In vivo delivery of engineered synthetic DNA-encoded SARS-CoV-2 monoclonal antibodies for pre-exposure prophylaxis in non-human primates

ABSTRACT

COVID-19 remains a major public health concern. Monoclonal antibodies have received emergency use authorization (EUA) for pre-exposure prophylaxis against COVID-19 among high-risk groups for treatment of mild to moderate COVID-19. In addition to recombinant biologics, engineered synthetic DNA-encoded antibodies (DMAb) are an important strategy for direct in vivo delivery of protective mAb. A DMAb cocktail was synthetically engineered to encode the immunoglobulin heavy and light chains of two different two different Fc-engineered anti-SARS-CoV-2 antibodies. The DMAbs were designed to enhance in vivo expression and delivered intramuscularly to cynomolgus and rhesus macaques with a modified in vivo delivery regimen. Serum levels were detected in macaques, along with specific binding to SARS-CoV-2 spike receptor binding domain protein and neutralization of multiple SARS-CoV-2 variants of concern in pseudovirus and authentic live virus assays. Prophylactic administration was protective in rhesus macaques against signs of SARS-CoV-2 (USA-WA1/2020) associated disease in the lungs. Overall, the data support further study of DNA-encoded antibodies as an additional delivery mode for prevention of COVID-19 severe disease. These data have implications for human translation of gene-encoded mAbs for emerging infectious diseases and low dose mAb delivery against COVID-19.

KEYWORDS:

• SARS-CoV-2
• monoclonal antibody
• DNA
• DNA-encoded monoclonal antibody
• prevention
• gene-encoded antibody
• DMAb
• macaque

Introduction

The coronavirus disease 2019 (COVID-19) pandemic continues to have a tremendous impact on global public health, with >770 million cases reported worldwide since 2020 [1]. COVID-19 disease is caused by the betacoronavirus severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), with clinical manifestations ranging from mild to severe or fatal disease and potential for development of long term post-acute sequalae of SARS-CoV-2. There remains an ongoing need for development of vaccines and therapeutics to respond to emerging SARS-CoV-2 variants of concern (VOC) and variants of interest, including preventative vaccines as well as preventative and therapeutic antibodies. The SARS-CoV-2 Spike surface glycoprotein receptor binding domain (RBD) is the major target for strain- specific neutralizing antibodies that can potently block interactions with the human angiotensin conserving enzyme 2 (ACE2) receptor, which is important for virus entry into host cells [2]. Several recombinant monoclonal antibody biologics received initial emergency use authorization (EUA) for treatment and for pre-exposure prophylaxis against COVID-19 [3]. However, as SARS-COV-2 evolves, mutations in the Spike protein RBD are resulting in diverging lineages that escape many of the neutralizing antibodies against ancestral strains [4–8]. All initial mAb biologics have lost EUA due to lack of neutralizing activity against the latest emerging Omicron lineage variants [9].

Although the tixagevimab plus cilgavimab mAb cocktail (Evusheld) is no longer recommended as pre-exposure prophylaxis, it serves as a model candidate for engineering gene-encoded platforms with potential for long-term in vivo expression of antibody. This antibody cocktail retained remarkable potency up until only the latest XBB lineage Omicron variants [5], supporting the potential benefit of a two-antibody cocktail against evolving viral pathogens. An updated version of Evusheld (AZD5156/AZD3152) that includes cilgavimab continues to be studied in clinical trials (clinicaltrials.gov NCT05648110). The original Evusheld tixagevimab plus cilgavimab cocktail comprises two RBD-binding mAbs, AZD8895 and AZD1061, which were originally isolated from a high-throughput screen of neutralizing mAbs isolated from SARS-CoV-2 convalescent individuals (parental antibodies COV2-2196 and COV2-2130, respectively) [10] and bind to non-competing epitopes on the RBD [11]. AZD8895 and AZD1061 contain the L234F/L235E/P331S (“TM”) Fc modification which reduces Fc receptor and complement C1q binding to mitigate potential risk for antibody-mediated enhancement [12] and includes additional Fc engineering incorporating the half-life extending M252Y/S254T/T256E (“YTE”) to improve pharmacokinetic durability following a single administration [13]. AZD8895 and AZD1061 have been shown to act synergistically to confer protection against SARS-CoV-2 in mouse, hamsters, and non-human primate (NHP) models [14, 15]. In people, tixagevimab plus cilgavimab displayed an extended half-life of approximately 90 days and reduced the risk of developing symptomatic COVID-19 in high-risk populations [16]. Cilgavimab (AZD1061) is being evaluated in ongoing human trials in combination with new mAb AZD3152 to expand broad protection across ancestral and emerging VOC.

In addition to biological efficacy, additional considerations are important for global delivery of mAb biologics against infectious diseases. During the COVID-19 pandemic, access to potentially life-saving mAb biologics faced multiple challenges for broader population coverage including high doses (in milligrams/kilogram), shelf-stability, temperature stability and distribution barriers for low-/middle-income countries and resource-limited settings. Therefore, additional strategies that can further facilitate mAb uptake and global availability would be valuable for infection control. We previously described in vivo administration of plasmid DNA-encoded antibodies (DMAbs) expressing synthetically engineered mAb heavy chain and light chain genes [17–19]. DMAb DNA is delivered directly in vivo via intramuscular administration, employing host cell machinery to initiate transcription, translation, folding, and secretion of fully functional antibody directly into circulation. In vivo co-administration of DMAbs encoding modified anti-SARS-CoV-2 mAbs COV2-2196 (DMAb-2196) and COV2-2130 (DMAb-2130), resulted in durable serum expression kinetics and neutralization of SARS-CoV-2 live virus and Spike pseudotype viruses against multiple VOCs including wild type, alpha, beta, delta, iota lineages, with protective efficacy in mouse and hamster infection models against the USA-WA1/2020 strain [20]. These previous studies highlight that in vivo delivery of gene-encoded antibodies does not follow in vitro predicted expression and can vary significantly between different variable heavy (VH) and variable light (VL) chain families in mice. In these studies, it was also shown that additional design, half-life extension, delivery optimizations, and formulation will be essential for achieving clinically translatable expression. Taken together, these data underscore the importance for additional study in larger animal models including understanding in vivo delivery of different VH and VL genes in macaques, a model closer to humans.

Here, we describe preclinical evaluation of a wild type anti-SARS-CoV-2 DMAb cocktail expressing DMAb-2196 and DMAb-2130 with and without half-life modification (YTE) in non-human primates (NHP). This is the first study in macaques evaluating pharmacokinetics of a nucleic acid encoded DMAb cocktail expressing a SARS-COV-2 antibody cocktail in vivo. We demonstrate a two-antibody DMAb cocktail, delivered via four separate plasmids, can express functional antibody in vivo at microgram (ug) levels. This consistent delivery was achieved through a combination of plasmid formulation, administration regimen, and modified side port needle delivery designed to enhance in vivo injection-site transfection. We demonstrate the protective efficacy of this 2-DMAb cocktail formulation a macaque infection model to show the ability of in vivo delivered DMAbs to protect against SARS-CoV-2-related disease pathology in the lungs. These findings are of importance given the continued evaluation of cilgavimab as part of updated COVID-19 mAb cocktails and provide support for in vivo delivery of gene-encoded antibodies using public clonotypes and similar VH and VL genes in humans against SARS-CoV-2 and other pathogens.

