key: cord-0754067-l3gtwyt6 authors: Denninger, Viola; Xu, Catherine K.; Meisl, Georg; Morgunov, Alexey S.; Fiedler, Sebastian; Ilsley, Alison; Emmenegger, Marc; Malik, Anisa Y.; Piziorska, Monika A.; Schneider, Matthias M.; Devenish, Sean R. A.; Kosmoliaptsis, Vasilis; Aguzzi, Adriano; Fiegler, Heike; Knowles, Tuomas P. J. title: Microfluidic Antibody Affinity Profiling Reveals the Role of Memory Reactivation and Cross-Reactivity in the Defense Against SARS-CoV-2 date: 2022-03-30 journal: ACS Infect Dis DOI: 10.1021/acsinfecdis.1c00486 sha: 81b89b294df0e440c4dd9952bb47c54623b64bb9 doc_id: 754067 cord_uid: l3gtwyt6 [Image: see text] Recent efforts in understanding the course and severity of SARS-CoV-2 infections have highlighted both potentially beneficial and detrimental effects of cross-reactive antibodies derived from memory immunity. Specifically, due to a significant degree of sequence similarity between SARS-CoV-2 and other members of the coronavirus family, memory B-cells that emerged from previous infections with endemic human coronaviruses (HCoVs) could be reactivated upon encountering the newly emerged SARS-CoV-2, thus prompting the production of cross-reactive antibodies. Determining the affinity and concentration of these potentially cross-reactive antibodies to the new SARS-CoV-2 antigens is therefore particularly important when assessing both existing immunity against common HCoVs and adverse effects like antibody-dependent enhancement (ADE) in COVID-19. However, these two fundamental parameters cannot easily be disentangled by surface-based assays like enzyme-linked immunosorbent assays (ELISAs), which are routinely used to assess cross-reactivity. Here, we have used microfluidic antibody affinity profiling (MAAP) to quantitatively evaluate the humoral immune response in COVID-19 convalescent patients by determining both antibody affinity and concentration against spike antigens of SARS-CoV-2 directly in nine convalescent COVID-19 patient and three pre-pandemic sera that were seropositive for common HCoVs. All 12 sera contained low concentrations of high-affinity antibodies against spike antigens of HCoV-NL63 and HCoV-HKU1, indicative of past exposure to these pathogens, while the affinity against the SARS-CoV-2 spike protein was lower. These results suggest that cross-reactivity as a consequence of memory reactivation upon an acute SARS-CoV-2 infection may not be a significant factor in generating immunity against SARS-CoV-2. D espite the increasing availability of diagnostic tests, antibody treatments, and vaccines, the COVID-19 pandemic still presents a major challenge to governments and health care systems all over the world, requiring a functional assessment of the immune response to SARS-CoV-2. Research focuses increasingly around the role of acutely increased production of pro-inflammatory cytokines and the recall of memory T-and B-cells, which were formed during previous infections with other coronaviruses. 1−4 In contrast to SARS-CoV-2, common human coronaviruses (HCoVs), although omnipresent and recurrent, generally cause mild disease only. 5 HCoVs can be detected throughout all agegroups with most individuals seroconverting during childhood. 6, 7 Reinfection subsequently leads to antibody affinity maturation, resulting in a natural buildup of protective memory T-and B-cells as well as long-lived plasma cells (LLPCs) that in some cases can persist up to a year. 8, 9 In the event of the emergence of new but highly homologous coronavirus species, memory B-cells might recognize the antigens of the new species, reactivate, and accelerate the immune response to the new virus, prompting the production of cross-reactive antibodies. 10 Although immunity derived from previous infections with similar viruses is generally advantageous, cross-reactive antibodies can cause antibody-dependent enhancement (ADE) of the viral infection in a subset of individuals. 10, 11 In these cases, cross-reactive antibodies, although able to recognize and bind the new viral antigens, not only fail to neutralize the virus due to sub-neutralizing concentrations, insufficiently high affinity, or binding to an inappropriate region, but also, more importantly, facilitate its reproduction via an Fc-receptormediated uptake into the host cell. Cross-reactivity-mediated ADE has been shown to elevate the disease severity of infections with many other viruses, including flaviviruses (especially Dengue and Zika viruses), influenza as well as two other coronaviruses MERS and SARS-CoV, and is therefore a key consideration in the development of vaccines and therapies for COVID-19. 