Neutralization of SARS-CoV-2 and Intranasal Protection of Mice with a nanoCLAMP Antibody Mimetic

Abstract

Intranasal treatment, combined with vaccination, has the potential to slow mutational evolution of viruses by reducing transmission and replication. Here, we illustrate the development of a SARS-CoV-2 receptor-binding domain (RBD) nanoCLAMP and demonstrate its potential as an intranasally administered therapeutic. A multi-epitope nanoCLAMP was made by fusing a pM affinity single-domain nanoCLAMP (P2710) to alternate epitope-binding nanoCLAMP, P2609. The resulting multimerized nanoCLAMP P2712 had sub-pM affinity for the Wuhan and South African (B.1.351) RBD (K_D < 1 pM) and decreasing affinity for the Delta (B.1.617.2) and Omicron (B.1.1.529) variants (86 pM and 19.7 nM, respectively). P2712 potently inhibited the ACE2:RBD interaction, suggesting its utility as a therapeutic. With an IC₅₀ = 0.4 ± 0.1 nM obtained from neutralization experiments using pseudoviral particles, nanoCLAMP P2712 protected K18-hACE2 mice from SARS-CoV-2 infection, reduced viral loads in the lungs and brains, and reduced associated upregulation of inflammatory cytokines and chemokines. Together, our findings warrant further investigation into the development of nanoCLAMPs as effective intranasally delivered COVID-19 therapeutics.

While vaccines remain the most important weapon in the fight against the COVID-19 pandemic, there is an unmet need for the development of intranasally administered pre- and post-exposure prophylactic treatments to prevent infection.^1,2 Along with vaccination, therapeutic monoclonal antibodies were one of the initial envisioned strategies in treating SARS-CoV-2 infection. Four intravenous-delivered anti-SARS-CoV-2 mAb products (bamlanivimab plus etesevimab, casirivimab plus imdevimab, sotrovimab, and bebtelovimab) and one antibody combination (tixagevimab/cilgavimab) for intramuscular use received Emergency Use Authorizations for the treatment of outpatients with mild to moderate COVID-19 and were based on the reduction of viral load.³⁻⁶ These systemic neutralizing antibodies mainly target the receptor-binding domain (RBD) of the trimeric Spike protein⁷ that decorates the surface of the coronavirus and plays a pivotal role during viral entry by first binding to Ace2 displayed on epithelial cells.⁸ Since the main port of entry of SARS-CoV-2 into the body is the ciliated epithelium lining the nose with one of the highest concentrations of Ace2 receptors of any tissue in the body, the deployment of countermeasures to that area has the potential to block the initial attachment of the virus to Ace2 receptors, pre-empting infection. Importantly, this strategy does not allow the virus to replicate and mutate in the host, as it does even in many vaccinated individuals whose immune systems effectively eliminate the virus. Since intramuscular injection of the current vaccines induces mainly IgG protection in the lower respiratory tract, there is new interest in nasal or mucosal vaccinations to promote IgA antiviral activity at the point of entry, which is hypothesized to reduce replication comparatively.⁹ In combination with a nasal vaccine that produces IgA protection in the nasal passage, nasal treatments that block the initial attachment of the virus to host cells could drastically reduce transmission, replication, and, by extension, mutation.

Several non-protein-based nasal treatments for COVID-19 are under development and include povidone iodine, nitric oxide, ethyl lauroyl arginate hydrochloride, and astodrimer sodium.¹ Protein-based biologics employed as nasal treatments include an engineered IgM with prophylactic efficacy when delivered intranasally in rodents.¹⁰ Monoclonal antibodies are the most widespread affinity protein therapeutics due to their high specificity, affinity, and safety, with RBD targeting neutralizing monoclonal antibodies in development since the start of the pandemic.^6,7,11−13 The use of therapeutic anti-spike monoclonal IgG, while effective, is challenging due to possible dramatic change of activity of anti-SARS-CoV-2 mAbs against specific variants and subvariants. Encouragingly, however, a broadly neutralizing antibody (35B5) was shown recently to neutralize all known variants of concern when administered intranasally.¹³ While monoclonal antibodies remain successful protein therapeutics, they are expensive to produce and are prone to degradation and aggregation in extreme conditions.¹⁴ Furthermore, their development requires immunization of an animal, or they must be isolated from convalescent patients, adding valuable time to their development. The cost for a single dose of intranasal mAb, which could possibly require several milligrams of antibody, is likely to be prohibitive for mass distribution, especially in developing countries. Therefore, low-cost antibody mimetics are emerging as an attractive alternative as nasal prophylactics.

Nanobodies and some antibody mimetics are being developed as alternatives to traditional antibodies for nasal prophylactics.¹⁵⁻²¹ These single-domain binding proteins can be produced cheaply in microbial hosts, have better engineering and stability profiles, and have the potential to address the high cost associated with mAb nasal prophylactics. A purely synthetic, potent miniprotein was developed using structural design alone to produce picomolar binders to the spike RBD.¹⁷ The well-characterized ankyrin repeat domain-based Darpins are also being developed as intranasal prophylactics.²¹ Wu et al.¹⁹ reported a potent bispecific nanobody that protects mice against SARS-CoV-2 infection via intranasal administration. The nanobody NIH-CoVnb-112 was effectively nebulized and demonstrated effective reductions in viral burden and lung pathology in a Syrian hamster model of COVID-19. More recently, Wu et al.¹⁶ demonstrated short-term instantaneous prophylaxis and treatment in hAce2 mice infected with the delta variant of SARS-CoV-2 with nanobody NB-22-Fc. Encouragingly, a single dose of intranasal Nb22 protected mice even when administered 7 days prior to infection with the Delta variant.

While nanobodies dominate this field, their development typically originates in immunized Camelidae, followed by cloning and development of the binding domain (the nanobody), adding time to their development. Furthermore, many camelid-derived nanobodies contain disulfide bonds, making them susceptible to reducing environments and forcing them to be produced in the periplasm ofEscherichia colior in secretion expression systems, and precluding their engineering with sulfhydryl reactive probes and reagents. Recently, a new class of antibody mimetics called nanoCLAMPs (nano-CLostridial Antibody Mimetic Proteins) has been described,²² with attractive properties for the development of an intranasal treatment. These 15 kDa protein binders can be screened from a synthetic, naïve phage display library of 1 × 10¹⁰ variants for high-specificity, low-nM-affinity binders to targets in 6 weeks and produced and purified cheaply from the cytosol ofE. coli with yields above 200 mg L^–1. They are naturally devoid of cysteines, are easily refolded following chemical denaturation with 6 M GuHCl, 0.1 N NaOH, or DMF when conjugated to solid support, and have high thermal stability (T_m > 65 °C), highlighting their stability in harsh environments. Importantly, they are easily multimerized, making them well-suited countermeasures to quickly mutating targets, since escape mutants must simultaneously mutate two epitopes.

Here, we describe the development of nanoCLAMPs with low nanomolar affinity for distinct epitopes on the Wuhan RBD, the affinity maturation of one of them, and finally their fusion into a multi-epitope-binding, sub-picomolar affinity nanoCLAMP, P2712. We characterized this protein’s ability to inhibit human Ace2:RBD binding in vitro and further investigated this activity by demonstrating the neutralization potential of P2712 using pseudovirus bearing the SARS-CoV-2 spike trimer. Finally, we demonstrated for the first time the efficacy of nanoCLAMPs as potential intranasally delivered therapeutics using a SARS-CoV-2-susceptible K18-hACE2 mouse model. Reduced viral load in lungs and brain and reduced upregulation of inflammatory cytokines and chemokines underline the potential of nanoCLAMPs as effective intranasally delivered COVID-19 therapeutics. An outline of this technology to cover other mutations is provided.

