key: cord-0979986-l0g911pu authors: Luo, Y. R.; Yun, C.; Chakraborty, I.; Wu, A. H. B.; Lynch, K. L. title: A Label-Free SARS-CoV-2 Surrogate Virus Neutralization Test and a Longitudinal Study of Antibody Characteristics in COVID-19 Patients date: 2021-01-20 journal: nan DOI: 10.1101/2021.01.19.21250137 sha: 983ff68df0597531e6f236189051f6797aa3e0c4 doc_id: 979986 cord_uid: l0g911pu Background. The laboratory-based methods to measure the SARS-CoV-2 humoral response include virus neutralization tests (VNTs) to determine antibody neutralization potency. For ease of use and universal applicability, surrogate virus neutralization tests (sVNTs) based on antibody-mediated blockage of molecular interactions have been proposed. Methods. A label-free surrogate neutralization assay (LF-sVNT) was established on the TFI label-free immunoassay analyzer to measure neutralizing antibody titers in serum samples. The LF-sVNT analyzes the binding ability of RBD to ACE2 after neutralizing RBD with antibodies in serum. Results. The LF-sVNT neutralizing antibody titers (IC50) were determined from serum samples (n=246) from COVID-19 patients (n=113), as well as the IgG concentrations and the IgG avidity indices. There is considerable variability in the kinetics of the IgG concentration and neutralizing titer between individuals, however, there is an initial rise, plateau and then in some cases a gradual decline at later timepoints after 40 days post-symptom onset. The IgG avidity, in the same cases, plateaued after the initial rise and did not show a corresponding decline with IgG concentration and neutralizing ability. Conclusions. The LF-sVNT can be valuable tool in clinical laboratories for the assessment of the presence of neutralizing antibodies to COVID-19. This study is the first to provide longitudinal neutralizing antibody titers beyond 200 days post symptom onset. Despite the decline of IgG concentration and neutralizing antibody titer, IgG avidity index increases, reaches a plateau and then remains constant up to 8 months post-infection. The decline of antibody neutralization potency can be attributed to the reduction in antibody quantity rather than the deterioration of antibody avidity, a measure of antibody quality. Since the start of the COVID-19 pandemic in early 2020, much research has focused on the kinetics and magnitude of the immune response and measurable correlates of acquired immunity. SARS-CoV-2 infection can be detected indirectly by measuring the host immune response. Most immunocompetent individuals with symptomatic infections develop detectable SARS-CoV-2 antibodies within 2 weeks of symptom onset (1-3). The SARS-CoV-2 IgG antibody response is more robust in severe cases of COVID-19 at all time-points after seroconversion (1) . However, the antibody neutralization potency, depicted as neutralizing antibody titers, cannot be directly obtained from the antibody concentrations (2). Intraindividual variation in the quantity of antibodies produced indicates that serological responses may not be equivalent in terms of future protection. Early reports suggest that the SARS-CoV-2 IgG concentrations can wane over time, however, it is unclear if the antibodies that persist are capable of neutralizing the virus (4). While the antibody quantity may decline, the quality of remaining IgG antibodies, as determined by the measurement of antibody avidity, or functional affinity, increases over time post symptom onset (5). It is still unknown how antibody production following vaccination will compare to that of acquired disease. Further studies are needed to determine how durable the humoral immune response is following acquired disease and vaccination. . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 20, 2021. ; https://doi.org/10.1101/2021.01.19.21250137 doi: medRxiv preprint Laboratory-based methods used to measure the SARS-CoV-2 humoral response include qualitative and quantitative methods for total antibody or antibody subclasses (IgG, IgM, IgA) (1, 4) , IgG avidity (5), and antibody neutralization potency. Plaque reduction neutralizing tests, also known as conventional virus neutralization tests (cVNTs), measure SARS-CoV-2 neutralizing antibody titer and involve the use of live pathogens and target cells (6,7). Experiments using pandemic pathogens like SARS-CoV-2 impose special safety requirements and cannot be implemented in most clinical laboratories, limiting the widespread availability of testing. Methods using pseudovirus (pVNTs) have been published, however, these can take days to obtain results (8-10). For ease of use and universal applicability, surrogate virus neutralization tests (sVNTs) based on antibody-mediated blockage of molecular interactions have been proposed (10-15). An