12−17 Cross-reactivity of antibodies is traditionally assessed using enzyme-linked immunosorbent assays (ELISAs), where the measurement of the antibody-antigen interaction relies on the binding of antibodies to an immobilized interaction partner, i.e., the antigen or a capture-antibody. As a result, the reported antibody titer, or EC 50 , depends on the combination of affinity, avidity, and concentration of the antibodies in question. These assays therefore cannot readily resolve the fundamental parameters of antibody responses, namely affinity and concentration, 18 which are crucial for monitoring the effectiveness of humoral immunity after natural infection and after vaccination. Here, we have used microfluidic antibody affinity profiling (MAAP) 18−20 to quantify the cross-reactivity of antibodies against SARS-CoV-2 S1, S2 and the receptor-binding domain (RBD) in nine convalescent COVID-19 patient samples and three pre-pandemic controls and compared our results to standard ELISA assays. While MAAP matched the ELISA data qualitatively, crucially, this in-solution technology enabled the simultaneous and independent determination of antibody affinities and concentrations to reveal the interindividual variation of antibody responses to SARS-CoV-2 not detectable by ELISA. Interestingly, testing all 12 serum samples in a reciprocal approach for antibodies against the corresponding antigens of two common HCoVs, namely HCoV-NL63 and HCoV-HKU1, revealed high antibody affinity and low concentrations across all samples tested. This indicates the presence of matured antibodies against both common HCoVs rather than Figure 1 . Microfluidic antibody affinity profiling (MAAP) of nine convalescent COVID-19 patients and three pre-pandemic sera. (A) Probability density plots of MAAP against fluorescently labeled SARS-CoV-2 RBD, spike S1, and spike S2. The graphs show the affinity (K D ) and the molar concentration of antibody binding sites for each convalescent serum sample. No binding could be detected in the three pre-pandemic sera. Points correspond to the maximum a posteriori values in the two-dimensional posterior probability distribution, and shaded regions correspond to the probability density. Gray shaded regions indicate the area of nonbinding for samples with [antibody binding sites] 90 nM, only one prepandemic serum (pre-pandemic serum 2) exhibited higher levels of antibody binding sites against the S1 subunits of HCoV-NL63 and HCoV-HKU1, indicative of an increased antibody production likely in response to a recent infection with both viruses. Intriguingly, with K D values <3.6 nM, all sera contained antibodies with very high affinities against the two HCoV-NL63 antigens. Furthermore, affinities against HKU1 RBD were in the sub-nanomolar range throughout all samples that were subjected to the MAAP assay. The high specificity toward HCoV-NL63 and HCoV-HKU1 epitopes indicates that these antibodies were not produced in response to SARS-CoV-2, but were rather derived from memory immunity. Based on our observations as well as previously published reports, 4,6,31 we hypothesize that the antibodies we detected with the use of our MAAP assay are part of a general background immunity of high affinity circulating serum antibodies against HCoV-NL63 and HCoV-HKU1 that were produced in response to previous infections. Pre-Pandemic Sera Show No Significant Cross-Reactivity against SARS-CoV-2 Spike S1, S2, and RBD. Given the ubiquitous prevalence and high reinfection rate of common HCoVs, 32 it is likely that most individuals will have gained an immunological memory in the form of memory Tand B-cells that will produce antibodies upon reactivation, as well as long-lived plasma cells (LLPs), constitutively secreting high-affinity antibodies but at concentrations comparatively lower than those in response to an acute infection. 