mRNA vaccine-induced SARS-CoV-2 spike-specific IFN-γ and IL-2 T-cell responses are predictive of serological neutralization and are transiently enhanced by pre-existing cross-reactive immunity

Abstract

The contributions of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)-specific T cells to vaccine efficacy and durability are unclear. We investigated relationships between mRNA vaccine-induced spike-specific interferon- gamma (IFN-γ) and interleukin-2 (IL-2) T-cell responses and neutralizing antibody development in long-term care home staff doubly vaccinated with BNT162b2 or mRNA-1273. The impacts of pre-existing cross-reactive T-cell immunity on cellular and humoral responses to vaccination were additionally assessed. Mathematical modeling of the kinetics of spike-specific IFN-γ and IL-2 T-cell responses over 6 months post-second dose was bifurcated into recipients who exhibited gradual increases with doubling times of 155 and 167 days or decreases with half-lives of 165 and 132 days, respectively. Differences in kinetics did not correlate with clinical phenotypes. Serological anti-spike IgG, anti-receptor binding domain (RBD) IgG, anti-spike IgA, and anti-RBD IgA antibody levels otherwise decayed in all participants with half-lives of 63, 57, 79, and 46 days, respectively, alongside waning neutralizing capacity (t_1/2 = 408 days). Spike-specific T-cell responses induced at 2-6 weeks positively correlated with live viral neutralization at 6 months post-second dose, especially in hybrid immune individuals. Participants with pre-existing cross-reactive T-cell immunity to SARS-CoV-2 exhibited greater spike-specific T-cell responses, reduced anti-RBD IgA antibody levels, and a trending increase in neutralization at 2-6 weeks post-second dose. Non-spike-specific T-cells predominantly targeted SARS-CoV-2 non-structural protein at 6 months post-second dose in cross-reactive participants. mRNA vaccination was lastly shown to induce off-target T-cell responses against unrelated antigens. In summary, vaccine-induced spike-specific T-cell immunity appeared to influence serological neutralizing capacity, with only a modest effect induced by pre-existing cross-reactivity.

Importance: Our findings provide valuable insights into the potential contributions of mRNA vaccine-induced spike-specific T-cell responses to the durability of neutralizing antibody levels in both uninfected and hybrid immune recipients. Our study additionally sheds light on the precise impacts of pre-existing cross-reactive T-cell immunity to severe acute respiratory syndrome coronavirus 2 on the magnitude and kinetics of cellular and humoral responses to vaccination. Accordingly, our data will help optimize the development of next-generation T cell-based coronavirus vaccines and vaccine regimens to maximize efficacy and durability.

Keywords: SARS-CoV-2; T-cell immunity; cross-reactivity; humoral immunity; hybrid immunity; mRNA vaccines.

Fig 1

An overview of the LTCH staff cohort throughout the study.

Fig 2

A representative ELISpot comparing spike and non-spike-specific IFN-γ (blue) and IL-2 (red) T-cell responses between non-cross-reactive and cross-reactive vaccinees. Spot counts are provided as IFN-γ/IL-2 spot-forming cells (SFC) per 250,000 PBMC. DMSO = negative controls; N = nucleocapsid; E = envelope; M = membrane; S1 + S2 = full spike; NSP = non-structural protein; CEF + CEFTA + SEB = positive controls.

Fig 3

(A) Scatterplot comparisons of N-, E-, M-, and NSP-specific T-cell responses between CR (blue circles) and HI (red squares) vaccine recipients. (B) Pie chart comparisons of the distribution of IFN-γ, IL-2, and dual IFN-γ/IL-2 T-cell responses to SARS-CoV-2 nucleoprotein (red), envelope (yellow), membrane (blue), and NSP (green) peptide masterpools between HI and CR vaccine recipients of BNT162b2 or mRNA-1273 at 2–6 weeks and 6 months post-second dose.