Results

Expressed DMAbs bind to human and non-human primate Fc gamma receptors

Binding kinetics to human Fc receptors were previously described for tixagevimab plus cilgavimab mAbs AZD8895 and AZD1061 with the hIgG1 G1m3 allotype, with and without YTE and the TM modifications [21]. Synthetically engineered hIgG1 DMAb-2130 and DMAb-2196 were optimized for enhanced expression through codon optimization and RNA secondary structure modifications, as previously described [20]. Variants were generated by site-directed mutagenesis in order to preserve sequence homology and confirmed by sequencing. We confirmed the binding affinities of synthetically engineered hIgG1 DMAb-2130 and DMAb-2196 variants for both human and monkey Fc gamma receptors using surface plasmon resonance (SPR). Wild type G1m3 allotype, YTE half-life extended variant, and a TM variant with Fc receptor abrogating function DMAbs were assayed. The properties of each Fc variation and equilibrium dissociation constants (K_D) for human Fc gamma receptor I (hFcγRI/CD64), human Fc gamma receptor III (hFcγRIII /CD16), cynomolgus macaque Fc gamma receptor I (cyFcγRI/CD64) and rhesus macaque Fc gamma receptor III (rhFcγRIII/CD16) are listed in Table 1. Dissociation curves, including control purified human IgG and macaque IgG are shown in Supplementary Figure 1 (A–D). Overall, binding of expressed DMAbs was comparable for both human and monkey FcγR.

Table 1. xxx.

Construct	Construct	hFcgRI binding (KD)	hFcgRIII Binding (KD)	mFcgRI Binding (KD)	mFcgRIII Binding (KD)
DMAb-2130	Wild type G1m3 hIgG1	6.4	1214	4.2	643
DMAb-2130-YTE	G1m3 hIgG1 + YTE	11.3	678	4..5	627
DMAb-2130-TM	G1m3, abrogated Fc effector activity	Negligible	No binding	Negligible	No binding
DMAb-2196	Wild type G1m3 hIgG1	4.8	1221	3.0	320
DMAb-2196-YTE	G1m3 hIgG1 + YTE	10.4	1110	5.2	718
DMAb-2196-TM	G1m3, abrogated Fc effector activity	Negligible	No binding	Negligible	No binding

In vivo pharmacokinetic expression of half-life extended DMAbs in cynomolgus macaques

In vivo expression of synthetic DNA encoding half-life extended DMAb-2130-YTE and DMAb-2196-YTE was evaluated in cynomolgus macaques following sequential administration on day 0 and day 3 (Group 1, n = 5) or a single administration on day 0 (Group 2, n = 5) (Figure 1(a)). A three-injection sequential administration regimen was first described in [18]. Here, we evaluated single and two injection regimens with optimized needle delivery. DMAb pharmacokinetics were monitored over time and quantified by an optimized ELISA for detection of human IgG in macaque sera. DMAb human IgG expression was detected in all animals from both study groups. Group 1 animals had a mean Cmax expression of 4050 ng/mL ± 1712 ng/mL on day 15 post-DMAb administration (mean of animals in Figures 1(b, c)). Group 2 animals had a mean Cmax expression of 2768 ng/mL ± 1775 ng/mL on day 35 post-DMAb administration (Figure 1(d)). The overall duration of expression in cynomolgus macaques was 63 days. Interestingly, NHP M16912 in Group 1 expressed measurable levels of dMAb in the serum for >500 days (Figure 1(c)). MHC genotyping revealed NHP M16912 to be heterozygous for M2 and recM1M3 haplotypes.

Figure 1. Human IgG antibody pharmacokinetic expression levels in the serum of cynomolgus macaques following administration of the DMAb-2130-YTE and DMAb-2196-YTE cocktail in cynomolgus macaques. (a) Overview of study. (b) Expression levels in Group 1 (n = 4), receiving two DMAb administrations. (c) Expression levels in Group 1 NHP M16912. (d) Expression levels in Group 2 (n = 5), receiving a single DMAb administration. (e) Detection of 2196 Fab by RBD K444A ELISA. (f) Detection of 2130 Fab by RBD F486A ELISA. (g) Detection of 2196 and 2130 Fab by RBD K444A and in NHP M16912.

As the cynomolgus macaques were administered both DMAbs simultaneously, the individual concentrations and specificities of DMAb-2130-YTE and DMAb-2196-YTE were quantified by ELISA coated with SARS-CoV-2 Spike RBD proteins containing amino acid dropout mutations in the epitopes for each mAb clone [14] (Figure 1(e–g)). A SARS-CoV-2 RBD with a K444A mutation was used to quantify DMAb-2196-YTE expression (Figure 1(e)). A SARS-CoV-2 RBD with a F486A mutation was used to quantify DMAb-2130-YTE expression (Figure 1(f)). Binding of NHP16912 to K444A and F486A is included as separate panel Figure 1(g) for clarity. Recombinant protein COV2-2130-YTE and COV2-2196-YTE were used as standards. All animals expressed both DMAbs. Group 1 mean Cmax expression levels of 4099 ng/mL ± 2366 ng/mL and 2545 ng/mL ± 847 ng/mL were detected for DMAb-2196-YTE and DMAb-2130-YTE, respectively. Group 2 Cmax expression levels of 3020 ng/mL ± 1603 ng/mL and 1855 ng/mL ± 920 ng/mL were detected for DMAb-2196-YTE and DMAb-2130-YTE, respectively. Binding to K444A or F486A was detected at all time points for NHP M16912 (Figure 1(g)).

To further confirm in vivo expression and RBD binding results from specific DMAb expression, anti-idiotypic antibodies M54-2E9 and M16H5.1 against each mAb Fab (Fab COV2-2130 and Fab COV2-2196, respectively) were used to quantify the specific amount of each DMAb in circulation in Group 1 and Group 2 (Supplementary Figure 2A) and Group 1 M16912 (Supplementary Figure 2B). Recombinant protein COV2-2130-YTE and COV2-2196-YTE were used as standards. The expression levels detected by anti-idiotype ELISA were consistent with those observed by total human IgG ELISA and RBD ELISAs. Positive correlations were observed between total IgG ELISAs and RBD ELISAs, as well as RBD ELISAs and anti-idiotype ELISAs demonstrating the reliability of DMAb detection in macaques (Supplementary Figure 2C).

Sera from DMAb-administrated cynomolgus macaques bind to SARS-CoV-2 RBD proteins

Sera collected on day 21 post-DMAb administration were screened against a panel of SARS-CoV-2 RBD proteins from VOC representing diverging lineages and subvariants (ancestral, Alpha, Beta, Delta, Omicron BA1.12.1, BA2.75, BA.5). ELISA plates were coated with stabilized 6P RBD proteins and serial dilutions of day 21 NHP sera were assayed for binding (Supplementary Figure 3). Overall, sera from DMAb administered animals bound to the RBD proteins. A decrease in binding was observed against Omicron variants, consistent with the decline reported with the tixagevimab and cilagavimab antibodies.

