Salivary, serological, and cellular immune response to the CoronaVac vaccine in health care workers with or without previous COVID-19

Abstract

We investigated the anti-SARS-CoV-2 post-vaccine response through serum and salivary antibodies, serum antibody neutralizing activity and cellular immune response in samples from health care workers who were immunized with two doses of an inactivated virus-based vaccine (CoronaVac) who had or did not have COVID-19 previously. IgA and IgG antibodies directed at the spike protein were analysed in samples of saliva and/or serum by ELISA and/or chemiluminescence assays; the neutralizing activity of serum antibodies against reference strain B, Gamma and Delta SARS-CoV-2 variants were evaluated using a virus neutralization test and SARS-CoV-2 reactive interferon-gamma T-cell were analysed by flow cytometry. CoronaVac was able to induce serum and salivary IgG anti-spike antibodies and IFN-γ producing T cells in most individuals who had recovered from COVID-19 and/or were vaccinated. Virus neutralizing activity was observed against the ancestral strain, with a reduced response against the variants. Vaccinated individuals who had previous COVID-19 presented higher responses than vaccinated individuals for all variables analysed. Our study provides evidence that the CoronaVac vaccine was able to induce the production of specific serum and saliva antibodies, serum virus neutralizing activity and cellular immune response, which were increased in previously COVID-19-infected individuals compared to uninfected individuals.

Figure 1

Serum IgG antibodies against SARS-CoV-2 spike protein. Serum from COVID-19 vaccinated (VAC) (n = 80; triangles) patients, those who were recovered from COVID-19 and vaccinated (REC/VAC) (n = 22; squares) and negative control (UI/UV) individuals (n = 13; circles) were analysed to measure the IgG antibodies anti-SARS-CoV-2 S1 and S2 proteins. Scatter plots show lines at the median with interquartile ranges. The dashed line represents the cut-off value for the test (33.8 BAU/mL). One-way ANOVA and the nonparametric Kruskal–Wallis test were used to compare the study groups. Asterisks denote statistical significance between the groups (***p < 0.001; ****p < 0.0001).

Figure 2

Neutralization of SARS CoV-2 lineages B, Gamma and Delta with serum from previously infected and/or vaccinated individuals, according to their VNT100. Serum obtained from COVID-19 vaccinated individuals (VAC) (n = 80; triangles), those recovered from COVID-19 and vaccinated (REC/VAC) (n = 21; squares) and negative control (UI/UV) individuals (n = 13; circles) were analysed to determine the virus neutralization titre (VNT₁₀₀) to the reference SARS-CoV-2 lineage B (A); Gamma (B) and Delta (C) variants. Scatter plots show lines at the median with interquartile ranges. The dashed lines represent the cut-off value for the test (20 VNT₁₀₀). One-way ANOVA and the nonparametric Kruskal–Wallis test were used to compare the study groups. Asterisks denote statistical significance between the groups (*p < 0.05; ***p < 0.001; ****p < 0.0001). Bar graphs represent the percentage of responders to each variant (D).

Figure 3

Salivary antibodies against SARS-CoV-2 spike protein. Saliva samples from COVID-19 vaccinated (VAC) (triangles), COVID-19 recovered vaccinated (REC/VAC) (squares) and negative control (UI/UV) individuals (circles) were analysed to determine the IgA (A) VAC (n = 42), REC/VAC (n = 22) and UI/UV (n = 13) and to determine the IgG (B) VAC (n = 35), REC/VAC (n = 22) and UI/UV (n = 13) antibodies against SARS-CoV-2 spike protein. Scatter plots show lines at the median with interquartile ranges. The dashed lines represent the cut-off value of 0.7 for IgA and 7 RU/mL for IgG. One-way ANOVA and the nonparametric Kruskal–Wallis test were used to compare the study groups. Asterisks denote statistically significant differences between the groups (**p < 0.01; ****p < 0.0001).

Figure 4

IFN-gamma production by T cells stimulated with SARS-CoV-2 pooled OPPs. PBMCs from vaccinated individuals (VAC) (triangles) (n = 72); COVID-19 recovered vaccinated individuals (REC/VAC) (squares) (n = 21) and uninfected/unvaccinated donors (UI/UV) (circles) (n = 13) were incubated for 18 h with a mixture of grouped SARS-CoV-2 peptide pools (M + N + S) at a final concentration of 1 μg/mL. The logarithmic scale represents the percentage of T cells producing IFN-γ. Scatter plots show lines at the median with interquartile ranges. IFN-γ expression by total lymphocytes (CD3+ T cells) (A); CD4+ T (B) and CD8+ T-lymphocytes (C) was analysed by intracellular staining. One-way ANOVA and the nonparametric Kruskal–Wallis test were used to compare the study groups. Asterisks denote statistically significant differences between the groups (*p < 0.05; **p < 0.01; ***p < 0.001).

Figure 5

Integrated data representation of analysed samples. (A) Hierarchical clustering heat maps without a reorganization of samples and features based on the values regarding the serology IgG, Virus Neutralization Titre (VNT₁₀₀) B, VNT₁₀₀ Gamma, VNT₁₀₀ Delta, salivary IgG, salivary IgA and %IFNg CD3+ for a total of 57 individuals: 27 in the VAC group (COVID-19 vaccinated), 18 in the REC/VAC group (COVID-19 recovered vaccinated individuals), and 12 in the UI/UV group (negative control)). (A) The values are shown as rectangles containing different colours corresponding to the levels indicated by the scale bar on the right. Each column represents each individual, and each line represents each variable (aerology IgG, VNT₁₀₀ B, VNT₁₀₀ Gamma, VNT₁₀₀ Delta, salivary IgG, salivary IgA and %IFNγ CD3+). The colours on the top represent each block of the three different groups (VAC—grey, REC/VAC—black and UI/UV—white). (B) Heat map based on the group averages. The average values are shown as rectangles containing different colours corresponding to the levels indicated by the scale bar on the right. Each line represents each variable (serology IgG, VNT₁₀₀ B, VNT₁₀₀ Gamma, VNT₁₀₀ Delta, salivary IgG, salivary IgA and %IFNγ CD3+), and each column represents each group. The colours on the top represent each block of the three different groups (VAC—grey, REC/VAC—black and UI/UV—white). The variables were normalized to a 0–100 scale by subtracting the minimum and dividing by the maximum of all the observations. The minimum and maximum values observed here were considered for each variable; a = 0 and b = 100. Second, all the normalized data were log-transformed (base 10). Then, the data were submitted to the Metaboanalyst 5.0 platform.