Results

High-Affinity SARS-CoV-2 RBD-Binding Single-Domain Antibody Mimetic nanoCLAMP P2710

With the aim to obtain nanoCLAMPs that potently neutralize SARS-CoV-2, the synthetic nanoCLAMP phage display library NL-21 (1 × 10¹⁰ variants) was screened for binders to the SARS-CoV-2 Wuhan strain (Wuhan-Hu-1) spike protein receptor binding domain (RBD). After 3 rounds of phage panning, we screened 96 random clones by semELISA and identified 41 unique nanoCLAMPs specific to recombinant RBD and ranked them according to off-rate using biolayer interferometry (BLI) with recombinant RBD (data not shown). Clones with poor monodispersity were carefully excluded, as soluble aggregates artificially increased the apparent affinity due to increased avidity. For this reason, we purified the monodisperse monomer of each nanoCLAMP by size exclusion chromatography, and tested the monodispersity prior to kinetic and functional analyses (Figure S1). The nanoCLAMP P2632 was determined to be the lead candidate with a K_D = 3.3 nM for recombinant Wuhan-RBD determined by BLI on a streptavidin sensor chip to which biotinylated-Wuhan-RBD was immobilized (Figure S2). NanoCLAMP P2632 was affinity matured by constructing and screening a saturation mutagenesis library for three additional rounds of phage panning, resulting in nanoCLAMP P2710, with K_D ≤ 1 pM for the Wuhan-RBD (Figure S1).

The low K_D is notably due to the nondetectable dissociation rate, k_off, measured by BLI, where single-digit pM and fM affinities cannot be reliably determined. nanoCLAMP P2710 has one of the highest affinities for RBD of the reported single-domain antibody mimetics (without multimerizing), with comparable affinity to some of the nanobodies from immunized llamas reported by Güttler et al.²³ While the affinity of nanoCLAMP P2710 to the U.K. variant (Alpha-RBD, B.1.1.7) remains comparable to that of the Wuhan-RBD, affinities toward the S. African RBD (Beta-RBD, B.1.351), Delta-RBD (B.1.617.2), and Omicron-RBD (B1.1.529) are markedly lower, indicating that P2710 affinity was affected by mutations in the RBD.

High-Affinity Multimeric RBD-Binding Single-Domain Antibody Mimetic nanoCLAMP P2712

Escape mutants pose a challenge for therapeutic approaches dependent on inhibiting RBD:Ace2 interaction. Given the prevalence of circulating RBD mutations and their impact on affinity, we hypothesized that fusing two or more nanoCLAMPs targeting different epitopes on the RBD would result in a higher affinity binder and reduce the effects of single RBD mutations on affinity. The affinity of nanoCLAMP P2609 for Wuhan-RBD was in the nanomolar (K_D = 7.4 nM) range (Figure 1) and had similar affinity and binding kinetics for B1.1.7, B1.351, and 1.617.2, suggesting that this binder contacted regions not perturbed by mutations. However, nanoCLAMP P2609 did not bind to the RBD of B1.1.529 at the concentrations tested. To test the hypothesis that P2710 and P2609 bind to different epitopes, epitope binning was performed using BLI. Immobilization of biotinylated nanoCLAMP P2710 to a streptavidin-modified BLI sensor resulted in strong binding to Wuhan-RBD, as expected (Figure 2A). Addition of soluble nanoCLAMP P2710 resulted in no additional binding, indicating that the epitope for P2710 was occupied, as expected; however, the addition of soluble P2609 resulted in a clear increase in signal, indicating that P2609 indeed binds to a different epitope than P2710, as both nanoCLAMPs could bind RBD at the same time (Figure 2A). Based on these data, P2710 was fused with P2609 via a 13-residue GS linker, resulting in nanoCLAMP P2712 (6His-P2710-linker-P2609). P2712 expressed at high levels as expected and was purified to homogeneity (Figure S3). The thermal stability of the individual nanoCLAMPs and the P2712 fusion was tested using differential scanning fluorimetry (Figure S4). The thermal stability of P2710 and P2609 (T_m = 72 and 76 °C, respectively) was only slightly diminished in the fusion, with a T_m = 67 °C.

Figure 1.

Binding affinities of nanoCLAMPs to SARS-CoV-2 RBD from VOCs. BLI sensorgrams showing binding of nanoCLAMPs P2710, P2609, and P2712 (fusion of P2609 and P2710) to recombinant SARS-CoV-2 RBDs from different variants of concern. Black lines depict binding data and red lines depict 1:1 binding model fit, except in cases of P2710 vs. Beta-RBD, Delta-RBD, and Omicron-RBD, which was modeled with heterogeneous ligand fit. Analyte concentrations were 10, 3.33, 1.11, and 0.37 nM for all except: P2712 vs. Omicron-RBD, P2710 vs. Delta-RBD, and P2609 vs. Delta-RBD where 30, 10, 3.33, and 1.11 nM were used, and P2710 vs. Omicron-RBD with 60, 30, 15, and 7.5 nM.

Figure 2.

Epitope binning of P2710 and P2609 to develop P2712 and Ace2:RBD inhibition by P2712. (A) BLI sensorgram of a streptavidin-coated sensor showing loading with biotinylated P2710 (red) or Buffer (black), baseline establishment, binding of Wuhan-RBD, and competitive binding of P2609 (green), P2710 (blue, dotted), or buffer (black). (B) Protein A-coated wells were coated with Wuhan RBD-Fc and incubated with biotinylated ACE2 that had been incubated with serial dilutions of nanoCLAMPs. Remaining ACE2 was detected with streptavidin-HRP and developed with a TMB-ELISA substrate. Data are represented as mean ± SD, with n = 3 biological replicates. IC₅₀ values were calculated using SoftMax Pro nonlinear regression 4 parameter fit.

We next tested whether the dual epitope binder P2712 would bind with high affinity to all the variants in this study. P2712 displayed low-pM binding for Wuhan, UK, and S. African variants, and a K_D of 85 pM and 17 nM for the Delta and Omicron variants, respectively (Figure 1). Since P2609 had constant affinity for the Wuhan, U.K., and S. African variants, we anticipated it would retain its affinity for the Delta and Omicron variants. Interestingly, while it retained similar affinity for the Delta variant, its binding was nearly abolished by the Omicron variant (Figure 1). We were surprised to see that P2710 retained some binding to the highly mutated Omicron variant, imparting at least nM affinity to the P2712 fusion for the Omicron variant. In summary, P2712 retains P2710 interaction affinity with B.1.1.529 and furthermore shows increased affinity to B1.351 and 1.617.2 variants (compared to both P2710 and P2609 alone) while affinities for Wuhan and B1.1.7 remained comparable to that of nanoCLAMP P2710. Based on these results, P2712 was chosen for further characterization.

Although the mutational drift of SARS-CoV-2 was causing a decline in the affinity of P2712 for the RBD, we tested whether nanoCLAMPs might be good candidates for intranasal therapeutics or prophylactics, reasoning that demonstration of the proof of principle could enable future therapeutic developments with different nanoCLAMPs, perhaps to more conserved regions of the Spike protein. With this aim in mind, we began testing the neutralization potential of P2712.

