key: cord-0777661-6cjnucx2
authors: Gilboa, Tal; Cohen, Limor; Cheng, Chi‐An; Lazarovits, Roey; Uwamanzu‐Nna, Augusta; Han, Isaac; Griswold, Kettner; Barry, Nick; Thompson, David B.; Kohman, Richie E.; Woolley, Ann E.; Karlson, Elizabeth W.; Walt, David R.
title: A SARS‐CoV‐2 Neutralization Assay Using Single Molecule Arrays
date: 2021-11-05
journal: Angew Chem Int Ed Engl
DOI: 10.1002/anie.202110702
sha: dc8179e424145428475319cd43aa73ebeedd427d
doc_id: 777661
cord_uid: 6cjnucx2

Coronavirus disease 2019 (COVID‐19) manifests with high clinical variability and warrants sensitive and specific assays to analyze immune responses in infected and vaccinated individuals. Using Single Molecule Arrays (Simoa), we developed an assay to assess antibody neutralization with high sensitivity and multiplexing capabilities based on antibody‐mediated blockage of the ACE2‐spike interaction. The assay does not require live viruses or cells and can be performed in a biosafety level 2 laboratory within two hours. We used this assay to assess neutralization and antibody levels in patients who died of COVID‐19 and patients hospitalized for a short period of time and show that neutralization and antibody levels increase over time. We also adapted the assay for SARS‐CoV‐2 variants and measured neutralization capacity in pre‐pandemic healthy, COVID‐19 infected, and vaccinated individuals. This assay is highly adaptable for clinical applications, such as vaccine development and epidemiological studies.

Thec oronavirus disease 2019 (COVID- 19) , which is caused by the novel severe acute respiratory syndrome coronavirus 2( SARS-CoV-2), has led to high morbidity and mortality globally.S erological assays for detecting SARS-CoV-2 antibodies are of critical importance for several reasons.F irst, serological assays provide abetter understand-ing of the virus-induced host immune response. [1, 2] Second, screening convalescent plasma for immunoglobulin content can improve the efficacyofantibody therapy. [3] Third, there is an urgent need to evaluate novel vaccines by detecting the levels and types of immunoglobulins produced in response to different vaccines,d oses,a nd schedules. [4] Finally,s erological assays can identify asymptomatic individuals with previous exposure to SARS-CoV-2.

There are currently two main serological assay formats. Thefirst format is based on the enzyme-linked immunosorbent assay (ELISA). ELISA can provide relative quantification of antibody levels and can be implemented at the point-of care;however,itcannot distinguish between neutralizing and non-neutralizing antibodies,l imiting its utility.T he second format is the live-virus neutralization assay. [5] [6] [7] This assay detects neutralizing antibodies in abiological sample,such as blood, and is currently the gold standard method for determining antibody efficacy.H owever,i tr equires use of infectious SARS-CoV-2 virus particles,w hich pose safety considerations and require handling in abiosafety level (BSL) 3lab.Asaresult, there have been efforts to develop simpler antibody neutralization assays. [8] Fore xample,t he pseudovirus neutralization assay overcomes the need for live virus and can be conducted in aB SL-2 lab. [9, 10] However,t he pseudovirus neutralization assay still has several challenges. [9] First, generating pseudoviruses is ac omplex process and requires many steps including packaging,p urifying,a nd titrating the recombinant virus.Second, the assay requires use of live cells, which adds additional complexity and reduces assay reproducibility.F inally,d espite being aB SL-2 assay,u sing al ive cell that has been engineered to be sensitive to infection has biosafety implications.T he presence of as ingle replicationcompetent viral particle in the sample can cause amplification of the virus.Having apurely protein-based assay circumvents these challenges.

