Several studies have demonstrated that the SARS-CoV-2 variant-of-concern B.1.1.529 (Omicron) exhibits a high degree of escape from Ab neutralization. Therefore, it is critical to determine how well the second line of adaptive immunity, T cell memory, performs against Omicron. To this purpose, we analyzed a human cohort (n = 327 subjects) of two- or three-dose mRNA vaccine recipients and COVID-19 postinfection subjects. We report that T cell responses against Omicron were largely preserved. IFN-γ–producing T cell responses remained equivalent to the response against the ancestral strain (WA1/2020), with some (∼20%) loss in IL-2 single or IL-2+IFN-γ+ polyfunctional responses. Three-dose vaccinated participants had similar responses to Omicron relative to post–COVID-19 participants and exhibited responses significantly higher than those receiving two mRNA vaccine doses. These results provide further evidence that a three-dose vaccine regimen benefits the induction of optimal functional T cell immune memory.
The B.1.1.529 (Omicron) SARS-CoV-2 variant was first reported to the World Health Organization by the South African government in 2021 (1) and was subsequently designated a variant of concern. Omicron has accumulated a large number of mutations compared with the original Wuhan strain; that is, a total of 60, of which 50 were nonsynonymous (2). Thirty-two of these mutations reside in the spike protein, which is the sole target of most SARS-CoV-2 vaccines. This raised concerns about the potential for increased transmissibility and escape from vaccine protection. Considerable evidence points to increased transmissibility of Omicron as compared with previous variants of concern, including the highly transmissible B.1.617.2 (Delta) variant. A cohort analysis of household transmission in England showed that 2-fold more Omicron index cases gave rise to a secondary household case in comparison with Delta (3). Omicron is now the dominant SARS-CoV-2 variant in many parts of the world (4, 5). Early studies demonstrated a substantial reduction in Ab neutralizing capacity against Omicron (Ref. 6 and A. Wilhelm, M. Widera, K. Grikscheit, T. Toptan, B. Schenk, and C. Pallas, manuscript posted on medRxiv, DOI: 10.1101/2021.12.07.21267432) which was maintained best in sera from individuals who experienced infection and were subsequently vaccinated (A. Rössler, A., L. Riepler, D. Bante, D. von Laer, and J. Kimpel, manuscript posted on medRxiv, DOI: 10.1101/2021.12.08.21267491). However, the spread of Omicron in South Africa has been followed by a decrease in hospitalizations (7) and death rates (8). Given the large degree of escape from Ab responses, T cell–mediated immunity could be essential to prevent Omicron-induced severe COVID-19. Depletion of T cells in convalescent macaques resulted in impaired immunity against rechallenge with SARS-CoV-2, suggesting a significant role for T cells in the context of subprotective Ab titers (9). Human clinical studies also demonstrated a link between SARS-CoV-2 T cell responses and reduced disease severity (10, 11). Thus, it is paramount to determine how well T cell immunity against Omicron is preserved. We and others have shown that T cell immunity against Delta, Gamma, and other variants was largely preserved in mRNA vaccine recipients (Ref. 12 and Jergović, M., J. L. Uhrlaub, M. Watanabe, C. M. Bradshaw, L. M. White, B. J. LaFleur, T. Edwards, R. Sprissler, M. Worobey, D. Bhattacharya, and J. Nikolich-Žugich, manuscript posted on bioRxiv, 10.1101/2021.07.22.453287). However, earlier variants contained drastically lower numbers of mutations in the spike protein. To assess T cell immune memory to the Omicron variant, we have analyzed a large cohort of participants recruited into the Centers for Disease Control and Prevention–sponsored AZ HEROES research study (13). Participants in this cohort included subjects sampled after SARS-CoV-2 infection; fully vaccinated subjects (two doses of mRNA vaccine); subjects postinfection receiving two doses of mRNA vaccine; participants receiving three doses of mRNA vaccine; and pre-pandemic controls. We have measured polyfunctional T cell responses in PBMCs to overlapping peptide pools corresponding to SARS-CoV-2 spike proteins of USA-WA1/2020 (ancestral strain) and the Omicron variant. We report that T cell responses to spike protein of the Omicron variant are largely preserved. We found no waning of T cell IFN-γ responses upon stimulation with Omicron peptide pools compared with the original strain, with a slight reduction in IL-2–producing and polyfunctional IFN-γ+IL-2+ double-positive cells and IFN-γ+IL-2+granzyme B (GrB)+ triple-positive cells. Similar results were obtained in a recently published study by Tarke et al. (14). Additionally, we report that T cell responses in participants receiving two doses of mRNA vaccines were lower than those postinfection, postinfection plus two vaccine doses, and in those receiving three vaccine doses, suggesting that three doses of mRNA should be considered as an optimal, full vaccination regimen to induce robust T cell immunity.
