A single dose, BCG-adjuvanted COVID-19 vaccine provides sterilising immunity against SARS-CoV-2 infection

Abstract

Global control of COVID-19 requires broadly accessible vaccines that are effective against SARS-CoV-2 variants. In this report, we exploit the immunostimulatory properties of bacille Calmette-Guérin (BCG), the existing tuberculosis vaccine, to deliver a vaccination regimen with potent SARS-CoV-2-specific protective immunity. Combination of BCG with a stabilised, trimeric form of SARS-CoV-2 spike antigen promoted rapid development of virus-specific IgG antibodies in the blood of vaccinated mice, that was further augmented by the addition of alum. This vaccine formulation, BCG:CoVac, induced high-titre SARS-CoV-2 neutralising antibodies (NAbs) and Th1-biased cytokine release by vaccine-specific T cells, which correlated with the early emergence of T follicular helper cells in local lymph nodes and heightened levels of antigen-specific plasma B cells after vaccination. Vaccination of K18-hACE2 mice with a single dose of BCG:CoVac almost completely abrogated disease after SARS-CoV-2 challenge, with minimal inflammation and no detectable virus in the lungs of infected animals. Boosting BCG:CoVac-primed mice with a heterologous vaccine further increased SARS-CoV-2-specific antibody responses, which effectively neutralised B.1.1.7 and B.1.351 SARS-CoV-2 variants of concern. These findings demonstrate the potential for BCG-based vaccination to protect against major SARS-CoV-2 variants circulating globally.

Fig. 1. Single immunisation with BCG:CoVac vaccine induces rapid development of anti-SARS-CoV-2 spike antibodies and IFN-γ-secreting T cells.

a C57BL/6 mice (n = 4/group) were vaccinated subcutaneously with PBS, BCG, BCG^SpK, Alum^SpK or BCG:CoVac and whole blood collected at day 14, 28 and 42. b, c Spike-specific IgG1 and IgG2c titres in plasma were determined by ELISA and estimated by the sigmoidal curve of each sample interpolated with the threshold of the negative sample ±3 standard deviations. The dotted line shows the limit of detection. d At 14 days post-vaccination PBMCs were restimulated ex vivo with 5 μg/mL of SARS-CoV-2 spike and cytokine production determined by flow cytometry. Representative dot plots of CD44⁺ CD4⁺ T cells and CD44⁺ CD8⁺ T cells expressing IFN-γ with average ± SEM (gating strategy in Supplementary Fig. 3). e Frequency of circulating CD4⁺ and CD8⁺ T cells expressing IFN-γ or f CD4⁺ T cells expressing IL-17 or TNF per 10⁶ PBMCs. Data presented as mean ± SEM and is representative of two independent experiments. Significant differences between groups compared to BCG^Spk *p < 0.05, **p < 0.01, ***p < 0.01 or Alum^{Spk †}p < 0.05, ^††p < 0.01, ^†††p < 0.001 were determined using one-way ANOVA.

Fig. 2. BCG:CoVac vaccination promotes expansion of T follicular helper cells and spike-specific B cells in mice.

a–c C57BL/6 mice (n = 5/group) were vaccinated subcutaneously with PBS, BCG, BCG^SpK, Alum^SpK or BCG:CoVac and 7 days after immunisation B and T cell response assessed by multicolour flow cytometry in the draining lymph node. Shown are representative dot plots of spike-specific germinal centre B cells (a, CD19⁺MHCII⁺GL7⁺CD38⁻), plasma B cells (b, CD19⁺MHCII⁺CD138⁺) and T follicular helper T cells (c, CXCR5⁺BCL6⁺) with average ±SEM (gating strategy in Supplementary Figs. 4 and 5). d–f The total number of d spike⁺ GC B cells, e spike⁺ plasma cells and f T follicular helper cells. Data presented as mean ± SEM and is representative of two independent experiments. Significant differences between groups *p < 0.05, **p < 0.01, ***p < 0.01 were determined by one-way ANOVA.

Fig. 3. BCG:CoVac induces high-titre neutralising antibodies against live SARS-CoV-2 that correlate with the production of antigen-specific IgG2c.

a–e Plasma from vaccinated mice (from Fig. 1) were tested for neutralising activity against live SARS-CoV-2 infection of VeroE6 cells. a Neutralising antibody (NAb) titres (IC₅₀) were calculated as the highest dilution of plasma that still retained at least 50% inhibition of infection compared to controls. NAb titres from PCR confirmed SARS-CoV-2-infected individuals (COVID) were determined using the same method. b, c Spearman correlations of spike-specific IgG2c or IgG1 titres and NAbs after Alum^SpK or BCG:CoVac vaccination. d, e Correlation of IgG2c or IgG1 titres and NAbs after vaccination with BCG^SpK. The dotted line shows the limit of detection. Data presented as mean ± SEM and is representative of two independent experiments. Significant differences between groups compared to BCG^Spk **p < 0.01, ***p < 0.0 or Alum^{Spk †}p < 0.05, ^†††p < 0.001 was determined by one-way ANOVA.

Fig. 4. A single dose of BCG:CoVac protects against severe SARS-CoV-2 infection.

a Male K18-hACE2 mice (n = 4/group) were immunised with sham (PBS), BCG or BCG:CoVac 21 days prior to challenge with 10³ PFU SARS-CoV-2. Disease outcomes were assessed 6 days later. b Clinical scores at day 6 post-infection. c Percentage of initial body weight loss in K18-hACE2 mice. Viral titres in lung homogenates (d) or bronchoalveolar lavage fluid (BALF) (e) were determined using plaque assay. The dotted line represents the limit of detection. f Total inflammatory cells in bronchoalveolar lavage fluid (BALF). g Total number of inflammatory cells in stained histological sections of lungs. h Cytokine/chemokine quantification in lung homogenates. i Six weeks after immunisation mice were challenged with M. tuberculosis H37Rv by aerosol (~100 CFU) and four weeks later the bacterial load was assessed in the lungs and presented as log₁₀ of the mean CFU ± SEM. Significant differences between groups *p < 0.05, **p < 0.01 were determined by one-way ANOVA.

Fig. 5. Heterologous boosting of BCG:CoVac-primed mice results in augmented SARS-CoV-2-specific IgG2c titres and neutralising antibodies.

a C57BL/6 mice (n = 3–4/group) were vaccinated (as in Fig. 1) and at day 21 mice were boosted with Alum^Spk. b Spike-specific IgG2c titres in plasma were determined by ELISA estimated from the sigmoidal curve of each sample interpolated with the threshold of the negative sample ±3 standard deviations. c Neutralising antibody (NAb) titres (IC₅₀) were calculated as the highest dilution of plasma for all groups that still retained at least 50% inhibition of infection compared to controls. The dotted line shows the limit of detection. d, e NAb titres against the B.1.1.7 or B.1.351 SARS-CoV-2 variants were also determined using plasma from either d Alum^SpK or e BCG:CoVac-vaccinated mice. Data presented as mean ± SEM and is representative of two independent experiments. Significant differences between groups compared to BCG^Spk *p < 0.05, **p < 0.01, ***p < 0.01 or Alum^{Spk †}p < 0.05, ^††p < 0.01, ^†††p < 0.001 were determined by one-way ANOVA.