Low-dose mRNA-1273 COVID-19 vaccine generates durable memory enhanced by cross-reactive T cells

Abstract

Vaccine-specific CD4⁺ T cell, CD8⁺ T cell, binding antibody, and neutralizing antibody responses to the 25-μg Moderna messenger RNA (mRNA)–1273 vaccine were examined over the course of 7 months after immunization, including in multiple age groups, with a particular interest in assessing whether preexisting cross-reactive T cell memory affects vaccine-generated immunity. Vaccine-generated spike-specific memory CD4⁺ T cells 6 months after the second dose of the vaccine were comparable in quantity and quality to COVID-19 cases, including the presence of T follicular helper cells and interferon-γ–expressing cells. Spike-specific CD8⁺ T cells were generated in 88% of subjects, with equivalent memory at 6 months post-boost compared with COVID-19 cases. Lastly, subjects with preexisting cross-reactive CD4⁺ T cell memory exhibited stronger CD4⁺ T cell and antibody responses to the vaccine, demonstrating the biological relevance of severe acute respiratory syndrome coronavirus 2–cross-reactive CD4⁺ T cells.

Response to low-dose mRNA-1273 vaccination over 7 months.

Immunological memory of antibodies, CD4⁺ T cells, and CD8⁺ T cells was examined after low-dose mRNA vaccination. Levels of spike-specific immune memory were then compared to immune memory observed after natural infection with SARS-CoV-2 or after full-dose vaccination. Robust immune memory comparable to natural infection but lower than after full-dose vaccination was observed. Increased vaccinee age correlated with reduced antibody levels but had no effect on cellular immune memory. Immune memory was enhanced by preexisting cross-reactive T cells. D1, day 1.

Fig. 1.. Spike antibodies induced by 25-μg mRNA-1273 vaccination.

Participants received two injections of the 25-μg mRNA-1273 vaccine, 28 days apart. PBMC samples were collected on day 1, day 15 ± 2 (2 weeks after first dose), day 43 ± 2 (2 weeks after second dose), and day 209 ± 7 days (6 months after second dose). (A) Longitudinal anti-SARS-CoV-2 spike IgG binding titers, (B) longitudinal anti-SARS-CoV-2 RBD IgG binding titers, and (C) longitudinal SARS-CoV-2 spike pseudovirus neutralizing titers (PSV). (D) Comparison of anti-spike IgG, anti-RBD IgG, and PSV neutralizing titers induced by two doses of 25-μg mRNA-1273 vaccine at day 209 ± 7 (n = 33) and COVID-19 convalescent donors at 170 to 195 days PSO (n = 14). Dotted green lines indicate the limit of quantification (LOQ). The bars in (A), (B), (C), and (D) indicate the geometric mean titers (GMTs) and geometric SD for anti-spike IgG (endpoint titer, ET), anti-RBD IgG (ET), and PSV neutralizing titers, respectively. Data were analyzed for statistical significance using Wilcoxon signed-rank test [(A), (B), and (C)] and Mann-Whitney U test (D). NS, nonsignificant. Background-subtracted and log data analyzed in all cases.

Fig. 2.. mRNA-1273 vaccination induces durable and multifunctional spike-specific CD4⁺ T cell responses.

(A) Longitudinal spike-specific CD4⁺ T cells in mRNA-1273 vaccinees measured by AIM. Spike-specific CD4⁺ T cells quantified by AIM (surface OX40⁺CD137⁺) after stimulation with spike megapool (MP) in mRNA-1273 vaccinees (see fig. S2 for gating strategy). (B) Comparison of spike-specific AIM⁺ CD4⁺ T cell frequencies between 25-μg mRNA-1273 vaccine at day 209 ± 7 (red circles, n = 32) and COVID-19 convalescent donors at 170 to 195 days PSO (yellow circles, n = 14). (C) Quantitation of spike-specific circulating T follicular helper (cT_FH) cells (CXCR5⁺OX40⁺surface CD40L⁺, as percentage of CD4⁺ T cells) after stimulation with spike MP. Representative examples of spike-specific cT_FH cells (red), overlaid on total CD4⁺ T cells, at days 15 ± 2 and 209 ± 7. (D) Spike-specific CD4⁺ T cells expressing intracellular CD40L (iCD40L) and producing IFNγ, TNFα, IL-2, or granzyme B (GzB) in mRNA-1273 vaccinees. (E) Longitudinal spike-specific CD4⁺ cytokine⁺ T cells expressing iCD40L or producing IFNγ, TNFα, IL-2, or GzB in 25-μg mRNA-1273 vaccinees (see fig. S4 for gating strategy). Dotted green lines indicate the limit of quantification (LOQ). White, day 1; light gray, day 15 ± 2; dark gray, day 43 ± 2; red, day 209 ± 7. The bars in (A) to (E) indicate the geometric mean and geometric SD in the analysis of the spike-specific CD4⁺ T cell frequencies. (F) Longitudinal multifunctional spike-specific CD4⁺ T cells in mRNA-1273 vaccinees. Proportions of multifunctional activity profiles of the spike-specific CD4⁺ T cells from mRNA-1273 vaccinees evaluated on days 1, 15 ± 2, 43 ± 2, and 209 ± 7. The blue, green, yellow, orange, and red colors in the pie charts depict the production of one, two, three, four, or five functions, respectively (see figs. S4 and S6 for details). Data were analyzed for statistical significance using Wilcoxon signed-rank test [(A), (C), (D), and (E)] and Mann-Whitney U test (B). Background-subtracted and log data analyzed in all cases.

Fig. 3.. mRNA-1273 vaccination induces multifunctional spike-specific CD8⁺ T cells.

