Evaluation and correlation between SARS-CoV-2 neutralizing and binding antibodies in convalescent and vaccinated subjects.

Since the first detection of a novel Coronavirus (SARS-CoV-2) in December 2019 in Wuhan (China), it has become crucial to assess and quantize the human humoral immune response after SARS-CoV-2 natural infection and/or vaccination. Having well standardized and reliable serological assays able to accurately measure the total IgG antibodies response as well as the neutralization dynamics, play a pivotal role for the evaluation of "second" and "third" vaccines generation and in monitoring the effect in case of reinfection in the human population caused by the original strains or new SARS-CoV-2 variants. In the present study we reported that both symptomatic convalescent and vaccinated donors showed the presence of different levels of neutralizing antibodies. In addition, vaccinated subjects presented high levels of anti-S antibodies, whereas the complete absence of anti-N antibodies, whereas convalescent patients presented high levels of both anti-S and anti-N antibodies. The evaluation of the correlation between SARS-CoV-2 neutralizing and binding antibodies in convalescent and vaccinated subjects revealed that the IgG anti-S log-values were significantly higher in the vaccinated group respect to convalescent subjects. In addition, the level of binding antibodies recognizing the S protein shows a positive linear regression when compared to neutralizing titres in both the two groups evaluated.

At the end of 2019, a novel beta-coronavirus was identified for the first time in Wuhan City, Hubei province in China and named severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) (Rodriguez-Morales et al., 2020). Since its first detection, this new pathogen has spread rapidly throughout the country reaching all continents with the exception of Antarctica and causing an ongoing pandemic with about 230.000.000 of confirmed cases and 4.700.000 deaths worldwide. In addition to the virus isolated in Wuhan (wild-type strain), novel SARS-CoV-2 variants, some of which identified as variants of concerns (VOCs) for to their significant impact on transmissibility, severity and/or immunity, which probably could modify the epidemiological situation (https://www.ecdc.europa.eu/en/covid-19/variants-concern, n.d), have been developing over the course of the pandemic. These additional variants of SARS-CoV-2 have furtherly raised the global effort for the development of an effective vaccine as well as acute antiviral drugs for the treatment of medium-to-severe stages of coronavirus disease 2019 (COVID-19). To date immunization represents the best strategy to prevent further morbidity and mortality. In the recent months major advances have been done in setting, improving and validating different serological assays for better understanding the humoral response after SARS-CoV-2 infection. Serological assays could be crucial to monitor the disease incidence in a population, allowing the identification of the proportion of individuals exposed and to determine the level of neutralizing antibodies necessary to provide some degree of protection against reinfection by the virus (Kellam and Barclay, 2020). Indeed, with the development and the successive massive administration of new SARS-CoV-2 vaccines, it has become essential to have reliable serological tests able to provide clear information on neutralization capability, avidity, abundance and decay over time of such antibodies. Different classes of antibodies have a pivotal role in the antibody-mediated immunity. Immunoglobulin M (IgM) are generally the first class to be secerned, representing almost 10% of all serum antibodies and showing to have quite high avidity towards the antigen. Immunoglobulin G (IgG), due to the affinity maturation process, is the last class to appear in the antibody-mediated immune response (Kellam and Barclay, 2020). Due to this accurate process of maturation via somatic mutation IgG antibodies present high affinity towards the antigen which results in an elevated neutralization capacity inhibiting viral infection. They represent almost 75% of all serum antibodies and are associated with the long-lasting immunity. IgA are the main responsible for mucosal immunity as a dimer, even if they are present also at systemic level in monomeric form. The majority of serological assays designed and currently available are able to detect antibodies, mainly IgG and IgM in serum/plasma samples, directed towards the Spike (S) protein, the S receptor-rinding domain (RBD) or the Nucleoprotein (N) of SARS-CoV-2. The S protein, in particular the RBD, is the main target of neutralizing antibodies due to its intrinsic biological functions in mediating the viral attachment, fusion, entry and transmission in host cells expressing the angiotensin converting enzyme 2 (ACE2) (Yan et al., 2020). On the contrary, even if the N protein is involved in many important functions associated with viral RNA packaging, transcription and replication, the majority of antibodies elicited against this epitope are not neutralizing. This may be due to the fact that N is not involved in the first step of attachment/entry of the viral particles into the target cells. To date serological assays can be distinguished into two main groups: those detecting all antibodies able to bind the antigen of interest (binding antibodies) and those able to detect functional neutralizing antibodies. Solid-phase immunoassays, such as enzyme-linked immunosorbent assay (ELISA), Electrochemiluminescence Immuno-Assay (ECLIA) and Chemiluminescent Immuno-Assay (CLIA) are the widely used tests in order to detect binding antibodies in human and animal serum/plasma samples. They present many advantages, including high throughput, possibility of automation, do not require the use of dangerous reagents and/or live pathogens and are cheap. One of the main disadvantages is that they are not able to provide information about the functionality of the antibodies detected. On the other hand, we have the neutralization assays. They are more labor-intensive tests requiring the use of live authentic SARS-CoV-2 viruses and for this reason they need to be strictly performed inside Biosecurity level 3 laboratories by highly qualified personnel; but to date, these are the only assays able to provide the information regarding the neutralizing