BNT162b2-induced memory T cells respond to the Omicron variant with preserved polyfunctionality – Nature Microbiology

BNT162b2-induced memory T cell responses against the Omicron spike protein

The present study enrolled HCWs without previous SARS-CoV-2 infection (seronegative for SARS-CoV-2 nucleocapsid) who were immunized with 2 (n = 20) or 3 (n = 20) doses of the BNT162b2 messenger RNA vaccine. We also recruited and longitudinally investigated individuals with prior SARS-CoV-2 infection who were immunized with 2 doses of BNT162b2 vaccine (n = 20).

We performed intracellular cytokine staining (ICS) assays for interferon-γ (IFN-γ), interleukin-2 (IL-2) and tumour necrosis factor (TNF) in peripheral blood mononuclear cells (PBMCs) that were stimulated using overlapping peptide (OLP) pools for the spike protein of the SARS-CoV-2 wild-type (WT) (lineage B) strain and the Omicron variant (Fig. 1a). We compared the frequency of IFN-γ-producing CD4+ T cells against the WT spike protein versus the Omicron spike protein and found that the average frequency against the Omicron spike protein was 22% (P < 0.001) and 10% (P < 0.05) lower in PBMCs from HCWs with 2 and 3 doses of the vaccine, respectively (Fig. 1b, left). Nevertheless, BNT162b2-induced memory CD4+ T cells maintained substantial responses to the Omicron spike protein. Notably, the two- and three-dose vaccinated groups did not exhibit significantly different responses, although a third vaccination dose is known to boost the neutralizing activities of nAbs against the Omicron variant7,12. The frequency of TNF-producing CD4+ T cells against the Omicron spike protein, compared to the WT spike protein, was decreased by 14% (P < 0.005) and 6% (non-significant) in the two- and three-dose vaccinated groups, respectively (Fig. 1b, middle). Similarly, the frequency of IL-2-producing CD4+ T cells against the Omicron spike protein, compared to the WT spike protein, was decreased by 21% (P < 0.001) in the two-dose vaccinated group and by 3% (non-significant) in the three-dose vaccinated group (Fig. 1b, right).

Fig. 1: Cytokine-secreting functions of BNT162b2-induced memory T cells against the WT and Omicron spike proteins.
figure 1

ICS was performed to examine the frequency of CD4+ or CD8+ T cells responding to the WT and Omicron spike proteins. a,d, Representative flow cytometry plots showing cytokine-producing cells among CD4+ (a) or CD8+ (d) T cells. b,c,e,f, The frequency of cytokine-producing CD4+ (b,c) or CD8+ (e,f) T cells against the WT and Omicron spike proteins in HCWs with 2 (n = 20) or 3 (n = 20) doses of the BNT162b2 vaccine (b,e) and in COVID-19-recovered individuals with 2 BNT162b2 vaccine doses (n = 20) (c,f). Data are presented as the mean ± s.d. The percentage numbers in each graph indicate the mean decrease in the frequency of cytokine-producing cells against the Omicron spike protein compared to the WT spike protein. P values were calculated with a two-tailed, matched-pairs Wilcoxon signed-rank test (WT versus Omicron) or two-tailed, unpaired Mann–Whitney U-test. NS, not significant.

Source data

We also performed a longitudinal analysis among individuals with prior SARS-CoV-2 infection and two-dose BNT162b2 vaccination at three time points: before vaccination (ten months after recovery from mild-to-moderate COVID-19) and at one and three months after the second vaccination. Two-dose BNT162b2 vaccination significantly increased the frequency of IFN-γ-producing CD4+ T cells against both the WT and Omicron spike protein (Fig. 1c, left). The frequency of IFN-γ-producing CD4+ T cells against the Omicron spike protein, compared to against the WT spike protein, was reduced by 26% (P < 0.01) before vaccination, 5% (non-significant) at 1 month after the second vaccination and 9% (P < 0.05) at 3 months after the second vaccination. Similarly, against the Omicron spike protein compared to the WT spike protein, the frequency of TNF-producing CD4+ T cells decreased by 19% (P < 0.05), 7% (non-significant) and 19% (P < 0.01) at the 3 time points (Fig. 1c, middle); the frequency of IL-2-producing CD4+ T cells decreased by 3% (non-significant), 8% (P < 0.05) and 11% (P < 0.005) at the 3 time points (Fig. 1c, right). These findings indicated that vaccine-induced memory T cells substantially responded to the Omicron spike protein.

The observed CD8+ T cell responses were relatively weak and IL-2-producing cells were not detected among CD8+ T cells (Fig. 1d). Therefore, we analysed IFN-γ and TNF production from CD8+ T cells and found that memory CD8+ T cells exhibited substantial responses to the Omicron spike protein. Both the IFN-γ and TNF responses were non-significantly decreased against the Omicron spike protein compared to the WT spike protein (Fig. 1e,f) and did not significantly differ between the two- and three-dose vaccinated groups (Fig. 1e). Vaccination induced significantly increased responses against both WT and Omicron spike protein among individuals with prior infection (Fig. 1f).