Inhibition of ACE-2:RBD Interaction by nanoCLAMP P2712

Intranasal delivery of prophylactics and therapeutics against respiratory viruses is an attractive option, because it delivers the agent directly to the viral point of entry and potentially blocks the virus from binding to and entering epithelial cells. It also has the potential to limit systemic exposure, which could drastically reduce off-target effects. We hypothesized that the small size and high thermal, proteolytic, and chemical stability of nanoCLAMPs might make them good candidates for intranasally formulated therapeutics and prophylactics. As a first step to evaluate the therapeutic potential of nanoCLAMP P2712 and take advantage of its high affinity for the RBD, we tested in a competitive ELISA whether P2712 could inhibit the binding of the Wuhan-RBD domain of the SARS-CoV-2 spike protein to the extracellular domain of the recombinant human ACE2 receptor (Figure 2B). A non-RBD-binding nanoCLAMP, P2588, was included as a negative control and showed no inhibition of the interaction at any concentration tested, as expected. Although P2632 and P2609 both have similar affinity for Wuhan-RBD, it is clear that P2632, and its affinity-matured derivative P2710, inhibits the RBD:ACE2 interaction more potently than P2609, with IC₅₀ values of 12 and 8 nM, respectively. The fact that P2609 poorly inhibits RBD:ACE2 binding suggested that this binder might interact outside the region affected by the mutations in the SARS-CoV-2 variants of concern. P2712, the fusion of P2710 and P2609, was the most potent inhibitor of the interaction, with an IC₅₀ of 4 nM, and was further investigated for its neutralization potential.

Cytotoxicity of nanoCLAMP 2712 to Vero E6 Cells

The cell toxicity of nanoCLAMP P2712 was established after 72 h of incubation on Vero E6 cells, one of the most widely used cell lines for the proliferation and isolation of severe acute respiratory syndrome coronaviruses, as they contain an abundance of ACE2 receptors.²⁴ This receptor is expressed in lung epithelial cells as well as endothelial cells lining the nasal tract, arteries, veins, capillaries, small intestine, testes, renal tissue, and cardiovascular tissue and is required for SARS-CoV-2 virus entry into the host cell. Infection of the host cell also relies on priming of the SARS-CoV-2 spike protein by the transmembrane serine protease (TMPRSS2).²⁵ The cytotoxicity was evaluated using cell viability assessment by the resazurin assay, based on the conversion of nonfluorescent dye to a fluorescent molecule by mitochondrial and cytoplasmatic enzymes. nanoCLAMP P2712 showed some toxicity to Vero E6 cells at higher concentrations when incubated for 72 h (Figure 3A). A comparable dose-dependent reduction of viability of Vero E6 cells was observed with the single epitope-binding nanoCLAMPs P2609 and 2710 (Figure S5). The cytotoxicity effect was different on normal human bronchial epithelium (NHBE) cells when evaluated after 72h (Figure 3B), and nanoCLAMP P2712 proved to be nontoxic up to 100 nM nanoCLAMP P2712. These results indicate indeed a cell-dependent cytotoxicity when 72 h was chosen as end point. The occurrence of cytotoxicity in the nanomolar range could be an issue of concern and an issue to be addressed for future research of clinical translation. However, as shown in Figure 4, P2712 neutralizes pseudovirus with an IC₅₀ of 0.31 nM (0.01 μg mL^–1), a range in which cell viability was near 100%. It was concluded that P2712 was sufficiently nontoxic at relevant concentrations to proceed.

Figure 3.

Cell viability of nanoCLAMP 2712 using Vero E6 and NHBE cells: Vero E6 cells (A) or normal human bronchial epithelium (NHBE) cells (B) were grown in 96-well plates (15 × 10³ cells/well) with 100 μL of culture medium containing increasing concentration of P2712 for 72 h. Bars represent the viability of cells treated with P2712 at different concentrations, expressed as percentage of viability, and are the mean value of three independent experiments with each treatment performed in triplicate. Negative control: 0 nM P2712.

Figure 4.

Neutralization and inhibition potency of nanoCLAMPs. (A) Neutralization of SARS-CoV-2 pseudotyped lentivirus infection of ACE2-expressing HEK293T cells with nanoCLAMPs P2632 and P2609, which bind separate epitopes; the affinity matured version of P2632 (P2710) and the dual-epitope-binding fusion of P2710 and P2609 (P2712). (B) Neutralization of U.K. (B.1.1.7), S. African (B.1.351) and Omicron (B1.1.529) PSV by P2712 in comparison to Wuhan variant (similar to Figure 4a, black line).

nanoCLAMP P2712 Potently Neutralizes SARS-CoV-2 Pseudovirus

To test whether the nanoCLAMPs would recognize the RBD in the context of the entire envelope Spike trimer, we tested their performance in a pseudovirus (PSV) neutralization assay utilizing lentivirus pseudotyped with a Wuhan SARS-CoV-2 Spike trimer (Figure 4A). P2609 did not neutralize the pseudovirus in the concentration range tested in agreement with Figure 2B, in which P2609 only weakly inhibited the Ace2:RBD interaction. Full dose–response neutralization curves were generated for P2632, P2710, and P2712, with improving neutralization IC₅₀s of 5.4, 1.3, and 0.33 nM, respectively. Encouraged by these results, we tested P2712’s ability to neutralize other pseudovirus variants (Figure 4B). The B1.1.7 variant was neutralized with an IC₅₀ of 0.21 nM in line with the Wuhan variant. P2712 only neutralized B1.351 and B1.1.529 pseudoviruses at higher concentrations.

Intranasal Administration of P2712 Is Highly Efficacious as a Therapeutic Treatment of K18-hACE2 Transgenic Mice Infected with Wuhan-Hu-1 SARS-CoV-2

In a proof-of-principle study, we next assessed the efficacy of nanoCLAMP P2712 on an authentic SARS-CoV-2 challenge model using K18-hACE2 transgenic mice. Ten mice per group (8-to-10-week-old female) were administered a single, daily 30 μL intranasal dose of 2.25 mg kg^–1 P2712 or an equimolar dose of negative control nanoCLAMP P2570 (corresponding to 1.12 mg kg^–1 with a 30 μL dose) for 3 days, 6 h before infection to day 2 after infection (Figure 5A). The mice were intranasally challenged with 2.5 × 10⁴ plaque-forming units (PFU) of SARS-CoV-2 per mouse, and changes of body weight (Figures 5B), survival (Figure 5C), and clinical score (Figure 5D) were monitored until the study end point at day 14. At day 5, four mice were sacrificed for tissue analysis.

Figure 5.