Sera from DMAb-administrated cynomolgus macaques neutralize authentic SARS-CoV-2 and SARS-CoV-2 pseudoviruses

The functional ability of macaque-expressed DMAbs to neutralize SARS-CoV-2 was assessed in live virus and pseudotype virus assays performed on macaque serum samples collected on day 0 and day 15 post-administration. Neutralization infectious dose 50 (ID50) titers were determined against SARS-CoV-2 USA-WA1/2020 (WT lineage) and hCoV-19/USA/PHC6582021 (Delta lineage) viruses for Groups 1 and 2, in which all animals exhibited significant neutralizing activity by day 15 (Figure 2(a), top).

Figure 2. SARS-CoV-2 authentic virus and pseudovirus titers in the serum of cynomolgus macaques following administration of DMAb-2130-YTE and DMAb-2196-YTE. Neutralization assays were performed on sera samples collected on day 15 post-DMAb administration. (a) Neutralization ID50 titers against authentic SARS-CoV-2 USA-WA1/2020 (WT lineage) and hCoV-19/USA/PHC658/2021 (Delta lineage). Graphs display animals all together (i, ii) and by individual group (iii, iv). (b) SARS-CoV-2 pseudovirus neutralization titers against WT, Beta, and Delta lineage pseudoviruses. Graphs display animals all together (i-iii) and by group (iv-vi). IC50 neutralization titers for (c) Live virus and (d) Pseudovirus. Omicron BA.2 and BA2.12 pseudovirus (e) ID50 titersand (f) IC50 values. Individual animals are displayed on the graphs, along with the geometric mean titer and geometric standard deviation. (*) p < 0.05 and (**) p < 0.01.

Pseudovirus neutralization titers were determined using aliquots of sera from the day 0 and day 15 time points. Three pseudotype viruses expressing SARS-CoV-2 Spike proteins were used in these assays: (1) Wild type (WT, Wuhan-Hu-1), (2) Beta lineage, and (3) Delta lineage. Pseudovirus neutralization ID50 titers increased from day 0 to day 15, demonstrating serum neutralization capacity following expression of DMAbs (Figure 2(b), top). Overall IC50 values for live virus and pseudovirus neutralization assays were shown in Figure 2(c) and (d)). In a pseudotype virus assay performed at Inovio, sera from macaques had activity against BA.2 and BA 2.12 pseudoviruses (Figure 2(e, f)). The mean values from live virus and pseudovirus assays are summarized in Supplementary Tables 1–3. IC50 values were calculated against the Day 15 serum expression levels reported in Figure 1. Additionally, purified DMAb from macaque sera demonstrated neutralizing antibody activity against BA.4 pseudovirus (Supplementary Figure 4).

Detection of anti-DMAb antibodies in cynomolgus macaques

As human IgG is xenogenic to non-human primates, the presence of monkey anti-drug antibodies (anti-DMAb antibodies/ADA) can reduce mAb levels in serum and therefore ADA was detected by ELISA. Strong ADA responses were detected in all animals except M16912, the NHP that continues to maintain expression levels long-term (Supplementary Figure 5). A decline in DMAb expression was observed when endpoint titers of host antibodies against DMAb exceed >10⁴. M16912 developed low ADA levels over time, however these low levels remained stable for several months and below 10⁴.

Prophylactic DMAb-administration in a rhesus macaque model of SARS-CoV-2 infection

DMAb administration was next evaluated in a SARS-CoV-2 rhesus macaque challenge model [22]. The single dose regimen based on cynomolgus macaque Group 2 was selected for the rhesus macaque study as there were no significant differences between the groups when NHP M16912 is separated from the comparative analyses. In the rhesus macaque study, we evaluated the contribution of YTE half-life extension to protection against SARS-CoV-2 infection. Rhesus macaque groups include a control group (no DMAb, Group 1, n = 6), DMAb-2130-YTE and DMAb-2196-YTE with half-life extension modification (DMAb-YTE, Group 2, n = 6/group), or DMAb-2130 and DMAb-2196 with wild type human IgG1 Fc without the YTE modification (DMAb-WT, Group 3, n = 6/group). Rhesus macaques were administered DMAb on day −14 before challenge with SARS-CoV-2 infectious virus, followed by necropsy on day 7 (Figure 3(a)).

Figure 3. Overview of serum expression and neutralizing activity of DMAbs in rhesus macaques. (a) Schematic overview of DMAb delivery in rhesus macaques. (b) Human IgG antibody expression levels in the serum with no DMAb, DMAb-YTE, or DMAb-WT cocktails. (c) Neutralization ID50 titers against authentic SARS-CoV-2 USA-WA1/2020 (WT lineage). (d) Pseudovirus neutralization ID50 titers against WT lineage. Group 1 (control) is depicted in grey circles, Group 2 (DMAb-YTE) is depicted in orange triangles, Group 3 (DMAb-WT) is depicted in purple squares. Lines represent the mean and error bars represent the standard deviation.

In vivo pharmacokinetic expression of DMAbs with and without half-life extension in rhesus macaques

DMAb expression levels were assayed in rhesus macaque sera collected over time (Figure 3(b)). Control animals had a mean ± standard deviation C_max expression of 35 ng/mL ± 8 ng/mL 21-days post-DMAb administration (Day +7). DMAb-YTE animals had a mean ± standard deviation C_max expression of 1441 ng/mL ± 712 ng/mL 15-days post-DMAb administration (Day +1). DMAb-WT animals had a mean ± standard deviation Cmax expression of 2134 ng/mL ± 1000 ng/mL on 21-days post-DMAb administration (Day +7). Samples from day −14 were unavailable, therefore the anti-idiotype ELISA was performed on day −11 and day 0 confirming that circulating human IgG is from in vivo expression resulting from DMAb administration (Supplementary Figure 6A, B).

Neutralizing activity of serum from DMAb-administered rhesus macaques against authentic SARS-CoV-2 virus and SARS-CoV-2 pseudoviruses

The functional activity of in vivo expressed DMAbs was assessed in live virus SARS-CoV-2 and Spike pseudovirus neutralization assays. Neutralization titers were detected against authentic SARS-CoV-2 USA-WA1/2020 (WT) and hCoV-19/USA/PHC6582021 (Delta lineage) live viruses and pseudoviruses (Figure 3(c, d), Supplementary text and Supplementary Figure 7). Challenge day IC50 values were calculated based on total DMAb IgG levels quantified by ELISA were (Supplementary Figure 7). Mean WT live virus and WT pseudovirus ID50 titers and IC50 values are summarized in Supplementary Tables 4 and 5.

Prophylactic administration of DMAbs protect against SARS-CoV-2-associated disease in rhesus macaques

To evaluate whether a single 2 mg intramuscular administration of the SARS-2 DMAbs could protect animals against a SARS-COV-2 infection all animals were challenged on day 0 with 2.8 × 10⁶ TCID50 SARS-CoV-2 nCoV-WA1-2020 via four routes of inoculation (intratracheal, intranasal, oral, and ocular), as previously described [22, 23]. Macaque weights were recorded and compared with DMAb expression data per kilogram and estimated blood volume (Supplementary Figure 8). Animals were monitored for 7 days following challenge, until necropsy (Supplementary Table 6). Signs of disease were recorded for each animal, with greater signs of disease observed in control Group 1 NHPs, compared to Group 2 or Group 3 animals (Figure 4(a)). Several animals had scores greater than 0 on day 0 prior to challenge as the daily scoring criteria includes the amount of food, water intake, and feces and it is common to observe animals with scores of 3 at baseline.

