key: cord-0950003-sbv79ljq
authors: Piano Mortari, E.; Russo, C.; Vinci, M. R.; Terreri, S.; Fernandez Salinas, A.; Piccioni, L.; Alteri, C.; Colagrossi, l.; Coltella, L.; Ranno, S.; Linardos, G.; Agosta, M.; Albano, C.; Agrati, C.; Castilletti, C.; Meschi, S.; Romania, P.; Roscilli, G.; Pavoni, E.; Camisa, V.; Santoro, A.; Brugaletta, R.; Magnavita, N.; Ruggiero, A.; Cotugno, N.; Amodio, D.; Cioffi Degli Atti, M. L.; Giorgio, D.; Russo, N.; Salvatori, G.; corsetti, t.; Locatelli, F.; Perno, C. F.; Zaffina, S.; Carsetti, R.
title: Highly-specific memory B cells generation after the 2nd dose of BNT162b2 vaccine compensate for the decline of serum antibodies and absence of mucosal IgA
date: 2021-06-09
journal: nan
DOI: 10.1101/2021.06.08.21258284
sha: 9ee0de9d8c90fb0ab07a02ad3892fc4a972e5566
doc_id: 950003
cord_uid: sbv79ljq

Specific memory B cells and antibodies are reliable read-out of vaccine efficacy. We analyzed these biomarkers after one and two doses of BNT162b2 vaccine. The second dose significantly increases the level of highly-specific memory B cells and antibodies. Two months after the second dose, specific antibody levels decline, but highly specific memory B cells continue to increase thus predicting a sustained protection from COVID-19. We show that although mucosal IgA is not induced by the vaccination, memory B cells migrate in response to inflammation and secrete IgA at mucosal sites. We show that first vaccine dose may lead to an insufficient number of highly specific memory B cells and low concentration of serum antibodies thus leaving vaccinees without the immune robustness needed to ensure viral elimination and herd immunity. We also clarify that the reduction of serum antibodies does not diminish the force and duration of the immune protection induced by vaccination. The vaccine does not induce sterile immunity. Infection after vaccination may be caused by the lack of local preventive immunity because of the absence of mucosal IgA.

The SARS-CoV-2 pandemic continues spreading and causing COVID-19 disease and death. Hospitals are under extreme pressure and the economic damage is unprecedented. Vaccination is the safest and most effective tool to achieve a protective response in most individuals. Effective vaccines against SARS-CoV-2 have been rapidly developed and are being administered with the aim of preventing 5 COVID-19, stopping viral circulation and terminating the pandemic. Phase III trials have demonstrated the high efficacy of two-dose regimens 1-3 , but because of the low number of vaccine doses available in Europe, it has been suggested that initially only the first dose may be administered to the whole population in the emergency phase. The first dose may be able to prevent the most severe forms of disease, thus avoiding hospitalization and death 4 . According to current epidemiological and political 10 opinions, the second dose may be delayed. Concerns have been raised that changing the immunization schedule proved to be effective by the authorization studies, because of results obtained in the real world on the effectiveness of the first dose 5 and on the consideration that incomplete immunity may favour the emergency of viral variants 6 . From the epidemiological point of view, it has been predicted that a onedose regimen will foster antigenic evolution and weaken the protective potentials of the available 15 vaccines and of those still in development, all based on the sequence of a virus that may have changed under the selective pressure of an immune response unable to completely clear the virus 7 . Several Variants of concern (VoC) have been identified and have spread throughout the world. New variants are

Another still open question is the duration of protective immunity after vaccination. Most attention has 20 been concentrated on the level of specific antibodies, which increase in response to the vaccine, but decline after a few months. The reduction of specific antibody levels generates considerable anxiety in the population, because it is perceived as incipient loss of protection.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint 4 Antibody decline is, however, a normal feature of the response to every vaccination 9 . Our most important protection from reinfection is represented by the synergistic action of memory B cells (MBCs) and memory plasma cells. Specific antibodies present in the serum are continuously secreted by memory plasma cells and face the pathogen at its first entry 10 . At the same time, MBCs migrate in response to chemokines to the inflamed tissues and reinforce protection by locally secreting antibodies. If the 5 antibodies produced by MBCs and plasma cells have a high affinity for the pathogen this is immediately eliminated 9 . T cells play a fundamental role by helping B cells to produce high-affinity antibodies and memory B cells 11,12 and, most importantly, by eliminating virus infected cells 13 .