SARS-CoV-2 infection in nanoCLAMP-treated K18-hACE2 mice with viral burden and viral titer. (A) 8-to-10-week-old female K18-hACE2 mice were infected with SARS-CoV-2 and treated with PBS, mock (P2570) or therapeutic (P2712) nanoCLAMPs as indicated on depicted scheme. (B) Weight change as a function of time and represented as mean per therapeutic group. (C) Survival follow-up (Mantel–Cox test) was daily evaluated until 14 days post infection. (D) Clinical score is monitored and represented daily until end point of experiment at 14 days post infection as mean per therapeutic group (means from 10 individuals for each group treated with 2.25 mg kg^–1 nanoCLAMP P2712 or molar equivalent of P2570. (E–H) SARS-CoV-2 viral burden and titer (qRT-PCR, TCID50) in infected and treated K18-hACE2 animals by PBS, mock compound (P2570), or therapeutic compound (P2712) at 5 days post infection. (E) Viral RNA level detected by RT-qPCR in lungs. (F) Viral charge detected by TCID50 in lungs of SARS-CoV-2-infected K18-hACE2. (G) Viral RNA level detected by RT-qPCR in the brain. (H) Viral charge detected by TCID50 in brains of SARS-CoV-2-infected K18-hACE2 mice (each plotted dot represents the viral burden and titer at day 5 post infection for an individual animal, and bars represent median and standard deviation). The data analysis was carried out using GraphPad. P values were obtained by two-tailed unpaired Student’s t test (*P < 0.05; **P < 0.05).

At day 8 post infection (dpi) (before any mice had died), mice treated with the negative control, nanoCLAMP P2570, lost over 20% of their initial body weight and showed increasing clinical score (conjunctivitis, lethargy, hunched appearance, respiratory difficulties). All these mice succumbed to infection or were humanely euthanized due to exceeding humane end points previously defined. In contrast, the nanoCLAMP P2712 group showed reduced body weight loss (around 5–10%) and limited clinical score compared to nanoCLAMP P2570. Only 40% of the mice in the P2712 group succumbed to infection or reached humane end point and were humanely euthanized. These mice showed over 20% body weight loss and a high global clinical score. The remaining 60% showed no major changes in body weight and low grade of clinical score, suggesting a protection effect of P2712 on associated pathology.

After viral challenge with an infectious dose of 2.5 × 10⁴ PFU, the level of viral RNA present in lungs and brains was assessed by RT-qPCR at day 5 post infection. Viral charge was significantly decreased in lung and brain homogenates of mice treated with the P2712 antibody compared to controls (Figure 5E,G). In parallel, viral charge in lungs and brains was assessed by TCID₅₀ measurements (Figure 5F,H), with P2712-treated tissue titers dropping from 10³ to 100 PFU mg^–1 in the lung and 10⁵ to 100 PFU in the brain, compared to controls. Overall, the viral charge in K18-hACE2 mouse lungs was attenuated in those having received the P2712 treatment. Based on qRT-PCR data, the genome equivalent per ng of RNA follows the same tendency. These results directly support the efficacy of the P2712 treatment in preventing viral replication along the course of infection in the Sars-CoV-2/K18-hACE2 model.

Extensive changes in cytokine profiles are documented to be associated with COVID-19 disease progression. The protein levels of 20 different cytokines and chemokines in lungs, brain, and sera were assessed using a multiplex assay (Figure 6A). At day 5 post infection, the cytokine response was nearly identical between mice treated with PBS and mock nanoCLAMP P2570, with high levels of pro-inflammatory cytokines and growth factors such as G-CSF, IL-1B, IP-10, M-CSF, MIG, and MIP-1a in lungs and brains. The therapeutic nanoCLAMP P2712 reduced secretion of pro-inflammatory cytokines in the lung and brain, coinciding with the protective effect observed with the clinical score. In plasma samples, this difference is less pronounced and all groups secreted lower levels of pro-inflammatory cytokines. However, compared with serum of PBS-treated infected K18-hACE2 animals, a significant reduction of IL-6 production after nanoCLAMP P2712 treatment was observed. These control data are consistent with cytokine profiling of serum from human patients with COVID-19 and transcriptional analysis of the BAL fluid of human patients, which showed that elevated levels of pro-inflammatory cytokines correlate with disease severity.

Figure 6.

Heat maps of cytokine levels and histopathological analysis of SARS-CoV-2 infection in K18-hACE2-treated mice after day 5 post infection. (A) Systemic and local cytokine response to SARS-CoV-2 infection in the lung, brain, and serum of K18-hACE mice treated with PBS, mock compound (nanoCLAMP P2570), or therapeutic compound (nanoCLAMP P2712) at 5 days post infection. For each cytokine, the fold change was calculated compared with mock-infected animals and the log2 [fold change] was plotted in the corresponding heat map. (B) Hematoxylin and eosin (HES) staining of lung sections from infected and treated K18-hACE2 mice following intranasal mock (P2570) or therapeutic nanoCLAMP (P2712) treatment. (C) Lung density analysis of the HES-stained histological section for representative samples. The measure of lung tissue density in response to SARS-CoV-2 infection shows no significant difference of lung inflammatory environment remodeling between nanoCLAMP P2570 and P2712 treatment. The data analysis was carried out using GraphPad. P values were analyzed by two-tailed unpaired Student’s t test. HES staining (hematoxylin, eosin). Nucleus: purple, cytoplasm: pink, muscles: pink-orange, collagen: yellow.

Transcriptional analysis of the BAL fluid of human patients with Covid-19 showed that elevated levels of pro-inflammatory cytokines correlated with disease severity.²⁶⁻²⁸ Overall, our data suggest that, in the context of the inflammatory response to SARS-CoV-2 in the lungs of K18-hACE2 mice, many cytokines and chemokines are induced, with many having down-regulation patterns of expression after nanoCLAMP P2712 treatment supporting its protective effect in resolution of inflammatory responses.

Hematoxylin and eosin (HES) staining is the most widely used stain in histology and medical diagnosis. This staining method involves application of hemalum and eosin that color cell nuclei blue and cytoplasm pink to red. The overall patterns of coloration from the stain show the general distribution of cells and provides a histological overview of a tissue structure. Analysis of HES lung sections from K18-hACE2 mice infected and treated with nanoCLAMP (mock and treatment) showed an inflammatory process at 5 dpi (Figure 6B). The lung density analysis on the HES-stained histological section shows peri-bronchial and peri-vascular inflammation accompanied by inflammatory foci in alveolar tissues characterized by local inflammatory cell accumulation and a thickening of the alveolar walls. The measure of lung tissue density in response to SARS-CoV-2 infection showed no decrease or resolution of remodeling between mice treated with nanoCLAMP P2570 and P2712. This finding corroborates with sustained production of pro-inflammatory cytokines and chemokines at 5 days post infection in the lung, even though cytokines were down-regulated in P2712-treated mice, compared with K18-hACE2-infected PBS and mock-treated animals. In the future, the immune cell infiltrates should be analyzed in order to better understand the immune response progression and resolution in the lung upon nanoCLAMP treatment in a dose–response manner.

Our results suggest that neurotropic invasion of SARS-CoV-2 through the olfactory route appears to be relatively plausible, in line with other published works.²⁹⁻³³ The other possibility is for a hematogenous route of viral brain entry, crossing the blood–brain barrier, which can effectively explain neurovascular dysfunctions. The neurological symptoms in COVID-19 may also arise as a consequence of a “cytokine storm”-induced neuroinflammation, or a regional or systemic pathology, associated with the disease. The occurrence of multiple routes of SARS-CoV-2 entry into the brain or multiple mechanisms involved in the pathogenesis of the neurological symptoms cannot be ruled out.³⁴ Currently, direct evidence of SARS-CoV-2-specific neuropathogenicity is limited.³⁵ We observed that the viral content in the brain was significantly reduced after nanoCLAMP administration. Future studies should include whole-body animal images after drug administration to illustrate spatiotemporal drug and viral distributions.