Figure 4. Clinical scores, viral loads, and histopathological analysis. (a) Clinical scores in control and DMAb-administered rhesus macaques following SARS-CoV-2 challenge. (b) Viral loads in nasal washes and bronchioalveolar lavage (BAL) following SARS-CoV-2 challenge detected by subgenomic qPCR. (b) (i) Nasal swab viral loads from individual animals are shown. (ii) Nasal swabs on day 5 post-challenge. (c) (i) BAL washes from individual animals are shown. (ii) BAL on day 3 post-challenge. Bars represent the mean and standard deviation. p < 0.05 is considered significant. LOD = assay limit of detection. (d) Photomicrographs of histological sections of the lung (top row HE, 100x) and anti-SARS-CoV-2 IHC (middle row 100x; bottom row 400x). Lesions are minimal to mild and demonstrate increased alveolar septal thickening and septal and alveolar inflammation. Immunoreactivity is scattered and sparse when present and appears specific to type I and type II pneumocytes.

Viral load detection

Shedding and viral loads were measured by qPCR of SARS-CoV-2 subgenomic RNA (Figure 4(b)). Nasal swabs were collected on exam days to compare shedding between the groups. Shedding was largely unaffected between the groups at early times post-infection viral RNA was significantly reduced in the nasal swabs of animals receiving DMAb-YTE on Day 5 post-challenge (Figure 4(b), p = 0.0385). BALs were collected during exams performed 1-, 3-, and 5-days post-challenge. Viral loads detected in the BALs were lower in the DMAb-YTE and WT group at 5 days post-challenge with only 3 animals in the treated groups having evidence of viral RNA at this time. However, significant differences were only detected in the DMAb-YTE BAL on day 3 post-challenge (Figure 4(c) ii, p = 0.0035).

Lung histology

The six non-treated control rhesus macaques (CoV521, 522, 523, 530, 531, 532) developed gross and histologic lung lesions typical of minimal to mild SARS-CoV-2 infection in rhesus macaques i.e. alveolar septal thickening and alveolar and septal inflammation as has been previously described [23]. Histological analysis of the lungs detected evidence of SARS-CoV-2 related pneumonia and perivascular inflammation (Figure 4(d) and Supplementary Figure 9). The six macaques administered the HC plasmid mAb cocktail (CoV524-529) developed lesions of SARS-CoV-2 but were, in most cases, less frequent and less severe than those of the non-treated controls. Likewise, the six macaques administered the HC TM plasmid mAb cocktail (CoV533-538) also developed lesions of SARS-CoV-2 but were, in most cases, less frequent and less severe than the non-treated controls. There was no appreciable difference in the lesions associated with the two treatment groups. The anti-SARS-CoV-2 immunohistochemistry results resemble the histology results with a reduction in the frequency and number of anti-SARS-CoV-2 immunoreactivity in the two treatment groups when compared to the non-treated control groups. Often the number of immunoreactive cells was so low that only a few could be detected in an entire tissue section. There was no difference noted between the two treatment groups. Post-SARS-CoV-2 challenge, DMAb was detected in bronchoalveolar lavage (BAL) of 2/6 and 4/6 NHPs in the DMAb-YTE and DMAb-WT groups, respectively (Supplementary Figure 10).

Detection of anti-DMAb antibodies in rhesus macaque sera

Low levels of ADA were detected on day 7 post-challenge (day 21 post-DMAb administration). ADA endpoint titers were detected in 1/6, 3/6, and 6/6 NHPs for naïve, DMAb-YTE, and DMAb-WT, respectively and ranged from 1:100 to 1:12,800 (Supplementary Figure 11). Although 6/6 NHPs in the DMAb-WT group had detectable ADA endpoint titers, specifically against DMAb-2196, only 2/6 were above 1:10⁴. ADA levels did not correlate with disease outcome post-challenge.

Discussion

Along with preventative vaccination, prophylactic administration of recombinant mAb have the potential to provide critical protection to individuals at high-risk for severe COVID-19 including those who do not induce strong immune responses following immunization and immune compromised patients. In parallel, gene-encoded mAbs are promising as an additional modality for in vivo mAb delivery, with the potential to overcome stability, storage, and transport limitations and to expand uptake in wider global populations. This would be of tremendous value during a pandemic like COVID-19 or against seasonal infections that affect millions of people worldwide. Plasmid DNA delivered of gene-encoded antibodies, DMAbs, could therefore have potential value alongside traditional recombinant mAb biologics for prevention and disease control. In prior work, we showed reduction in viral loads in rhesus macaques administered a single anti-Zika virus DMAb [18]. The betacoronavirus family is distinct from flaviviruses, including differences in antibody-associated protective mechanisms. Here, we demonstrate the ability of a two-antibody DMAb cocktail targeting betacoronavirus SARS-CoV-2.

In vivo delivery of two plasmid-encoded DMAbs based on parent mAbs COV2-2130 and COV2-2196 to cynomolgus and rhesus macaques resulted in systemically circulating mAb that functionally blocked RBD binding to the ACE2 receptor and neutralized pseudovirus and authentic SARS-CoV-2. Building on prior work in small animal models [20], in the current manuscript we demonstrate that improved in vivo DMAb expression in both cynomolgus and rhesus macaques can be achieved with a two-plasmid formulation for each mAb, combined with delivery regimen and new side port needle delivery to enhance injection site transfection. This formulation enables DNA-encoded VH and VL (VH1-58/VK3-20 and VK3-15/VK4-1) genes on four separate plasmids to self-assemble in the cell and secrete functional antibody into systemic circulation. DMAb-2196 eWe further show that low serum levels of DMAbs with or without inclusion of half-life extension afford moderate protection against signs of SARS-CoV-2-associated disease in macaques.

Prior studies by Loo et al demonstrated prophylactic efficacy of the tixagevimab plus cilgavimab cocktail (AZD7442, “TM” Fc modification) against SARS-CoV-2 challenge in rhesus and cynomolgus macaques [21]. The DMAb-2130/DMAb-2196 cocktail expressed in vivo at serum levels comparable to the lower 0.4 mg/kg range dose of AZD7742. DMAb levels did not decline in vivo over the 7-day course of SARS-CoV-2 infection, maintaining neutralization capacity over time. This supports the concept that the gene-encoded antibody produced in vivo is not immediately consumed during viral infection. Body weight along with estimated rhesus macaque blood volumes [24] indicate that total circulating anti-SARS-COV-2 DMAb ranged from 573 to 1488 ug per macaque (Supplementary Figure 8). However, there was no correlation between weight, serum DMAb levels, and signs of disease. We show that DMAb-administered rhesus macaques inoculated with 2.8 × 10⁶ TCID50/ml, a log higher than evaluated with precursor mAbs COV2-2130, COV2-2196, and the AZD7442 cocktail [14, 21], affords protective benefits at low serum DMAb levels. Furthermore, there was no evidence of enhanced disease associated with these low serum levels of either the WT or YTE-modified DMAbs. Nevertheless, additional research studies to understand the contributions of antibody effector-engagement following natural infection versus monoclonal antibody delivery and, specifically, with gene-encoded antibodies should continue to be pursued as newer VOC continue to emerge.

