1. Introduction
SARS-CoV-2 infection is triggered by the fusion of the viral membrane with the plasma membrane of the target cells, which is initiated by the binding of the receptor-binding domain (RBD) of the viral spike protein S1 to the angiotensin-converting enzyme 2 (ACE2) receptor on the cell membrane [
1,
2,
3,
4]. The virus-neutralizing activity of antibodies involved in protection against infection depends on the factors that can compete with RBD/ACE2 binding: one is the quantity of antiviral spike protein S1 antibodies and the other is the RBD binding-affinity/avidity strength of the antibodies [
5,
6].
Affinity indicates the binding interaction of a monoclonal antibody that binds to a single antigen epitope, and avidity indicates the sum of the binding affinities of each polyclonal antibody that binds to corresponding antigen epitopes. This indicates that avidity measurement is more suitable than affinity for measuring antigen affinity of polyclonal antibodies in blood. The conventional avidity measurement is based on the removability of antibodies bound to antigen epitopes by washing with chaotropic agents such as urea for a short time under nonequilibrium conditions [
7,
8] and has provided a large number of evidence regarding the avidity maturation of IgG to SARS-CoV-2 antigen, and their ability to protect against infections induced by several steps of vaccination much higher than natural infection [
7,
8,
9,
10,
11,
12]. In addition, the importance of high avidity index for protective immunity has been established for many infections with viruses and other microbes [
8].
To measure avidity index, enzyme-linked immunosorbent assay (ELISA) is conventionally used in the absence and presence of various concentrations (4–8 M) of chaotropic agents over various short-time treatments (3–10 min) in the washing buffer to remove low avidity IgG [
7,
8,
9,
10,
11,
12]. This method is called chaotropic avidity measurement in this paper. However, there are limitations to the measurement of avidity index in the assay system described above. Treatment of urea (which gently destructs the three-dimensional structure of proteins and modifies the conformation of antibodies and antigens in the reaction system) does not show the direct strength of binding interaction between native antigen and antigen-specific antibodies. Furthermore, in the above measurement system, the avidity index values under nonequilibrium conditions vary depending on the antibody concentration in the reaction mixture even for the same urea concentration and it is difficult to evaluate or compare the binding interactions reported by various research groups.
To overcome such limitation and to accurately determine the protective immunity of the antibodies, we used a new parameter in this study to measure the antigen binding avidity of the variable region of the IgG antibodies. The binding avidity assay was first used in the field of food allergy diagnosis by the densely carboxylated protein (DCP) microarray [
13] by us as a method to increase the diagnostic performance of antigen-specific IgE levels in the deferential diagnosis of food allergy and anaphylaxis, even when it is difficult to diagnose by food challenge testing, and its potential usefulness has been proven [
13,
14,
15,
16]. The DCP microarray uses a new surface chemistry of carboxylated arm technology to cover the surface of a glass slide or diamond-like carbon-coated chip to immobilize proteins and DNA at very high density of 0.94–7.82 × 10
9 molecules mm
−2, as described in detail previously [
17]. The binding avidity assay was also found to be useful in the evaluation of IgG antibody’s ability to protect against infection in the field of infectious diseases.
Accumulating evidence suggests that the infectious pathogenic antigen-capturing potency of antibodies, which reflects their ability to protect against infection, is not homogeneous, and that this is true even for the same subtype of antigen-specific antibodies. This means that the traditional methods used for measuring antibody titer, i.e., concentration of antibody with a secondary antibody that recognizes the constant region of the antibody (see illustration in
Figure 1A), cannot evaluate the overall protective immunity of the antibodies, because such assays do not measure the strength of the binding reaction to the antigen carried by the variable region of the antibody. As shown in
Figure 1B, the RBD antigen binding avidity of antibodies is quantitatively measured by the competitive binding inhibition of the antibodies between the immobilized SARS-CoV-2 S1 protein containing the RBD on the DCP microarray and the serially diluted soluble RBD protein in the test sample, and the IC50 (nM) value is determined after washing to remove unbound anti-RBD antibodies. IC50, the half-maximal inhibition concentration, represents the concentration of RBD protein required for 50% binding inhibition. Using this method, the antigen binding avidity was defined as the level of 1/IC50 (nM). We then used antigen binding avidity antibody titer (ABAT) as an index of protective immunity, which is expressed by the antibody quantity multiplied by the quality (antigen binding avidity, 1/IC50). Low IC50 values represent high binding avidity IgG while high IC50 values represent low binding avidity IgG (
Figure 1B).