The route of inoculation has been shown to influence the immune response and disease outcome upon viral infection.³⁶ These studies and observations are important to understand whether we should further use the drug as an aerosol or a nasal spray.³⁸ Due to the fact that nasal sprays are generally expected to act locally in the nasal cavity, they may act directly on disease pathogenesis and local inflammation preventing viral spread and replication and furthermore viral transmission. One study directly compared intranasal instillation with aerosol inoculation of mice with influenza virus and concluded that aerosol delivery resulted in a more robust infection, pulmonary cell infiltration and inflammation, and morbidity. K18-hACE2 transgenic mice exposed to aerosolized SARS-CoV-2 developed robust respiratory infection, anosmia, and signs of airway obstruction; however, mice infected intranasally did not experience fatal neuroinvasion.³⁹ Moreover, when compared to intranasal inoculation, aerosol exposure resulted in a more severe lung pathology, inflammation, and fibrin deposition.³⁷

Conclusions

This work has demonstrated the potential of neutralizing nanoCLAMPs as candidates for intranasally delivered SARS-CoV-2 prophylactics and therapeutics. Overall, our data suggest that intranasally delivered nanoCLAMP P2712 treatment can protect K18-hACE2 mice against SARS-CoV-2 infection in vivo, reduce viral titers in brain and lung, and also mitigate the inflammatory response in these tissues. While these preliminary results are encouraging, the potential toxicity and immunogenicity of intranasally administered nanoCLAMP must be explored in more detail in future animal studies. Furthermore, pharmacokinetic and dose–response analyses are warranted to better characterize the distribution and mode of action of the nanoCLAMP in vivo.

This proof-of-concept work suggests a path forward for identifying prophylactic and/or therapeutic intranasally delivered nanoCLAMP candidates against future SARS-CoV-2 variants by combining individual nanoCLAMPs to more conserved regions of the spike. The rapid generation, facile and low-cost production, and high stability of nanoCLAMPs should make them a useful addition to the growing arsenal of intranasally administered treatments to prevent initial viral host infection.

Indeed, as the pandemic wore on, increasing numbers of mutations in the RBD were observed, resulting in a decline in P2712’s affinity for the RBD. Because P2710 retained partial affinity to the Omicron-RBD (Figure 1), we reasoned that fusing two nanoCLAMP P2710s (to form P2715) might result in a nanoCLAMP with enhanced neutralization potential for the Omicron variant. As expected, we observed potent neutralization of the Omicron pseudovirus (Figure S6) for P2715, with an IC₅₀ of 7.8 nM. This supports the hypothesis that nanoCLAMPs to various epitopes on the target can be judiciously combined to develop intranasal therapeutics and prophylactics against future variants of concern.

Materials and Methods

Identification of nanoCLAMPs Specific to SARS-CoV-2 Spike RBD

Nectagen’s nanoCLAMP phage display library NL-21 was panned as described²² against SARS CoV 2 Spike protein RBD (Sino Biologicals, S1RBD-Fc, cat#40592-VO2H) for rounds 1 and 2, and then biotinylated SARS CoV 2 S protein RBD, His, Avitag (Acro Biosystems, cat# SPD-C82E9), for round 3 (both Wuhan strain). The nanoCLAMPs displayed in this library consist of the nanoCLAMP scaffold with three randomized binding loops, V, W, and Z, which contain three, seven, and five random residues, respectively. Loop V residues are composed of all amino acids except Cys, Glu, Lys, Met, Gln, and Trp. Loops W and Z are composed of all amino acids except Cys. NL-21 contains over 1.1E10 variants as assessed by the electroporation efficiency of the plasmid library.

For the first round of panning, 2.7 L of 2xYT medium with 2% glucose and 100 mg mL^–1 carbenicillin (2xYT/Glu/CB) was inoculated with 3.6 mL of the NL-21 library glycerol stock (OD₆₀₀ = 75), to an OD₆₀₀ of approximately 0.1, for over 5× coverage of the library size, and grown at 37 °C, 250 rpm until the OD₆₀₀ reached 0.52. The library was infected by adding helper phage VCSM13 (Stratagene, Cat#200251) to 750 mL of culture at an MOI of 20 phage cell^–1 and incubating at 37 °C, 100 rpm for 30 min, and then 250 rpm for an additional 30 min. The cells were pelleted at 7500 × g for 10 min, and the media discarded. The cells were resuspended in 1.2 L 2xYT/CB, 70 μg mL^–1 kanamycin (KAN), and incubated 15 h at 30 °C, 250 rpm. The cells were combined, and 100 mL was centrifuged at 10,000 × g for 10 min. The phage containing supernatant was transferred to clean tubes and precipitated by adding 37.5 mL of 5X PEG/NaCl (20% polyethylene glycol 6000/2.5 M NaCl) and incubated on ice for 25 min. The phage was pelleted at 13,000 × g for 25 min and the supernatant discarded. The phage was resuspended in 10 mL of 20 mM NaH₂PO₄, 150 mM NaCl, pH 7.4 (PBS), and then centrifuged at 15,000 × g for 15 min to remove insoluble material. The phage was precipitated a second time by adding one-fourth volume of 5× PEG/NaCl, incubated on ice for 5 min, and pelleted at 13,000 × g, for 10 min at 4 °C. The phage pellet was resuspended in 3 mL of PBS and quantified by absorbance at 268 nm (A₂₆₈ = 1 for a solution of 5 × 10¹² phage mL^–1). Two sets of 100 μL of either Protein A Magnetic Beads (Thermo PI88846) (rounds 1 and 2) or Dynabeads MyOne Streptavidin T1 (Thermo Fisher Scientific) (round 3) magnetic beads slurry were washed 2 × 1 mL with PBS-T (PBS with 0.05% Tween 20), applying magnet in between washes to remove the supernatant, and then blocked in 1 mL of 2% dry milk solution in PBS with 0.05% Tween 20 (2% M-PBS-T) for 1 h, rotating at room temperature. To preclear the phage against beads alone, 1 mL of phage was prepared at a concentration of 2 × 10¹³ phage mL^–1 in 2% M-PBS-T, the block removed from the first set of beads, and the phage added to the beads and incubated 1 h, rotating. The magnet was applied, and the precleared phage was removed and transferred to a clean tube. The magnet was applied, and this step was repeated two times to ensure no carryover of beads bound to phage to the next step. Fc-RBD (rounds 1 and 2) was added to the precleared phage to 100 nM of final concentration and incubated, rotating for 1 h. Block was removed from the second set of beads, and the phage/Fc-RBD mix was added to the Protein A magnetic beads to precipitate the Fc-tagged RBD and bound phage. The beads were washed 8× with PBS-T, 1 mL each, vortexing between each step and applying the magnet. The washed beads were eluted with 800 μL of 0.1 M glycine, pH 2.0, 10 min rotating, the magnet applied, and the eluate transferred to 72 μL of 2 M Tris base to neutralize. The neutralized phage was then added to 9 mL XL1-blue E. coli, which had been grown to OD600 = 0.435 and placed on ice. The cells were infected at 37°C, 45 min, 175 rpm, and then expanded to 100 mL 2xYT/Glu/CB and incubated overnight at 30°C, 250 rpm.²²

The overnight cultures were harvested by measuring the OD₆₀₀, centrifuging the cells at 10,000 × g for 10 min and then resuspending the cells to an OD₆₀₀ of 75 in 2xYT/18% glycerol. To prepare phage for the next round of panning, 5 mL of 2xYT/Glu/CB was inoculated with 5 μL of the 75 OD₆₀₀ glycerol stock and incubated at 37 °C, 250 rpm, until the OD₆₀₀ reached 0.5. The cells were superinfected at 20:1 phage:cell, mixed well, and incubated at 37 °C, for 30 min at 150 rpm and then 30 min at 250 rpm. The cells were pelleted at 5500 × g, 10 min, the glucose containing media discarded, and the cells resuspended in 10 mL of 2xYT/CB/KAN and incubated overnight at 30 °C, 250 rpm. The overnight phage prep was processed as described above. The phage was then prepared at A₂₆₈ = 0.8 in 2% M-PBS-T, and the panning and preclearing continued as described, except in the second and third rounds, the target concentration was reduced 10× per round. In round 3, the target was switched to biotinylated RBD and the MyOne T1 streptavidin beads were used. Washes after phage capture were also increased in the third round, to 12 washes.