In addition to the 10-fold higher challenge dose used in this study, the DMAb cocktails evaluated in this study retained wild type IgG binding to FcγR. The contribution of antibody Fc effector functions to protection against COVID-19 disease and potential impact on disease enhancement was a critical question early during SARS-CoV-2 biologic development [25]. To-date, the evidence of antibody-dependent enhancement associated with COVID-19 disease pathology in individuals with natural infection, vaccination, or receiving mAb biologics is not definitive (reviewed in [26]). The DMAb cocktails encoding a wild type G1m3 hIgG1 Fc alone or with additional half-life modification “YTE” did not lead to enhancement of disease in a rhesus macaque SARS-CoV-2 infection model. DMAb administered animals were protected against perivascular pneumonia and SARS-CoV-2-associated pneumonia in the lungs. The DMAb cocktail encoding YTE afforded a significant reduction in nasal and BAL viral loads, with fewer histopathological signs of disease, compared to control animals. Inclusion of YTE trended towards better protection than the wild type hIgG1 Fc, however this was not statistically significant. YTE has been previously shown to improve translocation into the lungs [27], which is one possible hypothesis for this difference, however in this study we detected DMAb-YTE in the BAL of only 2/6 animals compared with 4/6 animal administered DMAbs without YTE. We previously compared the difference in half-life between WT-Fc and YTE-Fc in mice with both recombinant mAb and DMAb [18]. However, due to the short duration of the rhesus challenge study, it was not possible to observe differences in half-life between DMAb-YTE and DMAb-WT. Differences will likely be more apparent in long-term studies or during treatment approaches. Additional studies assessing formulation, in vivo administration schedules, and contributions of half-life and effector function to protective efficacy will all be informative for clinical translation of gene-encoded antibodies.

Plasmid DNA delivery has been employed in gene therapy and infectious disease vaccine clinical trials, Phases 1-3. Alternatively, mRNA delivery via lipid nanoparticles (mRNA-LNP) has greatly advanced due to COVID-19 vaccines and is being employed in various gene therapy approaches. Although an early concern, genome integration of DNA has been evaluated in various studies indicating potential low-risk [28–30]. Importantly, the US Food and Drug Administration (FDA) has released guidelines outlining specific conditions for human use and safety evaluation to mitigate theoretical concerns for genome alteration [31]. Likewise, there is a low probability for mRNA integration and studies are ongoing to quantify this risk as this technology advances [32]. Both nucleic acid platforms have been employed for in vivo delivery of gene-encoded antibodies (reviewed in [33]). mRNA-LNP encoding a Chikungunya virus (CHIKV) mAb, mRNA-1944, was previously evaluated in macaques [34] and humans [35], with rapid peak expression was observed by 48 h. A single infusion of mRNA-1944 declined by day 25 and repeat intravenous injections were necessary to achieve longer expression. mRNA-encoded mAbs achieve peak expression levels within hours, resulting in several weeks of expression due to rapid mRNA elimination. Plasmid-encoded DMAbs build over days to reach Cmax, however expression continues over a longer time frame in animal studies [17–20, 36]. Plasmid DNA encoded mAb delivery therefore represents a potential approach to increase in vivo antibody half-life through single dose, transient, intramuscular injection.

In previous studies, in vivo delivery of DMAbs expressing different gene-encoded VH/VL family genes (VH3-30/VK3-20 [18]), VH1-8/VL2-28, and (VH4-59/VL3-21 [19]) expressed in macaques with a duration for up to 35 days. In the current study, we show a significant improvement in in vivo expression kinetics with a SARS-CoV-2 DMAb cocktail expressing public clonotype VH1-58/VK3-20 (DMAb-2196) and VK3-15/VK4-1 (DMAb-2130), extending up to 63 days in cynomolgus macaques. The full expression profile is likely longer, however DMAbs induced monkey anti-human antibodies/ADA in cynomolgus and rhesus macaques, likely due to non-self recognition of the human IgG1 DMAb (Supplementary Figure 6). While development of ADA was observed to be more rapid in 2/5 macaques in the two-dose regimen, it is known that human mAbs can be immunogenic in monkeys and poorly correlates with ADA development in people [37, 38]. It is therefore not possible to make conclusions about potential immunogenicity in people based on this data alone. Interestingly, in the current study, NHP #16912 that had no/low levels of ADA expressed DMAb for more than a year. This highlights the potential for long-term expression of DMAb in vivo and additional studies with species-match cynomologus or rhesus antibodies and detailed immune response characterization including ADA and modulation will be informative. In previous mouse studies, ADA development can be prevented through T cell depletion regimens [17] or immunodeficient models such as nude mice [36, 39] to evade immune recognition. Alternatively, we have observed long-term expression of a full mouse DMAb (fully mouse Fab with a mouse IgG2a Fc) in mice with re-administration [40], supporting the concept that a human species-match DMAb will have potentially more durable expression and may be less likely to induce ADA.

Importantly, the SARS-CoV-2 DMAb cocktail delivery in the current study takes advantage of optimized in vivo delivery parameters including side port needles to improve injection site DNA transfection with fewer injections. Here, we showed that a shortened 1 or 2 injection regimen, along with lower total DNA dose (2-4 mg) achieved higher expression levels of the SARS-CoV-2 DMAb cocktail compared to previous work with the anti-Zika DMAb. Given the expression profile observed with this SARS-CoV-2 DMAb cocktail, further studies are of interest to understand the full pharmacokinetics of different VH/VL families in large animal models to support human translation. There are significant efforts to increase the half-life of recombinant monoclonal antibodies with modifications (reviewed in [41]). FDA-approved recombinant anti-respiratory syncytial virus (RSV) mAbs palivizumab and the newer nirsevimab have demonstrated a strong safety profile in infants for at least a year. Furthermore, the SARS-CoV-2 tixagevimab plus cilgavimab mAb cocktail, related to DMAb-2196 and DMAb-2130, displayed over 1 year of serum levels in people [21]. Additional studies with recombinant and gene-encoded antibodies to understand the impact of long-term antibody in serum will be informative, particularly for seasonal infections like influenza and coronaviruses like SARS-CoV-2.

Further strategies to increase gene-encoded antibody expression through VH/VL sequence optimization, plasmid design and delivery, and modulation strategies will be beneficial for clinical translation. We previously performed tissue studies following administration of an anti-Zika DNA encoded antibody as part of IND-enabling data for a first-in-human phase 1 study (clinicaltrials.gov NCT03831503). Accumulation or off-target DNA expression in other tissues other than the injection site was not observed. Additional work studying bioavailability will improve understanding of potential tissue and organ distributions and interactions. These questions will be important for further development of all gene-encoded nucleic acid and viral vector antibody platforms.

Accessibility to vaccines, biologics, and other interventions has been a critical concern during the COVID-19 pandemic. In addition to rapidly declining antibody durability following natural infection and vaccination, along with emerging variants of concern, prophylactic strategies are of critical need for immune compromised and aging populations. DMAbs therefore serve as an additional approach for prophylactic protection. Continued research to understand the long-term expression and biology of gene-encoded antibodies will add a new dimension to nucleic acid delivery of biologics. The data described here provides support for DMAb translation to the clinic and data regarding use of low dose mAb with wild type effector functions for prevention against COVID-19 disease. SARS-CoV-2 antibodies share public clonotypes, therefore this gene-encoded approach can be adapted for in vivo delivery of new mAb candidates with broad activity against SARS-CoV-2 VOC and other emerging infectious diseases.