9 If crossreactive to SARS-CoV-2 antigens, these memory-derived antibodies could not only impact the accuracy of diagnostic tests for SARS-CoV-2 infections but could also present a risk for ADE, and thus potentially affect the severity of the immune response after infection. It is therefore crucial to further examine the ability of anti-HCoV antibodies to recognize and bind SARS-CoV-2 antigens. To probe for the cross-reactive potential of these antibodies against the corresponding SARS-CoV-2 epitopes, we performed MAAP assays against the three SARS-CoV-2 antigens, spike S1, spike S2, and RBD, in the three pre-pandemic sera. Interestingly, affinity profiling did not reveal any detectable binding of pre-existing antibodies to SARS-CoV-2 RBD, spike S1, or spike S2 in the three pre-pandemic sera ( Figure 1A and Table S1 ). To further investigate, we then assessed cross-reactivity directly via a competition assay based on the principle described in Schneider et al. 19 and Fiedler et al. 20 ( Figure 3A ). For this approach, a quantity of each serum corresponding to 20 nM antibody binding sites was mixed with 10 nM labeled HCoV-NL63 RBD to allow at least 80% of RBD was mixed either with buffer or with patient serum, incubated for 1 h, and subjected to microfluidic diffusional sizing to determine the size of the free labeled RBD (R h,free ) and the size of the immune-complex (R h,complex ). For binding competition, 250 nM unlabeled SARS-CoV-2 RBD was added to the mixture of 10 nM HCoV-NL63 RBD and patient serum and incubated for 1 h before measuring the hydrodynamic radius (R h ). If the serum contains cross-reactive antibodies, a size decrease is observed, as the antibodies will bind to the excess of unlabeled RBD (left box). If the antibodies in the serum are not cross-reactive, they remain bound to the labeled HCoV-NL63 RBD, and the R h of the immuno-complex will stay constant (right box). (B) SARS-CoV-2−HCoV-NL63 cross-reactivity competition assay in convalescent COVID-19 sera and pre-pandemic sera. the labeled antigen to be bound by antibodies. After equilibration, the mixture was subjected to microfluidic diffusional sizing, and complex formation was confirmed by the increase of the apparent R h . For competition, an excess of 250 nM unlabeled SARS-CoV-2 RBD was added to the HCoV-NL63 RBD−antibody complex, and R h was measured after the establishment of binding equilibrium. If antibodies bound to the RBD of HCoV-NL63 were cross-reactive, the excess unlabeled SARS-CoV-2 RBD would outcompete the binding, resulting in an unbound HCoV-NL63 RBD and therefore a decrease in R h compared to that of the antibody complexed HCoV-NL63 RBD. Using this approach, we again tested the nine convalescent COVID-19 as well as the three HCoV pre-pandemic sera. We found that even upon addition of excess unlabeled SARS-CoV-2 RBD, the R h did not decrease in any of the samples, indicating that all antibodies stayed tightly bound in a complex with labeled HCoV-NL63 RBD, and thus were not crossreactive with SARS-CoV-2 RBD ( Figure 3B ). Two convalescent COVID-19 serum samples however, serum 7 and serum 8, presented an interesting exception, as their complex size increased upon addition of unlabeled SARS-CoV-2 RBD. This was likely due to unsaturated antibody binding sites that were still free to bind some of the SARS-CoV-2 RBD protein but with a low affinity unable to outcompete the interaction with labeled HCoV-NL63 RBD (Tables S1 and S2) . Additionally, we observed greater variations of initial HCoV-NL63−antibody complex sizes within the pool of convalescent COVID-19 serum samples (R h = 5.9−9.6 nm) compared to the pre-pandemic sera (R h = 7.3−8.0 nm). This observation is in line with the fact that our cohort derived from recently COVID-19-recovered patients, which contained antibody types of different sizes, namely IgG and IgM. Based on these results, even though we cannot fully exclude cross-reactivity against the SARS-CoV-2 RBD, we would expect cross-reactive antibodies to exhibit at least a 10-fold lower affinity to the SARS-CoV-2 RBD region compared to its original antigen. To further test for specificity of convalescent serum antibodies against the RBD subunit of SARS-CoV-2, we additionally employed the reciprocal approach, where the fluorescently labeled RBD subunit of SARS-CoV-2 was incubated with patient serum for immune-complex formation ( Figure S2 ). Only the positive control, an access of unlabeled SARS-CoV-2 RBD, was able to outcompete the binding of its labeled counterpart. In contrast, R h,complex did not decrease upon addition of access