Qualitative semELISA of Individual Clones following Panning

At the end of the last panning round, individual colonies were plated on 2xYT/Glu/CB agar plates following the 45 min 150 rpm recovery at 37 °C of the infected XL1-blue cells with the eluted phage. The next day, 95 colonies were inoculated into 400 μL of 2xYT/Glu/CB in a 96-deep-well culture plate and grown overnight at 37 °C, 300 rpm, to generate a master plate, to which glycerol was added to 18% for storage at −80 °C. To prepare an induction plate for the ELISA, 5 μL of each master-plate culture was inoculated into 400 μL of fresh 2xYT/0.1% glucose/CB medium and incubated for 2.75 h at 37 °C, 300 rpm. IPTG was then added to 0.5 mM and the plates incubated at 30 °C with 300 rpm shaking overnight. Because the phagemid contains an amber stop codon, some nanoCLAMP protein is produced without the pIII domain, even though XL1-blue is a suppressor strain, resulting in the periplasmic localization of some nanoCLAMP, of which some percentage is ultimately secreted to the media. The media can then be used directly in an ELISA assay (soluble expression-based monoclonal enzyme-linked immunosorbent assay: semELISA). After the overnight induction, the plates were centrifuged at 1200 × g for 10 min to pellet the cells. Streptavidin-coated microtiter plates (Thermo Fisher) were rinsed three times with 200 μL of PBS and then coated with biotinylated target proteins at 2 μg mL^–1with 100 μL well^–1 and incubated 1 h. For blank controls, a plate was incubated with 100 μL/well PBS. The coating solution was removed, and the plates blocked with 2% M-PBS-T. The block was removed and 50 μL of 4% M-PBS-T added to each well. At this point, 50 μL of each induction plate supernatant was transferred to the blank and protein-coated wells and pipetted 10 times to mix, and incubated for 1 h. The plates were washed four times with 200 μL of PBS-T and the plates dumped and slapped on paper towels in between washes. After the washes, 75 μL of 1:2000 dilution anti-FLAG-HRP (Sigma A8592) in 4% M-PBS-T was added to each well and incubated for 1 h. The anti-FLAG-HRP was discarded and the plates washed as before. The plates were developed by adding 75 μL of TMB Ultra substrate (Thermo Fisher) and analyzed for positive signals compared to controls. Positive clones were then grown up from the master plate by inoculating 1 mL of 2xYT/2% glucose/100 μg mL^–1 CB with 3 μL glycerol stock and incubated for at least 6 h at 37 °C, 250 rpm. The cells were then pelleted and the media discarded. Plasmid DNA was prepared from the pellets using the QIAprep Spin Miniprep Kit, and the sequences determined by Sanger sequencing at GENEWIZ (South Plainfield, New Jersey).

Affinity Maturation of nanoCLAMP P2632

A saturation mutagenesis library was constructed using p2632 as template, amplifying the plasmid using mixtures of degenerate primer sets to introduce a single mutation into each variable residue in a loop using the degenerate codon NNK. Degenerate and wild-type primers were mixed such that the library should contain 29% single mutants, 43% double mutants, 22% triple mutants, and 6% wild type. The amplification of the plasmid and incorporation of variable regions were carried out as described previously.³⁸ The plasmid library was electroporated into TG1 E. coli and panned for three rounds as described. Because the base of this library already had affinity for the target, the concentration of the target protein in round 1 was lowered to 0.1 nM and was lowered in the two following rounds to 0.01 and 0.001 nM, respectively. Also, the wash stringency was increased by increasing the washes from 8 to 12 and including a 7 h wash in round 2 on the second to last wash. Following the third round of panning, individual clones were assessed by semELISA and positives were characterized as described above for affinity by BLI and monodispersity by SEC.

nanoCLAMP Expression and Purification

Positive clones identified in the semELISA (above) were subcloned into a pET expression vector containing an N-terminal MGSS-6His tag and either a C-terminal avitag, or no C-terminal tag, depending on the application, and transformed into chemically competent NEc1 E. coli, a BL21(DE3) derivative (Nectagen, Inc.). Plasmids were purified using Qiagen Miniprep kits (Qiagen) and sequence verified by Sanger sequencing (GENEWIZ). Glycerol stocks of the plasmids in NEc1 cells were prepared for seeding expression cultures.

Glycerol stocks of NEc1 cells harboring nanoCLAMP expression vectors were used to inoculate 3 mL starter cultures of 2xYT/2% glucose (Glu)/100 mg mL^–1 carbenicillin (CB) and grown overnight at 37 °C, 250 rpm. The overnight cultures were diluted 1:100 into 35 mL of Novagen Overnight Express Instant TB Medium/1% glycerol/CB and incubated 24 h, 30 °C, 250 rpm. Cells were pelleted and lysed with 6 M GuHCl, 20 mM Tris, pH 8 (QCB, pH 8), and insoluble material removed by centrifugation at 15,000 × g, 20 min, 15 °C. The cleared lysate was incubated with NiSeph6 FF (Cytiva) for >1 h rotating, at room temperature, and then transferred to 2 mL columns. The columns were washed with 6 × 1 mL QCB, pH 8, and then refolded with 11 mL of 20 mM MOPS, 150 mM NaCl (MBS), and 1 mM CaCl₂, pH 8. The nanoCLAMPs were eluted with MBS, 1 mM CaCl₂, and 250 mM imidazole, pH 8, and the buffer was exchanged to remove the imidazole using Zeba 7 MWCO desalting columns and normalized to 1 mg mL^–1 in MBS, 1 mM CaCl₂, pH 6.5. This procedure was scaled up as necessary to produce larger quantities of protein. For all functional studies, monodisperse monomeric fractions were collected from a Superdex 75 size exclusion column (described below) and pooled.

To generate biotinylated nanoCLAMPs, expression constructs bearing a C-terminal avitag were transformed into chemically competent NEc1/BirA cells (Nectagen, Inc.), containing a chloramphenicol selectable plasmid bearing an IPTG-inducible biotin ligase (BirA), and expressed as described above, except that the autoinduction culture was supplemented with 50 μM biotin. Prior to eluting the refolded nanoCLAMPs, biotinylation was driven to completion by adding recombinant Mal-BirA (Maltose binding protein with BirA fused to the C-terminus) to 1 μM, ATP to 10 mM, MgCl₂ to 10 mM, and d-biotin to 150 μM in MBS, pH 8, at a volume of 1.5 mL and incubating with rotation at 30 °C for 2 h. The resin was washed with 5 mL of MBS, pH 8, and then 1 mL of MBS, pH 8, 1 mM CaCl₂ to remove Mal-BirA. The protein was eluted as described above. Determination of percent biotinylation was measured by incubation of the protein with streptavidin and separating on 12% NuPAGE Bis–Tris (MES buffer) gel and estimating the percentage bound to streptavidin. The protein was buffer exchanged and concentration normalized, as described above.