We also analysed the geometric mean fluorescence intensity (MFI) for each cytokine in cytokine+ cells. Our results showed that the geometric MFI did not differ between T cells responding to the WT versus Omicron spike protein in the CD4+ or CD8+ T cell population (Extended Data Fig. 1).

Omicron spike protein-responding memory T cell responses in individual donors

Next, we analysed the percentages of positive responders for each cytokine. A positive response was defined as when the percentage of cytokine+ cells after OLP stimulation was at least threefold higher than in the negative controls for each sample19,20,21. Our analysis of CD4+ T cells from HCWs revealed an IFN-γ+ response against the WT spike protein in 90% (18 of 20) of two-dose vaccinated donors and 60% (12 of 20) of three-dose vaccinated donors (Fig. 2a, left). The corresponding percentages of IFN-γ+ responses against the Omicron spike protein were reduced to 80% (16 of 20 two-dose vaccinated donors) and 55% (11 of 20 three-dose vaccinated donors); however, these percentages did not notably differ from the percentages of positive responders against the WT spike protein. Similar results were observed in analyses of TNF- or IL-2-producing CD4+ T cells (Fig. 2a, middle and right).

Fig. 2: Percentage of positive responders against the WT and Omicron spike proteins.
figure 2

ad, The percentage of positive responders was determined on the basis of the percentage of cells positive for each cytokine in the CD4+ (a,b) or CD8+ (c,d) T cell population in HCWs with 2 (n = 20) or 3 (n = 20) doses of the BNT162b2 vaccine (a,c) and in COVID-19-recovered individuals with 2 BNT162b2 vaccine doses (n = 20) (b,d). A positive response was defined as when the percentage of cytokine+ cells after OLP stimulation was more than threefold higher than the percentage of cytokine+ cells in the negative controls for each sample.

Source data

In individuals with prior infection, a two-dose BNT162b2 vaccination increased the percentage of IFN-γ+ responders to 100% against both WT and Omicron spike protein and these responses were maintained for 3 months (Fig. 2b, left). Similar results were obtained in analyses of TNF- or IL-2-producing CD4+ T cells (Fig. 2b, middle and right).

In our analysis of CD8+ T cell responses, the percentages of positive responders were relatively low compared to our analysis of CD4+ T cell responses. However, the percentages of IFN-γ+ or TNF+ responders did not notably differ between CD8+ T cell responses against the WT versus Omicron spike protein in HCWs (Fig. 2c) or in individuals with prior infection (Fig. 2d).

We also calculated the ratio of the frequency of cytokine-producing T cells against the Omicron spike protein to the frequency of cytokine-producing T cells against the WT spike protein for each cytokine in each individual. In the CD4+ T cell population, this ratio ranged from 0.5 to 1.5, irrespective of the cytokines in most cases (Fig. 3a). However, this ratio was <0.1 for IFN-γ in 2 donors and for TNF in a single donor. Within the CD8+ T cell population, this ratio was <0.5 for IFN-γ and TNF in several donors (Fig. 3b). These findings indicate that although BNT162b2-induced memory T cells generally exhibited a substantial response to the Omicron spike protein, they could not respond to the Omicron spike protein in certain individuals.

Fig. 3: The ratio of the response against the Omicron spike protein to the response against the WT spike protein.
figure 3

a,b, The ratio of the frequency of cytokine-producing T cells against the Omicron spike protein to the frequency of cytokine-producing T cells against the WT spike protein was determined for each cytokine in each individual. CD4+ (a) and CD8+ (b) T cells were analysed in HCWs with two or three doses of the BNT162b2 vaccine and in COVID-19-recovered individuals with two BNT162b2 vaccine doses. Positive responders for each cytokine against the WT spike protein were included in this analysis. Data are presented as mean values ± s.d.

Source data

Polyfunctionality of BNT162b2-induced memory T cells against the Omicron spike protein

We further focused on the polyfunctionality of the memory T cells that responded to the WT or Omicron spike protein. Polyfunctional T cells, which simultaneously exert several effector functions, play a crucial role in host protection during viral infection22,23,24,25. Representative flow cytometry dot plots showed that a proportion of CD4+ T cells simultaneously produced IFN-γ, TNF and IL-2 in response to both WT and Omicron spike protein OLPs (Fig. 4a). Notably, the triple-positive population exhibited significantly higher geometric MFI for each cytokine compared to the double- and single-positive populations and the double-positive population had a significantly higher geometric MFI than the single-positive population (Fig. 4b and Extended Data Fig. 2). We additionally analysed polyfunctional (triple-positive or double-positive) CD4+ T cells from two- and three-dose vaccinated HCWs. Against the Omicron spike protein compared to the WT spike, the average percentage of polyfunctional cells decreased by 8% (P < 0.05) in two-dose vaccinated donors and 2% (non-significant) in three-dose vaccinated donors (Fig. 4c). In our analysis of polyfunctionality using every possible combination of functions, we observed no significant difference between CD4+ T cells responding to the WT versus Omicron spike protein (Fig. 4d).