Methods

Animals

All animal treatments and procedures were performed at Bioqual (Rockville, Maryland) or Rocky Mountain Laboratories (RML) according to approved IACUC protocols, #SP2100122 and # 2021-044-E, respectively.

Cynomolgus macaques – Inovio

Cynomolgus macaques aged 3–4 years were housed and treated at Bioqual. Study group 1 animals (n = 5) were administered on Day 0 and Day 3 with 1 mg each of (DMAb-2196-YTE plus DMAb-2130-YTE) plasmid DNA (total 2 mg DNA/injection day, 4 mg total). Study group 2 (n = 5) were administered on Day 0 with 1 mg each of (DMAb-2196-YTE plus DMAb-2130-YTE). Dosing for both groups consisted of 500 µg per heavy chain and light chain construct for each DMAb (1 mg total DNA), with two injection sites for a total of 2 mg of DNA per animal per injection day (Figure 2(a)). DMAbs were formulated with 135 U/ml Hylenex human recombinant hyaluronidase at the time of administration, and DMAb constructs were delivered by IM injections using a Side Port needle (Inovio Pharmaceuticals). After injection, EP was delivered using the CELLECTRA™ 2000 Pulse Generator, 5P-IM applicator and IM, Side Port Array according to the user manual M12-006918-01_C. The EP parameters were OpBlock 0078: 1.0 Amp, 3 pulses, 52msec per pulse, 1 sec interval between pulses.

Cynomolgus macaques were monitored longitudinally, including serial blood collection for serum and isolation of peripheral blood mononuclear cells from NHP# M16912 to send for genotyping. Genotyping analysis was performed as fee-for-service at the Wisconsin National Primate Research Center Genetics Services (University of Wisconsin–Madison) using a Fluidigim Access Array system.

Rhesus macaques – RML

SARS-CoV-2 studies performed at RML were approved by the Institutional Biosafety Committee (IBC) and performed in the high biocontainment facility at RML, NIAID. Samples were removed by IBC-approved standard operating procedures. Studies followed institutional guidelines for animal use, the guidelines and basic principles in the NIH Guide for the Care and Use of Laboratory Animals, the Animal Welfare Act, USDA and the USPHS Policy on Humane Care and Use of Laboratory Animals. Rhesus macaques were singly housed in adjacent cages to enable social interactions. The animal room was climate-controlled with a fixed 12 h light–dark cycle. Animals were provided commercial chow provided twice daily with vegetables, fruit, and treats to supplement. Water was available ad libitum. Human interaction, manipulanda, toys, video and music were provided for enrichment.

Rhesus macaques ages 10–16 years of age were acquired from the NIH. Prior to immunization, the animals were sedated by IM injection of Ketamine HCl (8-20 mg/kg) or Telazol (2-6 mg/kg, IM). A clinical exam preceded the immunization and animals were continuously monitored until fully awake. Animals were monitored daily to assess any effects of the procedure.

Study group 1 animals (n = 6) served as the control animals and received 2 mg of a control plasmid (1 mg in each leg). Study group 2 (n = 6) were administered on Day −14 with 1 mg each of DMAb-2196-YTE + DMAb-2130-YTE plasmid DNA. Study group 3 (n = 6) were administered on Day −14 with 1 mg each of (DMAb-2196 + DMAb-2130). Dosing for both groups consisted of 500 µg per heavy chain and light chain construct for each DMAb (1 mg total DNA), with two injection sites for a total of 2 mg of DNA per animal (Figure 4(a)). DMAbs were formulated and administered in vivo according to the same procedures as the cynomolgus macaque study.

Rhesus macaques were challenged with 2.8 x10⁶ TCID50 SARS-CoV-2 in the biocontainment facility at RML by four routes including intratracheal (4 mL), intraocular (0.5 mL), intranasal (0.5 ml/nostril), and oral administration (1 mL), as previously described [22, 23]. Virus stock dilutions were prepared in sterile culture medium. Following challenge, animals were scored daily and clinical exams were performed on days −14, −11, −7, 0, 1, 3, 5 and 7. Clinical examinations include a check for physical signs of disease, and collection of the body weight, temperature, heart rate, blood pressure, and blood oxygen levels. Blood was collected for hematological and serum to look at chemistry. Sample collection for virological analysis, detailed below, included blood draws, swabs, and BAL collection. Rhesus macaques were necropsied on Day 7 post challenge and samples (blood, tissue, organs) were collected for analysis. Serum samples from infected rhesus macaques were sent to The Wistar Institute for analysis.

Surface plasmon resonance (SPR) analysis of 2130-based and 2196-based DMAbs variants to human and monkey Fcγ receptors

Binding kinetics of DMAb-2130 and DMAb-2196 to human (hFcγ) and monkey (mFcγ) Fcγ receptor was measured using a Biacore T200 surface plasmon resonance (SPR) system. Briefly, a mouse anti-His antibody was immobilized on 3 and 4 flow cells of a CM5 sensor chip using amine coupling to approximately 4000 response units (RU). The CM5 sensor chip surface was first activated by injecting of the 1:1 mixture of 400 mM N-ethyl-N-(3- dimethylaminopropyl)-carbodiimide hydrochloride and 100 mM N- hydroxysuccinimide (Cytiva) at 10 μl/min for 7 min. The anti-His antibody (Genscript) were diluted at 50 μg/ml in 10 mM sodium acetate, pH 4.5, and injected over the activated surface at 30 μl/min for 8 min. The excessive reactive group was blocked by injecting 1.0 M ethanolamine/HCl, pH 8.5, at 10 μl/min for 7 min. Flow cell 3 was treated similarly using amine coupling reagents but without injection of any Fc gamma receptor and used as reference cell. His-tagged Fc gamma receptors were captured with running buffer containing 10 mM HEPES (pH-7.4), 150 mM NaCl, 3 mM EDTA and 0.05% Tween20 on flow cell 4 to approximately 25 RU. DMAb-2130 and DMAb-2196 with YTE or TM modification (ranged from 0.97-1000 nM) diluted in running buffer and were injected over the immobilized histidine tagged Fc gamma receptor surfaces at 50 μl/min with 3 min association and 5 min dissociation. After each injection, the surface was regenerated by injecting 10 mM glycine at pH 1.5 at 30 μl/min for 30s. The binding kinetics, k_on (1/Ms), k_off (1/s), and K_D (M) were calculated from global fittings using a 1:1 kinetics binding model using Biacore T200 Evaluation software.

Quantitative ELISA for total human IgG

Human IgG expression in NHP serum was quantified by ELISA. 96-well plates (NUNC) were coated with 50 µl/well anti-human H + L antibody cross-adsorbed to monkey at a concentration of 5 µg/mL and incubated overnight at 4°C. The following morning, plates were washed four times in PBS + Tween20 (PBST) wash buffer using plate washer (BioTek) and blotted on paper towels. 200 μL of casein blocking buffer was added to the wells and plates were incubated for 1 h at room temperature. Plates were washed as above. NHP sera were diluted 1:10, 1:30, or 1:90 in casein and 50 µl/well was added in duplicate. Purified mAbs COV2-2130-YTE and COV2-2196-YTE were both run individually as standards. Plates were incubated for 1 h at room temperature. Plates were washed again as above. Goat anti-human IgG cross-adsorbed to monkey conjugated to HRP (Southern Biotech) was diluted 1:8000 in casein and 50 µL added per well. Plates were incubated at room temperature for 30 min. Plates were washed again as above. Plates were developed using 100 µl/well TMB Turbo (ThermoFisher) and the reaction was stopped after 10 min with 100 µl/well 2N sulfuric acid. Plates were read at both 450 and 570 nm.