(A) Longitudinal spike-specific CD8⁺ T cells in mRNA-1273 vaccinees measured by AIM (surface CD69⁺CD137⁺). (Left panel) Representative examples of flow cytometry plots of spike-specific CD8⁺ T cells compared with DMSO control (see fig. S2 for gating strategy). (Right panel) Spike-specific CD8⁺ T cells quantified. (B) Comparison of spike-specific AIM⁺ CD8⁺ T cell frequencies between 25-μg mRNA-1273 vaccinees at day 209 ± 7 (n = 32) and COVID-19 convalescent donors at 170 to 195 days PSO (n = 14). (C) Spike-specific CD8⁺ T cells producing IFNγ, TNFα, or IL-2 by intracellular cytokine staining (ICS) in 25-μg mRNA-1273 vaccinees. (D) Longitudinal spike-specific CD8⁺ cytokine⁺ T cells producing IFNγ, TNFα, IL-2, or GzB in 25-μg mRNA-1273 vaccinees (see fig. S4 for gating strategy). Dotted green lines indicate the LOQ. The bars in (A) to (D) indicate the geometric mean and geometric SD. White, day 1; light gray, day 15 ± 2; dark gray, day 43 ± 2; red, day 209 ± 7. (E) Multifunctional activity profiles of spike-specific CD8⁺ T cells from 25-μg mRNA-1273 vaccinees, evaluated for IFNγ, TNFα, IL-2, or GzB (see figs. S4 and S8 for details). The blue, green, yellow, and orange colors in the pie charts depict the production of one, two, three, or four functions, respectively. ND, nondetectable. Data were analyzed for statistical significance using Wilcoxon signed-rank test [(A), (C), and (D)] and Mann-Whitney U test (B). Background-subtracted and log data analyzed in all cases.

Fig. 4.. Spike-specific antibody and T cell responses induced by mRNA-1273 vaccination.

(A to H) Immune responses to 25-μg mRNA-1273 vaccination in three adult age groups: 18–55 (light blue symbols), 56–70 (dark blue), and over 70 years of age (black). [(A) to (C)] Anti-spike IgG, anti-RBD IgG, and PSV neutralizing titers. [(D) and (E)] Spike-specific CD4⁺ T cells by AIM (D) or ICS (E). (F) Spike-specific cT_FH cells. [(G) and (H)] Spike-specific CD8⁺ T cells by AIM (G) or ICS (H). Data were analyzed for statistical significance using nonparametric analysis of variance Kruskal-Wallis (KW) test and Dunn’s post-test for multiple comparisons. The P values plotted on the bottom show the KW test results, and the P values plotted on the top show the post-test analysis comparing age groups. (I) Anti-SARS-CoV-2 spike IgG, RBD IgG, and PSV neutralizing titers in 25-μg and 100-μg mRNA-1273 vaccinees at day 43 (2 weeks after second dose). (J) Spike-specific CD4⁺ AIM⁺ (left) and cytokine⁺ (right) T cells in mRNA-1273 vaccinees. (K) Spike-specific CD8⁺ AIM⁺ (left) and cytokine⁺ (right) T cells in mRNA-1273 vaccinees. Dotted green lines indicate the limit of quantification (LOQ). The bars indicate geometric mean and geomean SD. Data in (I) to (K) were analyzed for statistical significance using Mann-Whitney U test. Background-subtracted and log data analyzed in all cases.

Fig. 5.. Preexisting anti-spike immunity modulates T cell and antibody responses.

(A) Preexisting spike-specific CD4⁺ AIM⁺ T cells at day 1 (see Fig. 2 for details). (B) Memory phenotype of preexisting spike-specific CD4⁺ AIM⁺ T cells from (A). T_CM, central memory cells; T_EM, effector memory cells; T_EMRA, terminally differentiated effector memory cells. (C) Spike-specific CD4⁺ AIM⁺ T cells in mRNA-1273 vaccinees with (“preexisting,” blue) and without (“no preexisting,” white) preexisting cross-reactive spike-reactive memory CD4⁺ T cells evaluated on days 1, 15 ± 2, 43 ± 2, and 209 ± 7 after immunization (see fig. S13 for details). (D) Spike-specific CD4⁺ cytokine⁺ T cells in mRNA-1273 vaccinees with and without preexisting cross-reactive memory CD4⁺ T cells. (E) Proportions of multifunctional spike-specific CD4⁺ T cells in mRNA-1273 vaccinees with (“pre-existing”) and without (“no preexisting”) preexisting cross-reactive spike-reactive memory CD4⁺ T cells evaluated on days 1, 15 ± 2, 43 ± 2, and 209 ± 7 after immunization (see fig. S10 for details). (F) Spike-specific cT_FH cells (as percentage of CD4⁺ T cells), (G) anti-spike IgG, (H) anti-RBD IgG, and (I) SARS-CoV-2 PSV neutralizing titers in mRNA-1273 vaccinees without and with preexisting cross-reactive spike-reactive memory CD4⁺ T cells. Dotted green lines indicate the LOQ. The bars in (A) and (C) to (I) indicate the geometric mean and geomean SD in the analysis of the antibody levels or spike-specific CD4⁺ and CD8⁺ T cells in mRNA-1273 vaccinees with (“preexisting”) and without (“no preexisting”) pre-existing cross-reactive spike-reactive memory CD4⁺ T cells evaluated on days 1, 15 ± 2, 43 ± 2, and 209 ± 7 after immunization. The bars in (B) indicate the mean and SD in the analysis of the memory phenotype of spike-specific CD4⁺ T cells. Data were analyzed for statistical significance using Mann-Whitney U test. Background-subtracted and log data analyzed in all cases.