Cellular immune breadth of an Omicron-specific, self-amplifying monovalent mRNA vaccine booster for COVID-19

Abstract

Selecting a booster vaccine strategy that generates cellular immune breadth is crucial for effectively recalling cellular reservoirs upon infection with severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) variants. This post hoc analysis from a multicentre, randomized phase 3 study (CTRI/2022/10/046475) compared the cellular immune breadth induced by self-replicating mRNA (samRNA) vaccine GEMCOVAC-OM, encoding Omicron B.1.1.529 Spike protein, with the adenovector vaccine ChAdOx1 nCoV-19, encoding Wuhan variant Spike protein, when administered as a booster. GEMCOVAC-OM elicited significant expansion of memory B-cells (MBCs) specific to Omicron B.1.1.529, compared to ChAdOx1 nCoV-19. GEMCOVAC-OM also induced more B-cells reactive to Omicron XBB.1.5 and BA.2.86 Spike proteins. Additionally, GEMCOVAC-OM triggered higher frequencies of Omicron-Spike-specific T-cells, including stem cell, central, and effector memory subsets. In summary, while ChAdOx1 nCoV-19 showed some cross-reactivity, GEMCOVAC-OM induced a more targeted immune response. GEMCOVAC-OM offers a broader, longer-lasting immunity, making it a promising candidate for future vaccine development and global distribution.

Fig. 1. CONSORT diagram for cellular response analysis.

Cellular immune responses were assessed in 25% of the immunogenicity cohort.

Fig. 2. Omicron B.1.1.529 Spike protein-specific memory B-cell responses following ChAdOx1 nCoV-19 (n = 34) and GEMCOVAC-OM (n = 67) booster doses.

a Experimental design. Blood samples were collected before vaccine booster (day-1), day-29, and day-90. Representative dot plots showing gates for B.1.1.529 Spike⁺ non-naive B-cells, PBs, MBCs, and rMBCs. b B.1.1.529 Spike-specific non-naïve B-cells gated as the frequency of IgD^loCD27⁺B.1.1.529 Spike⁺ cells in CD3^-CD19⁺CD20⁺ B-cells. c B.1.1.529 Spike-specific PBs gated as the frequency of IgD^loCD27⁺B.1.1.529 Spike⁺ CD38⁺ cells in CD3^-CD19⁺CD20⁺ B-cells. d B.1.1.529 Spike-specific MBCs gated as the frequency of IgD^loCD27⁺B.1.1.529 Spike⁺CD38^+/lo cells in CD3^-CD19⁺CD20⁺ B-cells. e B.1.1.529 Spike-specific rMBCs gated as frequency of IgD^loCD27⁺B.1.1.529 Spike⁺CD38^+/loCD21⁺ cells in CD3^-CD19⁺CD20⁺ B-cells. The data were presented as median with interquartile range. Statistical comparisons within the cohorts and between the cohorts were performed by using Friedman test and Mann–Whitney test, respectively. p values are shown to represent statistical comparison within or between GEMCOVAC-OM and ChAdOx1 nCoV-19 booster cohorts.

Fig. 3. Cross-reactive B-cell responses against XBB.1.5 and BA.2.86 Spike protein following ChAdOx1 nCoV-19 (n = 34) and GEMCOVAC-OM (n = 67) booster doses.

a Experimental design. Cross-reactive B-cell responses against XBB.1.5 and BA.2.86 Spike protein were assessed in PBMCs from day-90. b Total XBB.1.5 Spike reactive B-cells gated as frequency in CD3^-CD19⁺CD20⁺ B-cells. c Total BA.2.86 Spike reactive B-cells gated as frequency in CD3^-CD19⁺CD20⁺ B-cells. d Pseudovirus neutralization titers at day-90 time points harboring the Spike protein of XBB.1.5. The data were presented as median with interquartile range. Statistical comparison was performed by using the Mann–Whitney test. p values are shown to represent statistical comparison within or between GEMCOVAC-OM and ChAdOx1 nCoV-19 booster cohorts. e, f Spearman correlation between B.1.1.529 Spike-specific B-cell subsets and B.1.1.529, XBB.1.5, and BA.2.86 Spike+ B-cells in ChAdOx1 nCoV-19 (e) and GEMCOVAC-OM (f) cohorts. In corrplot, the blue boxes represent positive correlations and the red boxes represents negative correlation. Significant correlations were represented as an asterisk in the boxes. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Fig. 4. MBC breadth against XBB.1.5 and BA.2.86 Spike protein following ChAdOx1 nCoV-19 (n = 34) and GEMCOVAC-OM (n = 67) booster doses.

