Lung directed antibody gene transfer confers protection against SARS-CoV-2 infection

Abstract

Background: The COVID-19 pandemic continues to be a worldwide threat and effective antiviral drugs and vaccines are being developed in a joint global effort. However, some elderly and immune-compromised populations are unable to raise an effective immune response against traditional vaccines.

Aims: We hypothesised that passive immunity engineered by the in vivo expression of anti-SARS-CoV-2 monoclonal antibodies (mAbs), an approach termed vectored-immunoprophylaxis (VIP), could offer sustained protection against COVID-19 in all populations irrespective of their immune status or age.

Methods: We developed three key reagents to evaluate VIP for SARS-CoV-2: (i) we engineered standard laboratory mice to express human ACE2 via rAAV9 in vivo gene transfer, to allow in vivo assessment of SARS-CoV-2 infection, (ii) to simplify in vivo challenge studies, we generated SARS-CoV-2 Spike protein pseudotyped lentiviral vectors as a simple mimic of authentic SARS-CoV-2 that could be used under standard laboratory containment conditions and (iii) we developed in vivo gene transfer vectors to express anti-SARS-CoV-2 mAbs.

Conclusions: A single intranasal dose of rAAV9 or rSIV.F/HN vectors expressing anti-SARS-CoV-2 mAbs significantly reduced SARS-CoV-2 mimic infection in the lower respiratory tract of hACE2-expressing mice. If translated, the VIP approach could potentially offer a highly effective, long-term protection against COVID-19 for highly vulnerable populations; especially immune-deficient/senescent individuals, who fail to respond to conventional SARS-CoV-2 vaccines. The in vivo expression of multiple anti-SARS-CoV-2 mAbs could enhance protection and prevent rapid mutational escape.

Keywords: COVID-19; respiratory infection; viral infection.

Figure 1

S-LV infection requires hACE2, which can be supplied to mouse lungs by rAAV in vivo gene transfer. (A) WT parental HEK293T/17 cells and HEK293T/17 cells expressing hTMPRSS2, hACE2 or both hACE2 and hTMPRSS2 as indicated, were infected with an S-LV expressing enhanced green fluorescent protein (eGFP). The percentage of S-LV transduced cells was evaluated by flow cytometry. The dotted line represents the limit of quantification. One-way analysis of variance (ANOVA), with Dunnett’s multiple comparisons test (ns and ***** represent p>0.05 and p<0.0001, respectively). (B) Lung immunohistochemistry for eGFP was assessed in BALB/c (9-week-old) mice 14 days after intranasal delivery of 1xD-PBS (control), 1E11 GC of rAAV9 or 7E10 GC rAAV6.2 vectors expressing eGFP (n=3/group). Scale bar=500 μm. (C) Lung sections were subjected to RNAscope in situ hybridisation analysis 14 days after intranasal delivery of 1xD-PBS (control), 1E11 GC of rAAV9 or 7E10 GC rAAV6.2 vectors expressing hACE2 (n=3/group); hACE2 vector-specific WPRE probe (red), alveolar type II cell specific Sftpb probe (green), 4′,6-diamidino-2-phenylindole (DAPI) stained nuclei (blue). AW, airway; P, parenchyma. Scale bar=125 μm.

Figure 2

In vivo delivery of hACE2 allows the SARS-CoV-2 mimic S-LV to infect the lungs of standard laboratory mice. (A) Experimental design for in vivo investigation of hACE2/hTMPRSS2 delivery via rAAV vectors to support S-LV infection. BALB/c mice (n=4–8/group) were transduced intranasally with the indicated rAAV vector(s) or vehicle (1xD-PBS). At 14 days post-rAAV delivery, as indicated, lungs were infected with 890 ng p24 of an S-LV expressing firefly luciferase. S-LV dependent in vivo luciferase bioluminescence was monitored for each animal 1, 2, 4, 7, 14, 21 and 38 days post-S-LV infection. (B) Representative in vivo bioluminescence images of mice pretreated 14 days previously with the indicated doses of rAAV.hACE2 and hTMPRSS2 vectors and, as indicated, subsequently challenged with 890 ng p24 of S-LV at day 0. Repeated bioluminescence imaging of S-LV dependent luciferase expression was performed at the indicated time points. Bioluminescence values (photons/s/cm²/sr) are presented as a pseudocolour scale as indicated. (C) Time-course of bioluminescence for the indicated treatment groups after infection with 890 ng p24 of S-LV. Symbols represent group mean±SD, n=4–8 per group. The dotted line indicates the mean naïve background signal. (D) Area under curve of bioluminescence values (photons/s/cm²/sr×days) for each animal in B and C was computed, symbols represent individual animals and group mean±SD (ANOVA, Dunnett’s multiple comparison against the unlabelled treatment group (1E11 GC rAAV9.hACE2); ns, * and **** represent p>0.05, <0.05 and <0.0001, respectively).