Materials and Methods
This study was approved by the University of Arizona Institutional Review Board (protocols 2102460536 and 1410545697) and the Oregon Health and Science University Institutional Review Board (protocol 00003007, pre-pandemic controls) and was conducted in accordance with all federal, state, and local regulations. Pre-pandemic samples included healthy community-dwelling individuals recruited in Tucson, AZ and Portland, OR prior to December 2019. All COVID-19 –positive participants tested positive by PCR before April 1, 2021 prior to emergence of the Delta variant in Arizona (15). All subjects >55 y of age were considered older adults and subjects <50 y of age were considered adult. We did not include subjects aged 50–55 y of age. Blood was drawn into BD Vacutainer blood collection tubes with sodium heparin. PBMCs were separated by Ficoll gradient separation and cryopreserved in FCS + 10% DMSO.
Cryopreserved PBMCs (5 × 106/sample) were thawed in prewarmed RPMI 1640 media supplemented with l-glutamine (Lonza) + 10% FCS. Thawed PBMCs were rested for 3–4 h at 37°C in X-VIVO 15 medium (Lonza) supplemented with 5% human AB serum. Cells were then stimulated with ∼1 nmol of peptide pool corresponding to spike of US-WA (wild-type) or Omicron (B.1.1.529) variant (16-mer peptide pools, overlapping by 10 aa (21st Century Biochemicals). Cell suspensions were transferred to precoated human IFN-γ, TNF-α, IL-2, GrB FLUOROSpot kits (Mabtech) and developed after 42 h according to the manufacturer’s instructions. Spots were imaged and counted using an Iris FLUOROspot reader (Mabtech).
PBMCs (1 × 106) were stimulated with ∼1 nmol of peptide pool corresponding to the spike protein of US-WA (wild-type) or Omicron (B.1.1.529) variant (21st Century Biochemicals) for 24 h in X-VIVO 15 medium (Lonza) supplemented with 5% human AB serum at 37°C. Unstimulated wells from each participant were used for subtraction of background. Cells were stained with surface Abs in PBS (Lonza) + 2% FCS for 1 h, stained with the Live/Dead fixable blue dye for 30 min (Thermo Fisher Scientific), and then fixed and permeabilized using a FOXP3 Fix/Perm kit (eBioscience). Samples were acquired using a Cytek Aurora cytometer and analyzed by FlowJo software (Tree Star).
GraphPad Prism v9 was used for statistical analysis. Upon inspection of data distribution by a Shapiro–Wilk normality test, differences between paired samples treated with different peptide pools were calculated by a two-tailed Wilcoxon rank test. Differences between the subject groups were calculated by and Kruskal–Wallis test with Dunn’s post hoc correction. For all figures, data are presented as mean ± SEM. A p value <0.05 was considered statistically significant (*p < 0.05, **p < 0.01, ***p < 0.001. ****p < 0.0001).
The data and materials that support the findings of this study are available from the corresponding author upon request.
Results and Discussion
A total of 359 participants were included in this study: 1) 210 received two doses of SARS-CoV-2 mRNA vaccines (2× VAX); 2) 54 were SARS-CoV-2 postinfection samples with no additional vaccination (PI); 3) 39 were vaccinated following SARS-CoV-2 infection (PI + VAX); 4) 24 received three doses of mRNA vaccines (3× VAX); and 5) 32 participants were recruited before September 2019 (demographics are shown in Supplemental Table I). We analyzed T cell responses of the participants to the spike protein of USA-WA1/2020 and Omicron variant by stimulating PBMCs with overlapping peptide pools spanning the spike protein and measuring production of T cell effector cytokines IFN-γ, IL-2, and GrB by FLUOROSpot. For each sample, the number of spot-forming units (SFU) of each cytokine was calculated by subtracting the SFU number measured in the unstimulated wells from the SFU number of stimulated wells (representative FLUOROSpot images in (Fig. 1A). The spike peptide pool induced a 5-fold increase in IFN-γ SFU in immunized (vaccinated or post–COVID-19) participants whereas no effect was evident in the pre-pandemic samples (Supplemental Fig. 1). When comparing samples from all immunized participants, the number of IFN-γ SFU between Omicron- and WA1/2020 peptide-stimulated wells was equal (Fig. 1B, left panel). However, the number of IL-2 SFU was reduced (Fig. 1B, middle panel) by ∼16.8%. There was no difference in GrB SFU between the peptide pools (Fig. 1B, right panel). Next, we analyzed the number of polyfunctional (double or triple cytokine-positive) T cells. We detected a decrease in IFN-γ IL-2 double-positive cells, which were decreased 23.3% when stimulated with Omicron peptides, but not IFN-γ GrB double-positive cells (Fig. 1C), whereas triple-positive (IFN-γ, IL-2, GrB) cells were 38.7% decreased with Omicron peptide pool stimulation (Fig. 1D); however, this was numerically the smallest population. Based on this large sample (n = 327 exposed/immunized participants) we can conclude that the decrease in T cell immunity to mutated Omicron peptide is measurable, but the response is ∼80% preserved. We next investigated the differences in T cell responses between the groups of study subjects. When stimulated with the WA1/2020 peptide pool, recipients of two doses of mRNA vaccine had significantly lower numbers of IFN-γ SFU (Fig. 2A, left panel) than all other groups. A similar trend was measured with the Omicron peptide pool, albeit without reaching