Quantitative ELISA for individual DMAb clones with epitope specific RBDs

Expression levels of individual DMAb clones was detected using recombinant SARS-CoV-2 RBD proteins K444A for quantification of DMAb-2196-YTE and DMAb-2196 and F486A for quantification of DMAb-2130-YTE and DMAb-2130. 96-well plates (NUNC) were coated with 50 µl/well each respective RBD at a concentration of 5 µg/mL and incubated overnight at 4°C. The following morning, plates were washed four times in PBS + Tween20 (PBST) wash buffer using plate washer (BioTek) and blotted on paper towels. 200 μL of casein blocking buffer was added to the wells and plates were incubated for 1 h at room temperature. Plates were washed as above. NHP sera were diluted 1:10, 1:30, or 1:90 in casein and 50 µl/well was added in duplicate. Purified mAbs COV2-2130-YTE and COV2-2196-YTE were both run individually as standards. Plates were incubated for 1 h at room temperature. Plates were washed again as above. Goat anti-human IgG cross-adsorbed to monkey conjugated to HRP (Southern Biotech) was diluted 1:8000 in casein and 50 µL added per well. Plates were incubated at room temperature for 30 min. Plates were washed again as above. Plates were developed using 100 µl/well TMB Turbo (ThermoFisher) and the reaction was stopped after 10 min with 100 µl/well 2N sulfuric acid. Plates were read at both 450 and 570 nm.

Quantitative ELISA for individual DMAb clones with anti-idiotypes

Expression levels of individual DMAb clones were also detected using anti-idiotypes specific for each antibody Fab. Anti-idiotype M16H5.1 was used to detect DMAb-2196-YTE and DMAb-2196 and anti-idiotype M54-2E9 for detection of DMAb-2130-YTE and DMAb-2130. 96-well plates (NUNC) were coated with 50 µl/well each respective anti-idiotype at a concentration of 5 µg/mL and incubated overnight at 4°C. Plates were washed and detected following the same protocol as the RBD quantification ELISA described above.

Pseudovirus neutralization assay (WT, Beta, Delta, BA.4 VOC SARS-CoV-2 pseudoviruses)

SARS-CoV-2 pseudovirus neutralization assays were established using huCHOAce2 cells (Creative Biolabs) plated in a 96-well format. Cells were resuspended in D10 media (DMEM supplemented with 10% FBS and 1X Penicillin–Streptomycin), plated (10,000 cells/well) and rested overnight in 37°C/5% CO2 conditions. The following day, sera were heat-inactivated and serially diluted in duplicate. Diluted serum samples were incubated with the indicated SARS-CoV-2 pseudoviruses (WT or Delta) for 90 min at RT and then transferred to the rested huCHOAce2 cell monolayer. Plates were incubated in 37°C/5% CO2 conditions for 72 hrs and lysed using the Britelite plus luminescence reporter gene assay system (Perkin Elmer Catalog). RLUs were measured using the Biotek plate reader. Using GraphPad Prism 8, nonlinear regressions were applied to duplicate RLU values for each sample to determine the best fit line. Neutralization titers (ID50) were then calculated, defined as the reciprocal dilution that yielded a 50% reduction in RLU compared to sample control wells; RLUs from cell-only control wells on each plate were subtracted as background prior to analysis. ID50s for each sample were also used along with the corresponding DMAb titer (ng/mL) to calculate inhibitory concentrations (IC50s = DMAb titer / ID50) that reflect the individual molecular potency of a test sample while controlling for expression levels.

SARS-CoV-2 pseudovirus neutralization assay (Omicron BA.1, BA.2, BA2.12 VOC pseudoviruses)

The following Omicron pseudovirus assays were performed at Inovio. Pseudovirus production: SARS-CoV-2 pseudovirus stocks expressing the Spike protein of Omicron BA.1, BA.2, BA.2.12 or BA.4 were produced using HEK 293 T cells transfected with Lipofectamine 3000 (ThermoFisher) using IgE-SARS-CoV-2 S plasmid variants (Genscript) co-transfected with pNL4-3. Luc.R-E- plasmid (NIH AIDS reagent) at a 1:8 ratio. 72 h post transfection, supernatants were collected, steri-filtered (Millipore Sigma), and aliquoted for storage at −80 °C. CHO cells stably expressing ACE2 (ACE2-CHOs) were used as target cells at 7000 cells/well. SARS-CoV-2 pseudovirus were tittered to yield greater than 30 times the cells only control relative luminescence units (RLU) after 72 h of infection. NHP sera were heat inactivated and serially diluted two-folds starting at 1:32 dilution. Sera were incubated with SARS-CoV-2 pseudovirus for 90 min at room temperature. After incubation, sera-pseudovirus mixture was added to ACE2-CHOs and allowed to incubate in a standard incubator (37°C, 5% CO₂) for 72 h. After 72 h, cells were lysed using Britelite Plus Reporter Gene Luciferase (PerkinElmer) and RLU was measured using an automated luminometer. Neutralization titers (ID₅₀) were calculated using GraphPad Prism 8 and defined as the reciprocal serum dilution at which RLU were reduced by 50% compared to RLU in virus control wells after subtraction of background RLU in cell control wells. Serum DMAb IC₅₀ titers were calculated by dividing the 50% inhibitory serum dilution by the serum IgG concentration of each individual sample.

NHP serum IgG concentration

Pooled sera from NHP M16912 (2 mL total) was added to ∼30 mL of Protein A binding buffer and purified on a Protein A column on the AKTA purification system (Cytiva). Fractions 1A5-1A11 were pooled after Protein A purification. The second smaller peak was also pooled for analysis (1B1-1B9) (peak 2). Peak 1 Total IgG yield = 24.914 mg in 700ul at 35.592 mg/ml, Peak 2 total IgG yield = 1.726 mg in 190 ul at 9.084 mg/ml.

Live virus neutralization assay

Live virus studies with SARS-CoV-2 isolates USA-WA1/2020 and hCoV-19/USA/PHC658/2021 were performed in the BSL-3 facility at the Wistar Institute. Both isolates were obtained from BEI Resources: USA-WA1/2020: The following reagent was deposited by the Centers for Disease Control and Prevention and obtained through BEI Resources, NIAID, NIH: SARS-Related Coronavirus 2, Isolate USA-WA1/2020, NR-5228. hCoV-19/USA/PHC658/2021: The following reagent was obtained through BEI Resources, NIAID, NIH: SARS-Related Coronavirus 2, Isolate hCoV-19/USA/PHC658/2021, (lineage B.1.617.2; Delta Variant), NR-55611, contributed by Dr. Richard Webby and Dr. Anami Patel. Vero cells (ATCC; CCL-81) were maintained in DMEM supplemented with 10% fetal bovine serum (FBS). Viral propagation and titration were achieved as previously described. Briefly, the USA-WA1/2020 or hCoV-19/USA/PHC658/2021 virus stock was serially diluted in DMEM with 1% FBS and transferred in replicates of 8 to previously seeded Vero cells and incubated for five days under 37°C/ 5% CO₂ conditions. Individual wells were then scored positive or negative for the presence of cytopathic effect (CPE) by examination under a light microscope. The virus titer (TCID50/mL) was calculated using the Reed-Munch method and the published Microsoft Excel-based calculator. For neutralization assays, Vero cells were seeded in DMEM with 1% FBS at 20,000 cells/well in 96 well flat bottom plates and incubated overnight. Samples were heat-inactivated at 56°C for 30 min and then serially diluted in triplicates. These were incubated for 1 hr at RT with 300 TCID50/mL of virus before the mixture was transferred to previously seeded Vero cells and incubated for 5 days. Viral titer (TCID50) was determined as described above.