sVNT measures the competitive inhibition of the interaction between a viral structural protein and angiotensin-converting enzyme 2 (ACE2), since ACE2 is the receptor of SARS-CoV-2 on host cells (16). Thus, sVNTs can be designed with compatibility for routine clinical laboratory settings. Like cVNTs, sVNTs detect neutralizing antibodies in an isotype-independent manner, offering a key advantage over antibody concentration assays. The SARS-CoV-2 spike protein receptor-binding domain (RBD) is the favored choice as the viral structural protein used in sVNTs because (1) it is the binding domain located on the spike protein responsible for viral entry into host cells (17); (2) it has better binding characteristics in comparison to spike protein S1 subunit and nucleocapsid protein (11); and (3) it is a highly specific target for antibodies and has less cross-reactive epitopes with other coronavirus (18,19). . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 20, 2021. Recently, open-access label-free technologies have emerged as a novel solution for next-generation immunoassays in clinical laboratories (20,21). These technologies can measure the time course of immunoreactions in real-time without attaching a reporter (enzyme, fluorophore, etc.); thus it provides fast measurement, simple operation with automation, and ease of assay development and troubleshooting. One such technology, thin-film interferometry (TFI), has been used for the routine therapeutic drug monitoring of the monoclonal antibody therapeutics and associated anti-drug antibodies in human serum. The same technology has been applied to the measurement of SARS-CoV-2 IgG avidity (5). This paper describes the development and validation of a novel label-free surrogate virus neutralization test (LF-sVNT) using TFI. The method was used to measure neutralizing antibodies in a cohort of COVID-19 patients and determine if they correlated with total SARS-CoV-2 IgG concentration or avidity. Serial serum samples collected from mild to severe COVID-19 patients were measured out to 8 months post symptom onset to determine the kinetics and durability of the neutralizing antibody response. All serum samples used in the analysis were remnant specimens obtained following routine clinical laboratory testing. The study protocol was approved by the Institutional Review Board of the University of California San Francisco. The committee deemed that written consent was not required for use of remnant specimens. . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 20, 2021. ; https://doi.org/10.1101/2021.01.19.21250137 doi: medRxiv preprint Individual and serial serum samples (n=246) from 113 patients diagnosed with COVID-19 (PCR-confirmed) were obtained for testing. The patients were 63% male and 75% Hispanic, with a median age of 51 years. Fifty-eight patients (51%) were hospitalized and 55 patients (49%) were outpatients. Of the hospitalized patients, 33 (57%) were admitted to the intensive care unit (ICU), 25 (43%) received mechanical ventilation, and 2 died while in the ICU. The sampling time span of the patients ranged from 5 to 225 days after symptom onset. Recombinant RBD and ACE2 were purchased from Sino Biological (Wayne, PA). A human monoclonal anti-RBD IgG1 antibody was obtained from Absolute Antibody (Oxford, UK), and a goat anti-human IgG antibody (anti-IgG) from Jackson Immunoresearch (West Grove, PA). The TFI label-free immunoassay analyzer and the sensing probes coated with RBD were manufactured by Gator Bio (Palo Alto, CA). The Pylon 3D fluorescence immunoassay analyzer was manufactured by ET Healthcare (Palo Alto, CA). A label-free surrogate neutralization assay (LF-sVNT) was established on the TFI label-free immunoassay analyzer to measure neutralizing antibody titers in serum samples. The LF-sVNT analyzes the binding ability of RBD to ACE2 after neutralizing RBD with antibodies in serum. The sensing probes in use were pre-coated with RBD. Each serum sample was diluted in a series (1:50, 1:100, 1:250, 1:500, 1:1000, 1:2000) in running buffer (PBS at pH 7.4 with 0.02% Tween 20, 0.2% BSA, and 0.05% NaN 3 ) for analysis. The LF-sVNT was carried out by dipping a sensing probe sequentially into a sample and reagents. The LF-sVNT protocol consisted of two . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 20, 2021. ; https://doi.org/10.1101/2021.01.19.21250137 doi: medRxiv preprint cycles: the first cycle included the steps of 1) dipping the sensing probe in running buffer for a baseline, 2) forming RBD-Ab immune complexes with a sample, 3) rinsing the sensing probe in running buffer, 4) forming RBD-ACE2 immune complex with 12 μ g/ml ACE2 in running buffer; the second cycle included only the first and last steps of the first cycle. Between the two cycles, the sensing probe was regenerated in 10 mM glycine at pH 2.0 to strip off the molecules bound to RBD. The concentration of ACE2 was set at 12 μ g/ml (0.14 μ M), several times higher than the reported affinity constant between RBD and ACE2 monomer 18.5 nM (22), to ensure the saturation of RBD on the sensing probe. The first cycle measured the binding ability of RBD to ACE2 after neutralization, and the second cycle provided the full binding ability of RBD without neutralization. In each cycle, the recorded time course of signals, as known as the sensorgram, was recorded. The readout measured the signal increase in the step of forming RBD-ACE2 immune complex, representing the quantity of the immune complex on the sensing probe. The neutralization index was calculated as the ratio of the readout in the first cycle to that in the second cycle, presented as a percentage, meaning the residual binding ability of RBD to ACE2 after neutralization. The illustration of the assay protocol and example sensorgrams are shown in Figure 1 . The precision of the LF-sVNT was verified using a spiked serum sample, i.e. a human monoclonal SARS-CoV-2 anti-RBD IgG at 20 μ g/ml in a negative serum sample. The spiked serum sample was measured by 5 sensing probes, and the measurement was repeated 3 times using the same sensing probes with regeneration in between. When measuring patient serum samples, the spiked serum sample was used as a positive control, and the results of the positive . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 20, 2021. To obtain the neutralizing antibody titer (IC50) for each serum sample, the neutralization indices were plotted against dilutions, and the points were fitted using a linear interpolation model. The reciprocal of the dilution resulting in a 50% neutralization index was defined as the neutralizing antibody titer. A label-free IgG avidity assay was also established using the TFI technology, as reported previously (5). The sensing probes and running buffer were the same as those in the LF-sVNT. Each serum sample was 10-fold diluted in running buffer for analysis. The IgG avidity assay protocol included the steps of 1) dipping the sensing probe in running buffer for a baseline, 1) forming RBD-IgG immune complex on the sensing probe, 2) dissociating loosely bound IgG using either running buffer or 3 M urea in running buffer, and 3) forming RBD-IgG-Anti-IgG immune complex using 10 μ g/ml anti-IgG in running buffer. The signal increase in the final step, which is proportional to the quantity of RBD-IgG-Anti-IgG immune complex on the sensing probe, was measured. The IgG avidity index was calculated as the ratio of the readout with the dissociation agent (urea) to the reference (running buffer), presented as a percentage. A . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 20, 2021. ; https://doi.org/10.1101/2021.01.19.21250137 doi: medRxiv preprint fluorescence IgG concentration assay was carried out on the Pylon 3D fluorescence immunoassay analyzer, as reported previously (1). A cVNT was used for method comparison with the LF-sVNT in a subset of serum samples spanning the range of the assay. A 2-fold dilution series of each serum sample was prepared in Hank's balanced salt media (BA-1). Each dilution was mixed with an equal volume of SARS-CoV-2 virus suspension and incubated for 60 min at 37°C with 5% CO 2 . The mixture was then added to a Vero cell suspension and incubated for 45 min. After incubation, a first layer of overlay (2X MEM with 1% agarose) was added to each well, and the plates were again placed in the incubator. After 24 hours, a second overlay with 0.05% neutral red dye was added, and the plates were incubated at 37°C. Viral plaques were counted the following day and the 50% cutoff was calculated based on negative control plaque counts. The PRNT50 titer reported is the reciprocal of the highest dilution of serum that inhibits ≥ 50% of the plaques relative to the control. The precision (% coefficient of variation) of the neutralization indices for dilutions 1:50, 1:100, 1:250, 1:500, 1:1000, 1:2000 across 5 sensing probes was 9.0%, 2.2%, 2.9%, 4.4%, 3.0%, 1.7%, and across 3 repeats (same sensing probes with regeneration in between) was 7.5%, 2.9%, 5.9%, 6.0%, 4.2%, 1.8%, respectively. The precision of the neutralization indices across different batches (n=8) was 9.8%, 4.7%, 9.5%, 8.3%, 3.8%, 2.3% for the same dilutions, respectively. . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. To further characterize the apparent decline in LF-sVNT neutralizing antibody titers and the correlation with IgG concentrations and avidity, all available paired serum samples from week 4 and 3-8 months post-symptom onset were measured (Figure 3 ). For 20 paired samples, IgG concentrations and LF-sVNT neutralizing antibody titers declined in all but 2 and 1 subjects, respectively. In contrast, the SARS-Cov-2 IgG avidity increased in all but 1 subject. For 9 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 20, 2021. subjects, 4 or more serial samples with at least one sample beyond 30 days post symptom onset were available for measuring the kinetics of the antibody response (Figure 4 ). There is considerable variability in the kinetics of the IgG concentration and neutralizing titer between individuals, however, there is an initial rise, plateau and then in some cases a gradual decline at later timepoints after 40 days post-symptom onset. The IgG avidity, in the same cases, plateaued after the initial rise and did not show a corresponding decline with IgG concentration and neutralizing ability. The clinical utility of serological testing for SARS-CoV-2 continues to be a subject of debate. To date, multiple qualitative and quantitative methods for SARS-CoV-2 total antibodies, antibody subclasses, antibody avidity and neutralization potency, have been developed and used to characterize the humoral response to active infection through convalescence. Despite the advancing knowledge from these studies, questions remain regarding lasting protection following infection and vaccination. A method that can rapidly provide data on SARS-CoV-2 antibody neutralization potency and is amenable to a high-throughput clinical laboratory setting may offer a diagnostic test to accurately determine protective immunity. This remains to be determined. Here we present an sVNT method using a novel label free technology, that correlated with absolute IgG antibody concentration and a cVNT, and can be performed with a rapid turn-around time. The LF-sVNT employed a sensing probe coated with RBD to mimic the surface of SARS-CoV-2 and ACE2 as an equivalent of host cells. Although the RBD-ACE2 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 20, 2021. This advantage of LF-sVNT eliminates any possible interference to the viral protein-ACE2 complex during these steps and decreases testing time compared to other published sVNT methods (10,11,14) . The LF-sVNT provides process efficiency for fully automated, random-access testing, compared to batch testing for most other sVNTs. The real-time monitoring of the RBD-ACE2 interaction can be used for quality assurance to reduce experimental error rate. The LF-sVNT can be easily modified to incorporate mutated variants of RBD. Thus, the LF-sVNT can be a valuable tool in clinical laboratories for the assessment of the presence of neutralizing antibodies to COVID-19. Recent reports provide evidence for a decline in SARS-CoV-2 neutralizing antibody titers in individuals sampled over 40 days after disease onset (4). Measurements were carried out with samples collected up to 94 days. Further assessment of antibody neutralization potency for a longer time frame is still necessary to determine the longevity of the neutralizing antibody response. In this study, the overall decline of neutralizing antibody titer was consistent with previous reports, however, this data is the first to provide longitudinal neutralizing antibody titers . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 20, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 20, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 20, 2021. . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 20, 2021. ; https://doi.org/10.1101/2021.01.19.21250137 doi: medRxiv preprint . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Longitudinal analysis of clinical serology assay performance and neutralising antibody levels in COVID19 convalescents Persistence and decay of human antibody responses to the receptor binding domain of SARS-CoV-2 spike protein in COVID-19 patients Decline of Humoral Responses against SARS-CoV-2 Spike in Convalescent Individuals Longitudinal Dynamics of the Neutralizing Antibody Response to Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Infection Decline in SARS-CoV-2 Antibodies After Mild Infection Among Frontline Health Care Personnel in a Multistate Hospital Network -12 States Immunological memory to SARS-CoV-2 assessed for up to 8 months after infection International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity The authors thank Dr. Raymond Goodrich and Lindsay Hartson at the Infectious Disease Research Center, Colorado State University for carrying out the cVNT. The study was funded by departmental discretionary funds available to the corresponding author. Experiment consumables were donated by Gator Bio. Gator Bio was not involved in any aspect of the study design or execution. All authors declare no competing interests.