Figure 3

S-LV mouse lung infection is limited by hACE2 availability. (A) Experimental design for in vivo investigation of S-LV dose-response. BALB/c mice (n=3/group) were dosed intranasal with 1E11 GC rAAV9.hACE2 to establish hACE2 expression. Fourteen days later, lungs were infected with 0–940 ng p24 of an S-LV expressing firefly luciferase. S-LV dependent in vivo luciferase bioluminescence was monitored for each animal 1, 2, 4 and 7 days post-S-LV infection. (B) Representative in vivo bioluminescence images of mice, treated as described in (A). Repeated bioluminescence imaging to monitor S-LV dependent luciferase expression was performed at the indicated time points. Bioluminescence values (photons/s/cm²/sr) are presented as a pseudocolour scale as indicated. (C) Time-course of bioluminescence after S-LV infection for the indicated treatment groups as described in (A). Symbols represent group mean±SD, n=4–8 per group. The dotted line indicates the mean naïve background signal. (D) Area under curve of bioluminescence (photons/s/cm2/sr×days) for each animal in C) was computed, symbols represent group mean±SD (ANOVA, Dunnett’s multiple comparison against the unlabelled treatment group (470ng p24); ns, **, *** and **** represent p>0.05, p<0.05, <0.001 and <0.0001, respectively).

Figure 4

NC0321 mAb expression by rAAV and rSIV.F/HN vectors. (A) Schematic of a codon-optimised, single open-reading frame, human IgG mAb cDNA. Regions encoded include IgG heavy and kappa light chain variable and constant regions, with each proceeded by a human growth hormone signal sequence (hGH SS) and joined by a Furin/2A (F2A) protein cleavage site. (B) The SARS-CoV-2 receptor binding domain (RBD) protein binding activity of the anti-SARS-CoV-2 mAbs NC0321 and CR3022 single open reading frame protein expressed from an rSIV.F/HN vector) was examined by ELISA. Binding activity of OD at 450 nm is proportional to the amount of antibody bound to the SARS-CoV-2 RBD protein. (C) The neutralising activity of anti-SARS-CoV-2 mAb NC0321 (single open reading frame protein expressed from an rSIV.F/HN vector) to block S-LV infection (multiplicity of infection 1) of an hACE2-293T cell line was examined. In (B and C) a single open reading frame anti-influenza mAb T1-3B acted as an isotype negative control; and CR3022 an anti-SARS-CoV-2 neutralising mAb, that can bind but not neutralise SARS-CoV-2 was used as a comparator. (D) Serum human IgG levels in mice were determined at day 7, 14 and 28 after transduction with the indicated doses of NC0321 expressing vector using ELISA (ANOVA, Dunnett’s multiple comparison against the unlabelled treatment group; *** and **** represent <0.001 and <0.0001, respectively). The levels of human IgG observed in naïve animals is indicated by the dotted line. (E) BALF of mice from (D) was collected at day 28 post-transduction with NC0321 expressing vector, and human IgG levels measured using ELISA; IgG levels in epithelial lining fluid (ELF) were computed by comparison of urea levels in BALF and serum (Kruskal Wallis, Dunn’s multiple comparison against the unlabelled treatment group; ns, **,*** and **** represent p>0.05, <0.01, <0.001 and <0.0001, respectively). Each dot represents an individual mouse and data are presented as endpoint titres (mean±SD).

Figure 5

Vector-mediated expression of NCO321 antibody protects against S-LV infection. (A) Experimental design to test efficacy of the COVID-19 Vectored Immunoprophylaxis strategy in vivo. Groups of mice were dosed as indicated (groups 1–7) under a single anaesthesia to establish hACE2 and NC0321 expression. Twenty-one days later, animals were infected with an S-LV expressing firefly luciferase (day 0) and, subsequently, S-LV dependent in vivo luciferase bioluminescence monitored for each animal on days 1, 2, 4 and 7. Serum and BALF samples were taken at indicated times for determination of NC0321 IgG levels. (B) Vectored delivery and expression of NC0321 to mouse lungs significantly inhibits S-LV infection. Groups 1–7 of BALB/c mice (n=10/group) were treated as indicated and after 21 days infected with 470 ng p24 of an S-LV expressing firefly luciferase (day 0). Representative images of in vivo bioluminescence are shown 7 days post-S-LV infection for each of the six treatment groups. Bioluminescence values (photons/s/cm²/sr) are presented as a pseudocolour scale as indicated. (C) Area under curve of bioluminescence (photons/s/cm²/sr×days) for each animal in B was computed. To aid visualisation, bioluminescence values were normalised such that T1-3B isotype control values were 100%. Control values were obtained from animals that were infected with S-LV but had not received rAAV9-hACE2. Group mean±SD is indicated (ANOVA, Dunnett’s multiple comparison against the unlabelled treatment group; ns, **, *** and **** represent p>0.05, <0.01, <0.001 and <0.0001, respectively). (D) Serum from mice 28 days post receiving rAAV8.NC0321 was collected, and limiting dilutions were made to measure the binding activity against SARS-CoV-2 Spike or receptor binding domain proteins of newly emerging SARS-CoV-2 variants as indicated. Negative control is binding activity observed with cell culture medium only.