV, as defined by a signal to background ratio greater than 2.00 but less than 10.00, was detected in thirteen PCR positive samples and no reactivity was seen with MERS-COV S protein ( Figure 2B ). In contrast, while all but one of the thirty presumed SARS-CoV-2 negative serum samples had antibodies to one or more of the endemic CoVs, none of these samples were positive for SARS-CoV-2 full-length S (nCoV(i)) and only one sample was weakly positive for RBD (nCoV(ii)) antibodies ( Figure 2C and 2D) . Weak reactivity to the extracellular S2 antigen, nCoV(iii), was detected in seven of the samples ( Figure 2C ).

Despite weak reactivity to S2 or RBD, none of the thirty presumed SARS-CoV-2 negative samples fulfilled criteria for a positive signal for SARS-CoV-2 antibodies.

Notably, all of the SARS-CoV-2 PCR positive and all but one of the presumed negative samples had varying levels of reactivity to the four endemic hCoVs ( Figure 2B and 2D). 

Qualitative results of thirty randomly selected SARS-CoV-2 PCR-positive samples analyzed by EDI, EUROIMMUN, and VaxArray assays were compared to the reciprocal FRNT50 titer (Figure 3 ). For each assay the negative cut-off for the Seven of the SARS-CoV-2 PCR positive samples tested negative for SARS-CoV-2 IgG on the VaxArray Coronavirus SeroAssay. As limited clinical information was available for these individuals, it is possible that their initial viral load may have been low, therefore leading to a suboptimal antibody response 19 , or that perhaps the PCR result was a false positive. All seven samples were negative on the EDI assay and out of the five samples tested on the EUROIMMUN assay, four were negative and one was borderline. The neutralizing titer of the three negative samples tested for neutralizing antibodies was <1/23, demonstrating very little antibody neutralizing activity.

When compared with the EDI and EUROIMMUN assays, the performance of VaxArray was more sensitive than both assays, was superior to EDI in specificity, and was similar in specificity to EUROIMMUN, at least for this sample set. The similarity with the EUROIMMUN ELISA is not surprising, given that both assays utilize the S antigen. The indicate that while the nCoV(i) signal was greater than 1.5, the sum of the three antigens was less than 6.18 so the sample was categorized as negative.

Representative slides of the SARS-CoV-2 PCR positive samples that were negative on the VaxArray platform (E). 

Each antigen is coated on a separate location of the multiplexed slide arranged in a 9x9 grid. Within this grid each antigen is spotted on to the slide in a 3x3 grid. S represents the full spike protein; S1, the first subunit of the spike protein; S2, the second subunit including the extracellular domain (ECD); and RBD, the receptor binding domain. Cut-off values were kit specific and borderline results were excluded J o u r n a l P r e -p r o o f

Estimation of US SARS-CoV-2 Infections, Symptomatic Infections, Hospitalizations, and Deaths Using Seroprevalence Surveys

Estimated SARS-CoV-2 Seroprevalence in the US as of

Assessing the extent of SARS-CoV-2 circulation through serological studies

Covid-19: US cases are greatly underestimated, seroprevalence studies suggest

SARS-CoV-2 antibody responses in children with MIS-C and mild and severe COVID-19. medRxiv

False negative RT-PCR and false positive antibody tests

Treatment of 5 Critically Ill Patients With COVID-19 With Convalescent Plasma

COVID-19 Antibody Tests Perform Well in Head-to-Head Comparison

Performance characteristics of five immunoassays for SARS-CoV-2: a head-to-head benchmark comparison

Multiplexed, microscale, microarray-based serological assay for antibodies against all human-relevant coronaviruses

Presence and short-term persistence of SARS-CoV-2 neutralizing antibodies in COVID-19 convalescent plasma donors. Transfusion

Severe Acute Respiratory Syndrome Coronavirus 2 from Patient with Coronavirus Disease, United States

Phylogenetic Analysis and Structural Modeling of SARS-CoV-2 Spike Protein Reveals an Evolutionary Distinct and Proteolytically Sensitive Activation Loop

Emerging coronaviruses: Genome structure, replication, and pathogenesis

Serologic cross-reactivity of SARS-CoV-2 with endemic and seasonal Betacoronaviruses. medRxiv