Increases in the maturation of the antigen binding avidity of antipathogen-specific IgG to the target epitopes are triggered by repeated vaccinations or by infections. This is due to B cell maturation over time that occurs in the germinal center, leading to maturation of antibodies and enhancing their binding avidity for their cognate antigen [
11].
Here, we describe our new method to evaluate the maturation of the protective immunity of anti-SARS-CoV-2-specific antibodies by measuring the RBD binding avidities and their ABATs. Our results showed that the increase in infection protection potency of anti-SARS-CoV-2 specific antibodies following SARS-CoV-2 infection or repeated COVID-19 messenger RNA (mRNA) vaccination, as expressed by ABAT values, can be classified into four groups: anti-SARS-CoV-2 specific antibodies below the limit of detection and three anti-SARS-CoV-2 specific antibodies with different levels of maturity of infection defense capacity; (i) the group of ABAT antibodies with an initial RBD binding avidity, (ii) the group of ABAT antibodies with low RBD binding avidity, and (iii) the group of ABAT antibodies with high RBD binding avidity. These ABAT antibodies correlated closely to the antiviral neutralizing activities and also with the clinical severity of hospitalized COVID-19 patients.
2. Materials and Methods
2.1. Study Population and Plasma Samples
Serum samples were obtained from 163 healthy vaccinated volunteers (hospital staff of National Hospital Organization Mie National Hospital) aged 21–71 years. All subjects had received the first dose of the BioNTech/Pfizer mRNA vaccine (COMIRNATY
®/BNT162B2) in March 2021, followed by second vaccination three weeks later, third vaccination at 38–40 weeks, and fourth vaccination at 71–74 weeks after the first vaccination. Blood samples were collected 3–4 weeks after each vaccination (
Table 1). We also included in this study 339 unvaccinated COVID-19 patients, aged 14 to 98 years, who had been admitted to three separate hospitals for acute SARS-CoV-2 infection, as confirmed by examination of pharyngeal swabs by polymerase chain reaction (PCR) test, between the period of April 2020 to August 2021 prior to the Omicron variant epidemic in two distant regions of Tokyo and Okinawa in Japan [
18,
19]. Based on the admission interview, these patients were determined to be not previously infected with SARS-CoV-2 and naïve to specific treatment for COVID-19, including remdesivir or favipiravir [
20,
21]. Disease severity was determined based on the COVID-19 severity classification developed by the Ministry of Health, Labor, and Welfare (MHLW) of Japan [
22]. Sera collected from the COVID-19 patients and vaccinated individuals were stored at −20 °C until analysis.
2.2. Ethics Statement
The study protocol, informed consent, and data collection of the healthy vaccinated volunteers and COVID-19 patients were reviewed and approved by the Institutional Review Board of the Tokushima University Hospital (#4067 and #4068-1), and a written informed consent was obtained from each participant.
2.3. Antigens and Reagents
HEK293 expressing recombinant His Tag SARS-CoV-2 S1 of the original Wuhan strain (Accession# QHD43416.1, Val16–Arg685) and S protein RBD (Arg319–Phe541) were purchased from ACROBiosystems (Newark, DE, USA). Antihuman IgG-Fc fragment antibody (goat polyclonal) was from Bethyl Laboratories, Inc., Montgomery, TX, USA).
