Immunogenicity of a new gorilla adenovirus vaccine candidate for COVID-19

Abstract

The coronavirus disease 2019 (COVID-19) pandemic caused by the emergent severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) threatens global public health, and there is an urgent need to develop safe and effective vaccines. Here, we report the generation and the preclinical evaluation of a novel replication-defective gorilla adenovirus-vectored vaccine encoding the pre-fusion stabilized Spike (S) protein of SARS-CoV-2. We show that our vaccine candidate, GRAd-COV2, is highly immunogenic both in mice and macaques, eliciting both functional antibodies that neutralize SARS-CoV-2 infection and block Spike protein binding to the ACE2 receptor, and a robust, T helper (Th)1-dominated cellular response. We show here that the pre-fusion stabilized Spike antigen is superior to the wild type in inducing ACE2-interfering, SARS-CoV-2-neutralizing antibodies. To face the unprecedented need for vaccine manufacturing at a massive scale, different GRAd genome deletions were compared to select the vector backbone showing the highest productivity in stirred tank bioreactors. This preliminary dataset identified GRAd-COV2 as a potential COVID-19 vaccine candidate, supporting the translation of the GRAd-COV2 vaccine in a currently ongoing phase I clinical trial (ClinicalTrials.gov: NCT04528641).

Keywords: COVID-19; SARS-CoV-2; gorilla adenovirus; immunogenicity; vaccine.

Graphical abstract

Figure 1

Phylogenetic analysis of gorilla adenovirus (GRAd)32 and seroprevalence in human sera (A) Phylogenetic analysis using adenoviral polymerase sequences identifies GRAd32 as a group C adenovirus. HAdV, human adenovirus; SAdV, simian adenovirus; GAdV, GRAd. (B and C) Neutralizing antibody (NAb) titers measured in sera collected from a cohort of 40 human healthy donors of US origin (B) and in 90 sera from Italian volunteers (C) enrolled in a GRAd-COV2 phase 1 study (C, filled symbols = younger adults 18−55 years; empty symbols = older adults 65−85 years). Data are expressed as the reciprocal of serum dilution resulting in 50% inhibition of SEAP activity. Horizontal black dotted lines indicate assay cut-off (titer of 18). Red dotted lines indicate NAb titer of 200, which is reported to potentially impact vaccine immunogenicity. Red continuous lines indicate geometric mean. Statistical analysis of the datasets was performed by unpaired, two-tailed Mann-Whitney analysis. The tables show the absolute numbers and the percentage of sera with NAb titers to Ad5 or GRAd32 below cut-off (<18), between 18 and 200 (<200), and above 200 (>200).

Figure 2

Productivity of GRAd32 backbone variants in 2 L bioreactor Suspension-adapted packaging HEK293 cells were seeded at 5 × 10⁵ cells/mL and infected at MOI 200. The titer of virus contained in the bulk cell lysates collected at different time points (hours) after infection was measured by qPCR.

Figure 3

Expression of Spike antigens in GRAd32 vectors (A–D) Whole cell FACS analysis of HeLa cells not infected or infected with 150 MOI of GRAd32b (A and C) or GRAd32c (B and D). 48 h after infection, cells were stained with anti-S-2 antibody (A and B) or with human soluble his-tagged ACE2 receptor and anti-His antibody (C and D), both then detected with goat anti-mouse IgG mAb conjugated with Alexa Fluor 647. The histograms of gated live cells corresponding to cells infected with the vector-encoded wild-type (WT) Spike protein are labeled in red and with the stabilized S-2P in blue; the gray line represents non-infected cells. (E and F) Quantification of expression levels as detected by anti-S-2 (E) or ACE2 (F) binding, expressed as the mean fluorescence intensity (MFI).

Figure 4

Humoral response to SARS-CoV-2 Spike induced in BALB/c mice 2 and 5 weeks (w2 and w5) after immunization with GRAd-COV2 (A) Spike-binding total IgG titers in sera from mice immunized with 1 × 10⁹ vp of GRAd-COV2. IgG titers were measured in sera collected 2 or 5 weeks post-immunization by ELISA on recombinant full-length Spike or RBD. Data are expressed as endpoint titer. The main and error bars indicate geometric mean and 95% confidence interval (CI). (B) The ratio between IgG2a and IgG1 titers measured in week 5 sera by ELISA on full-length Spike in mice vaccinated with either 1 × 10⁹ vp of GRAd-COV2 (n = 13) or with two injections 2 weeks apart of 2.5 μg Spike protein formulated in alum adjuvant (n = 3). (C) SARS-CoV-2 NAbs induced by GRAd-COV2 and detected in sera at week 5 post-immunization by SARS-CoV-2 (2019-nCoV/Italy-INMI) microneutralization assay on VERO E6 cells (CPE, black symbols plotted on left y axis) or by SARS-CoV-2-pseudotyped virus neutralization assay (gray symbols plotted on right axis). Neutralizing titers are expressed as IC₅₀ or IC₉₀, or the reciprocal of serum dilution achieving 50% or 90% neutralization, respectively. Horizontal lines in (B) and (C) represent geometric mean.