ability of the antibodies present in a given sample. Serum samples from 18 mRNA-Comirnaty (Pfizer-BioNTech) vaccinated volunteers were collected through an internal survey from VisMederi Srl (Siena, Italy) and the SMILE trial (Screening and Multiple Intervention on Lung Epidemics, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy) at least two weeks after the second dose. This mRNA vaccine, administered as intramuscular injection, is able to provide the genetic instruction to human cells for building the SARS-CoV-2 S protein, which than get released into the body provoking a response from the immune system. Forty-three samples from convalescent patients were kindly provided by the University of Milan (UNICORN study), Fondazione IRCCS Istituto Nazionale Tumori also including samples of the commercial human Panel F (Cambridge Biosciences). A panel of pre-pandemic samples collected in 2015 as part of routine medical checks and research projects, stored as residual sera in compliance with Italian ethics law, were provided by the University of Siena and used as negative controls. All samples were received and analyzed anonymously. The study was conducted in accordance with the Declaration of Helsinki and informed consent to use the samples for research was obtained from all subjects involved in the study. Immunoglobulin levels towards the S and N antigens of SARS-CoV-2 were evaluated by using the Elecsys® enzyme immunoassay (Roche Diagnostic International Ltd), whereas neutralizing antibodies were evaluated through the in-house Micro-Neutralization (MN) assay as previously described (Manenti et al., 2020) using the wild type Wuhan strain. Briefly, 2-fold serial dilutions of inactivated sera samples have been incubated for 1 h with a standardized amount of SARS-CoV-2 live virus. After the incubation, the serum-virus mixture has been passed in a VERO E6 pre-seeded culture plate and incubated for three days. After the incubation time the plates have been inspected for the presence/absence of the cytopathic effect (CPE) by means of an inverted microscope. The MN titre is expressed as the last serum dilution able to prevent the CPE in the VERO E6 monolayer. The Elecsys® Anti-SARS-CoV-2 is an IVD immunoassay for the quantitative (S) and qualitative (N) detection of anti-SARS-CoV-2 antibodies in human and plasma samples. The assay uses a recombinant protein (spike or nucleocapsid) in a double-antigen sandwich format ables to detect high affinity antibodies. The aim of the present work was to dissect, evaluate and compare the immune response and in particular antibody levels of both binding and neutralizing antibodies by using an approved in vitro diagnostic (IVD) assay and a lab-based assay, respectively, in the two different groups (healthy vaccinated subjects vs naturally infected subjects). Specifically, our main interest was focused on evaluating whether, in presence of equal neutralizing titers, there were differences in the quantity of total antibodies elicited in the two groups. As already demonstrated by different studies (Rogers, 2020; McMahan et al., 2021) there is a close relation between the level of neutralizing antibodies and the protection from the COVID-19 disease. Even if, at present, no correlates of protection are available, neutralizing antibodies are used as surrogate in order to predict a certain level of protection from the disease.as In addition, other aspects that should be investigated more deeply, but that are under evaluation are the duration of protective immunity, the T cell mediated immunity and the memory B cells (Tan et al., 2020) (McMahan et al., 2021). All statistical analyses were performed using Software R version 4.0.3. In order to be included in the data analysis, all the results had to meet the internal quality control criteria of the assays. In case a sample did not meet the internal quality control it was re-tested and if the internal quality control was not met even after the retest, the sample was discarded from the statistical analysis. All samples derived from symptomatic convalescent and vaccinated donors showed the presence of neutralizing antibodies at different levels, apart one vaccinated subject showing neutralizing antibodies below the threshold of detection. All vaccinated subjects presented high levels of anti-S antibodies, whereas the complete absence of anti-N antibodies; on the other hand, convalescent patients presented high levels of anti-S and presence of anti-N antibodies, as expected. All pre-pandemic (control) samples showed negative results obtained from both binding and neutralization assays. These results support previous findings reporting a robust mRNA vaccine-mediated antibody response. Since we do not have precise indications regarding the time of sampling after natural infection for all subjects, it is not possible to compare the immune response in terms of absolute titres achieved from both cohorts as well as to evaluate the kinetics of decay of the immune response. What we did is the intra-cohort comparison between the total amount of binding antibodies and the neutralizing titres elicited after SARS-CoV-2 natural infection or vaccination. A linear regression analysis was performed on the log-transformed data to assess the relationship between IgG anti-S and MN Geometric Mean Titre (GMT) measurements. The slope of the regression line ina log-log approach can be interpreted in terms of responsiveness of the IgG anti-S outcome respect to the MN assay results. After log transformation of the data we assessed the normality assumption. The results of the Anderson-Darling test led us to not reject the null hypothesis of normality neither for the group of vaccinated subjects (A = 0.63, p value = 0.08) or for the group of convalescent subjects (A = 0.54, p value = 0.16). A t-test for independent samples was conducted comparing the IgG anti-S measurements between the two groups. The IgG anti-S log-values (mean = 7.05 U/ml, sd = 1.08 U/ml) were significantly higher in the vaccinated group (t = 5.54, degrees of freedom = 36.25, p value<0.0001) compared to the group of convalescent subjects (mean = 5.29 U/ml, sd = 1.24 U/ml). The level of binding antibodies recognizing the S protein shows a positive linear regression when compared to neutralization titres in both the two groups evaluated (Fig. 1 , Table 1 ), in accordance with previous studies (Luchsinger et al., 2020).