Cloning, Expression, and Purification of Dual Epitope nanoCLAMP P2712 and P2712avi

A dual epitope-binding nanoCLAMP (P2712) was constructed by fusing the nanoCLAMPs P2710 to P2609 in the format MGSS-6His-P2710-(GSlinker)-P2609 (see the Supporting Information for sequence) or with a C-terminal avitag for biotinylation, in the pET vector described above.

P2712 Purification from a 1 L Shake Flask for In Vivo Experiments

The p2712 plasmid was expressed in 1 L of autoinduction media as described above, and the pellet lysed with QCB, pH 8, using a polytron. Insoluble material was removed by centrifugation and the cleared supernatant applied to 5 mL of NiSeph6 FF, which was washed with 80 mL of QCB at pH 8, 25 mL of 8 M urea, and 20 mM Tris at pH 8 and then eluted in 8 M urea, 20 mM Tris at pH 8, and 250 mM imidazole. The protein was precipitated with ice-cold ethanol, pelleted at 15,000 × g, 20 min, 4 °C, and then resuspended with 8 M urea and 20 mM Tris, pH 8 (buffer A). The protein was applied to a 1 mL HiTrap Q HP column (Cytiva) and eluted with a gradient of buffer B (buffer A with 1 M NaCl) from 0 to 80% B. Fractions containing pure P2712 were pooled and applied to a second IMAC column, washed with 8 M urea and 20 mM Tris, pH 8, and then refolded with MBS at pH 8 and 1 mM CaCl₂, and finally eluted with refolding buffer with 250 mM imidazole. The monodisperse monomer was collected by size exclusion chromatography on a Superdex 75 column, as described above. Purity was estimated at over 95% by NuPAGE Bis–Tris (MES buffer) and staining with GelCode Blue (Thermo) (SI Appendix, Figure S5).

Endotoxin Removal from P2712 and P2570 (neg Control)

Endotoxin was removed from purified P2712 and P2570 using the method of Teodorowicz (2017),²⁰⁰ by extracting with Triton X-114 (Sigma) and removing traces of TX114 with Bio-Beads SM-2 (Bio-Rad). Finally, proteins were filtered through a 0.2 μm PES filter (Millex-GP SLGPR33RS) and then adjusted to 66.7 μM. Endotoxin levels of each preparation were measured by the LAL assay.

Analysis of Monodispersity by Size Exclusion Chromatography

Purified nanoCLAMPs were diluted to a final concentration of 0.18 mg mL^–1 in MBS, 1 mM CaCl₂, pH 6.5, centrifuged at 20,000 × g, for 2 min at 4 °C, and the supernatants were transferred to a clean tube. The samples were loaded into a 125 μL sample loop and injected onto a Superdex 75 10/300 GL column (GE Healthcare Life Sciences, Pittsburgh, Pennsylvania) equilibrated in MBS, 1 mM CaCl₂, pH 6.5, at a flow rate of 0.65 mL min^–1. The column was calibrated with Bio-Rad Gel Filtration Standard per manufacturer’s instructions.

Determination of Melting Temperature by Differential Scanning Fluorimetry

The melting temperature of purified nanoCLAMPs was determined using GloMelt Thermal Shift Protein Stability Kit (Biotium) per manufacturer’s instructions. Briefly, purified nanoCLAMPs were adjusted to 1 mg mL^–1 in MBS, 1 mM CaCl₂, pH 6.5, diluted in half with 2X GloMelt (Biotium), and aliquoted to a 386-well plate and sealed with an optical film. The plate was then heated in a QuantStudio 5 qPCR machine using a SYBR Green reporter with no passive reference. The heating profile was 25 °C for 2 min, ramp at 0.05 °C/s to 99 °C, and 99 °C for 2 min. T_m is defined as the inflection point in the unfolding curve.

Biolayer Interferometry of nanoCLAMPs

Kinetic analysis of interactions between nanoCLAMPs and RBD was carried out on an OctetRed96 using SAX streptavidin-coated sensor tips. The tips were transferred first to buffer (MBS, 1 mM CaCl₂, pH 6.5 + 1% BSA) for 300 s, and then to biotinylated RBD or nanoCLAMP at 2 μg mL^–1 in buffer for 180 s, and then to buffer for 300 s, and then to at least four dilutions of nanoCLAMPs or RBDs (depending upon orientation) in buffer (association) for 200 s, and then to buffer (dissociation) for 720 s. The cells were constantly vortexing at 1000 rpm at room temperature. The kinetics were fit to a 1:1 model and K_d calculated using global fit analysis. For high-throughput off-rate analysis, biotinylated RBD was immobilized on SAX tips as described and then incubated in a single concentration (800 nM) of nanoCLAMP using the parameters above.

For epitope binning experiments, SAX sensors were coated with biotinylated P2710avi at 15 μM, and then loaded with Wuhan RBD at 30 nM, and finally incubated with “competitor” nonbiotinylated nanoCLAMP at 30 nM, or buffer (neg control).

RBD/Ace2 Competitive Inhibition ELISA

Polystyrene protein A-coated microtiter plates (Thermo Fisher: 15132) were coated by incubation on a plate shaker at room temperature with 0.5 μg mL^–1 of recombinant SARS-CoV-2 Spike RBD-Fc (Sino Biological: 40592-V02H). Negative control wells were incubated with 2% milk + PBS-T (20 mM NaH₂PO₄, 150 mM NaCl (PBS) + 0.05% Tween 20) or 0.5 μg mL^–1 recombinant SARS-CoV-2 Spike RBD-Fc. Plates were washed three times with PBS-T. Following 1 h of incubation with recombinant SARS-CoV-2 Spike RBD-Fc, preparations of 800 nM nanoCLAMP candidates were serially diluted 4-fold (i.e., 800, 200, 50, 12.5, 3.125, 0.781, and 0.195 nM) with Biotinylated Human ACE2/ACEH Protein, His, Avitag (ACROBiosystems: AC2-H82E6) containing dilution buffer at a final Biotinylated Human ACE2 concentration of 25 ng mL^–1. The solutions were incubated on a plate shaker at room temperature for 1 h. Biotinylated Human ACE2 containing dilution buffer was added to the negative control wells. Following the 1 h incubation, the wells were washed three times with PBS-T. All wells were then incubated with Streptavidin-HRP (Thermo Scientific: N504) at a 1:5000 dilution on a plate shaker at room temperature for 1 h. After incubation with SA-HRP, the plate was washed three times with PBS-T and developed with 1-Step Ultra TMB ELISA Substrate Solution (Fisher Scientific: 34028) for approximately 7.5 min. The reaction was stopped using 2N Sulfuric Acid solution. Absorbance was read using Molecular Devices, Kinetic Microplate Reader and analyzed using SoftMax Pro 5.4.

