key: cord-0933365-hewenznd
authors: Jayathilaka, Deshni; Jeewandara, Chandima; Gomes, Laksiri; Jayadas, Tibutius Thanesh Pramanayagam; Kamaladasa, Achala; Somathilake, Gayasha; Guruge, Dinuka; Pushpakumara, Pradeep Darshana; Ranasinghe, Thushali; Aberathna, Inoka Sepali; Danasekara, Saubhagya; Gunathilaka, Buddini; Kuruppu, Heshan; Wijewickrama, Ananda; Wijayamuni, Ruwan; Schimanski, Lisa; Tan, T K; Ogg, Graham S; Townsend, Alain; Malavige, Gathsaurie Neelika
title: Kinetics of immune responses to SARS-CoV-2 proteins in individuals with varying severity of infection and following a single dose of the AZD1222
date: 2022-01-27
journal: Clin Exp Immunol
DOI: 10.1093/cei/uxac009
sha: 83636ee1a72fd364919b660ad5bdc252699abe4d
doc_id: 933365
cord_uid: hewenznd

To characterize the IgG and IgA responses to different SARS-CoV-2 proteins, we investigated the antibody responses to SARS-CoV-2 following natural infection and following a single dose of AZD1222(Covishield), in Sri Lankan individuals. The IgG and IgA responses were assessed to S1, S2, RBD and N proteins in patients at 4 weeks and 12 weeks since onset of illness or following vaccination. Antibodies to the receptor binding domain of SARS-CoV-2 wild type (WT), alpha, beta and delta and ACE2 (Angiotensin Converting Enzyme 2) receptor blocking antibodies were also assessed in these cohorts. Those with mild illness and in vaccinees, the IgG responses to S1, S2, RBD and N protein increased from 4 weeks to 12 weeks, while it remained unchanged in those with moderate/severe illness. In the vaccinees, IgG antibodies to the S2 subunit had the highest significant rise(p<0.0001). Vaccinees had several fold lower IgA antibodies to all the SARS-CoV-2 proteins tested than those with natural infection. At 12 weeks, the Haemagglutination test (HAT) titres were significantly lower to the alpha in vaccinees and significantly lower in those with mild illness and in vaccinees to beta and for delta. No such difference was seen in those with moderate/severe illness. Vaccinees had significantly less IgA to SARS-CoV-2, but comparable IgG responses those with natural infection. However, following a single dose vaccinees had reduced antibody levels to the VOCs, which further declined with time, suggesting the need to reduce the gap between the two doses, in countries experiencing outbreaks due to VOCs.

The COVID-19 pandemic due to the SARS-CoV-2 virus, continues to cause significant mortality and morbidity and many countries are experiencing a worse situation, than experienced at the beginning of the pandemic 1 . Emergence of SARS-CoV-2 variants of concern such as the B. 1 . 1.7 (alpha) and more recently B. 1.617.2 (delta) has led to exponential increase of the number of COVID-19 cases and deaths in many countries [1] [2] [3] . While the higher income countries have vaccinated a large proportion of their population, resulting in lower case numbers, many lower income and lower-middle income countries are grappling with the increase in the case loads, overburdening of health care resources and the inability to secure adequate doses of COVID-19 vaccines 4 .

Although the duration of protection against re-infection from SARS-CoV-2 is not known, it has been shown that re-infection does occur, especially among older individuals, probably due to waning of immunity 5 . Re-infection has shown to occur particularly with certain variants such as P.1 (gamma) variant in Brazil despite a very high seroprevalence 6 , and also with B. 1 .351 (beta) due to escape natural and vaccine induced immunity 7 . Individuals who had experienced milder illness have shown to have reduced levels of neutralizing antibodies compared to those who had severe illness 8, 9 . Apart from the presence of neutralizing antibodies to the receptor binding domain (RBD), antibodies specific to S2 and N protein of SARS-CoV-2 are also detected in patients who have recovered from COVID-19 10 . However, the usefulness of antibodies directed against S1, S2 and N protein in preventing re-infection are not known. The IgG and IgA specific to S1, S2 have been detected in breast milk of infected mothers and therefore, possibly provide protection to the neonate 11 . Antibodies against the S2 subunit have been detected in unexposed individuals and S1, S2 and N protein A c c e p t e d M a n u s c r i p t 5 specific memory B cell responses have been detected in those who were infected with SARS-CoV-2 12 . Children and adolescents who were unexposed to SARS-CoV-2 were shown to have a higher frequency of pre-exiting IgG antibodies specific to S2, which were able to cross neutralize SARS-CoV-2 13 . The presence of high levels of cross-reactive antibodies to the S2 in children and adolescents have been speculated to reduce disease severity when infected with SARS-CoV-2 13, 14 . Although many studies have investigated the role of SARS-CoV-2 specific IgG responses, virus specific IgA was detected during early illness and was shown to be able to neutralize the SARS-CoV-2 virus to a greater extent than virus specific IgG 15 .