HCoV-NL63 RBD or HCoV-HKU1 RBD, again suggesting that convalescent serum antibodies against SARS-CoV-2 RBD were indeed specific and did not derive from cross-reactive pre-existing antibodies against HCoV-NL63 or HCoV-HKU1. Protective and adverse effects of antibody cross-reactivity can influence the humoral immune response of an individual to an acute infection, vaccination, or treatment. In this study, we have employed microfluidic antibody affinity profiling (MAAP) to quantitatively investigate cross-reactivity against SARS-CoV-2 and two common coronaviruses, HCoV-NL63 and HCoV-HKU1, in nine convalescent COVID-19 and three pre-pandemic sera. This enabled us to quantify the affinity and concentration of antibodies against different spike subunits of the three coronaviruses, allowing a differentiation between antibodies raised in response to an acute infection and those originating from immunological memory. Affinity profiling against different SARS-CoV-2 spike antigens revealed a wide range of antibody affinities and concentrations in all convalescent sera, especially against the RBD and the S1 subunit of SARS-CoV-2. In our cohort, this variability suggests a de novo synthesis of antibodies in response to a SARS-CoV-2 infection, further supported by the fact that we did not find evidence for significant cross-reactivity of antibodies against SARS-CoV-2 spike antigens in pre-pandemic, HCoV-seropositive sera. This contrasts with the antibody affinities and concentrations that were found when profiling the same 12 sera against spike S1 and RBD from HCoV-NL63 and HCoV-HKU1. With the exception of one pre-pandemic serum that was seropositive for all common HCoVs and thus showed elevated levels of antibodies against HCoV-NL63 and HCoV-HKU1, all sera tested contained antibodies with high affinity but at relatively low antibody concentrations. Consistent with epidemiological studies on infection patterns of HCoV-NL63 and HCoV-HKU1, 33 this suggests the existence of a low amount of circulating antibodies against the RBD and the S1 subunit of the two HCoVs, constitutively secreted by longlived plasma cells that were formed during previous HCoVinfections, rather than being the result of the cross-reactivityinduced recall of memory B-cells. 31 With the ability to break down the humoral immune response into affinity and concentration of serum antibodies, MAAP generates an individual fingerprint for each sample, providing crucial insights into the immunological landscape of de novo and reoccurring infections. This unique quantitative approach therefore delivers new prospects to monitor immune induction following vaccination or a treatment more precisely, helping to evaluate the efficacy of antibody production and to assess the potential risks of humoral memory. ■ METHODS Sequence Alignment. Sequence alignment of coronavirus full-length spikes was performed using the Clustal Omega Sequence alignment tool. Input sequences were HCoV-NL63 (UniProt ID: Q6Q1S2), HCoV-HKU1 (UniProt ID: Q5MQD0), and SARS-CoV-2 (UniProt ID: P0DTC2). Origin of Serum Samples. All serum samples from COVID-19 convalescent patients as well as the pre-COVID sera were purchased from BioIVT. Pre-COVID sera were seropositive for the following coronaviruses: serum 1 (HCoV-229E, HCoV-NL63), serum 2 (HCoV-229E, HCoV-NL63, HCoV-OC43, HCoV-HKU1), and serum 3 (HCoV-229E). BioIVT sought informed consent from each subject or the subject's legally authorized representative and appropriately documented this in writing. All samples are collected under IRB-approved protocols. ELISA Assay. ELISA-based serology was carried out as previously described, 27, 34 with minor modifications. Highbinding 384-well plates (PerkinElmer, SpectraPlate 384 HB) were coated with 20 μL of 1 μg/mL spike ectodomain (Lausanne, EPFL SV PTECH PTPSP), RBD (Lausanne, EPFL SV PTECH PTPSP), or spike S1 domain (S1N-C52H4, ACROBiosystems) in PBS at 37°C for 1 h, followed by three washes with PBS 0.1% Tween 20 (PBS-T) using a BioTek EL406 plate washer and blocked with 40 μL of 5% milk in PBS-T for 1.5 h. Serum samples were serially diluted (range: 0.02−1.2 × 10 −6 ) in sample buffer (1% milk in PBS-T) and added to wells (20 μL/well). After sample incubation for 2 h at RT, the wells were washed five times with wash buffer and the presence of IgGs directed against above-defined SARS-CoV-2 