In this work, we developed aSARS-CoV-2 neutralization assay based on antibody-mediated blockage of the ACE2spike interaction without the need for live cells or viruses using Single Molecule Arrays (Simoa). Previously,w eh ave developed competitive Simoa assays,w hich are % 50 fold more sensitive than ac onventional competitive ELISA. [11] Here,weshow that acompetitive Simoa assay can be adapted and used to assess antibody neutralization in blood. This assay has high sensitivity and multiplexing capabilities.Weused this assay for two applications.F irst, we measured the neutralization capacity and antibody levels in hospitalized COVID-19 patients.W es how that there is an overall increase in neutralization capacity and antibody levels over time.Second, we developed amultiplex Simoa neutralization assay against the SARS-CoV-2 receptor binding domain and three variants, and measured the neutralization capacity in pre-pandemic healthy,C OVID-19 positive,and vaccinated individuals.T he Simoa-based SARS-CoV-2 neutralization assay presented here is ac ell-free,q uick, and highly sensitive alternative to live-virus neutralization assays and can be used for av ariety of clinical applications.The assay is versatile and can be easily adapted towards future detection of antibody neutralization for other infectious diseases.

Development of an Antibody Neutralization Assay Using Single Molecule Arrays (Simoa)

We developed ac ompetitive bead-based antibody neutralization assay using Single Molecule Arrays (Simoa). [12] In this assay format (Figure 1 ), SARS-CoV-2 spike protein is first conjugated to micron-sized beads,c reating as pikedecorated surface that resembles the virus.Then, biotinylated ACE2 and as ample containing antibodies against SARS-CoV-2 spike are incubated with the beads,f ollowed by addition of streptavidin coated beta-galactosidase,which can bind to the biotinylated ACE2. Theb eads are then resuspended in afluorogenic substrate and loaded onto an array of femtoliter sized wells in which each well can fit only one bead. Thew ells are sealed with oil and active wells are counted using the units of average enzyme per bead (AEB). [13] In the absence of anti-spike antibodies,the biotinylated ACE2 binds to the spike-coated beads,l eading to high signal. In the presence of anti-spike antibodies,t he biotinylated ACE2 competes with the antibodies for the binding sites on the bead and thus,a st he concentration of anti-spike antibodies in the sample increases,t he signal in the assay decreases.I ti s important to note that the timing of incubations must be kept constant across different samples in this competitive immunoassay since the amount of antibody as well as antibody avidity varies between the plasma samples. [11, 14] To determine the antibody neutralization capacity (NT50), we serially diluted each patient plasma sample and then determined the dilution factor that corresponded to 50 %loss in assay signal.

To develop the Simoa assay,w ef irst prepared beads coated with the SARS-CoV-2 spike protein (Methods section and Figure S1 ). We then added increasing concentrations of ACE2 and showed that the signal also increased, confirming that ACE2 interacts with the spike-coated beads ( Figure 2A ). We also tested beads coated with the SARS-CoV-2 nucleocapsid protein ( Figure S1 ) as ac ontrol, and observed that as the concentration of ACE2 increases,t he signal of the assay does not change,a sexpected ( Figure 2B ). As another validation, we tested the assay using aneutralizing anti-spike monoclonal antibody (40592-R001) and observed that as the concentration of the neutralizing antibody increases,the assay signal decreases to background level, as expected (Figure 2C ). To assess the potential cross-reactivity of the assay, we tested antibodies against other coronaviruses,i ncluding SARS-CoV and MERS-CoV ( Figure 2C ). Even at ah igh concentration (15 mgmL À1 )ofanti SARS-CoV (40069-R723) and anti MERS-CoV (CR-3022) antibodies,there is minimal neutralization capacity and the assay signal decreases by only 15 %a nd 30 %, respectively.T hese results suggest that this assay could be used for various applications,such as compar- Figure 1 . Simoa neutralization assay.The assay is based on antibodymediated blockage of the ACE2-spike interaction. Spike-coated beads, biotinylatedACE2, and patient plasma are mixed. After several washing steps, streptavidin conjugated beta-galactosidase (SbG) is added. The beads are then washed again, resuspended in resorufin b-D-galactopyranoside (RGP), and loaded into amicrowell array for imaging. As the concentration of anti-spike antibodies in the sample increases, the signal in the assay decreases.T odetermine neutralization capacity (NT50), the plasma sample is diluted and the signal at each dilution factor is normalized to the AEB of the highest dilution factor. Figure 2 . Developmento fthe Simoa neutralization assay for SARS-CoV-2 spike and nucleocapsid. A) As the concentration of ACE2 increases using spike-coated beads, so does the signal. B) As the concentration of ACE2 increases using nucleocapsid-coated beads, the signal does not increase. C) As the concentration of aneutralizing SARS-CoV-2 antibody increases, the assay signal decreases. As the concentration of SARS-CoV and MERS-CoV antibodies increases, the assay signal decreases slightly.All measurements were obtained in duplicate.