However, adults with severe illness had higher levels of SARS-CoV-2 specific IgA levels compared to adults with milder illness and children, which was shown to enhance disease severity in vitro by enhanced neutrophil activation and thus release of inflammatory mediators 16 . Therefore, although virus specific IgA is an important component of mucosal immunity, its role in protection vs disease pathogenesis is not clear. Further, the role of serum IgA, in contrast to mucosal IgA have not been studies extensively.

Currently there are several vaccines for COVID-19, which have shown to be safe and have high efficacy rates against the original Wuhan SARS-CoV-2 virus and variants of concern 17-19 . However, due to non-availability of adequate quantity of vaccines and also in order to vaccinate as many individuals as fast as possible, some countries have increased the gap between the two doses of vaccine such as AZD1222 to 12 or 16 weeks 20 . While there have been many studies characterizing the IgG and IgA responses to different SARS-CoV-2 proteins in individuals with natural infection, the induction of IgG and IgA to different viral proteins in vaccinees have not been extensively studied. It was recently shown that the mRNA vaccines induce high levels of both IgG and IgA antibodies against the spike protein 21 . However, there is limited data characterizing the IgG, IgA, ACE2-receptor blocking M a n u s c r i p t 6 antibodies in individuals with varying severity of natural infection over time, in comparison to those who have received a single dose of the AZD1222 vaccine. Therefore, in this study, we investigated the antibody responses in those with varying severity of natural infection and in those who received a single dose of the AZD1222 at 4 weeks and 12 weeks to the S1, S2, RBD and N proteins and also for SARS-CoV-2 variants of concern in a Sri Lankan population.

Patients confirmed of SARS-CoV2 infection based on the positive RT-PCR who were admitted to the National Institute of Infectious Diseases (NIID), Sri Lanka, were recruited following informed written consent. They were followed throughout their illness while they were in hospital and clinical disease severity was classified as mild, moderate and severe according to the WHO guidance of COVID-19 disease severity 22 . For this study we recruited two cohorts of patients (supplementary table 1) . Serum samples from the patient cohort 1 (n=30) was used to determine the IgG and IgA antibody levels at 4 weeks since onset of illness, the ACE2 receptor blocking antibody levels and the antibodies to RBD by the HAT assay for the wild type (WT) and SARS-CoV-2 variants. The duration of illness was defined from the day or onset of symptoms and not the day of PCR positivity or admission to hospital. Based on the WHO COVID-19 disease classification, 15 patients had mild illness and 15 patients had moderate/severe illness 22 . As all the patients in the first cohort could not be traced at 12 weeks, in order to carry out the above assays, we recruited a second cohort of patients. Based on the WHO COVID-19 disease classification, 14 patients had mild illness and 6 patients had moderate/severe illness 22 . A c c e p t e d M a n u s c r i p t 7 In order to compare the antibody responses following infection with one dose of the AZD1222 vaccine, we recruited 20 individuals 4 weeks following vaccination and same 20 individuals were followed at 12 weeks following vaccination. All 20 individuals who were included at 4 weeks following vaccination were included at 12 weeks following vaccination as well. We also included serum samples from individuals who had a febrile illness in 2017 and early 2018. Ethical approval was received by the Ethics Review Committee of Faculty of Medical Sciences, University of Sri Jayewardenepura. Informed written consent was obtained from patients.

Luminex assay to measure SARS-CoV-2 S1, S2, RBD and N specific IgA and IgG antibody responses SARS-CoV-2 S1, S2, RBD and N specific IgA and IgG antibody responses were measured separately using multiplex SARS-CoV-2 antigen panels IgG and IgA (Millipore). The assay was carried out according to manufactures instructions. The mean fluorescence intensity (MFI) was measured in each serum sample using MAGPIX® which was positively correlated with S1, S2, RBD and N specific IgG and IgA in serum.

The HAT was carried out as previously described 23 1.617.2 versions of the IH4-RBD reagent were produced as described 23 , but included the relevant amino acid changes introduced by site directed mutagenesis.