From the point of view of B-cell immunology, in order to support public health decisions and measures, the most important issues to be addressed are: 1. the affinity of MBCs and plasma cells generated after 10 the 1 st and 2 nd vaccine dose; 2. the longevity of vaccine-induced immune memory; 3. the ability of the vaccine to induce sterilizing immunity, thus preventing not only COVID-19, but also infection.

We addressed these three points in a study on the B-cell response to the first and second dose of BNT162b2 vaccine and a further evaluation of the established protection three months after.

We report the effects of the two-doses regimen of BNT162b2 vaccine in a population of Health Care 15 Workers (HCWs) of the Bambino Gesù Children Hospital IRCCS.

In 108 HCWs we measured specific MBCs by ELISpot, flow-cytometry and serum antibodies at different time points, before and after first and second dose and three months later. We found that one single dose results in immune memory lacking the precision indispensable to immediately clear the virus that is the basis for vaccine effectiveness. The second dose increases the frequency of vaccine responders and 20 specific IgG levels and generates highly specific MBCs. Three months later, serum antibodies significantly decline, but, in contrast, MBCs increase in frequency and affinity. Most importantly, the vaccine triggers a serological IgA response, but does not generate mucosal IgA. The lack of specific IgA . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint 5 strategically located at the virus site of entrance explains why the vaccine does not induce sterilizing immunity. 5 We included in our study 108 HCWs (Table 1 ) who had never experienced SARS-CoV-2 infection before, as demonstrated by molecular (Allplex2019-ncov, Seegene) and antibody assays (Elecsys®Anti-N, Roche) (data not shown).

The BNT162b2 vaccine was administered as prescribed, in two doses, 21 days apart. Serum and blood samples were collected at time 0 (T0), before the first dose; seven days after the first dose (T7); on the 10 day of the second dose (T21), seven days after the second dose (T28) and 3 months after the first dose (T90) (Fig. 1) .

Specific antibodies were measured in the serum, saliva and breast milk (Fig. 1 ). MBCs were identified by two different methods. By ELISpot, after polyclonal stimulation with CpG, we detected MBCs, based on their ability to secrete antibodies that bind the vaccine target Trimeric Spike on coated plates. We also 15 identified MBCs ex-vivo by flow-cytometry, thanks to the availability of biotin-labelled Trimeric Spike coupled to a very high brightness fluorescence dye (PE) and the same biotin-labelled trimeric Spike coupled to a moderate brightness fluorescent dye (BUV395) 14 . We were able to distinguish MBCs with low (PE single positive, S+) or high affinity (PE-BUV395 double positive, S++) for Trimeric Spike. 

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint 6 Serum antibodies are the most commonly used surrogate biomarkers of vaccine efficacy. We measured specific serum antibodies using different methods.

Anti-RBD antibodies of IgM isotype were measured by ELISA. We found that anti-RBD IgM was present at low levels at T0 and T7 reflecting the presence of natural 11,15 or cross-reactive antibodies 16, 17 .

The increase at T21 and T28 is the effect of the vaccination. At T90 anti-RBD IgM slightly declined 5 (Fig. 2a) .

Anti-RBD antibodies of all isotypes, evaluated by Elecsys®Anti-SARS-CoV-2-S (Roche, cut-off: 0.8 BAU/mL) were absent at T0 and T7, but already increased at T21. Their level, however, was much higher at T28 and decreased at T90 (Fig. 2b) . The geometric mean at T21 was 29.13 BAU/mL and became 1736 BAU/mL at T28, demonstrating the importance of the booster dose. At T90 the geometric mean 10 decreased to 819.8 BAU/mL. Similar results were obtained when we measured IgG against Trimeric Spike (DiaSorin, cut-off: 13 BAU/mL). The geometric mean values increased from 205.9 BAU/mL at T21 to 1660 BAU/mL at T28 and then decreased to 991.3 BAU/mL at T90 (Fig. 2c) .