ing and quantifying the neutralizing capacity of different therapeutic antibodies.

To assess whether the assay can be used in clinical samples,w et ested plasma from COVID-19 patients (Figure 3A )a nd pre-pandemic patients,i ncluding patients with various respiratory diseases,( Figure 3B and C) using the neutralization assay.A ll ten of the COVID-19 patients had antibodies with neutralizing capacity in their plasma (Figure 3A and Table S1 ). Forthe ten pre-pandemic controls,we could not calculate NT50 values,i ndicating that anti-spike antibodies are not present in the plasma ( Figure 3B ). We also could not calculate NT50 values in samples from patients with other documented respiratory infections prior to October 1 st , 2019 ( Figure 3C ), indicating that cross-reactivity is minimal. We note that the patients infected with other respiratory diseases have lower overall assay signals,c orresponding to higher neutralization, compared to the pre-pandemic healthy controls.P atients with different respiratory diseases likely have antibodies that minimally cross-react with antibodies against SARS-CoV-2. To evaluate the precision of our assay, we measured each sample in duplicate and calculated the coefficient of variation (CV) for each dilution factor (Figure 3D) . We found high agreement between technical replicates,w ith most CVs falling below 20 %. To assess runto-run variability,w em easured eight samples,i ncluding COVID-19 positive samples and pre-pandemic controls,i n two consecutive runs ( Figure S3 ). We observed high agreement between the NT50 values for the two measurements with low CVs (< 20 %). Overall, these results show that the Simoa neutralization assay can be used to assess antibody neutralization capacity in the plasma of COVID-19 patients with good precision.

We then sought to validate our Simoa assay using an orthogonal neutralization assay.W eo btained 28 plasma samples from patients with COVID-19 and measured the NT50 values using both Simoa and pseudovirus neutralization assays.W es how that overall, there is good correlation between our method and the pseudovirus neutralization assay across the different samples,with aPearson correlation coefficient of 0.92 ( Figure 4A ). We note that there may be additional mechanisms of neutralization that are not dependent on blocking the spike-ACE2 interaction. [15, 16] For example,a ntibodies can still bind to the spike protein and neutralize the virus by halting membrane fusion without disrupting the spike-ACE2 interaction. However,d ue to the high correlation between the assays,these other mechanisms are unlikely to be the dominant neutralization factors.W e then obtained serial samples from two different COVID-19 patients and measured antibody neutralization over time using both Simoa and pseudovirus neutralization assays and observed good agreement between the two methods (Fig-Figure 3 . Simoa neutralization assay validation using patient plasma samples. A) Plasma samples from ten COVID-19p atients were serially diluted and measured using the Simoa neutralization assay.B )Plasma samples from ten pre-pandemiccontrols were serially diluted and measured using the Simoa neutralizationa ssay.C)Plasma samples from ten pre-pandemic controls with respiratory illnesses were serially diluted and measured using the Simoa neutralization assay.All measurements were performed in duplicate. D) CVso fduplicate measurements for each of the sample dilutions presented in (A), (B), and (C). ure 4B). We also measured anti-spike IgG and IgM levels in these samples and show similar trends between antibody levels and antibody neutralization over time ( Figure 4B ).

We then evaluated the sensitivity of the Simoa neutralization assay by comparing the assay performance to two commercially available ELISA kits.Wefirst measured SARS-CoV-2 neutralizing and SARS-CoV non-neutralizing antibodies using these assays ( Figure 5A ). All three assays show good agreement, with expected signal decrease for the neutralizing antibody and minimal signal decrease for the non-neutralizing antibody.W et hen measured NT50 values for six COVID-19 patients with low neutralization capacity ( Figure 5B ). We show that the Simoa assay is substantially more sensitive than ELISA and enables quantification of neutralization capacity.This improved sensitivity is important for assessing the immune response in patients with COVID-19.

