Intramuscular mRNA BNT162b2 vaccine against SARS-CoV-2 induces neutralizing salivary IgA

Abstract

Intramuscularly administered vaccines stimulate robust serum neutralizing antibodies, yet they are often less competent in eliciting sustainable "sterilizing immunity" at the mucosal level. Our study uncovers a strong temporary neutralizing mucosal component of immunity, emanating from intramuscular administration of an mRNA vaccine. We show that saliva of BNT162b2 vaccinees contains temporary IgA targeting the receptor-binding domain (RBD) of severe acute respiratory syndrome coronavirus-2 spike protein and demonstrate that these IgAs mediate neutralization. RBD-targeting IgAs were found to associate with the secretory component, indicating their bona fide transcytotic origin and their polymeric multivalent nature. The mechanistic understanding of the high neutralizing activity provided by mucosal IgA, acting at the first line of defense, will advance vaccination design and surveillance principles and may point to novel treatment approaches and new routes of vaccine administration and boosting.

Keywords: BNT162b2 vaccine; SARS-CoV-2 neutralizing Abs; mucosal immunity; secretory IgA; secretory component.

Figure 1

BNT162b2 vaccinees mount serum anti–RBD–SARS-CoV-2 IgG and IgA, with IgG showing strong neutralization potential. (A) Independent ELISA measurements of anti-RBD IgG and of anti-RBD IgA in serum samples collected from pre-COVID (N = 51), BNT162b2 vaccinees (N = 17), and post-COVID-19 (N = 22) convalescents, as indicated. Convalescent samples were collected within 3–10 weeks post-recovery, as defined by the clinical definition in Israel at time of sample collection. BNT162b2 vaccinee samples for which kinetic samples were available, the value shown represents the peaks of individual responses based on their kinetic curve [see (B) and Table S7 ]. (B) Quantitative kinetic profile of anti-RBD IgG (blue) and IgA (red) in serum sampled (N = 76) in the vaccinee cohort (N = 18), plotted as a function of days, after first vaccine dose. See Table S1 for cohort and sampling details. Independent ordinate axes for IgG (left, blue) and IgA (right, red) highlight the restricted, relative nature of the comparison between isotypes in this experiment, as discussed in the text; see also Figure 2 for subsequent developments. Green arrows indicate timing of the second vaccine dose (the boost). (C) Serum neutralization assessed by SARS-CoV-2 spike-pseudotyped VSV-GFP-ΔG reporter assay on Vero-E6 cells. Neutralization is expressed as a percentage of pseudovirus-infected green cells without serum (total infection = 100%). Percentage of neutralization by sera of pool of four individual vaccinees is plotted as a function of the reciprocal values of sera dilutions displayed on a log2 scale, as indicated (filled circles, total serum, NT50 is reached on average at the dilution of ∼1:360, extrapolated by cross-section with the dashed line. Contribution of IgG to serum neutralization is evaluated by the depletion of the IgG isotype using anti-IgG specific magnetic beads (triangles). See Figure S1F for assessment of completeness and specificity of IgG depletion. Results of three experimental repeats are represented.

Figure 2

Quantitative ELISA measurement of anti-RBD IgG and IgA content in biological fluids. (A) Schematic representation of detection of anti-RBD IgG or IgA by indirect ELISA using isotype-specific HRP-conjugated secondary antibodies. OD values are not directly comparable between the isotypes because of the use of different secondary antibodies. (B) Schematic representation of sandwich capturing ELISA for selective quantification of total immunoglobulin isotypes (IgG vs. IgA). Yellow stars schematically represent molar equivalents of antibody quantities. We introduce pure IgA and IgG commercial references to transform the OD values to their molar equivalents, using the standard dilution curve in capture ELISA format. Implementing such a standard in every experiment allows for determining the antigen specific and total molar amounts of each isotype within the linearity range. We assume average molecular weight (MW) of IgG = 146 kDa and IgA = 150 kDa in circulation. (C) Molar measurement of anti-RBD IgG and IgA implementing the methodology described in (A, B) depicts stoichiometric ratios between the antigen-specific isotypes. (D) Percentage of antigen-specific anti-RBD out of total immunoglobulin isotype, as indicated. (E) Individual longitidual profiles of anti-RBD IgG (blue) and IgA (red) monitored in six vaccinated individuals up to 7 months after vaccination are inferred as picomole per ml of serum.