These variants were titrated in a control HAT with the monoclonal antibody EY-6A (to a conserved class 4 epitope 23, 24 ) and found to titrate identically with the original version so A c c e p t e d M a n u s c r i p t antibodies. The RBD-specific antibody titre for the serum sample was defined by the last well in which the complete absence of "teardrop" formation was observed.

The surrogate virus neutralization test (sVNT) 26 , which measures the percentage of inhibition of binding of the RBD of the S protein to recombinant ACE2 26 (Genscript Biotech, USA) was carried out according the manufacturer's instructions as previously described by us 9 .

Inhibition percentage ≥ 25% in a sample was considered as positive for NAbs.

Data was analysed by GraphPad Prism 9 version 9. 2.0 . The data was first tested for normality and homoscedasticity using Shapiro Wilk and Levene's tests and since the assumptions were violated, non-parametric tests were used for the analysis. Kruskal-Wallis test was used to determine the difference between the antibody levels between the three different groups (twotailed) followed by multiple comparisons using two-stage step-up procedure of Benjamini, Krieger and Yekutieli while controlling the false discovery rate (FDR) Mann-Whitney test (two tailed) was used to determine the differences between antibody levels between 4 weeks and 12 weeks in those with natural infection. Wilcoxon paired t-tests (two tailed) were used to determine the differences between antibody titres against S1, S2, RBD, N proteins and 

IgG responses to the S1, S2, RBD and N protein were measured in individuals with COVID-19 at 4 weeks and at 12 weeks since onset of illness and also in serum samples of 15 individuals who had a febrile illness in 2017 and early 2018. At 4 weeks since onset of illness, the highest magnitude of IgG antibody responses was seen for RBD in those with moderate/severe illness, whereas those with mild disease, had the highest responses to S2 ( Figure 1A , table 1). Those who had a febrile illness in year 2017 and 2018 (controls), also had high antibody levels to S2, but not for other proteins. There was no difference in the antibody levels to S2 in those with mild illness compared to the controls (p=0.213), although those with milder disease had significantly higher antibody levels to S1 (p=0.002) and RBD (p=0.0028) and N protein (p=0.0044), than the controls. Further, those with moderate/severe infection had significantly higher antibody titres compared to mild illness to all the proteins ( Figure 1A ). In those who received a single dose of the AZD1222 vaccine, the IgG responses to the S1 and S2 protein was similar, although the levels for the RBD was significantly higher (table 1) . As expected, the IgG responses to the N protein was very low, but even lower than for the controls. The antibody levels to S1 (p=0.0002), S2 (p=0.01), RBD (p=0.002) and N A c c e p t e d M a n u s c r i p t 10 (p<0.0001) proteins were found to be significantly different between the three groups of individuals at 4 weeks as resulted by Kruskal Wallis test ( Figure 1A ).

At 12 weeks since onset of illness, those with moderate/severe illness had the highest responses to N protein, whereas those with mild illness still had the highest responses to S2 ( Figure 1B) . At 12 weeks for all proteins, those with moderate/severe disease had significantly higher antibody levels than those with milder illness ( Figure 1B) . The antibody responses only to N protein (p=0.0137) was significantly different between the those with mild illness, moderate/severe disease and the vaccinees as resulted by Kruskal Wallis test ( Figure 1B) . From 4 to 12 weeks, the S1 and RBD specific antibodies rose in those with mild illness, although they were not significant (table 1) . Patients who had moderate/severe illness sustained the same levels of antibodies for all four proteins from 4 weeks to 12 weeks.

In the vaccinees, from 4 weeks to 12 weeks the IgG levels to S1 (p=0. (table 1) . IgA levels for S1 (p=0.004) and RBD (p=0.0262) were significantly higher than the control group in the vaccinees . However, at 4 weeks vaccinees had significantly lower IgA levels to all proteins compared to those who had moderate/severe infection ( Figure 1C ).

Significant differences of IgA responses were seen in those with mild illness, moderate/severe illness and vaccinees for S1 (p=0.001), S2 (p=0.0003), RBD (p=0.0003) and N protein (p=0.04) at 4 weeks as resulted by Kruskal Wallis test ( Figure 1C ).

There was no difference in IgA levels to any of the proteins at 4 weeks compared to 12 weeks in patients with mild illness or with moderate/severe illness (table 1) . However, at 12 weeks, no significant differences were seen between the three groups to S1, S2, RBD and N protein ( Figure 1D ).

Due to the lack of BSL-3 facilities to measure neutralizing antibodies, we used a surrogate test to measure the inhibition of binding of antibodies in patient sera to the ACE2 receptor 26 .