2.4. Anti-Spike S1 IgG Quantification
Serum titers of anti-spike S1-specific IgG were measured by ELISA using 96-well plates [
23] or DCP microarrays, as described in detail previously [
13,
14,
15,
16,
17]. RBD binding avidity was also measured using DCP microarrays. A 96-well plate (Nunc, Naperville, IL, USA) coated with S1 protein (0.1 μg/well) in ELISA Coating Buffer (Bethyl Laboratories) was used for ELISA assay. Anti-spike S1-specific IgG titers were also measured by fluorescence immunoassay on the DCP microarray. Briefly, S1 antigen (4.8 ng) in phosphate-buffered saline (PBS) spotted on the DCP microarray was incubated for 1 h with 8 μL of 1:20 to 1:6000 diluted serum, washed three times with TTBS (50 mM Tris-HCl, pH 7.5, containing 150 mM NaCl and 0.05% Tween 20), rinsed with deionized water, and then reacted with a HiLyte Fluor 555 (Dojindo Molecular Technologies, Inc., Kumamoto, Japan)-labeled secondary goat polyclonal antibody against human IgG-Fc fragment at a concentration of 30 ng/mL in dilution buffer (0.5% bovine serum albumin, 0.05% Tween 20 and 0.3 M KCl in PBS) for 1 h. The resulting images were acquired by scanning the microarrays with an InnoScan 710 (Innopsys, Carbonne, France), and converted to numerical fluorescence intensity. Using standard concentrations of purified human IgG (Fujifilm Wako Pure Chemicals, #143-09501, Osaka, Japan), the amounts of IgG bound to the microarrays were calculated and expressed as binding units (BUs). The detection limit is 200.0 BUg/mL under the present assay conditions. We reported previously the intra- and inter-assay coefficients of variation for the DCP microarray assay as follows: inter-assay: 5.70% to 13.5%, within-slide: 7.70% to 25.2%, and batch-to-batch: 2.7% to 24.4% [
17]. As shown in
Supplementary Figure S1, the S1-specific IgG titers analyzed by the DCP microarray correlated strongly with those determined by ELISA (
r = 0.8964).
2.5. RBD Binding Avidity Measurement of Anti-Spike S1 IgG Antibodies
The level of anti-S1-IgG in each serum sample bound on the DCP microarray was adjusted at fluorescence intensities between 5000 and 60,000 by dilution with PBS containing 0.3 M KCl and 0.05% Tween 20 before analysis of RBD binding avidity. RBD protein at 0, 2, 20, 200 ng as a binding competitor of anti-spike S1 IgG antibodies was then added to the reaction mixture (16 μL) of serum and then preincubated at 37 °C for 30 min. After the preliminary reaction, each reaction mixture (8 μL) was applied to an S1-immobilized DCP microarray and competitive binding assay was conducted under equilibrium conditions at 37 °C for 1 h. After washing three times with TTBS, rinsing with deionized water, followed by spin-down drying, the antibodies on the DCP microarray were detected by reaction with a HiLyte Fluor 555-labeled secondary antibody against human IgG-Fc fragment at a concentration of 300 ng/mL at 37 °C for 1 h. After the same washing and drying as described above, we measured the fluorescence intensity using a scanner, followed by calculation of IC50.
2.6. Neutralizing Activity of Antibodies against Pseudovirus
Lenti-X 293T cells and human ACE2 293T cells were cultured in DMEM (high glucose: 4.5 g/L), containing 4 mM L-glutamine and 3.7 g/L sodium bicarbonate, 10% fetal bovine serum, 100 units penicillin G sodium, 100 μg/mL streptomycin sulfate, and 1 mM sodium pyruvate. Lenti-X™ SARS-CoV-2 Packaging Single Shots (D614G Spike, Truncated) (Takara Bio Inc., Shiga, Japan) were prepared with Lenti-X 293T cells, according to the protocol provided by the manufacturer. The concentrations of pseudoviruses incorporating SARS-CoV-2 (mutation of D614G, deletion of the final 19 amino acids of the C-terminus) in the pLVXS-ZsGreen1-Puro Vector were measured using Lenti-X GoStix Plus (Takara Bio Inc.).