Fig 4

(A–C) comparisons of net spike-specific (A) IFN-γ, (B) IL-2, and (C) dual IFN-γ/IL-2 T-cell responses at 2–6 weeks and 6 months post-second dose. Squares and whiskers within each violin plot represent medians with interquartile ranges (IQRs). Refer to Table S4 for medians with IQRs for each group. Significant cross-sectional comparisons are indicated with black asterisks. Significant longitudinal comparisons are indicated with blue asterisks. Baseline is comprised of a group of seven uninfected LTCH staff. (D, E) Violin plot analysis of (D) growth and (E) decay rates in spike-specific IFN-γ and IL-2 T-cell responses across the entire cohort and between mRNA vaccine types. Squares and whiskers within each violin plot represent medians with IQRs. (F, G) Spike-specific, (F) IFN-γ, and (G) IL-2 T-cell responses as a function of time since dose 1 for n = 113 participants that could be analyzed longitudinally, separated by increasing (blue) and decreasing (yellow) trends. Refer to Table S5 for a summary of median decay and growth rates with half-lives and doubling times.

Fig 5

(A, B, G, H) Violin plot analysis of serological (A) anti-spike IgG, (B) anti-RBD IgG, (G) anti-spike IgA, and (H) anti-RBD IgA antibody levels at 2–6 weeks and 6 months post-second dose. Squares and whiskers within each violin plot represent medians with IQRs (see Table S6 for medians with IQRs). Significant cross-sectional comparisons are indicated with black asterisks. Significant longitudinal comparisons are displayed with blue asterisks for the entire cohort alone to avoid overcrowding the figure. Baseline is comprised of a group of seven uninfected LTCH staff. (C, D, I, J) Violin plot analysis of (C) anti-spike IgG, (D) anti-RBD IgG, (I) anti-spike IgA, and (J) anti-RBD IgA decay rates across the entire LTCH staff cohort after adjusting for sex, vaccine type, and immunological phenotype. (E, K) Mathematical modeling of decay rates in (E) anti-spike IgG (green), anti-RBD IgG (red), (K) anti-spike IgA (green), and anti-RBD IgA (red) antibody levels since dose 1 of BNT162b2 or mRNA-1273. (F, L) A histogram of individual decay rates in (F) anti-spike IgG (green), anti-RBD IgG (red), (L) anti-spike IgA (green), and anti-RBD IgA (red) antibody levels. Refer to Table S7 for a summary of median decay rates with half-lives for anti-spike and anti-RBD IgG and IgA antibody levels.

Fig 6

(A) Violin plot analysis of serological neutralizing capacity against ancestral SARS-CoV-2 at 2–6 weeks and 6 months post-second dose. Squares and whiskers within each violin plot represent medians with IQRs. Refer to Table S8 for medians with IQRs for each group. Significant cross-sectional comparisons are indicated with black asterisks. Significant longitudinal comparisons are displayed with blue asterisks for the entire cohort alone to avoid overcrowding the figure. Baseline is comprised of a group of seven uninfected LTCH staff. (B) Violin plot analysis of decay rates in serological neutralizing capacity across the entire LTCH staff cohort and following adjustments for sex, vaccine type, and immunological phenotype. (C) Mathematical modeling of decay in serological neutralizing capacity since dose 1 of BNT162b2 or mRNA-1273. (D) A histogram of individual decay rates in serological neutralizing capacity. Refer to Table S9 for a summary of median decay rates with half-lives.

Fig 7

(A) Spike-specific T-cell responses at 2–6 weeks post-second dose plotted against serological anti-spike IgG, anti-RBD IgG, and anti-RBD IgA antibody levels at 6 months post-second in uninfected and HI LTCH staff. (B) Serological neutralizing capacity at 6 months plotted against spike-specific T-cell responses at 2–6 weeks or 6 months post-second dose. Black circles = uninfected; red squares = HI.

Fig 8

A longitudinal comparison of net non-spike-specific IFN-γ, IL-2, and dual IFN-γ/IL-2 T-cell responses from 2 to 6 weeks to 6 months post-second dose in NCR (n = 30), CR (n = 48), HI (n = 20), and N+, PCR(−) (n = 7) vaccine recipients. Non-spike-specific T-cell responses represent cumulative responses to nucleoprotein, envelope, membrane, and non-structural protein. Violin plots represent median responses with interquartile ranges. (B) Longitudinal pairwise comparisons (n = 7) of net CEF-, CEFTA-, and SEB-specific IFN-γ, IL-2, and dual IFN-γ/IL-2 T-cell responses from baseline (red circles) to 2–6 weeks post-second dose (blue squares). (C) Longitudinal pairwise comparisons (n = 105) of net CEF-, CEFTA-, and SEB-specific IFN-γ, IL-2, and dual IFN-γ/IL-2 T-cell responses from 2 to 6 weeks (red circles) to 6 months post-second dose (blue squares).