This was shown to be 100% specific in the Sri Lankan population, with none of the sera of individuals collected in 2017 and 2018, giving a positive response 9 . The ACE2 blocking antibodies were significantly higher in those with moderate to severe illness, when compared to those with mild illness at 4 weeks (p=0.0306) and at 12 weeks (p=0.0342) as reported previously (Figure 2 A c c e p t e d M a n u s c r i p t 14 and at 12 weeks (p=0.02) for WT At 4 weeks moderate/severe illness patients had significantly higher antibody titres to WT compared to those who had mild illness (p=0.0139) and those who were vaccinated (p=0.0005) ( Figure 4A) . This difference was also seen for the B.1. 1.7 (alpha) at 4 weeks between those with mild illness and moderate/severe illness (p=0.0381) and with those with moderate/severe illness and a single dose of the vaccine (p=0.0003). However, at 12 weeks those with a single of vaccine, had significantly low antibody titre compared to mild illness (p=0.0339) and moderate/severe illness (p=0.013) ( Figure 4B) . Similarly, antibody titres against B.1.617.2 (delta) too were higher in patients with moderate/severe illness compared to mild illness (p=0.0072) and vaccinated individuals with a single dose (p<0.0001). At 12 weeks those with moderate/severe illness had higher antibody titre only against vaccinated individuals (p=0.031) ( Figure 4D ). However, there was no difference between the antibody titres to the B. 1 .351 (beta) between those with mild, moderate/severe illness and vaccinees at 4 weeks and 12 weeks (p=0.02) ( Figure 4C ).

In this study we investigated the kinetics of IgG and IgA responses to S1, S2, RBD and N protein, ACE2 receptor blocking antibodies and antibodies against SARS-CoV-2 variants, in individuals at 4 and 12 weeks following natural infection and in those who had a single dose of the AZD1222. Based on the Luminex assays for IgG and IgA levels to S1, S2, RBD and N, IgG antibodies to these proteins following vaccination were increased significantly from 4 weeks to 12 weeks. In mild illness, although not significant, antibodies for S1 and RBD rose from 4 weeks to 12 weeks. In the vaccinees, the most significant rise was seen for the S2 subunit, while in those with mild illness the rise was seen for IgG antibodies for the RBD. In those with moderate/severe illness while there was no change in the IgG responses from 4 to A c c e p t e d M a n u s c r i p t 15 12 weeks, but the responses to the N protein had increased although this was not significant.

Unexpectedly, the antibodies against N proteins were also increased from 4 to 12 weeks, possibly due to asymptomatic infection in some individuals after a single dose of AZD1222 vaccine. Therefore, the kinetics of antibody responses to S1, S2, RBD and N appears to vary based on the severity of natural infection and also appeared to be different in vaccinees.

Interestingly, blood samples of those who had a febrile illness in 2017 and 2018 also gave IgG and IgA high responses the S2 subunit, suggesting the presence of S2 subunit cross reactive antibodies, in these donors as previously seen in other studies 13, 14 . Following a single dose of the AZD1222 vaccine, the antibodies against S2 appears to continue to rise from 4 to 12 weeks, possibly due to stimulation of pre-existing cross reactive memory B cell responses to the S2 subunit 14 .

SARS-CoV-2 specific IgA antibodies have shown to be generated during early illness and have potent neutralizing ability 15 . IgA antibodies to the RBD have shown to develop earlier than IgG and while some studies have shown that serum IgA does not associate with clinical disease severity 15 , patients who developed severe disease were shown to have higher levels of virus specific IgA 29 . Serum IgA was shown to activate neutrophil, thereby leading to production of increased levels of inflammatory mediators leading to disease pathogenesis 16 . We found that at 4 weeks of illness, those with moderate/severe illness had significantly higher serum IgA to S1, S2, RBD and N compared to those with mild illness, but these high levels of IgA declined except for S2 protein and there were no differences between these two groups at 12 weeks since onset of illness. Vaccinees had several fold lower IgA antibodies to all the SARS-CoV-2 proteins tested than those with mild and moderate/severe illness at 4 weeks and 12 weeks. The importance of serum IgA in preventing re-infection is currently unknown and if those with lower IgA have reduced protection is currently unknown.