To measure the pseudovirus-neutralization activity of anti-S1-IgG antibodies, human ACE2 293T cells were seeded onto 96-well plates at 4 × 104 cells per well and infected with pseudovirus after 24 h incubation at 37 °C. Pseudovirus (100 ng/well) and serum (dilution: 25-fold to 210-fold) in the culture medium were incubated for 1 h at room temperature to infect the above human ACE2 293T cells in the presence of 6 μg/mL of polybrene followed by 72 h incubation. Before measurement, blood samples were classified into three categories of binding avidity (1/IC50) maturation according to the correlation between anti-SARS-CoV-2 IgG antibodies and ABAT values, described in detail later: initial binding avidity (0.001–0.004 nM), low binding avidity (0.005–0.009 nM), and high binding avidity (0.053–0.077 nM). After culture, the medium was aspirated, and the cells were fixed with 4% paraformaldehyde for 30 min at room temperature. After washing with PBS, the ZsGreen1-expressing infected cells were photographed with a fluorescence microscope (BZ-X700, KEYENCE Corp., Osaka, Japan) and counted with the analysis application software provided with the microscope. The number of infected cells in the absence of serum was set as 100%, and the number of infected cells after treatment with pseudovirus and neutralizing serum in each assay group was counted.
2.7. Statistical Analysis
Receiver operating characteristic (ROC) curve and area under the curve (AUC) statistical analyses were performed using JMP software, version 14.2.0 (SAS Institute Inc., Cary, NC, USA). Correlation analyses were performed using Excel software (Microsoft Corp, Redmond, WA, USA). r ≥ 0.7 was considered high correlation. Data were expressed as mean ± SEM.
4. Discussion
It has become clear in recent years that even the same concentration of antibodies has diverse binding potencies to antigens. The defense capability of IgG against infection is also diverse, and the antibody titer, which is estimated by using a secondary antibody that recognizes the constant region of IgG molecules and measures the concentration of specific IgG molecules, is insufficient for the full assessment of the diverse defense capability of the IgG. The ABAT parameter evaluated in this study assesses more accurately and directly the ability to capture antigens as an antibody function using the antigen binding avidity parameter. As shown in
Supplementary Figure S2, the RBD binding avidity (1/IC50 nM) correlated highly (
r = 0.811) with the virus-neutralizing activity of the antibodies using pseudovirus. The RBD binding avidity assay system uses the DCP microarray [
13,
14,
15,
16,
17], which has been certified in Japan for quantitative antigen-specific IgE analysis, based on the high-quantification data and reproducibility.
Assessment of the defense capability of anti-COVID-19 antibodies by ABAT has the following advantages compared to the pseudovirus-neutralization assay using the cell culture system. (i) Short processing time: The results of the conventional method of virus-neutralizing activity using pseudovirus and cultured cells become available after a minimum period of 3–5 days from the start of cell culture, whereas the results of the ABAT assay by DCP microarray are ready within about 5 h. (ii) Several hundred specimens can be processed simultaneously and rapidly in the ABAT assay, whereas this is practically difficult in the virus-neutralizing activity using the cell culture system. (iii) Using the DCP microarray, the ABAT assay provides a comparable unit of antigen binding avidity and quantitative measure (1/IC50 RBD antigen concentration), instead of the 50% viral neutralization activity represented by the dilution fold of the specimen. In other words, the ABAT assay allows comparison of different specimens.