statistical significance (Fig. 2A, right panel). Similarly, the number of IL-2 SFU was decreased in the 2× VAX group stimulated with WA1/2020 peptides (Fig. 2B). Double-positive (IFN-γ+IL-2+) and triple-positive (IFN-γ+IL-2+GrB+) SFU were also lower in the 2× VAX group (Fig. 2C). Interestingly, the magnitude of polyfunctional triple-positive cells to Omicron was the highest in the 3× VAX group (Fig. 2D), although the third dose of the vaccine did not increase the number of single- or double-positive cells reacting to Omicron pools (Fig. 2A–C). Jointly, these results imply that two doses of mRNA vaccine induced submaximal T cell immunity, and that at least three doses are required to achieve the same level of Ag-specific T cells found in postinfection subjects. We also compared T cell responses to each variant between subjects vaccinated with vaccines from different manufacturers (Pfizer and Moderna) and found that although there was no difference in IFN-γ T cell responses (Supplemental Fig. 2, left panel) within vaccine manufacturers, there was a higher number of IL-2+ SFU in samples from Moderna vaccinated subjects in response to either WA1/2020 or Omicron as compared with their Pfizer vaccinated counterparts (Supplemental Fig. 2, middle panel). The subjects grouped by vaccine manufacturer did not differ in the number of GrB SFU (Supplemental Fig. 2, right panel). Next, we investigated whether T cell responses were dependent on the age of the subject in our largest group (2× VAX). Given that IL-2 was most impacted by the mutations in Omicron peptides as compared with WA1/2020 (Fig. 1), we first analyzed the impact of age by IL-2+ SFU. We have detected a minimal, but measurable, negative correlation of IL-2 responses with age (r = −0.16, p = 0.02). There was no impact of age for double-positive (IFN-γ+IL-2+) SFU (r = −0.11, p = 0.09), whereas triple-positive (IFN-γ+IL-2+GrB+) SFU were negatively correlated to age of the subject (r = −0.24, p < 0.0001). Overall, the change in T cell responses with the age of the participant was minimal. Given that the 3× vaccinated group was on average 8 y older than the 2× VAX group (58 versus 50 y average age, Supplemental Table I) but had higher T cell responses than the 2× VAX group (Fig. 2), we conclude that age is not a major determinant in T cell responses to mRNA vaccines following administration of multiple vaccine doses.
To further investigate the phenotype of T cells responding to either WA1/2020 or Omicron, we performed flow cytometry on approximately half of the 2× VAX subject samples (randomly selected, n = 96). Expression of costimulatory molecules CD137 (4-IBB) and OX-40 is one of the common methods for enumerating Ag-specific T cells by flow cytometry, alongside intracellular staining for cytokines (16). Again, as with the FLUOROSpot assay, the spike peptide pool induced a robust increase in CD137- and OX-40–activated T cells compared with stimulated wells (Supplemental Fig. 3, representative flow cytometry in (Fig. 3A). As previously reported (17), we have detected more CD4 helper T cells than cytotoxic CD8 T cells responding (Fig. 3B). We did detect a slight decrease in the number of Ag-specific CD4 and CD8 T cells in response to Omicron peptides versus WA1/2020. However, the difference was not statistically significant (Fig. 3B). Given that CD4+ helper T cells represent the bulk of the response, we have additionally investigated the expression of Th polarizing cytokines IL-17a and IL-4 in the 2× VAX group. We found that both IL-17a (Fig. 3C) and IL-4 (Fig. 3D) were reduced in response to Omicron peptides, which could be interpreted as a positive finding, given that Th17 and Th2 responses would be expected to be counterproductive against viral infection.
Overall, our findings highlight resilience of T cell responses generated in response to a pre-Omicron infection and/or to Wuhan-derived spike protein-based vaccines in the face of the B1.1.529 Omicron variant, particularly as measured by the IFN-γ production, which was the largest component of the antiviral T cell response. We did measure decreased IL-2 and polyfunctional responses; however, these were generally reduced by <20–30%. Stratified analysis of subjects immunized by vaccines or prior infection clearly revealed that 2× VAX participants exhibited inferior responses to both Omicron and the ancestral strain relative to other groups, and in particular 3× VAX participants. Therefore, our results stress the need for a three-dose vaccine regimen to achieve robust T cell immunity.
We thank the personnel of the University of Arizona Health Sciences Biobank for expert processing and storage of human blood samples. We thank our Centers for Disease Control and Prevention colleagues Drs. Natalie J. Thornburg, Mark G. Thompson, and Julie Mayo Lamberte for logistical help, input into the study design, and comments on the manuscript.
This work was supported in part by U.S. Public Health Service Grant R37 AG020719 and the Bowman Professorship in Medical Sciences (to J.N.-Ž.), as well as by Centers for Disease Control and Prevention HEROES Project Award 75D30120C08379 (to J.L.B.).
The online version of this article contains supplemental material.
J.N.-Ž. is the cochair of the scientific advisory board of and receives research funding from Young Blood Institute, Inc. The other authors have no financial conflicts of interest.