antigens was detected using an HRP-linked anti-human IgG antibody (Peroxidase AffiniPure Goat Anti-Human IgG, Fcγ Fragment Specific, Jackson, 109-035-098, at 1:4000 dilution in sample buffer) at 20 μL/well. The incubation of the secondary antibody for 1 hour at RT was followed by three washes with PBS-T, the addition of TMB, incubation of 5 minutes at RT, and the addition of 0.5 M H 2 SO 4 . The plates were centrifuged after all dispensing steps, except for the addition of TMB. The absorbance at 450 nm was measured in a plate reader (PerkinElmer, EnVision), and the inflection points of the sigmoidal binding curves (p(EC 50 ) or −log(EC 50 ) values of the respective sample dilution) were determined using the customdesigned fitting algorithm previously reported. 27 The p(EC 50 ) values for all samples and antigens were visualized using the ggplot2 package in R. Fluorescent Labeling of Proteins. SARS-CoV-2 S1 protein (S1N-C52H4), SARS-CoV-2 S protein RBD (SPD-C52H3), and SARS-CoV-2 S2 protein (S2N-C52H5) were purchased from ACROBiosystems. HCoV-NL63 S1 protein (40600-V08H) and HCoV-HKU1 S1 protein (40021-V08H) were purchased from Sino Biological. HCoV-NL63 Spike RBD (10605-CV) and HCoV-HKU1 Spike RBD (10600-CV) were obtained from R&D Systems. All proteins were reconstituted according to the manufacturer's instructions. For fluorescent labeling, the protein was diluted into labeling buffer (1 M NaHCO 3 pH 8.3) to give a final concentration of 0.2 M NaHCO 3 , followed by the addition of Alexa Fluor 647 NHS ester (A2006, Thermo Scientific) at a molar dye-to-protein ratio of 3:1. Following incubation overnight at 4°C, the labeled proteins were purified by size exclusion chromatography (SEC) on an ÄKTA pure system (Cytiva) with phosphate-buffered saline (PBS) at pH 7.4 (P4417, Merck) as elution buffer. For SEC-purification of RBD proteins, a Superdex 75 Increase 10/300 GL column was used; the larger S1 and S2 proteins were purified using a Superdex 200 Increase 3.2/300 column (Cytiva). Pooled fractions that correspond to the labeled protein were concentrated, snap-frozen, and stored at −80°C, with 10% v/v glycerol as a cryoprotectant. Equilibrium Affinity Binding Curves of Recombinant Antibodies against S1 Proteins of HCoV-HKU1 and HCoV-NL63. For affinity measurements of anti-human coronavirus spike glycoprotein HKU1 (40021-MM07-100, Sino Biological) to HCoV-HKU1 S1 protein, the antibody was reconstituted according to the manufacturer's instructions and diluted into PBS containing 0.05% Tween 20 and 5% human serum albumin (HSA) to achieve a 2-fold concentration series ranging from 60 pM to 1 μM. Antibody dilutions were subsequently mixed in a 1:1 ratio with a 20 nM solution of Alexa Fluor 647 labeled HCoV-HKU1 S1 to obtain a final concentration of 10 nM. For cross-reactivity measurements, anti-SARS-CoV-2 neutralizing antibody (SAD-S35, ACROBiosystems) was reconstituted according to the manufacturer's instructions. A 4-fold dilution series ranging from 60 pM to 1 μM was prepared by diluting into PBS containing 0.05% Tween 20 and 5% human serum albumin (HSA). Antibody dilutions were subsequently mixed in a 1:1 ratio with a 20 nM solution of Alexa Fluor 647 labeled HCoV-HKU1 S1 or Alexa Fluor 647 labeled HCoV-NL63 S1, respectively, to obtain a final concentration of 10 nM. All samples were incubated for 1 h at 4°C prior to the measurement and kept at 4°C throughout the experiment. Measurements were performed in triplicate using the 1.5−8 nm size-range setting on the Fluidity One-W Serum (Fluidic Analytics). The binding affinity, K D , was generated by nonlinear least-squares fitting to eq 1, leaving all parameters to be fit but constrained to be >0. with R h , R h,free , and R h,complex being the average hydrodynamic radii of the reaction at equilibrium of the unbound labeled S1 protein and of the complex of labeled S1 with unlabeled antibody, respectively. The parameters [U] 0 and [L] 0 are total concentrations of unlabeled antibody and labeled S1, respectively, and n is the number of binding sites per unlabeled antibody. Microfluidic Antibody Affinity Profiling of Serum Antibodies (MAAP). Measurements of serum samples were performed as reported previously. 