Divergent SARS-CoV-2 Omicron-reactive T and B cell responses in COVID-19 vaccine recipients

Abstract

The severe acute respiratory distress syndrome coronavirus 2 (SARS-CoV-2) Omicron variant is spreading rapidly, even in vaccinated individuals, raising concerns about immune escape. Here, we studied neutralizing antibodies and T cell responses targeting SARS-CoV-2 D614G [wild type (WT)] and the Beta, Delta, and Omicron variants of concern in a cohort of 60 health care workers after immunization with ChAdOx-1 S, Ad26.COV2.S, mRNA-1273, or BNT162b2. High binding antibody levels against WT SARS-CoV-2 spike (S) were detected 28 days after vaccination with both mRNA vaccines (mRNA-1273 or BNT162b2), which substantially decreased after 6 months. In contrast, antibody levels were lower after Ad26.COV2.S vaccination but did not wane. Neutralization assays showed consistent cross-neutralization of the Beta and Delta variants, but neutralization of Omicron was significantly lower or absent. BNT162b2 booster vaccination after either two mRNA-1273 immunizations or Ad26.COV2 priming partially restored neutralization of the Omicron variant, but responses were still up to 17-fold decreased compared with WT. SARS-CoV-2-specific T cells were detected up to 6 months after all vaccination regimens, with more consistent detection of specific CD4⁺ than CD8⁺ T cells. No significant differences were detected between WT- and variant-specific CD4⁺ or CD8⁺ T cell responses, including Omicron, indicating minimal escape at the T cell level. This study shows that vaccinated individuals retain T cell immunity to the SARS-CoV-2 Omicron variant, potentially balancing the lack of neutralizing antibodies in preventing or limiting severe COVID-19. Booster vaccinations are needed to further restore Omicron cross-neutralization by antibodies.

Fig. 1.

Study design and detection of virus-specific binding antibodies and T-cells. (A) Diagram of the number of included participants and study groups. A total of N=423 participants were included for the analysis of binding antibodies and T-cell responses, responses were measured early and late after completion of the vaccination regimen. Binding antibodies were also assessed in N=23 convalescent participants. (B) Levels of binding S-specific antibodies early (28 days after second vaccination, or 56 days after Ad26.COV2.S vaccination) and late (6 months after completion of vaccination regimen) after vaccination or infection. LLoD is 4.81 BAU/ml, responder (resp) cut-off is 33.8 BAU/ml (dotted line). Geometric mean titers are indicated above the graph. (C) IFN-ɣ levels in plasma after stimulation of whole-blood with peptide pools spanning the S protein (Ag2, QIAGEN) early and late after ChAdOX-1 S, Ad26.COV2.S, or mRNA-1273 vaccination, ChAdOx-1 S responses were exclusively measured at 6 months). LLoD is 0.01 IU/ml, responder cut-off is 0.15 IU/ml. Geometric means are indicated above the graph. Comparisons of timepoints within study groups were performed by paired t test. White symbols represent samples selected for in-depth analyses. LLoD = lower limit of detection, S = Spike, BAU = binding arbitrary units, GMT = geometric mean titer, d = days, m = months, IFN-ɣ = interferon gamma.

Fig. 2.. SARS-CoV-2 variant-specific neutralizing antibody responses.

(A) Schematic overview of the infectious virus PRNT50 assay. Serum samples were pre-incubated with 400 plaque forming units infectious virus for 1 hour, transferred to Calu-3 cells, and N-positive plaques were counted after 8 hours. A PRNT50 (50% reduction of plaques) was calculated by a proportionate distance formula. (B-D) Levels of neutralizing antibodies in convalescent donors after 6 months (B), early after completion of the vaccination regimen (C), and late after completion of the vaccination regimen (D). N=15 participants were included per vaccination regimen. The lowest serum dilution tested was 1:20, undetectable PRNT50 values (<20) were set at a PRNT50 of 10. Geometric mean titers are indicated above the graphs, fold change reductions are indicated and calculated by dividing the D614G GMT by the Omicron GMT. Comparisons of VOC-specific responses within study groups were performed by Friedman test with multiple comparisons. WT = wildtype, d = days, m = months, PRNT50 = plaque reduction neutralization titer – 50%,

Fig. 3.. SARS-CoV-2 variant-specific T-cell responses.

(A) Schematic overview of AIM assay. PBMC were stimulated with different overlapping peptide pools for 20 hours, followed by measurement of up-regulation of activation markers by flow cytometry and cytokines in cell culture supernatant. (B) SARS-CoV-2-specific T-cells were detected by flow cytometry. (C-D) Up-regulation of AIM early and late after completion of the vaccination regimen on CD4+ (C) and CD8+ (D) T-cells. Percentages indicate the percentage of AIM+ T-cells after subtraction of observed background in a DMSO stimulation, bars indicate the means. N=49 participants were analyzed over various timepoints. Comparisons of VOC-specific responses within study groups were performed by Friedman test with multiple comparisons. LLoD = lower limit of detection, PBMC = peripheral blood mononuclear cells, AIM = activation-induced markers, d = days.

Fig. 4.. Variant-specific immune responses early after booster vaccination.

(A) Diagram of the number of included participants in the booster analysis. A total of N=24 participants were included for the analysis of binding antibodies, variant-specific neutralizing antibodies, and variant-specific T-cell responses. N=15 participants were primed with one shot Ad26.COV2.S, N=9 participants were primed with two shots mRNA-1273, all participants were analyzed throughout the panels. Responses were measured early after BNT162b2 booster vaccination. (B) Levels of binding S-specific antibodies early after booster vaccination. LLoD is 4.81 BAU/ml, responder (resp) cut-off is 33.8 BAU/ml (dotted line). Geometric mean titers are indicated above the graph. Comparisons of timepoints within study groups were performed by paired t test. (C) Levels of neutralizing antibodies in boosted donors. The lowest serum dilution tested was 1:20, undetectable PRNT50 values (<20) were set at a PRNT50 of 10. Geometric mean titers are indicated above the graphs, fold change reductions are indicated and calculated by dividing the D614G GMT by the Omicron GMT. (D) Up-regulation of AIM on CD4+ T-cells and CD8+ T-cells in boosted donors. Percentages indicate the percentage of AIM+ cells after subtraction of observed background in a DMSO stimulation, bars indicate the means. Comparisons of VOC-specific responses within study groups were performed by Friedman test with multiple comparisons. BAU = binding arbitrary units, GMT = geometric mean titer, PRNT50 = plaque reduction neutralization titer – 50%, LLoD = lower limit of detection, AIM = activation-induced markers, d = days.