Figure 5

T cell response to SARS-CoV-2 Spike induced in BALB/c mice 5 weeks after GRAd-COV2 immunization (A) IFN-γ ELISpot on splenocytes. Data are expressed as IFN-γ spot-forming cells (SFCs)/10⁶ splenocytes. Individual data points represent response to S-1 and S-2 pools’ stimulation compared to mock stimulation (DMSO) in each animal. (B) IFN-γ/IL-2/IL-4 + IL-13/IL-17 intracellular staining and FACS analysis on splenocytes. Data are expressed as the percentage of cytokine-secreting CD8 or CD4 T cells in response to S-1 and S-2 Spike peptide pools’ stimulation, obtained by summing reactivity to each of the 2 Spike peptide pools and subtracting 2 times the DMSO background. Main and error bars indicate geometric mean and 95% CI in both (A) and (B).

Figure 6

Humoral response to SARS-CoV-2 Spike induced in BALB/c mice by GRAd encoding for pre-fusion stabilized versus WT Spike (A) Full-length Spike-binding total IgG titers (black symbols, plotted on left y axis) and neutralizing titers on pseudotyped viruses (red symbols, plotted on right y axis) in week 5 sera from mice immunized with 1 × 10⁹ vp of GRAd encoding either WT Spike (open symbols) or S-2P (filled symbols). Data are expressed either as endpoint titer or as IC₉₀ neutralizing titer. Horizontal lines represent geometric mean. Two-tailed unpaired t test was used (∗∗p = 0.007, t = 3.16, degree of freedom [df] = 14). (B) Inhibition of binding between immobilized RBD and recombinant human ACE2 by dilution curves of week 5 sera from mice immunized with 1x10⁹ vp of GRAd encoding either WT Spike (red curves) or S-2P (blue curves). Data are expressed as percentage of neutralization activity relative to binding in the absence of serum. Inhibition curve from a naive, non-immunized mouse as control is shown in black.

Figure 7

Humoral response to SARS-CoV-2 Spike induced in cynomolgus macaques by GRAd-COV2 immunization Four cynomolgus macaques received a single 5 × 10¹⁰ vp intramuscular injection of either GRAd-COV2 (blue curves) or GRAd32c-S (red curves). (A–D) Serum samples obtained before (Pre) and 2, 4, 6, and 10 weeks post-immunization were tested by (A) full-length Spike-binding ELISA (total IgG endpoint titers), (B) RBD-binding ELISA (total IgG endpoint titers), (C) pseudotyped virus neutralization assay (neutralizing IC₅₀ titers), or (D) SARS-CoV-2 microneutralization assay (neutralizing IC₅₀ titer). (E) IC₅₀ titers measured by the SARS-CoV-2 microneutralization assay in a panel of COVID-19 convalescent human sera and in GRAd-COV2-immunized macaques 10 weeks post-immunization. Horizontal lines represent geometric mean.

Figure 8

T cell response to SARS-CoV-2 Spike induced in cynomolgus macaques upon GRAd-COV2 or GRAd32c-S immunization (A) IFN-γ or IL4 ELISpot on freshly isolated PBMCs 2 and 8 weeks post-immunization. Data are expressed as IFN-γ or IL4 SFCs/10⁶ PBMC. Individual data points represent a cumulative Spike T cell response, calculated by summing S-1 and S-2 peptide pools’ stimulation response corrected for the DMSO background of each animal. Bar represents geometric mean. (B) IFN-γ ELISpot response to individual S-1 and S-2 peptide pools in PBMC 2 weeks post-immunization. (C) Intracellular staining and FACS analysis on PBMC. Data are expressed as the percentage of either CD8- or CD4-secreting cytokine in response to S-1 and S-2 Spike peptide pools’ stimulation, obtained by summing reactivity to each of the 2 Spike peptide pools and subtracting 2 times the DMSO background. Horizontal lines represent geometric mean.