Thus, specific antibodies are already produced after the first vaccine dose. However, anti-RBD antibodies increased 60 times between T21 and T28, and anti-Trimeric Spike IgG raised 8 times after the second 15 dose. Three months after the first dose anti-RBD antibodies and anti-Trimeric Spike IgG were reduced by half (Fig. 2b,c) .

The anti-viral activity of IgG after the first and second dose and at T90 was measured by using, as surrogate virus neutralization assay, a specific chemiluminescence microparticle antibody assays (CMIA) on Abbott ARCHITECT® i2000sr 18 that detects IgG antibodies specific for RBD. The 20 geometric mean at day 21 was 559.3 AU/mL and increased 25 times at T28 (geometric mean 14129 AU/mL), thus confirming the importance of re-stimulation to obtain a significant amplification of the . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint protective response (Fig. 2d) . Surrogate-neutralization activity was reduced by 3.2 folds at T90 (geometric mean 4403; Fig. 2d ).

The reduction of antibody levels three months after vaccination reflects the normal kinetics of the serological immune response 19 .

There was no correlation between antibody levels measured by the different tests at early times after 5 vaccination (data not shown), whereas the correlation became significant at T90 (Fig. 2e) . As the commercial test used here measure slightly different responses (all isotypes against RBD, IgG specific for either Trimeric Spike or RBD), the correlation at T90 may reflect the loss short-lived plasma cells released in the early stages of the immune reaction and the progressive selection of useful IgG antibodies against trimeric Spike and RBD.

ELISpot to detect B cells able to secrete Trimeric Spike-specific IgM, IgG and IgA antibodies was performed in 74 HCWs. We found that a high frequency of IgM MBCs was able to bind Trimeric Spike already at T0 and their number slightly increased 7 days after the first dose (T7) (Fig. 3a) . This is an 15 expected result, because our pool of IgM MBCs includes a wide range of cross-reactive specificities from which high affinity MBCs will be shaped and selected by the germinal center (GC) reaction 12 .

In the GC, the mechanisms of somatic mutation and antigen-driven affinity-maturation improve the specificity of the antibodies and class-switching generates MBCs and plasma cells able to secrete switched antibodies that have the function of terminating infections and clearing the pathogen.

High specificity and affinity are the most important characteristics of protective MBCs, generated by the adaptive immune system in response to infection or vaccination.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint

In agreement with the timing of the GC reaction, IgG MBCs specific for Trimeric Spike significantly increased at T21 after the first dose and significantly more at T28, namely seven days after the second dose (Fig. 3a) . Importantly, we found that 20.8% of the HCWs did not have IgG-producing MBCs cells in the peripheral blood at T21, following the first dose. Seven days after the second dose (T28), however, IgG-producing MBCs became detectable in the great majority of the HCWs (92.4%) (Fig. 3a,b) , again 5 demonstrating the importance of the booster administration.

IgA-secreting MBC were less frequently detected in the peripheral blood of vaccinated individuals. We found that 58.4% of the HCWs produced IgA MBCs 21 days after the first dose; seven days after the second dose (T28), IgA-producing MBCs were found in the peripheral blood of 71.3% of the individuals ( Fig. 3a,b) . 10 An important finding of our study is that, in contrast to the decline of specific antibody levels in the serum, MBCs of IgM, IgG and IgA isotypes remained stable three months after the first dose (Fig. 3a,b) . . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint

MBCs can be detected in the peripheral blood of convalescent patients and immunized individuals using fluorescent Trimeric Spike and RBD 22 .

We chose to visualize biotin-labelled Trimeric Spike using two different streptavidin-fluorochromes, one Among Spike positive MBCs (S+ plus S++), we also detected RBD-specific cells with Streptavidin -FITC bound to biotin-RBD. After excluding aspecific binder B cells with streptavidin-PE-Cy7, we gated CD19+CD24+CD27+CD38-MBCs (Fig. 4a,b) and calculated the frequency of S+ and S++ MBCs and of RDB binder. We found that low affinity MBCs (S+) were already detectable at T0 (Fig. 4c ,d) and 15 were mostly of IgM isotype (Fig. 4e) , as expected, because of the structure and function of the large repertoire from which we shape and select our specific MBCs and in agreement with our ELISpot results ( Fig. 3a,b) . S+ MBCs slightly increase at T7 and more significantly at T21 and T28 (Fig. 4d ) remaining mostly of IgM isotype (Fig. 4e) .