a Experimental design. Cross-reactive MBC responses against XBB.1.5 Spike protein were assessed in PBMCs from day-1 (baseline) and day-90. Cross-reactive MBC responses against BA.2.86 Spike protein were assessed in PBMCs from the day-90 study time point. Representative dot plots showing gates for XBB.1.5 and BA.2.86 Spike⁺ non-naive B-cells, PBs, MBCs, and rMBCs. b XBB.1.5 Spike reactive non-naïve B-cells c XBB.1.5 Spike reactive PBs. d XBB.1.5 Spike reactive MBCs. e XBB.1.5 Spike reactive rMBCs. f BA.2.86 Spike reactive non-naïve B-cells. g BA.2.86 Spike reactive PBs. h BA.2.86 Spike reactive MBCs i, BA.2.86 Spike reactive rMBCs. The data were presented as median with interquartile range. Statistical comparison was performed by using the Mann–Whitney test to compare XBB.1.5 reactive B-cell subsets. and BA.2.86 reactive B-cell subsets. p values are shown to represent statistical comparison within or between GEMCOVAC-OM and ChAdOx1 nCoV-19 booster cohorts. j, k Spearman correlation between B.1.1.529 Spike-specific B-cell subsets and XBB.1.5 and BA.2.86 Spike reactive B-cell subsets in ChAdOx1 nCoV-19 (j) and GEMCOVAC-OM (k) cohorts. In corrplot, the blue boxes represent positive correlations and the red boxes represents negative correlation. Significant correlations were represented as an asterisk in the boxes. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. l, m tSNE maps display the distribution of Spike⁺ B cells (red dots) in B.1.1.529 (blue), XBB^.1.5 (yellow), and B.2.86 (orange) B-cell experimental clusters having CD27, CD38, and CD21 memory markers spaces in the ChAdOx1 nCoV-19 (l) and GEMCOVAC-OM (m) cohorts.

Fig. 5. Omicron Spike protein-specific memory T-cell responses following ChAdOx1 nCoV-19 (n = 35) and GEMCOVAC-OM (n = 70) booster doses.

a Experimental design. Omicron B.1.1.529 specific memory T-cell responses were assessed using AIMs assay in the PBMCs from day-90. b Representative dot plots showing gates for CD4⁺ T-cells, CD8⁺ T-cells, CD137⁺Ox40⁺, or AIMs⁺ CD4 T-cells. c Total AIMs⁺ CD4⁺ T-cells. d Representative dot plots showing gates for CD137⁺CD69⁺ or AIMs⁺ CD8⁺ T-cells. e Total AIMs⁺ CD8 T-cell. f Representative dot plots showing gates for central memory CD4⁺ T-cells (CD4⁺ T_cm), effector memory CD4⁺ T-cells (CD4⁺ T_em), terminal effective RA⁺ (CD4⁺ T_emRA) and Naive CD4⁺ T-cells (CD4⁺ T_naive). g–j Total CD4⁺ T_cm (g), CD4⁺ T_em (h), CD4⁺ T_emRA (i), and CD4⁺ T_naive (j), cell frequency as % of grandparent CD4⁺ T-cell gate. k Representative dot plots showing gates for central memory CD8⁺ T-cells (CD8⁺ T_cm), effector memory CD8⁺ T-cells (CD8⁺ T_em), terminal effective RA⁺ (CD8⁺ T_emRA) and Naive CD8⁺ T-cells (CD8⁺ T_naive). l–o Total CD8⁺ T_cm (l), CD8⁺ T_em (m), CD8⁺ T_emRA (n), and CD8⁺ T_naive (o), cell frequency as % of grandparent CD8⁺ T-cell gate. p Representative dot plots showing gates for stem T-cells (CD4⁺ T_scm). q Total CD4⁺ T_scm cell frequency as % of CD4⁺ T-cell gate. r Representative dot plots showing gates for CD8⁺ stem T-cells (CD8⁺ T_scm). s Total CD8⁺ T_scm cell frequency as % of CD8⁺ T-cell gate. t, u tSNE maps display distribution of Omicron B.1.1.529 Spike-specific memory T-cell sub-populations in CD4⁺ T-cell and CD8⁺ T-cell clusters in the ChAdOx1 nCoV-19 (t) and GEMCOVAC-OM (u) cohorts. v Representative dot plots showing gates for follicular helper CD4⁺ T-cells (CD4⁺ T_FH). w Total CD4⁺ T_FH cell frequency as % of CD4⁺ T-cell gate. The data were presented as median with interquartile range. Statistical comparison was performed by using the Mann–Whitney test. p values are shown to represent statistical comparison within or between GEMCOVAC-OM and ChAdOx1 nCoV-19 booster cohorts.