Figure 3

Detection of anti-RBD IgA in resting saliva of BNT162b2 vaccinees and characterization of its neutralizing potential. (A) Longitudinal assessment of molar quantities of anti-RBD IgA in saliva of vaccinees compared with naïve individuals is presented in picomole per ml and time-categorized as indicated. The molar expression in saliva is corrected to bi-valent for the comparison to circulatory immunoglobulins. (B) Saliva neutralization assessed by SARS-CoV-2 spike-pseudotyped VSV-GFP-ΔG reporter assay on Vero-E6 cells. Neutralization is expressed as a percentage of pseudovirus-infected green cells without incubation with saliva (total infection = 100%). Percentage of independently measured neutralization by five naïve individuals vs. four vaccinees are plotted as a function of the reciprocal values of sera dilutions displayed on a log2 scale, as indicated. Each neutralization curve was tested in three biological replicates. Standard deviation represents difference between individuals in each group. The NT50 of vaccinees saliva is achieved on average at the dilution of ∼1:60, extrapolated by cross-section with the dashed line. The specific neutralization NT50 value is reached at dilution of ∼1:20 and represents “vaccine-added” neutralization, corrected to the basal innate neutralization of naïve individuals, that is probably the consequence of innate proteolytic and mucus (lectin) presence in naïve saliva. (C) The values of anti-RBD IgA and IgG in picomole per ml of five naïve saliva samples and four saliva samples of vaccinees, used in neutralization assay described in (B) are shown.

Figure 4

The association of salivary anti-RBD IgA with the secretory component governs the prominent neutralization activity in vaccinees. (A) Depletion of IgA from saliva samples of vaccinees completely abrogates the specific neutralization activity of vaccinees saliva. Saliva neutralization was assessed by SARS-CoV-2 spike-pseudotyped VSV-GFP-ΔG reporter assay on Vero-E6 cells. The magnitude of neutralization is expressed as a percentage of pseudovirus-infected green cells without incubation with saliva (total infection = 100%). Percentage of measured neutralization by saliva pool of five vaccinees is plotted as a function of the reciprocal values of the saliva dilutions displayed on a log2 scale. The NT50 of saliva pool is reached on average at the dilution of ∼1:60 (extrapolated by the cross-section with the dashed line). Depletion of IgA results in abrogation of vaccine-induced neutralization activity (squares), whereas IgG-depleted saliva pool coincides with the non-depleted pool (triangles vs. circles). Depletion is achieved using anti-IgA and anti-IgG–specific magnetic beads. Results of three experimental repeats are represented. Analyses of completeness of isotype depletion and of its specificity are presented in Figure S4A . (B) Schematic outline of the detection of anti-RBD IgA in serum and SC-associated anti-RBD IgA in saliva samples. Illustrated are the expected differences between the circulatory monomeric IgA and the salivary mucosal dimeric/polymeric IgA, covalently bridged by J-chain and associated with pIgR. Left panel shows non-discriminative detection of both isoforms by anti-IgA secondary HRP-conjugate. Right panel shows the selective quantitative determination of dimeric/polymeric secretory IgA in saliva, but not in the serum, using anti-SC mouse monoclonal Ab, followed by anti-mouse secondary detection. (C) Molar quantification of dimeric secretory IgA. We introduce as reference standard using commercial secretory dimeric IgA purified from human colostrum to transform the OD values to their molar equivalents. We assume average molecular weight (MW) of dimeric secretory IgA = 424 g/mole. We considered IgA in its dimeric form because it is mostly found in that form at the mucosal surfaces and in mucosal secretions. (D) Analysis of dimeric anti-RBD SC-IgA in saliva (upper panel) vs. monomeric anti-RBD IgA in serum (lower panel) after vaccination, measured by quantitative ELISA. The molar expression in saliva is corrected to bi-valence to simplify the comparison to circulatory immunoglobulins.