Since the protection potential of anti-COVID-19 antibodies could be quantitatively expressed by the RBD binding avidity (1/IC50 nM) and the ABAT value, it is possible to classify the antibodies into three categories based on differences in the binding avidity maturation (see
Figure 4). A high RBD binding avidity and ABAT values of antibodies with the highest protection capacity against infection could be set as goal target values for vaccination and protective antibodies in the infected patients. By using these data, it could be possible to estimate the need for additional vaccination(s) and assess clinical outcome in patients with severe infection. In patients with high RBD binding avidity on admission, none was severely ill and 77.8% were mildly ill at 2–3 weeks of hospitalization or at discharge (
Figure 7). In contrast, 77.2% and 56.2% of patients with low and initial RBD binding avidity on admission had moderate and severe illness, respectively, at 2–3 weeks of hospitalization or at discharge. Thus, the RBD binding avidity and ABAT well reflect the severity of illness in hospitalized patients. A similar relationship between avidity index analyzed by chaotropic avidity and severity of SARS-CoV-2 patients was reported by Georg Bauer [
8].
Maturation of the RBD binding avidity of antibodies induced by SARS-CoV-2 infection for 2–3 weeks was weaker than those at 3–4 weeks after the first vaccination with BioNTech/Pfizer mRNA, and approximately 32–36% of the patients showed neither an increase in antibody titer nor RBD binding avidity. These results suggest that the immunogenicity of the live SARS-CoV-2 virus is lower than the mRNA vaccine. Most patients showed low and initial RBD binding avidity, and high RBD binding avidity was noted in only 6% of these patients. Based on the above findings in the infected patients, we recommend mRNA vaccination of the patients before or after discharge to achieve antiviral antibody ABAT levels that could prevent rehospitalization.
As shown in
Figure 2B, the speed of maturation of the RBD binding avidity, which reflects the ability to protect against infection, varied greatly among individuals. Although most of the vaccinated individuals (90.2%) achieved 1/IC50 of 0.02 nM or higher at 3–4 weeks after the second vaccination, some (9.8%) showed slow maturation that did not reach 0.02 nM. However, even among vaccinated subjects who showed slow maturation, all achieved 1/IC50 of 0.02 nM or higher at 6–8 months after vaccination without booster vaccination. These findings suggest that the second and subsequent vaccinations can be applied after a long interval of 6–8 months, anticipating the maturation of the pre-existing immunity. This result supports the concept of “longer vaccination interval improves SARS-CoV-2 neutralization” reported recently [
24].
Our study has certain limitations. First, we could not follow up changes in antibody titer, RBD binding avidity, and ABAT values after 2–3 weeks of hospitalization; thus, future long-term follow-up studies are planned. Second, the number of study subjects needs to be increased in the future to determine the effect of the underlying disease on the analyzed parameters. Third, the present study was conducted before the emergence of SARS-CoV-2 Omicron variants, but it is necessary to investigate the protective potential and cross-reactivity of SARS-CoV-2 wild-type antibodies against antibodies generated during infection with SARS-CoV-2 Omicron variants, using antibody titers, RBD binding avidity, and ABAT values.
In conclusion, we described in this report the development of a new method to measure the antigen binding avidity/affinity of antibodies involved in infection defense capacity. Conventional methods of determination of antibody titer, which measure antigen-specific antibody levels, do not provide the diverse antigen binding potencies, which are well found among antibodies. In this study, the antigen binding capacity expressed by the variable region of the antibody was measured by the competitive binding inhibition (IC50) between the antigen immobilized on the DCP microarray and the soluble antigen added to the test sample. Thus, the quantity × quality (binding avidity) of antibodies, ABAT, was successfully used to quantify the ability of the antibody to protect against infection. The values of RBD binding avidity correlated strongly and significantly with cell-based virus-neutralizing activity. In addition, the new method allows the monitoring the maturation process of RBD binding avidity, as well as the classification of antibodies into three categories based on the differences in antibody maturation: initial, low, and high ABAT. Furthermore, once high RBD binding avidity was achieved, it was maintained for at least 6–8 months regardless of the subsequent decrease in antibody levels. These characteristics may prove useful in formulating effective vaccination programs.