Fig. 1

Log-log regression of anti-S antibody levels and GMT of the MN titres. The anti-S antibody levels were averaged over the levels of the GMT.

Table 1

Summary of log-log regression analysis in the groups of vaccinated subjects and non-vaccinated convalescent subjects.

Group	Parameter	Estimate	Std.Error	t-stat	Pr(>|t|)	Significance	Adjusted R-squared
Vaccinated	Intercept	37.468	0.576	6.505	0.000629	***	0.83
	log(GMT)	0.7214	0.1208	5.971	0.000989	***
Non-vaccinated convalescent	Intercept	29.364	0.8219	3.573	0.0117	*	0.64
	log(GMT)	0.7237	0.1967	3.679	0.0103	*

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘’ 1.

Interestingly, as is possible to observe from Fig. 2 a significative difference in the average of the total amount of binding antibodies respect to the neutralization titers between convalescent and vaccinated subject was observed. Even if the sample size of this preliminary study is small, our results seem to indicate that at comparable neutralization titres, the vaccinated group has a much higher level of total IgG against the whole S protein. Holding the neutralizing titre as a fix point, the level of binding antibodies elicited after natural infection was lower respect to that induced by vaccination. Only for two vaccinated subjects we found quite low neutralizing titres, but they showed detectable levels of binding antibodies. This finding could be explained by a low avidity/affinity maturation of antibodies respect to SARS-CoV-2 S protein.

Fig. 2

Anti-S antibodies in the groups of vaccinated subjects and non-vaccinated convalescent subjects; vaccinated subjets (light grey column) presented a significantly higher average concentration (log(U/mL)) of binding IgG antibodies respect to convalescent subjects (dark grey column).

These results confirm furtherly the data available regarding the significant relationship between S-specific binding assay, such as ELISA or ECLIA, and neutralization results. The regression lines are substantially parallel with slopes equal to 0.72, which implies that the fractional response of anti-S antibodies to a fraction change in GMT is similar in both groups. The differences found in the absolute quantity of anti-S antibodies in vaccinated subjects respect to convalescents may be due to the fact that the concentration of S antigen presented to the immune system after SARS-COV-2 natural infection results to be inferior than the actual time of availability of the antigen after vaccination. This is in line with the observation that higher IgG levels as well as avidity indices were found in patient with more several disease respect to mild/paucisymptomatic patients (Moura et al., 2021). If our study characterized the antibody response distinguishing between total anti-S antibody and neutralizing antibodies after SARS-CoV-2 mRNA vaccination and natural infection, it presented several limitations. The main limit is represented by the small size of our cohort and by the non-homogeneous sample size within the groups, which render necessary to confirm such results on a larger number of subjects also considering disease characteristics and humoral immune response induced by different SARS-CoV-2 variants. Due to the objective difficulties to have, especially for the convalescent group, the exact time of infection for all subjects, we lack time-dependent information which could have given indication about the kinetics of the immune response. Another interesting point would be the evaluation of antibody levels over time in the two groups as well as the antibody avidity index in both groups.