Double positive Trimeric Spike-specific MBCs (S++) significantly increased at T21 and even more at 20 T28, seven days after the second dose (Fig. 4c,d) . Most of them expressed switched isotypes (IgM-) (Fig.   4e ).

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint At T90 we found that S+ MBCs significantly declined, whereas the frequency of S++ MBCs increased 4 fold (Fig. 4c,d) . These results reflect the ongoing maturation of the germinal center response leading to the progressive selection of MBCs that optimally bind the antigen and are mostly of switched isotypes.

At T90 the majority of S+ MBCs were still of IgM isotype (Fig. 4e) .

Among all MBCs able to bind Trimeric Spike (S+ plus S++), we identified RBD+ cells that represent a 5 minority of the MBCs generated by vaccination. RBD+ MBCs significantly increased at T21 and even more at T28. Their frequency remained stable at T90 (Fig. 4f,g) . RBD+ MBCs were almost only found in the S++ MBC pool (Fig. 4h) , indicating the need of the germinal center mutation and selection mechanisms to generate this new specificity.

The production of IgA is of great interest, because IgA is the main immunoglobulin for protection at mucosal sites, including the upper airways, the site of SARS-CoV-2 entry.

We measured Spike-specific IgA in the serum of immunized HCWs. We found that after the first dose 22% of the HCWs had not produced Spike-specific IgA, but all of them had specific IgA at T28 in the 15 serum. The level of IgA strongly decreased at T90 (geometric mean 12.75 at T28 and 2.6 at T90) indicating that most of the IgA plasma cells generated in response to the vaccine were short-lived ( Fig.   5a ).

Antibodies can reach mucosal sites by transudation from the serum or be locally produced after MBCs have migrated to mucosal sites and have become plasma cells. The latter mechanism is indispensable for 20 the secretion of IgA into breast milk 23 . Pro-inflammatory cytokines and chemokines are physiologically increased in the lactating breast 24,25 and switched MBCs primed to secrete antibodies can be found in . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint maternal milk. Sixteen HCWs were vaccinated after pregnancy, during lactation (Table 3) . Seven days after the second vaccine dose, all of them had detectable IgA specific for trimeric Spike in breast milk ( Fig. 5b) confirming the ability of vaccine-induced MBCs to home to the inflammatory environment of the lactating mammary gland and locally produce IgA.

Saliva can be considered a proxy of mucosal immunity containing secretory IgA produced by local 5 plasma cells 26 . In order to investigate whether serum IgA is associated with mucosal protection after vaccination, we measured IgA in the saliva of 43 HCWs that had the highest IgA values in the serum at day 28.

We found that two to three weeks after the second dose, IgA was undetectable in the saliva of all vaccinated individuals tested (Fig. 5c) , demonstrating that the vaccination does not result in the presence 10 of significant levels of IgA in the oral mucosa. Specific IgA was detectable in the saliva of 39 HCWs previously infected with COVID-19 ( Fig. 5c ), indicating that, as published before, the disease induce mucosal immunity 27 .

Twenty-one of the 3511 vaccinated HCWs of our Institute had a positive swab after completion of the two-dose vaccine cycle. All of them were related to familiar clusters. All the infected HCWs had 15 responded to the vaccine with serum antibody production before the positive nasopharyngeal swab. The infection was either asymptomatic or mild, without signs of pneumonia. We studied the immune response of 6 HCWs infected after vaccination (Table 4 ). Salivary IgA became detectable 0-6 days after the first positive swab and increased further after 6 days from the infection (Fig. 5c ). Specific IgA in the saliva was 10 times lower than IgA in the serum.

Two of the HCWs who had a positive swab after the second vaccine dose were included in our MBC study. From the first one we compared the samples obtained at the T28, at T70, and at T118 when the HCW returned to work. During the infection, caused by a virus variant (P.1), the initial symptoms were . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint mild, with fever, cough and muscle pain. Most of the MBCs were of low affinity (S+) at T28 and T70. S++ MBCs were increased 23 times in the last sample (Fig. 5d ).