19 For the MAAP measurements, varying fractions of COVID-19 convalescent or pre-COVID serum samples were combined with Alexa Fluor 647 labeled antigen of varying concentrations ranging between 1 and 250 nM. Buffer was added to give a constant volume of 20 μL for each sample, followed by incubation for 1 h at 4°C. For MAAP against SARS-CoV-2 RBD, HCoV-NL63 RBD or HCoV-HKU1 RBD, PBS containing 0.05% Tween 20 (pH 7.4) was used as a buffer. For profiling of serum against SARS-CoV-2 S1, SARS-CoV-2 S2, HCoV-NL63 S1, and HCoV-HKU1 S1, the buffer additionally contained 5% human serum albumin (HSA). Because of limited sample availability, only a subset of the nine convalescent serum samples could be affinity profiled against HCoV-HKU1 RBD and HCoV-HKU1 S1 ( Figure S1 and Table S3 ). Following incubation, the hydrodynamic radius, Rh, of the serum antibody-antigen complex was determined by microfluidic diffusional sizing using a Fluidity One-W Serum. Additionally, each serum was measured separately at a concentration of 100, 50, and 25% to allow correction of titration data points for the autofluorescence of the individual serum. The binding affinities and concentrations of antigen-specific antibodies in serum samples were determined by monitoring the fraction of labeled antigen that diffused into the distal chamber of the microfluidic device. Following background correction of samples to account for autofluorescence signal arising from the serum, this fraction can be determined by Given that we monitor a mixture of free and bound (B) labeled species (L), I d , and I u can be expressed as where k is a constant that relates the concentration of L to the observed fluorescence intensity, [B] is the equilibrium concentration of bound L, and ρ f and ρ b are the fractions of free and bound labeled antigen that diffuse across to the distal channel, respectively. f d can thus be expressed as follows By solving the binding equilibrium equation, [B] is given by where [U] is the total concentration of antibody binding sites in the serum, and α is the fractional concentration of serum in the binding measurement. Bayesian inference was used to determine K D and [U], with a prior that was flat in logarithmic space for K D and [U], and flat in linear space for ρ f and ρ b . The likelihood was a Gaussian with mean f d and its standard deviation defined as the standard deviation of all measurements of each labeled antigen. As serum samples were derived from convalescent patients potentially containing different classes of antibodies with variable numbers of binding sites (IgG = 2, IgM = 10), antibody concentration is expressed in terms of binding sites rather than molecules. SARS-CoV-2 RBD and HCoV-NL63-RBD Competition Assay. Briefly, 10 nM Alexa Fluor 647 labeled HCoV-NL63 RBD in PBS containing 0.05% Tween 20 was combined with serum sample corresponding to a final anti-NL63 RBD antibody concentration of 20 nM (based on previous MAAP of each serum sample). For competition, unlabeled SARS-CoV-2 RBD was added at a concentration of 250 nM. Samples were incubated 1 h at 4°C prior to equilibrium binding measurements using the Fluidity One-W Serum. Labeled HCoV-NL63 RBD in buffer and a mixture of labeled HCoV-NL63 RBD with serum only served as controls. For the reciprocal approach, convalescent sera were incubated with 10 nM Alexa Fluor 647 labeled SARS-CoV-2 RBD in PBS containing 0.05% Tween 20. For competition, unlabeled SARS-CoV-2 RBD, unlabeled HCoV-NL63 RBD, or unlabeled HCoV-HKU1 RBD was added at a concentration of 250 nM. Samples were incubated for 1 h at 4°C prior to equilibrium binding measurements using the Fluidity One-W Serum. Labeled SARS-CoV-2 RBD in buffer and a mixture of labeled SARS-CoV-2 RBD with serum only served as controls. Due to limited serum availability, the reciprocal assay was performed only in a subset of the nine convalescent sera (sera 4, 6, 7, and 8 for SARS-CoV-2 RBD and HCoV-NL63 RBD; sera 4, 7, and 8 for HCoV-HKU1 RBD). All measurements were performed in triplicate. For background correction, serum was diluted in PBS containing 0.05% Tween 20 at the same v/v concentration as used in the assay and measured separately. Background subtraction was applied to individual datapoints obtained in the assay. Quality Control. As part of constructing the dataset used in this research, we developed a series of procedures that we jointly refer to as a "quality control (QC) pipeline." First, they serve as simple sanity checks to ensure that data processing