To assess the utility of the Simoa assay in clinical settings, we compared the neutralization capacity with antibody levels in 130 samples from patients who were admitted to the hospital with COVID-19. Our sample cohort consisted of patients hospitalized with COVID-19 for less than ten days who did not require treatment in the intensive care unit (ICU) and hospitalized patients who died of COVID-19. We measured the neutralization capacity and antibody levels of IgG and IgM isotypes against the receptor binding domain (RBD), S1, spike,a nd nucleocapsid, for at otal of eight antibody measurements,a spreviously described. [17] We then compared the NT50 values to the antibody levels and observed that as the antibody levels against RBD,S 1, and spike increased, so do the NT50 values ( Figure 6A ). On the other hand, antibodies against nucleocapsid show poor correlation with NT50 ( Figure S4 ). Theneutralization capacity varied by approximately three orders of magnitude,with an NT50 range of % 5t o% 3000. Thec apability of this assay to measure aw ide range of sample dilution factors,a long with the high assay sensitivity,allows us to quantify neutralization capacities throughout disease progression.

We assessed antibody levels over time in the two patient groups and observed that, overall, the group of patients who died of COVID-19 had higher antibody levels compared to the group of patients with ashorter period of hospitalization ( Figure S5 ). When we evaluated antibody neutralization over time in the two patient groups ( Figure 6B ), we observed that the antibody levels and neutralization capacities increase over the first 20 days post symptom onset, which is consistent with previous reports. [18] Our results demonstrate the utility of Simoa assays to comprehensively characterize the immune response by profiling both antibody levels and neutralization capacity for various clinical applications. Simoa has higher sensitivity than ELISA. A) Neutralizing SARS-CoV-2 and non-neutralizing SARS-CoV antibodiesw ere serially diluted and measured using Simoa and two ELISA neutralization assays. B) NT50 for six patients amples were measured using Simoa and two ELISA neutralization assays. Error bars represent the standard deviation of two measurements.

In response to the on-going emergence of SARS-CoV-2 variants,w ea dapted the Simoa assay to am ultiplex assay format to assess antibody neutralization capacities against three SARS-CoV-2 variants of concern (VOCs), as classified by the US Department of Health and Human Services. [19, 20] Thef irst VOCh as an N501Y mutation (B.1.1.7, Alpha variant), the second VOCh as K417N,E 484K, and N501Y mutations (B.1.351, Beta variant), and the third VOCh as T478K and L452R mutations (B.1.617.2, Delta variant) on the receptor binding domain (RBD) epitope.Amajor advantage of Simoa compared to traditional ELISA is the ability to multiplex and thereby measure multiple target analytes in one sample.I na ddition to interrogating more than one target, multiplexing is advantageous as it increases throughput and saves precious patient sample volume.