In contrast, the second HCW had both S+ and S++ MBCs at T28 and T90. A nasopharyngeal swab resulted positive 20 days after T90 and the MBCs pool remained stable in the next three days, and strongly increased 2 weeks later (Fig. 5d) . The infection was asymptomatic in this case. 5 Of 4 HCWs that were not included in the initial study and had a positive swab between T60 and T80, we obtained a sample when they returned to work after a paucisymptomatic disease. Their S++ were in higher frequency compared to the HCWs in the initial study group (median=0.08%, in Fig. 4d) confirming that MBCs generated by the vaccine increase in response to the virus (Fig. 5e) . A further evidence of the function of MBCs is the increase of serum specific IgG after the positive swab in 5 out 10 of 6 HCWs (Fig. 5f ). HCW8 was found to have a positive swab during a routine control, in the absence of symptoms. The nasopharyngeal swab became negative after two days. High levels of salivary IgA (overflow) were detected at day 7 (the only sample available). Retrospectively we found that HCW8 had antibodies against the viral nucleoprotein already after the complete vaccine cycle, in February. We hypothesize that a previous asymptomatic and undetected infection had resulted in a powerful mucosal 15 protection able to eliminate the virus in two days and prevent the systemic response.

The aim of vaccination is to generate immune memory, namely the pool of cells, including high-affinity 20 MBCs and long-lived plasma cells with their antibodies that are able to prevent reinfection. Specificity and rapidity of action are the indispensable properties of protective immune memory. Differently from memory T cells, MBCs have the ability to improve their specificity by repeated steps of somatic mutation . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint 13 and selection in the GC. MBCs with high affinity for Trimeric Spike produce antibodies able to neutralize the virus. Monoclonal antibodies obtained after cloning from these MBCs can be used for cure and prevention of SARS-CoV-2 infection 28 .

Few days after a vaccine cycle, specific antibodies peak in the serum thanks to the activity of short and long-lived plasma cells 9 . The decay of short-lived plasma cells 9 causes the reduction of serum antibodies. 5 As no vaccination before has been under such a constant and severe scrutiny by the public and media as this against SARS-CoV-2, the physiological drop of serum titers is seen as a sign of vaccine failure and loss of protection. In response to this anxiety, the necessity of further vaccine booster doses has been suggested.

Although three months after the first dose serum specific antibody levels decline, we found that MBCs . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint

In agreement with the ELISpot data, MBCs with low affinity for Trimeric Spike (S+), mostly of IgM isotype are present at all time-points ( Fig. 4c and d) , whereas high affinity MBCs become detectable at T21 and significantly increase further at T28. They are mostly of switched isotypes ( Fig. 4c and d) . At T90 the frequency of high-affinity MBCs is further increased, suggesting that the germinal center reaction continues its rounds of selection of the best specificity for months after the second vaccine dose. vaccinated HCWs suggests that, in strict terms, the vaccine is not sterilising, because it is unable to generate preventive mucosal immunity.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint

The lack of direct mucosal protection explains why vaccinated individuals can have a positive nasopharyngeal swab. In most cases, however, the infection remains asymptomatic or mild. Part of their defence may be due to Trimeric Spike-specific IgG and IgA antibodies exudated to the tissues. An important role may also be played by MBCs that migrate to the infected areas. We show that MBCs generated by the vaccine are able to migrate to mucosal tissues and locally produce IgA antibodies in 5 two different situations, both characterized by the production of attracting inflammatory cytokines. First we show that Spike-specific IgA can be detected in the breast milk of women vaccinated after pregnancy ( Fig. 5b) . Secondly, we found that HCWs with a positive swab after the complete cycle of vaccination, produce salivary IgA 2-4 days after the first positive swab (Fig. 5c) . suggests that vaccine-induced MBCs Real-world efficacy of one and two doses of RNA vaccines has been just reported by a CDC study.