Fig. 4: Polyfunctionality of BNT162b2-induced memory CD4+ T cells.
figure 4

a, Representative flow cytometry plots showing polyfunctional cells among CD4+ T cells. b, Representative polychromatic dot plots for IFN-γ and TNF. T cells positive for a given number of functions are marked in different colours to show their distribution in the dot plots for IFN-γ and TNF. c, Percentage of polyfunctional T cells among any cytokine-producing CD4+ T cells from HCWs with 2 (n = 18) or 3 (n = 12) doses of the BNT162b2 vaccine. d, Detailed analyses of polyfunctionality are presented with every possible combination of functions (n = 18 for 2× vaccination; n = 12 for 3× vaccination). e, Percentage of polyfunctional T cells among any cytokine-producing CD4+ T cells from COVID-19-recovered individuals with 2 doses of the BNT162b2 vaccine (n = 18 for pre-vaccine; n = 20 for 1 month post 2× vaccine; n = 20 for 1 month post 2× vaccine). f, Pie graphs representing the fraction of cells positive for a given number of functions among CD4+ T cells with any type of function before vaccination and one month after the second vaccination. Permutation test = 10,000 permutations. g, Detailed analyses of polyfunctionality are presented with every possible combination of functions (n = 20). CD4+ T cell IFN-γ responders against the WT spike protein were included in this analysis. ce,g, Data are presented as mean values ± s.d. P values were calculated with a two-tailed, matched-pairs Wilcoxon signed-rank test (WT versus Omicron) or two-tailed, unpaired Mann–Whitney U-test.

Source data

In individuals with prior infection, two-dose BNT162b2 vaccination significantly increased the percentage of polyfunctional cells against both the WT and Omicron spike protein (Fig. 4e). At all three time points, the percentage of polyfunctional cells did not significantly differ between CD4+ T cells responding to the WT versus Omicron spike protein. Polyfunctionality analysis using pie graphs confirmed that two-dose vaccination was associated with a significant increase in the percentage of polyfunctional cells against both the WT and Omicron spike protein (Fig. 4f). In our analysis of polyfunctionality using every possible combination of functions, we found no significant difference between the CD4+ T cells responding to the WT versus Omicron spike protein (Fig. 4g).

We also analysed the polyfunctionality of BNT162b2-induced memory CD8+ T cells. In the HCW groups with two and three doses of the vaccine, the percentage of polyfunctional cells was not significantly decreased against the Omicron spike protein compared to the WT spike (Fig. 5a). In our analysis of polyfunctionality using every possible combination of functions, we observed no significant difference between CD8+ T cells responding to the WT versus Omicron spike protein (Fig. 5b). In individuals with prior infection, at all three time points, the percentage of polyfunctional cells did not significantly differ between CD8+ T cells responding to WT versus Omicron spike protein (Fig. 5c). The pie graphs also confirmed that the polyfunctional responses did not significantly differ against WT versus Omicron spike protein (Fig. 5d); when we analysed polyfunctionality using every possible combination of functions, we again found no significant difference between CD8+ T cells responding to the WT versus Omicron spike protein (Fig. 5e). Taken together, these results demonstrate that BNT162b2-induced memory CD4+ and CD8+ T cells exhibit preserved polyfunctionality against the Omicron spike protein.

Fig. 5: Polyfunctionality of BNT162b2-induced memory CD8+ T cells.
figure 5

a, Percentage of polyfunctional T cells among any cytokine-producing CD8+ T cells from HCWs with two or three doses of the BNT162b2 vaccine. b, Detailed analyses of polyfunctionality are presented with every possible combination of functions. c, Percentage of polyfunctional T cells among any cytokine-producing CD8+ T cells from COVID-19-recovered individuals with two doses of the BNT162b2 vaccine. d, Pie graphs representing the fraction of cells positive for a given number of functions among CD8+ T cells with any type of functions before vaccination and one month after the second vaccination. Permutation test = 10,000 permutations. e, Detailed analyses of polyfunctionality are presented with every possible combination of functions. CD8+ T cell IFN-γ responders against the WT spike protein were included in this analysis. ac,e, Data are presented as mean values ± s.d. P values were calculated with a two-tailed, matched-pairs Wilcoxon signed-rank test (WT versus Omicron) or two-tailed, unpaired Mann–Whitney U-test.

Source data

#BNT162b2induced #memory #cells #respond #Omicron #variant #preserved #polyfunctionality #Nature #Microbiology

Leave a Comment

Your email address will not be published. Required fields are marked *