We first developed four singleplex Simoa neutralization assays:one assay against the wild-type (WT) RBD,and three against the variants ( Figure S6 ). We used these assays to quantify the neutralization capacity of aS ARS-CoV-2 monoclonal neutralizing antibody (40591-MM43) against each of the four RBD epitopes.A st he concentration of antibody increased, the assay signal decreased for each variant ( Figure 7A ), indicating that this antibody had neutralization potential against SARS-CoV-2 VOCs.W et hen developed afour-plex Simoa assay,which can measure all four of these targets simultaneously,and measured the neutralization capacity of the antibody.T he assay signals for the singleplex and multiplex assays showed excellent agreement for all four targets ( Figure 7A ). We also tested the singleplex and multiplex assays using asecond monoclonal neutralizing antibody (40592-R001) against each of the four variants ( Figure S7 ). Then eutralizing antibody had as imilar neutralization capacity for the Alpha and Delta variants and the WT RBD. [21] [22] [23] Then eutralizing antibody had as ubstantially lower neutralization capacity ( % 9f old) against the Beta variant, which could be due to the E484K mutation that has been shown to escape serum antibody responses. [24, 25] Next, we assessed the neutralization capacity in plasma from eight healthy pre-pandemic controls,e ight COVID-19 patients who had seroconverted, and eight vaccinated individuals (mRNA-1273 vaccine,5 6days post 1 st dose) (Figure 7B and Figure 7C ). We measured these samples using both singleplex and multiplex Simoa assays to ensure agreement between the two assay formats (Figures S8 and S9 ). As expected, no neutralization was found in the healthy pre- pandemic controls.T he COVID-19 patients had overall higher neutralization capacity against the WT,A lpha, and Delta variants compared to the Beta variant. No neutralization against the Beta variants was observed for some of the patients.Onthe other hand, the neutralization capacity for all the variants was high in the vaccinated individuals,suggesting that this vaccine can lead to production of neutralizing antibodies against the variants at the time at which the samples were collected. [26] We demonstrate that this multiplex Simoa assay can be rapidly adapted to detect the neutralization capacity against novel SARS-CoV-2 variants.Particularly,i th as been recently shown that the levels of neutralizing antibodies are related to vaccine efficacy. [27, 28] We anticipate similar adaptability of the assay for future infectious disease outbreaks.

In this study,wedeveloped asimple assay for determining antibody neutralization using Single Molecule Arrays (Simoa). This assay shows high sensitivity and multiplexing capabilities,e nabling high resolution profiling of neutralization capacity for several targets.A dditionally,t he assay can complement traditional serology assays that measure antibody levels by quantifying the antibodies that inhibit the viral antigen-ACE2 interaction. More specifically,a ntibodies that have higher off-rates,orlower affinities,tospike than ACE2 will be displaced. As ar esult, the information obtained from this assay closely resembles the biological immune response, in which antibodies with lower dissociation constants for spike,orhigher affinities than ACE2, demonstrate substantial neutralization.

We demonstrated the utility of this assay for two applications.F irst, we used both the neutralization assay and the Simoa serology assay to characterize the immune response in hospitalized COVID-19 patients.W eobserved an overall increase in neutralization capacity over time in our sample cohort. We found that patients who died of COVID-19 had higher levels of antibodies than patients who were hospitalized for as horter period. These results are in agreement with previous reports. [29] [30] [31] [32] However,t he relationship between disease severity and the production of neutralizing antibodies is presently unclear. Future studies can use this assay format to determine the antibody potencybymeasuring the neutralization capacity and levels of antibody isotypes for predicting disease outcome. [31] Second, we developed amultiplex Simoa assay to assess the neutralization capacity against SARS-CoV-2 variants.W em easured the neutralization capacity in pre-pandemic healthy,C OVID-19 positive,a nd vaccinated individuals against WT RBD and three variants. Thea bility to quantify neutralization capacity against the different variants can be used to guide vaccine development and epidemiological studies.T he Simoa neutralization assay we developed here has high reproducibility and specificity, with minimal cross-reactivity.The assay is highly versatile and is broadly applicable to future detection of antibody neutralization for SARS-CoV-2 and other infectious diseases.

Acceptedm anuscript online: September1 7, 2021 Version of record online

Thea uthors would like to thank Kathleen Bellon for her assistance with the experiments.F unding for this work came from agenerous donation from Barbara and Amos Hostetter and the Chleck Foundation. This work was also largely funded through ag rant from the Massachusetts Consortium for Pathogen Readiness.F unding was also provided by the Synthetic Biology Platform at the Wyss Institute for Biologically Inspired Engineering.

David Walt has afinancial interest in Quanterix Corporation, ac ompany that develops an ultra-sensitive digital immunoassay platform. He is an inventor of the Simoa technology, af ounder of the company and also serves on its Board of Directors.D r. Waltsi nterests were reviewed and are managed by BWH and Partners Healthcare in accordance with their conflict of interest policies.Keywords: COVID-19 ·m ultiplexing ·n eutralization assays · single molecule arrays