Among HCWs, one dose had 80% efficacy whereas two doses resulted in 90% efficacy 4 . In a different setting, in individuals over 60, after the first dose of BNT162b2 vaccine there was a significant decrease, . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint but only of about 33% in the rate of positive tests for the coronavirus. The efficacy increased to 60% after two doses 5 . Many factors contribute to vaccine efficacy including age and viral circulation, but from the immunological point of view two doses are more protective than one. The real-world efficacy of one vaccine dose, demonstrated after day 14 post-immunization, can be explained by the performance of lowaffinity MBCs and antibodies able to prevent severe COVID-19. Nevertheless, our data show that two For all these reasons we should not worry about the antibody decline in the serum, observed a few months after vaccination or natural infection. Our MBCs are our most important defense weapon that ensure local and systemic protection after re-encounter with the antigen. In conclusion, as any other vaccine . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint produced before, also RNA-vaccines appear to be able to generate a typical immunological memory that is predictably stable and durable. Medians are plotted and statistical significance was determined using Wilcoxon matched-pairs signedrank test. *p < 0.05, **p < 0.01, ***p < 0.001. Medians are plotted and statistical significance was determined using Wilcoxon matched-pairs signedrank test. *p < 0.05, **p < 0.01, ***p < 0.001.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint Medians are plotted and statistical significance was determined using Wilcoxon matched-pairs signedrank test or Chi-Square test. *p < 0.05, **p < 0.01, ***p < 0.001. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint

x-axis indicates that after T28, the time of the first positive NPS was variable (between day 60 and day 110 after first dose) and longer compared to the time between the positive and negative NPS (2 to 20 days).

Medians are plotted and statistical significance was determined using Wilcoxon matched-pairs signedrank test or unpaired Mann-Whitney t-test. *p < 0.05, **p < 0.01, ***p < 0.001.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. 

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint 

Mothers Age (mean ± SD) 33.36 ± 4.17

Mothers' breast milk 7 not vaccinated 16 vaccinated . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint 

We thank the study participants for their dedication to this project. Marco Scarsella and Ezio Giorda for the assistance with the FACs acquisition. Sandra Martino for technical coordination, Patrizia Palomba, Authors declare that they have no competing interests.

All data are available in the main text or the supplementary materials.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint

Ethical approval was obtained from the Ethics Committee at OPBG, Bambino Gesù Children Hospital.

According to the guidelines on Italian observational studies as established by the Italian legislation about Human subjects 10 We analysed peripheral blood from 108 HCWs of the Bambino Gesù Children Hospital (Table 1 and 2).

We included in the study all HCWs that were vaccinated by two injections of BNT162b2, 30 μg, 21 days apart.

Blood and serum samples were collected at defined intervals after vaccination. Samples were available before first dose (T0), and at 7 (T7), 21 (T21), 28 (T28) ant 90 (T90) days after the first dose. All HCWs 15 had a negative SARS CoV-2 status by molecular (Allplex2019-ncov, Seegene) and antibody assays (Elecsys®Anti-N, Roche).

At the time of blood sampling, none of the subjects had an acute infection or were taking any medication known to alter immune function (such as steroids or statins).

Given the limiting amount of blood available for the study, we randomly selected 74 HCWs samples for 20 the ELISpot and 34 for FACs (Table 2) . Serology was performed in all samples and surrogate neutralization in 70 samples and only at T21, T28 and T90 (Table 2) .

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint Breast milk was obtained from 23 HCWs, of which 16 were vaccinated during lactation (Table 3) . Saliva samples were collected by aspirating with a 1 mL syringe from the oral cavity at least 30 minutes after after eating, drinking, smoking, or chewing gum. Saliva was immediately transfered to an eppendorf tube and frozen at -20 o C until use.

Cell isolation and cryopreservation. 5 Heparinized PBMCs were isolated by Ficoll Paque™ Plus 206 (Amersham PharmaciaBiotech) densitygradient centrifugation and immediately frozen and stored in liquid nitrogen until use. The freezing medium contained 90% Fetal Bovine Serum (FBS) and 10% DMSO.

As we published before 41 in order to induce antibody production from MBCs, we polyclonally stimulated 

The B cell (enzyme-linked immunospot) ELISpot assay was performed with 5 days stimulated-PBMCs for the detection of MBCs.