SARS-CoV-2 spike binding assay

96-well Flat-Bottom Half-Area plates (Corning) were coated overnight with one of the following SARS-CoV-2 6P VOC trimers: B.1.1.7 (alpha), B.1.351 (Delta), BA.1 (Omicron), BA.5 (Omicron) or BA2.75 (Omicron) at 1ug/mL, followed by blocking with blocking buffer containing 5% milk/1x PBS/0.01% Tween-20 at room temperature for 1 h. NHP Sera from DMAb administered NHPs was serially diluted 3-fold starting at 1:20 with dilution buffer (5% milk/1x PBS/0.01% Tween-20) and added to the plate, followed by a room temperature incubation for 1 h. Anti-YTE Ab (provided by AstraZeneca) was added to the plates at a constant concentration of 1ug/mL diluted with the dilution buffer and incubated at RT for 1 h. The plates were then incubated at room temperature for 1 h with anti-mouse secondary antibody (Cat# A90-231P, Bethyl Laboratories) at 1:10,000 dilution. The plates were further developed by adding TMB substrate (ThermoFisher) followed by quenching with 1M H₂SO₄. The plates were washed four times between each step using the wash buffer (PBS and 0.05% Tween-20). Absorbances at 450 and 570 nm were recorded with using a BioTek plate reader. GraphPad Prism 9 was used to analyze the curves.

ELISA to detect anti-drug antibodies

Monkey anti-drug antibodies against the human IgG were detected by ELISA. 96-well plates (NUNC) were coated with 50 μl/well COV2-2130-YTE and COV2-2196-YTE at a concentration of 5 μg/mL and incubated overnight at 4°C. The following morning, plates were washed four times in PBS + Tween20 (PBST) wash buffer using plate washer (BioTek) and blotted on paper towels. 200 μL of casein blocking buffer was added to the wells and plates were incubated for 1 h at room temperature. Plates were washed as above. NHP sera were diluted two-fold starting at 1:100 in casein and 50 μl/well was added in duplicate. Plates were incubated for 1 h at room temperature. Plates were washed again as above. Mouse anti-monkey IgG HRP (Southern Biotech) was diluted 1:8000 in casein and 50 μL added per well. Plates were incubated at room temperature for 30 min. Plates were washed again as above. Plates were developed using 100 μl/well TMB Turbo (ThermoFisher) and the reaction was stopped after 10 min with 100 µl/well 2N sulfuric acid. Plates were read at both 450 and 570 nm.

Histology and immunohistochemistry

Tissues were fixed in 10% Neutral Buffered Formalin with two changes, for a minimum of 7 days according to IBC-approved SOP. Tissues were processed with a Sakura VIP-6 Tissue Tek, on a 12-hour automated schedule, using a graded series of ethanol, xylene, and PureAffin. Embedded tissues were sectioned at 5 μm and dried overnight at 42°C prior to staining with hematoxylin and eosin. Specific staining was detected using SARS-CoV/SARS-CoV-2 nucleocapsid antibody (Sino Biological cat#40143-MM05) at a 1:1000 dilution. The tissues were processed for immunohistochemistry using the Discovery Ultra automated stainer (Ventana Medical Systems) with a ChromoMap DAB kit (Roche Tissue Diagnostics cat#760–159). Stained slides were analyzed by a board-certified veterinary pathologist blinded to the study groups.

Statistical analysis

Statistical analysis was performed in GraphPad Prism 9.4.1 (La Jolla, CA). Graphs display individual animals or mean values (arithmetic or geometric means, as required). Error bars represent the standard deviation/geometric standard deviation. A paired non-parametric two-way analysis with a Wilcoxon signed rank test was performed for analysis of live virus and pseudovirus neutralization assays. A paired non-parametric two-way ANOVA, corrected for multiple comparisons and a Dunnett's post hoc test were performed for all necropsy gross pathology and tissues scores, as well as viral loads. Data was considered statistically signification with a p value < 0.5.

Author contributions

D.B.W. and A.P. conceived the project and secured funding. A, P. K.R., E.M.P., D.B.W., T.R.F.S, C.S, H.F. designed and supervised studies. A.P., K.R., E.M.P., F.F, S.B, A.J.G, B.S., M.L., J.C., N.C., V.M., B.J.S., D.F., A.R.A., J.L., B.N., P.W.H., S.N.W., E.N.G., A.K., A.G., generated reagents, performed laboratory experiments, and analysis. J.R.F., K.R., D.W.K., M.T.E., T.R.F.S. provided valuable materials and programmatic support/guidance. A.P., K.R., D.B.W. prepared the original manuscript. All authors edited, reviewed, and approved the manuscript.

Acknowledgements

The authors would like to thank The Wistar Institute Core facilities for assistance with SPR. The authors are grateful to staff of the Rocky Mountain Veterinary Branch for animal care, veterinary service and pathology investigation. This work was approved for public release, distribution unlimited. Additional funding includes Award T32 CA09171 (to E.N.G.), the WW Smith Charitable Trust (to D.B.W.), The Jill and Mark Fishman Foundation (to D.B.W.). The views, opinions and/or findings expressed are those of the authors and should not be interpreted as representing the official views or policies of the Department of Defense or the US Government. Funding sources were not involved in the design of this study, collection and analyses of data, decision to submit or preparation of the manuscript.

Disclosure statement

D.B.W. has received grant funding, participates in industry collaborations, has received speaking honoraria, and has received fees for consulting, including serving on scientific review committees and board series. Remuneration received by D.B.W. includes direct payments and stock or stock options. D.B.W. also discloses the following paid associations with commercial partners: GeneOne (consultant), Geneos (advisory board), AstraZeneca (advisory board, speaker), Inovio (BOD, SRA, Stock), Sanofi (advisory board) and BBI (advisory board). A.J.G, B.S., V.M., B.J.S., B.N., A.G., and T.R.F.S. are employees of Inovio Pharmaceuticals and as such receives salary and benefits, including ownership of stock and stock options. J.R.F, K.R, and M.T.E are employees of and hold or may hold stock in AstraZeneca. All other authors declare no completing interests.

Additional information

Funding

This work was funded by the Defense Advanced Research Projects Agency and the Joint Program Executive Office for Chemical, Biological, Radiological and Nuclear Defense (JPEO-CBRN) (Award HR0011-21-9-0001 to D.B.W.). Work performed by the Laboratory of Virology and Rocky Mountain Veterinary Branch is funded by the Intramural Research Program, NIAID, NIH.