Enhanced immune responses following heterologous vaccination with self-amplifying RNA and mRNA COVID-19 vaccines

Abstract

The optimal vaccination strategy to boost responses in the context of pre-existing immune memory to the SARS-CoV-2 spike (S) glycoprotein is an important question for global public health. To address this, we explored the SARS-CoV-2-specific humoral and cellular immune responses to a novel self-amplifying RNA (saRNA) vaccine followed by a UK authorised mRNA vaccine (BNT162b2) in individuals with and without previous COVID-19, and compared these responses with those who received an authorised vaccine alone. 35 subjects receiving saRNA (saRNA group) as part of the COVAC1 clinical trial and an additional 40 participants receiving an authorised SARS-CoV-2 vaccine only (non-saRNA group) were recruited. Antibody responses were measured by ELISA and a pseudoneutralisation assay for wildtype, Delta and Omicron variants. Cellular responses were measured by IFN-ƴ ELISpot and an activation induced marker (AIM) assay. Approximately 50% in each group had previous COVID-19 prior to vaccination, confirmed by PCR or antibody positivity on ELISA. All of those who received saRNA subsequently received a full course of an authorised vaccine. The majority (83%) of those receiving saRNA who were COVID-19 naïve at baseline seroconverted following the second dose, and those with previous COVID-19 had an increase in antibody titres two weeks following saRNA vaccination (median 27-fold), however titres were lower when compared to mRNA vaccination. Two weeks following the 2nd authorised mRNA vaccine dose, binding and neutralising antibody titres were significantly higher in the saRNA participants with previous COVID-19, compared to non-saRNA, or COVID-19 naive saRNA participants. Cellular responses were again highest in this group, with a higher proportion of spike specific CD8+ than CD4+ T cells when compared to those receiving the mRNA vaccine only. These findings suggest an immunological benefit of increased antigen exposure, both from natural infection and vaccination, particularly evident in those receiving heterologous vaccination with saRNA and mRNA.

Fig 1. Simplified study schematic.

Fig 2. Anti-spike versus anti-nucleocapsid IgG.

2a. Baseline (pre-vaccination) anti-S and anti-N IgG in COVID-19 experienced (blue, n = 37) and COVID naïve (red, n = 38) participants as measured by ELISA. 2b. Anti-S and anti-N IgG two weeks following the 2^nd UK authorised vaccine dose in COVID-19 experienced (n = 36) and naïve (n = 33). LOD; level of detection.

Fig 3. Binding and neutralising antibody responses.

3a. Antibodies against SARS-CoV-2 Wuhan-hu-1 spike protein as measured by ELISA at baseline and two weeks following dose 1 and 2 of saRNA and mRNA vaccines in saRNA (COVAC1) participants (red and blue) and at baseline and two weeks following dose 1 and 2 of mRNA vaccines in non-saRNA participants (orange and green). 3b. Neutralising antibodies against SARS-CoV-2 Wuhan-hu-1 measured using pseudovirus at baseline and two weeks following the 2^nd dose of saRNA and mRNA vaccines in saRNA participants (red and blue) and two weeks following the 2^nd dose of mRNA vaccine in non-saRNA participants (orange and green). 3c. Neutralising antibodies against SARS-CoV-2 Wuhan-hu-1, Delta and Omicron BA.1 and BA.4/5 using pseudovirus two weeks following the second mRNA vaccine dose and the differences between groups. 3d. Fold decrease in neutralisation against Delta and Omicron BA.1 and BA.4/5 compared to Wuhan-hu-1 (y axis inverted) 3e. Correlation between binding antibodies against SARS-CoV-2 Wuhan-hu-1 spike protein (ELISA) and neutralisation (pseudovirus) against Wuhan-hu-1, Delta and Omicron BA.1 and BA.4/5 two weeks following the second mRNA vaccine dose. Number of samples included in the analysis indicated on graphs. Geometric mean titres (GMT) and standard deviation (sd) are shown. Differences between groups determined by Kruskall-Wallis and tested using Mann-Whitney. Median fold decrease within groups by Wilcoxon matched pairs signed rank test. Correlations by Spearman’s rank correlation coefficient. -, no exposure; +, single exposure; ++, two exposures; LOD, level of detection; WT, wildtype; Δ, Delta. Significant values displayed: **** p<0.0001; *** p<0.001; ** p<0.01; * p<0.05; ns, non-significant.

Fig 4. Cellular responses.

4a. IFN-γ spot forming units (SFU) per million cells (ELISpot) from PBMC stimulated with SARS-CoV-2 spike peptide pools. 4b. Representative flow cytometry plots of SARS-CoV-2 specific CD4+ cells (CD40L+OX40+) in a vaccinated subject. 4c. Percentage of SARS-CoV-2 spike specific AIM+ non-naïve CD4+ cells (CD40L+OX40+) at baseline and two weeks following the second licensed vaccine dose in non-saRNA participants (orange and green) and two weeks following the second saRNA and mRNA vaccine doses in saRNA (COVAC1) participants (red and blue) 4d. Representative flow cytometry plots of SARS-CoV-2 specific CD8+ cells (CD69+CD137+) in a vaccinated subject. 4e. Percentage of SARS-CoV-2 spike specific AIM+ non-naïve CD8+ cells (CD69+CD137+) at baseline and two weeks following the second mRNA vaccine dose in non-saRNA participants (orange and green) and two weeks following the second saRNA and mRNA vaccine doses in saRNA participants (red and blue). 4f. Antigen specific CD8+:CD4+ ratio (geometric mean shown). 4g. Correlation between IFN-γ ELISpot and % antigen specific CD4+ and CD8+ T cells. Graphs are all on a logarithmic scale. Differences between groups determined using Kruskall-Wallis followed by Mann-Whitney for individual comparisons (displayed); Correlation by Spearman’s rank correlation coefficient; -, no exposure; +, single exposure; ++, two exposures; Significant values displayed: *** p<0.005 ** <0.01 *<0.05.