Booster COVID-19 mRNA vaccination ameliorates impaired B-cell but not T-cell responses in older adults

Abstract

Age-associated differences in the effect of repetitive vaccination, particularly on memory T-cell and B-cell responses, remain unclear. While older adults (aged ≥65 years) exhibited enhanced IgG responses following COVID-19 mRNA booster vaccination, they produced fewer spike-specific circulating follicular helper T cells-1 than younger adults. Similarly, the cytotoxic CD8⁺ T-cell response remained diminished with reduced PD-1 expression even after booster vaccination compared with that in younger adults, suggesting impaired memory T-cell activation in older adults. In contrast, although B-cell responses in older adults were weaker than those in younger adults in the primary response, the responses were significantly enhanced upon booster vaccination, reaching levels comparable with that observed in younger adults. Therefore, while booster vaccination ameliorates impaired humoral immunity in older adults by efficiently stimulating memory B-cell responses, it may less effectively enhance T-cell-mediated cellular immunity. Our study provides insights for the development of effective therapeutic and vaccine strategies for the most vulnerable older population.

Keywords: COVID-19 article type: original research article; booster vaccination; cTfh; immune aging; immunological memory; mRNA vaccine; memory B cells; memory T cells.

Figure 1

Study design. Timeline of the vaccinations and blood collections. Participants received the first dose of BNT162b2 on day 0, the second dose around day 21, and the third dose approximately nine months after the first dose. The sampling points were set before the first vaccination (Pre), seven days after the first dose to one day before the second dose (Post1), two weeks after the second dose ± four days (Post2), three months after the first dose ± two weeks (3mo-Post1), and six months after the first dose ± two weeks (6mo-Post1). The sampling points in the follow-up study were set at two weeks ± four days after the third dose (Post3) and six months ± four weeks after the third dose vaccination (6mo-Post3).

Figure 2

Anti-RBD antibody responses to third-dose vaccination are comparable between adults and older adults. (A) Concentrations of anti-RBD IgG and IgM antibodies. Each dot represents the serum antibody titer of each subject. Center lines and error bars indicate the median and interquartile range (IQR), respectively. Dashed and dotted lines indicate cutoffs and limits of detection (LODs), respectively. Sample numbers were as follows: 107 (Adults) and 108 (Older adults) (Pre), 106 and 108 (Post1), 105 and 108 (Post2), 105 and 106 (3mo-Post1), 105 and 106 (6mo-Post1), 70 and 36 (Post3), and 65 and 37 (6mo-Post3), respectively. (See also Supplementary Figure 1 ). (B) Anti-RBD IgG antibody concentrations six months after the first dose (6mo-Post1) and six months after the third dose (6mo-Post3). ****p < 0.0001 (Wilcoxon matched-pairs signed rank test). (C) The ratio of the anti-RBD IgG titer after the third vaccination (Post3) to that after the second dose (Post2). (D) Correlation between anti-RBD IgG titers (AU/mL) after the second dose (Post2) and after the third dose (Post3). Spearman’s rank correlation (r_s) was used to identify correlations between two variables, and regression lines were drawn based on linear regression analysis. Blue, red, and black characters represent the results of adults, older adults, and both groups, respectively. (E) Neutralization capacity of sera collected after the second dose (Post2) and the third dose (Post3) against RBD of ancestral, Delta, and Omicron strains. Center lines and error bars indicate the median and IQR, respectively. (A, C, E) Blue and red represent adults and older adults, respectively. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns, not significant (Mann–Whitney test).

Figure 3

Spike-specific cTfh1 cells are fewer in older adults than in adults after third-dose vaccination. (A) Absolute numbers of AIM⁺cTfh1 cells in blood. Each dot represents the value of each subject. Bar graphs indicate medians, error bars indicate IQRs. Blue and red represent adults and older adults, respectively. Sample numbers were as follows: 5 (Adults) and 5 (Older adults) (Pre), 32 and 35 (Post1), 31 and 35 (Post2), 20 and 22 (3mo-Post1), 31 and 35 (6mo-Post1), 28 and 18 (Post3), and 23 and 14 (6mo-Post3), respectively. (See also Supplementary Figure 2 ). *p < 0.05, ns, not significant (Mann–Whitney test). (B) Comparison of the numbers (x10³ cells/L) of AIM⁺cTfh1 cells at six months (6mo-Post1) and after the third dose (Post3). ***p < 0.001, ****p < 0.0001 (Wilcoxon matched-pairs signed rank test). (C) Correlations between AIM⁺cTfh1 cell numbers (x10³ cells/L) at the indicated time points. (D) Correlations between AIM⁺cTfh1 cell numbers (x10³ cells/L) and anti-RBD IgG concentrations (AU/mL) at the indicated time point. (C, D) Spearman’s rank correlation (r_s) was used to identify correlations between two variables, and regression lines were drawn based on linear regression analysis. Blue, red, and black characters represent the results of adults, older adults, and both groups, respectively.