A c c e p t e d M a n u s c r i p t 16 Although the IgG antibodies to S1, S2 and the RBD rose from 4 to 12 weeks in the vaccinees, the ACE2 receptor blocking antibodies, which were shown to correlate with neutralizing antibodies significantly decreased 26 . The HAT assay, which also measures antibodies to the RBD and has shown to correlate well with the ACE2 receptor blocking assay and with neutralizing antibodies 23, 28 , also showed that the RBD binding antibodies decreased from 4 to 12 weeks in the vaccinees. This suggests that although ACE2 receptor blocking antibodies are reduced, the antibodies that bound to S2 region increases which neutralize the SARS- These levels further declined at 12 weeks following vaccination, to VOCs, showing that a single dose of the AZD1222 was likely to offer less protection against VOCs. In fact, it has been shown that one dose of AZD1222 is only 33% effective in preventing symptomatic disease with B.1.617.2 (delta), 3 weeks following the first dose 30 . The efficacy of a single dose against B.1.617.2 (delta) is likely to decline further by 12 weeks, as the antibodies to RBD further wanned. However, the efficacy of two doses of AZD1222 against hospitalization was 92%, while for Pfizer-BioNTech was 96% 31 . Therefore, in countries which have outbreaks due to VOCs, especially B. 1.617.2 (delta) , it would be prudent to A c c e p t e d M a n u s c r i p t 17 reduce the gap between the two doses to increase efficacy as currently carried out in many countries. Interestingly, although those with mild or moderate/severe illness also had a marked reduction in antibodies to the RBD of B. 1 In summary, we have investigated the kinetics and differences in IgG and IgA antibody responses to the S1, S2, RBD and N in those with varying severity of infection and vaccinees who received a single dose of AZD1222, which showed that vaccinees had significantly less IgA to SARS-CoV-2, but comparable IgG responses those with natural infection. However, following a single dose vaccinees had reduced antibody levels to the VOCs, which further declined with time, suggesting the need to reduce the gap between the two doses, in countries experiencing outbreaks due to VOCs.

All data are available within the manuscript, figures and the tables. Individual data points are shown in all figures. Source data are provided with this paper.

None of the authors have any conflicts of interest. M a n u s c r i p t 22 

Tan CW, Chia WN, Qin X, Liu P, Chen were measured by the surrogate virus neutralizing test following natural infection at 4 weeks in those with mild illness (n=15) and moderate/severe illness (n=15) and at 12 weeks in those with mild (n=14) and moderate/severe illness (n=6). Antibodies were also measured at 4 weeks (n=20) and 12 weeks (n=20) in vaccinees following a single dose of AZD1222. The

Kurskal-Wallis test was used to determine the difference between the antibody levels between the three different groups (two-tailed) followed by multiple comparisons using two- A c c e p t e d M a n u s c r i p t 

John Hopkins University; 2021

Increased mortality in community-tested cases of SARS-CoV-2 lineage B.1.1.7

SARS-CoV-2 variants of concern and variants under investigation in England

Coronavirus Pandemic (COVID-19)

Assessment of protection against reinfection with SARS-CoV-2 among 4 million PCR-tested individuals in Denmark in 2020: a population-level observational study

Resurgence of COVID-19 in Manaus, Brazil, despite high seroprevalence

Evidence of escape of SARS-CoV-2 variant B.1.351 from natural and vaccine-induced sera

Disease severity dictates SARS-CoV-2-specific neutralizing antibody responses in COVID-19

Difference in levels of SARS-CoV-2 S1 and S2 subunits-and nucleocapsid proteinreactive SIgM/IgM, IgG and SIgA/IgA antibodies in human milk

Infection Includes Broad Reactivity to the S2 Subunit

Preexisting and de novo humoral immunity to SARS-CoV-2 in humans

Crossreactive antibody immunity against SARS-CoV-2 in children and adults

IgA dominates the early neutralizing antibody response to SARS-CoV-2

Humoral signatures of protective and pathological SARS-CoV-2 infection in children

National Study Group for C-V. Effectiveness of the BNT162b2 Covid-19 Vaccine against the B.1.1.7 and B

COVID-19 vaccine coverage in health-care workers in England and effectiveness of BNT162b2 mRNA vaccine against infection (SIREN): a prospective, multicentre, cohort study

COVID-19 Vaccines vs Variants-Determining How Much Immunity Is Enough

An updated look at the 16-week window between doses of vaccines in BC for COVID-19

Human IgG and IgA responses to COVID-19 mRNA vaccines

linical management of severe acute respiratory infection when novel coronavirus 2 -n o infection is suspected: interim guidance

A haemagglutination test for rapid detection of antibodies to SARS-CoV-2

A pneumonia outbreak associated with a new coronavirus of probable bat origin

Antibody and T cell responses to a single dose of the AZD1222/Covishield vaccine in previously SARS-CoV-2 infected and naïve health care workers in Sri Lanka

A c c e p t e d M a n u s c r i p t