96-well Multiscreen filter plates (Millipore) were pre-wetted with 35% ethanol for ≤ 1 min and washed 20 with ddH2O (5 x 200 µl/well). Wells were coated with 50 µl PBS containing either isotype-specific AffiniPure F(ab')2 Fragment Goat anti-Human antibody (Jackson ImmunoResearch Laboratories) (anti-. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint IgM final concentration 2.5 µg/mL, anti-IgG 15 µg/mL, anti-IgA 10 µg/mL); or Trimeric Spike at a final concentration of 1 µg/ml to detect specific responses. The plates were then incubated overnight (ON) at 4ºC. After washing with PBS (5 x 200 µl/well) and blocking with PBS + 1% BSA + 5% Sucrose (5 x 200 µl/well) for 2 h at RT, plates were filled with sterile RPMI + 10% FBS (100 µl/well) for 1 h at 37ºC.

After 5 days of stimulation (as previously described), cells were harvested and washed once with RPMI 5 before counting. Different concentration of PBMCs were used depending on the following conditions: total IgM, IgG and IgA: 1.5 x 10 4 PBMCs/well; specific IgM: 5 x 10 4 PBMCs/well; specific IgG and IgA: 1-5 x 10 5 PBMCs/well. Four steps of two-fold serial titrations were performed for each condition.

Plates were incubated ON at 37ºC. Cells were discarded and after washing with PBS + 0,05% Tween (5

x 200 µl/well), plates were incubated for 1 h at 37 ºC with peroxidase-conjugated goat anti-human IgM, is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint 30 with a calibrator. Results were evaluated by calculating the ratio between the extinction of samples and the extinction of the calibrator. The ratio interpretation was as follows: <0.8 = negative, ≥0.8 to <1.1 = borderline, ≥1.1 = positive 43 .

To detect IgM anti RBD we developed an in-house ELISA. 96-well plates (Corning) were coated for 1 h at 37°C with 1 μg/mL of purified SARS-CoV-2 RBD protein (Sino Biological). After washing with 5 PBS 1×/0.05% Tween and blocking with PBS 1×/1% BSA, plates were incubated for 1 h at 37°C with diluted sera (1:100). After washing again, plates were incubated for 1 h at 37°C with peroxidaseconjugated goat anti-human IgM antibody (Jacksons ImmunoResearch Laboratories). The assay was developed with o-phenylen-diamine tablets (Sigma-Aldrich) as a chromogen substrate. Absorbance at 450 nm was measured, and IgM concentrations were calculated by interpolation from the standard curve 10 based on serial dilutions of monoclonal human IgM antibody against SARS-CoV-2 Spike-RBD (Invivogen) 43 .

Serum samples were tested by two different methods and analytical platforms. Qualitative detection of antibodies direct against the nucleocapsid (N) protein and semi-quantitative detection of total antibodies 15 (IgA, IgM, IgG) directed against the RBD of the virus Spike (S) protein of SARS-CoV-2 were tested by an electro-chemiluminescence sandwich immunoassay (ECLIA), using Elecsys-anti SARS-CoV-2 and Elecsys-anti SARS-CoV-2 S (Roche Diagnostics) test on a Cobas e801 analyzer following the manufacturer's instructions. Results for anti-N antibodies are expressed as "present" or "absent" on the basis of a cut-off index (COI) COI ≥ 1.0 and COI < 1.0, respectively. Detection and quantification were 20 automatically calculated for each sample in U/mL, equivalent to the Binding Arbitrary Unit (BAU)/mL of the first WHO International Standard for anti-SARS-CoV-2 immunoglobulins. Titer was interpreted as absent when < 0.8 U/mL (< 0.8 BAU/mL) and as present when ≥ 0.80 U/mL (≥ 0.8 BAU/mL). When . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint antibodies titer was higher than 250 U/mL (250 BAU/mL), the instrument automatically executed a 20fold dilution, ranging the upper limit of quantification to 5000 U/mL (5000 BAU/mL). The quantitative determination of anti-Trimeric spike protein specific IgG antibodies to SARS-CoV-2 was run on LiaisonXL platform by a new generation of chemiluminescence immunoassay (CLIA) TrimericS IgG assay (DiaSorin) with a quantification range between 4.81 BAU/mL and 2080 BAU/mL (dilution factor 5 1:20).