Figure 4

Spike-specific CD8⁺ T-cell responses are lower in older adults after the third dose. (A) Absolute numbers of AIM⁺CD8⁺ T cells in blood. Each dot represents the value of each subject. Bar graphs indicate medians, error bars indicate IQRs. Blue and red represent adults and older adults, respectively. (B) Comparison of the numbers of AIM⁺CD8⁺ T cells at six months after the first dose (6mo-Post1) and after the third dose (Post3). *p < 0.05, ****p < 0.0001 (Wilcoxon matched-pairs signed rank test). (C) The ratio of the AIM⁺CD8⁺ T cell number after the third vaccination (Post3) to that six months after the first dose (6mo-Post1). (D) Correlation between AIM⁺CD8⁺ T cell numbers at the indicated time point. Spearman’s rank correlation (r_s) was used to identify correlations between two variables, and regression lines were drawn based on linear regression analysis. Blue, red, and black characters represent the results of adults, older adults, and both groups, respectively. (E) Proportion of participants with SI of AIM⁺CD8⁺ T cells ≤2 at each time point. (F) Geometric mean fluorescence intensity (gMFI) of PD-1 on AIM⁺CD8⁺ T cells at each time point. Blue and red represent adults and older adults, respectively. Dots and error bars indicate medians and IQRs, respectively. (A, C, F) *p < 0.05, **p < 0.01, ns, not significant (Mann–Whitney test).

Figure 5

Spike-specific memory B cells (MBCs) are induced to comparable levels in adults and older adults after third-dose vaccination. (A) Representative flow-cytometric data of ancestral spike-binding MBCs at each time point. (B) Absolute numbers of ancestral spike- or Omicron spike-binding MBCs in blood at each time point. Sample numbers were as follows: 5 (Adults) and 5 (Older adults) (Pre), 32 and 35 (Post1), 31 and 35 (Post2), 31 and 35 (3mo-Post1), 31 and 34 (6mo-Post1), 28 and 18 (Post3), and 23 and 14 (6mo-Post3), respectively. (C) Frequency of activated (CD21^–CD27⁺) cells in Ancestral spike (AS)-binding MBCs. (D) Frequency of AS-binding activated MBCs after the second dose (Post2) and third dose (Post3). **p < 0.01, ****p < 0.0001 (Wilcoxon matched-pairs signed rank test). (E) Ratio of AS-binding MBCs after the third vaccination (Post3) to that six months after the first dose (6mo-Post1). (F) MFIs of AS-BV421 or OS-BV785 among spike-binding MBCs. Dots and error bars indicate medians and IQRs, respectively. (G) Absolute numbers of AS- or OS-binding IgG⁺ MBCs in blood at each time point. (H) Correlation between the number (x10⁴ cells/L) of AS-binding MBCs after the third dose (Post3) and the number of AS-binding MBCs (x10⁴ cells/L) or AIM⁺cTfh1 cells (x10³ cells/L) at the indicated time point. (I) Correlation between the number of AS-binding IgG⁺ MBCs (x10⁴ cells/L) and the concentration of anti-RBD IgG (AU/mL) at the indicated time point. (B, C, G) Each dot represents the value of each subject. Bar graphs indicate medians, error bars indicate IQRs. (B, C, E–G) Blue and red represent adults and older adults, respectively. *p < 0.05, **p < 0.01, ns, not significant (Mann–Whitney test). (H, I) Spearman’s rank correlation (r_s) was used to identify correlations between two variables, and regression lines were drawn based on linear regression analysis. Blue, red, and black characters represent the results of adults, older adults, and both groups, respectively.