Evaluation of the systemic and mucosal immune response induced by COVID-19 and the BNT162b2 mRNA vaccine for SARS-CoV-2

Background Currently used vaccines to protect from COVID-19 mostly focus on the receptor-binding domain (RBD) of the viral spike protein, and induced neutralizing antibodies have shown to be protective. However, functional relevance of vaccine-generated antibodies are poorly understood on variants-of-concern (VOCs) and mucosal immunity. Methods We compared specific antibody production against the S1 subunit and the RBD of the spike protein, the whole virion of SARS-CoV-2, and monitored neutralizing antibodies in sera and saliva of 104 BNT162b2 vaccinees and 57 individuals with natural SARS-CoV-2 infection. Furthermore, we included a small cohort of 11 individuals which received a heterologous ChAdOx1-S/BNT162b2 prime-boost vaccination. Results Vaccinated individuals showed higher S1-IgG antibodies in comparison to COVID-19 patients, followed by a significant decrease 3 months later. Neutralizing antibodies (nAbs) were poorly correlated with initial S1-IgG levels, indicating that these might largely be non-neutralizing. In contrast, RBD IgGAM was strongly correlated to nAbs, suggesting that RBD-IgGAM is a surrogate marker to estimate nAb concentrations after vaccination. The protective effect of vaccine- and infection-induced nAbs was found reduced towards B.1.617.2 and B.1.351 VOCs. NAb titers are significantly higher after third vaccination compared to second vaccination. In contrast to COVID-19 patients, no relevant levels of RBD specific antibodies were detected in saliva samples from vaccinees. Conclusions Our data demonstrate that BNT162b2 vaccinated individuals generate relevant neutralizing antibodies, which begin to decrease within three months after immunization and show lower neutralizing potential to VOCs as compared to the original Wuhan virus strain. A third booster vaccination provides a stronger nAb antibody response than the second vaccination. The systemic vaccine does not seem to elicit readily detectable mucosal immunity.


77
Starting from the pandemic spread of the coronavirus disease in December 2019 , 78 global research efforts were made to identify effective vaccine candidates. Vaccines based on 79 vectors, inactivated viruses and mRNA were licensed, whereas the latter comprises a novel 80 immunization technology. However, the kinetics of SARS-CoV-2 antibody production and the 81 persistence of humoral immunity following such a vaccination over time is of great interest for 82 national health services and the management of the pandemic. Recent studies suggested that 83 SARS-CoV-2 specific antibody production following vaccination with the BNT162b2 84 (BioNTech/Pfizer) mRNA vaccine is comparable to seroconversion following recovery from 85 COVID-19.
[1] During SARS-CoV-2 infection, the median time to detect circulating antibodies 86 has been reported at 11 days after the onset of symptoms and this period will be affected by 87 initial disease severity. [2] In patients with milder symptoms, some antibodies wane rapidly, 88 especially IgG against the nucleocapsid protein, whereas our group and others have shown that 89 antibodies to the spike protein and neutralizing antibodies remain detectable much longer. [3][4][5] 90 We performed a prospective study after the second BNT162b2 vaccination and assessed the 91 longevity of vaccine-induced antibodies for three and eight months in follow-up visits and 92 compared it with the humoral immune response after a third booster vaccination. Furthermore, . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 30, 2022. ;https://doi.org/10.1101https://doi.org/10. /2022 BNT162b2 immune responses 4 93 we analyzed effectiveness and kinetic of BNT162b2-induced nAbs against the variants of 94 concern B.1.351 (Beta) and B.1.617.2 (Delta) and. [6,7] and the impact of the homologous 95 BNT162b2 booster vaccination. Moreover, we also dissected the SARS-CoV-2 specific 96 antibody production in individuals that were vaccinated with a heterologous ChAdOx1-S vector 97 vaccine (AstraZeneca) and BNT162b2 mRNA vaccine (BioNTech/Pfizer) prime-boost regime. 98 In addition to systemic immunity, mucosal immune responses are considered to be critically is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 30, 2022. ;https://doi.org/10.1101https://doi.org/10. /2022 BNT162b2 immune responses 5 107

108
Study design and human samples 109 The study included two cohorts of vaccinated individuals that were followed prospectively and 110 previously bio-banked samples from COVID-19 patients (Fig 1, Table 1 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Commercial assays for the detection of antibodies against S1 and Nucleocapsid