The SARS-CoV-2 IgG II kit, a chemiluminescence microparticle antibody assays detecting antibodies against the RBD of SARS-CoV-2, (CMIA, IgG antibody concentrations expressed as arbitrary units, AU/mL ≥ 50 are considered positive, ARCHITECT® i2000sr Abbott Diagnostics) was used according to manufacturer's protocols.

To detect SARS-CoV-2 specific B cells, biotinylated protein antigens were individually multimerized with fluorescently labelled streptavidin at 4°C for one hour 22 . Biotinylated trimeric SARS CoV-2 Spike (S1) was purchased from R&D systems. RBD were generated in-house and biotinylation was performed using EZ-Link TM Sulfo-NHS-LC-Biotin reaction kit (ThermoScientific) following the manufacturer's 15 standard protocol and dialyzed overnight against PBS. Biotinylated Spike was mixed with streptavidin BUV395 (BD) and streptavidin PE (BD) at 25:1 ratio and 20:1 ratio respectively. Biotinylated RBD was mixed with streptavidin FITC (BD) at 2.5:1 ratio. Streptavidin PE-Cy7 (BD) was used as a decoy probe to gate out SARS-CoV-2 non-specific streptavidin-binding B cells. The antigen probes prepared individually as above were then mixed in Brilliant Buffer (BD). ~5x10 6 previously frozen PBMC samples 20 were prepared and stained with 85μL antigen probe cocktail containing 100ng Spike per probe (total 200ng), 27.5ng RBD and 20ng SA-PE-Cy7 at 4°C for 30min to ensure maximal staining quality before surface staining with antibodies as listed in table S1 was performed in Brilliant Buffer at 4°C for 30min.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint SARS-CoV-2 specific memory B cells were defined as CD19+CD24+CD27+CD38-Spike+ (S+) or CD19+CD24+CD27+CD38-Spike++ (S++).

Stained PBMC samples were acquired on FACS LSRFortessa (BD). At least 4x10 6 cells were acquired and analyzed using FlowJo10.7.1 (BD). Phenotype analysis of antigen-specific B cells was performed only in subjects with at least 10 cells detected in the respective antigen-specific memory B cell gate. In order to further discriminate between B.1.351 and P.1, the presence of SNP H655Y and A701V were further investigated by means of home-made sanger sequencing protocol. In brief, spike region encompassing amino acids 605-900 was retrotranscribed, amplified and sequenced by using two different 

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint

Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine

An mRNA Vaccine against SARS-CoV-2 -Preliminary Report

Safety and efficacy of the ChAdOx1 nCoV-19 vaccine (AZD1222) against SARS-CoV-2: an interim analysis of four randomised controlled trials in Brazil, South Africa, and the UK

Interim Estimates of Vaccine Effectiveness of BNT162b2 and mRNA-1273 COVID-19 Vaccines in Preventing SARS-CoV-2 Infection Among Health Care Personnel, First Responders, and Other Essential and Frontline Workers -Eight

Covid-19: Reports from Israel suggest one dose of Pfizer vaccine could be less 20 effective than expected

Values were compared by the non-parametric Kruskal-Wallis test and, if not significant, the Wilcoxon matched pair signed-rank test was used for the comparison between time points of each subject. Pairwise comparisons were evaluated by the Mann-Whitney U test nonparametric test. The Chi-square test with 2 × 2 contingency tables was used to compare pie charts. A p-values < 0.05 were considered statistically significant. Statistical analyses were performed with GraphPad Prism 8.0 (GraphPad Software). 5 The relationship between variables was studied using a correlation test based on the Pearson's product moment correlation coefficient cor(x, y) and follows a t distribution with length(x)-2 degrees of freedom if the samples follow independent normal distributions. If there are at least 4 complete pairs of observation, an asymptotic confidence interval is given based on Fisher's Z transform (https://stat.ethz.ch/R-manual/R-devel/library/stats/html/cor.test.html). . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint 35 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint 36 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint 37 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint 38 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint 39 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint 40 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint 41 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)The copyright holder for this preprint this version posted June 9, 2021. ; https://doi.org/10.1101/2021.06.08.21258284 doi: medRxiv preprint