and model fitting were performed correctly. Second, they help identify problematic measurements, samples, or model fits. Based on this, decisions to re-collect or exclude parts of the data can be made. Finally, if assumptions about relationships in the dataset are appropriately tested, information can be shared between experiments and samples in a robust way, which improves model fitting. The detailed discussion of these methods is outside the scope of this publication; however, we are providing it in the form of a supplementary to this work to serve as a guide for anyone performing similar analyses (Supporting Material 1). When applying the described procedures to the present dataset, we identified a small number of outlier measurements. These were removed during the model fitting and data analysis step. Otherwise, the dataset passed all QC tests. The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsinfecdis.1c00486. Microfluidic antibody affinity profiling against HCoV-HKU1 spike S1 and RBD reveals high-affinity antibodies at even lower concentrations than against HCoV-NL63 ( Figure S1) ; reciprocal cross-reactivity competition assay in 4 (3) convalescent COVID-19 sera ( Figure S2 ); bestfit values and confidence intervals for all of the MAAP measurements against the SARS-CoV-2 spike antigens RBD, S1, and S2 (Table S1 ); best-fit values and confidence intervals for all of the MAAP measurements against the HCoV-NL63 spike antigens RBD and S1 (Table S2) ; best-fit values and confidence intervals for all of the MAAP measurements against the HCoV-HKU1 spike antigens RBD and S1 (Table S3) ; Quality Control (QC) pipeline (Supporting Material 1) (PDF) Immune Response to SARS-CoV-2 and Mechanisms of Immunopathological Changes in COVID-19 Older Adults Lack SARS CoV-2 Cross-Reactive T Lymphocytes Directed to Human Coronaviruses OC43 and NL63. Sci. Rep. 2020 T Cell Response to SARS-CoV-2 Infection in Humans: A Systematic Review Cross-Reactive Serum and Memory B Cell Responses to Spike Protein in SARS-CoV-2 and Endemic Coronavirus Infection Human Coronaviruses and Other Respiratory Infections in Young Adults on a University Campus: Prevalence, Symptoms, and Shedding Van Der Hoek, L. Human Coronavirus NL63 and 229E Seroconversion in Children First Infection by All Four Non-Severe Acute Respiratory Syndrome Human Coronaviruses Takes Place during Childhood Remembrance of Things Past: Long-Term B Cell Memory After Infection and Vaccination Humoral Immunity Due to Long-Lived Plasma Cells Antibody Dependent Enhancement Due to Original Antigenic Sin and the Development of SARS Molecular Mechanism for Antibody-Dependent Enhancement of Coronavirus Entry Cross-Reacting Antibodies Enhance Dengue Virus Infection in Humans Modulation of Dengue/ Zika Virus Pathogenicity by Antibody-Dependent Enhancement and Strategies to Protect against Enhancement in Zika Virus Infection. Frontier in Immunology The Possible Role of Cross-Reactive Dengue Virus Antibodies in Zika Virus Pathogenesis Neutralization of SARS-CoV-2 Omicron Variant by Sera from BNT162b2 or Coronavac Vaccine Recipients Role of Antibody-Dependent Enhancement (ADE) in the Virulence of SARS-CoV-2 and Its Mitigation Strategies for the Development of Vaccines and Immunotherapies to Counter COVID-19. Hum. Vaccines Immunother Microfluidic Antibody Affinity Profiling for In-Solution Characterisation of Alloantibody -HLA Antibody Affinity in Human Plasma Antibody Affinity Governs the Inhibition of SARS-CoV-2 Spike/ ACE2 Binding in Patient Serum Cross-Reactive Antibody Responses against SARS-CoV-2 and Seasonal Common Cold Coronaviruses Conformational Transition of SARS-CoV-2 Spike Glycoprotein between Its Closed and Open States Antibody-Dependent SARS Coronavirus Infection Is Mediated by Antibodies against Spike Proteins Absence of Severe Acute Respiratory Syndrome Coronavirus 2 Neutralizing Activity in Prepandemic Sera From Individuals With Recent Seasonal Coronavirus Infection Hosts and Sources of Endemic Human Coronaviruses Remembrance of Things Past: Long-Term B Cell Memory After Infection and Vaccination Seasonal Coronavirus Protective Immunity Is Short-Lasting Seasonal Coronavirus Protective Immunity Is Short-Lasting Anti-Prothrombin Autoantibodies Enriched after Infection with SARS-CoV-2 and Influenced by Strength of Antibody Response against SARS-CoV-2 Proteins