164
In-house developed enzyme-linked immunosorbent assays 165 Detection of SARS-CoV-2 inactivated whole virion (IWV) IgG-antibodies and SARS-CoV-2 166 RBD polyvalent IgGAM-antibodies was performed according to Rockstroh et al. 2021.[3] 167 Briefly, Nunc PolySorp plates were coated with 1.5 µl per well of inactivated SARS-CoV-2 wt 168 viral particles and 250 ng/well of RBD protein in 100 µl per well of carbonate coating buffer 169 (15 mM Na2CO3, 7 mM NaHCO3 pH 9.6) overnight at 4°C. RBD protein (AA residues 329-. CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 30, 2022. ; https://doi.org/10.1101/2022.01.29.22270066 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 30, 2022. were incubated for 48 h at 37°C with 5 % CO 2 until cytopathic effect (CPE) was visible. Virus 196 containing supernatants were centrifuged at 4000 g for 10 min at 4°C and then stored at -80°C 197 until use. Viral titers were determined using a focus-forming assay. All viral stocks were 198 sequenced to verify their spike protein sequences and expected mutation sites. CoV-2 focus forming units were stained using a monoclonal rabbit anti-S1 antibody (CR3022, 210 abcam, 1:1,000) and a secondary goat anti-rabbit IgG HRP-conjugated antibody (Dianova,  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Bind microtubes (Eppendorf, Hamburg, Germany) until further use. Samples were centrifuged 223 at 10,000 g for 5 min and the supernatant was collected. 25 µl of saliva was mixed with equal 224 amounts of LEGENDplex assay buffer and S1-or RBD-coated beads in a 5 ml polypropylene 225 FACS tube, sealed and incubated overnight at 7°C in the dark. Subsequently, the bead mixture  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 30, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 30, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint [3] SARS-CoV-2 neutralizing as well as S1, RBD and IWV binding antibodies were 291 detectable in most individuals three weeks after the first vaccination. Two weeks after the 292 second dose, nAb titers and binding antibodies to all tested antigens were significantly 293 increased. At this stage, nAb titers (Median FRNT 90 = 320) as well as IWV IgG antibody signals 294 were comparable to those induced by natural mild COVID-19 but did not reach the high levels 295 of patients having severe COVID-19 courses (4-and 1.3-fold decrease for nAb and IWV IgG 296 respectively) (Fig 2 A and E). In contrast, RBD IgGAM and S1 IgA antibody levels 297 corresponded to those induced after severe COVID-19 and were even significantly exceeded 298 by vaccine-induced S1 IgG antibodies (3.5 fold increase). For some individuals, IgG antibodies 299 against the inactivated whole virion of SARS-CoV-2 (IWV) were detected even before the first 300 vaccination. The correlation between BNT162b2-induced S1 IgG, RBD IgGAM antibody signals and nAb 314 titers using Spearman's rank coefficient is presented in S1 Fig. Herein, S1 IgG and RBD 315 IgGAM antibodies indicated the strongest correlation to SARS-CoV-2 nAbs (r=0.93) (Fig 3 A-316 B), whereas IWV and S1-IgA correlation coefficients ranged between r=0.722 and 0.817 (S1  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

330
Ratios of binding to neutralizing antibodies were calculated to compare the proportion of 331 neutralizing antibodies induced by vaccination to convalescent individuals in each test (Fig 3).

332
Vaccinees presented a significantly lower proportion of neutralizing to S1 binding antibodies 333 in comparison to the COVID-19 group (10-fold and 5-fold lower for S1 IgG and S1 IgA 334 respectively). Similarly, this ratio was lower for RBD binding IgGAM antibodies compared to 335 patients after severe COVID-19 but was found comparable to the mild COVID-19 group. For 336 IWV-IgG the neutralizing proportion of binding antibodies was higher than in patients with 337 mild COVID-19 courses but below the median of severe courses.

339
Long term kinetics of antibody titers 340 The kinetics of antibody abundance varied greatly, with a mean reduction of nAb by 3.3 fold at 341 P2 and 9.7 fold at timepoint P3 compared to timepoint P1 two weeks after second vaccination. 342 (Fig 4 A) RBD IgGAM ratios decreased 1.8 fold after three month and 4.0 fold after eight 343 month (Fig 4 C). S1 antibody concentrations fall 5.7 fold at timepoint P2 and 36.6 fold at P3 344 (Fig 4 C). Smallest differences were observed in the IWV IgG antibody signals with a 1.1 fold 345 mean reduction after three months and a 2.2 fold mean reduction after eight month (Fig 4 D). is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 30, 2022. ;https://doi.org/10.1101https://doi.org/10. /2022 348 90% (FRNT90). (B) RBD IgGAM signals were determined as sample/calibrator ratios (C) S1 IgG antibodies were 349 quantitatively measured in binding antibody units per milliliter (BAU/ml). (D-E) S1 IgA and IWV (SARS-CoV-2 350 inactivated whole-virion) titers were determined as sample/calibrator ratios. (F) Fold reduction of SARS-CoV-2 351 neutralizing antibody titers, RBD IgGAM, S1 IgG, S1 IgA and IWV IgG antibody signals in follow-up vaccine 352 sera calculated as ratio of value on timepoint P1 to timepoint P3.

354
Neutralizing effect of BNT162b2-induced antibodies on B.1.617.2 and B.1.351 VOCs 355 Neutralizing antibody titers were decreased towards VOCs in almost all tested individuals and 356 cohorts when compared with wildtype SARS-CoV-2 (Fig 5). A 5.1-and 11.5-fold mean titer   (Fig 6 A). The comparison between time point P3 and boost shows median 373 increase ranges between 76.3 fold and 138 fold (Fig 6 B). is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 30, 2022.  (Fig 7). is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 30, 2022. ;https://doi.org/10.1101https://doi.org/10. /2022  Long-term kinetics of SARS-CoV-2 antibody response after BNT162b2 vaccination 414 We have observed significant reduction of vaccine-induced antibody levels three months after 415 the second vaccination. Interestingly, S1 quant IgG and IgA antibodies decreased strongly, 416 whereas nAb levels and RBD-GAM dropped to a lesser extent.
[18] Thus, the S1 quant IgG 417 ELISA does not seem to be an optimal diagnostic choice for determining longevity of humoral 418 immune responses after SARS-CoV-2 mRNA vaccination. Therefore, we propose the use of 419 RBD-IgGAM determination as a rapid and simple surrogate marker to estimate the levels of 420 nAbs after SARS-CoV-2 mRNA vaccination. One explanation for a better correlation between 421 RBD-IgGAM and nAbs could be a better RB-Domain presentation in the vaccine antigen.

422
Considering that a broader nAb production against the RBD region was observed in vaccinees is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 30, 2022. ;https://doi.org/10.1101https://doi.org/10. /2022 As an observed decrease in SARS-CoV-2 antibody titers within a short diagnostic interval infections. Of note, the group size for the heterologous vaccination is rather small compared 449 with the homologous mRNA vaccinated one. Therefore, the data need to be re-confirmed with 450 a larger cohort.

451
The booster immunization led to a significantly stronger production of SARS-CoV-2 specific 452 nAbs compared to the threshold after second vaccination. Of note, the median increase was . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint antibody titers were even below the limit of detection of the FRNT assay.

456
The decline in the neutralizing serological effect in recovered patients is however more 457 pronounced. Here, we observed a 21.5-to 34.7-fold decrease in nAb titers for patients with mild 458 and a 34-to 75.3-fold reduction for patients with severe COVID-19 courses. In other studies, a  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 30, 2022. ; https://doi.org/10.1101/2022.01.29.22270066 doi: medRxiv preprint

478
BNT162b2 mRNA vaccination provided sustainable formation of SARS-CoV-2 neutralizing 479 antibodies in our studied cohort. To preserve detectable nAb titers after 6 months, a booster 480 vaccination should be considered, especially for the protection against variants of concern. A 481 heterologous vaccine regime involving ChAdOx1-S vector-based prime and BNT162b2 mRNA 482 vaccine boost even exceeded these titers of neutralizing antibodies and might thus feature 483 beneficial synergy. None of the studied vaccines induced detectable mucosal immune response.

497
The authors are indebted to the voluntary help of probands and patients to participate in this study. Furthermore,

498
we would like to particularly emphasize the efforts made by Ulrike Schmidt and Cathrin Crimmann, and all 499 participating doctors and nurses during the vaccination campaign. Moreover, we like to thank Ulrike Ehlert and 500 Steffen Jakob for their excellent technical assistance. We also thank Corinna Pietsch (Leipzig University Hospital) 501 and Klaus Überla (Erlangen-Nürnberg University) for supplying VOCs.

502
. CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 30, 2022. ; https://doi.org/10. 1101/2022