Pseudovirus Neutralization Assay

Lentivirus pseudotyped with Spike protein from SARS-CoV-2 Wuhan-Hu-1 and bearing a Luciferase-IRES-ZsGreen payload was prepared as described²⁰¹ and also purchased from Integral Molecular (Cat# RVP-701L). Lentivirus pseudotyped with U.K. Variant (B.1.1.7) and S. African Variant (B.1.351) Spike proteins were purchased from Integral Molecular (cat# RVP-706L and RVP-707L, respectively). SARS-CoV-2 Omicron XBB.1.5 pseudovirus was obtained from ProSci (Cat# 95-208). Ace2-expressing HEK293T cells (293T-hsACE2) were purchased from Integral Molecular (Cat# C-HA102). Briefly, 293T-hsAce2 cells were plated in a 96-well format at 2.0 × 10⁴ cells/well. On the day of infection, the pseudotyped viruses bearing the luciferase reporter were preincubated with the serially diluted nanoCLAMPs in DMEM containing 10% FBS and 100 units mL^–1 of penicillin and streptomycin at 37 °C for 1 h and applied to the cells. Luminescence was measured 72 h post infection using the Bright-Glo Assay System (Promega: E2610) when assaying in-house-generated Wuhan pseudovirus, or the Renilla-Glo Assay System (Promega: E2710) when assaying the Integral Molecular pseudoviruses. Luminescence was read using Molecular Devices, Kinetic Microplate Reader with 1000 ms integration time and analyzed using SoftMax Pro 5.4. Pseudotyped viruses were diluted to yield a signal-to-background ratio of approximately 500-fold.

In Vitro Experiments

Vero E6 cells (ATCC CRL-1586) were cultured in Dulbecco’s modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 1% l-glutamine, 1% antibiotics (100 U mL^–1 penicillin, and 100 μg mL^–1 streptomycin), in a humidified atmosphere of 5% CO₂ at 37 °C.

Normal human epithelium (NHBE) cells (Lonza, France) were cultured in PneumaCult-Ex Plus Medium (StemCEll Tech., Gtanbke, France). Cells were seeded in 96-well plates (2 × 10⁴ cells/well) 24 h before test. The culture medium was replaced with a fresh medium that contains increasing concentrations of nanoCLAMP for 72 h . The old medium was then aspirated, and cells were washed with PBS. The cell viability was evaluated using a resazurin cell viability assay. Briefly, 100 mL of the resazurin solution (11 μg mL^–1) in DMEM/10% FBS was added to each well and the plate was incubated for 4 h. The fluorescence emission of each well was measured at 593 nm (20 nm bandwidth) with an excitation at 554 nm (18 nm bandwidth) using a Cytation 5 Cell Imaging Multi-Mode Reader (BioTek Instruments SAS, France). Each condition was replicated three times, and the mean fluorescence value of cells treated the same way in the absence of nanoCLAMPswas taken as 100% cellular viability.

SARS-CoV-2 Infection and Treatment in Mice

Animal housing and experimental procedures were conducted according to the French and European Regulations (Parlement Européen et du Conseil du 22 Septembre 2010, Decret n° 2013–118 du 1er fevrier 2013 relatif à la protection des animaux utilisés a des fins scientifiques) and the National Research Council Guide for the Care and Use of Laboratory Animals (National Research Council (U.S.), Institute for Laboratory Animal Research (U.S.), and National Academies Press (U.S.), Eds., Guide for the Care and Use of Laboratory Animals, eighth ed. Washington, D.C: National Academies Press, 2011). The animal BSL3 facility is authorized by the French authorities (Agreement N° B 13 014 07). All animal procedures (including surgery, anesthesia, and euthanasia, as applicable) used in the current study were submitted to the Institutional Animal Care and Use Committee of the CIPHE and approved by the French authorities (APAFIS#26484-2020062213431976 v6). All CIPHE BSL3 facility operations are overseen by a biosecurity/biosafety officer and accredited by the Agence Nationale de Sécurité du Médicament (ANSM).

Animals

Heterozygous K18-hACE C57BL/6J mice (strain: 2B6. Cg-Tg (K18-ACE2)2Prlmn/J) were obtained from The Jackson Laboratory. All breeding, genotyping, and production of K18-hACE2 was performed at the CIPHE. The sample size was based on previous articles reporting the use of K18-hACE2 mice in SARS-CoV2 challenge experiments (10 animals per experimental group). Animals were housed in groups within cages and fed with standard chow diets.

Wuhan/D614 SARS-CoV-2 Virus Production

Vero E6 cells (CRL-1586; American Type Culture Collection) were cultured at 37 °C in DMEM supplemented with 10% FBS, 10 mM HEPES (pH 7.3), 1 mM sodium pyruvate, 1× nonessential amino acids, and 100 U mL^–1 penicillin/streptomycin. The strain Beta CoV/France/IDF0372/2020 was supplied by the National Reference Centre for Respiratory Viruses hosted by the Institut Pasteur (Paris, France). The human sample from which strain BetaCoV/France/IDF0372/2020 was isolated was provided by the Bichat Hospital, Paris, France. Infectious stocks were grown by inoculating Vero E6 cells and collecting supernatants upon observation of the cytopathic effect. Debris was removed by centrifugation and passage through a 0.22 mm filter. Supernatants were stored at 80 °C.

Infection Assay of K18-hACE2 Transgenic Mice

Intranasal virus and antiviral treatment was performed under anesthesia, and all efforts were made to minimize animal suffering. 8-to12-week old heterozygous K18-hACE C57BL/6J mice (strain: 2B6. Cg-Tg (K18-ACE2)2Prlmn/J) were inoculated with 2.5 × 10⁴ PFU Wuhan/D614 SARS-CoV-2 via intranasal administration of 30 μL. Daily treatments were administered intranasally at 2.25 mg kg^–1 (P2712, therapeutic) or 1.12 mg kg^-1 (P2570, mock, which is the molar equivalent to the P2712 dose) using the average weight of each group (30 μL volume). Mice were monitored daily for morbidity (body weight) and mortality (survival). During the monitoring period, mice were scored for clinical symptoms (weight loss, eye closure, appearance of the fur, posture, and respiration). Mice obtaining a clinical score defined as reaching the experimental endpoint were humanely euthanized.

Measurement of SARS-CoV-2 Viral Load by RT-qPCR and TCID50 (50% of Tissue-Culture Infective Dose)

For viral titration by RT-qPCR, tissues were homogenized with ceramic beads in a tissue homogenizer (Precellys, Bertin Instruments) in 0.5 mL of RLT buffer. RNA was extracted using the RNeasy Mini Kit (QIAGEN) and reverse transcribed using the High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific). Amplification was carried out using ONEGreen FAST qPCR Premix (Ozyme) according to the manufacturer’s recommendations. The number of copies of the SARS-CoV-2 RNA-dependent RNA polymerase (RdRp) gene in samples was determined using the following primers: forward primer catgtgtggcggttcactat, reverse primer gttgtggcatctcctgatga. This region was included in a cDNA standard to allow the copy number determination down to ≫100 copies per reaction. The copies of SARS-CoV-2 were compared and quantified using a standard curve and normalized to total RNA levels. An external control (mock-infected wild-type animal, nondetectable in the assay) and a positive control (SARS-CoV-2 cDNA containing the targeted region of the RdRp gene at a concentration of 10⁴ copies μL^–1 [1.94 × 10⁴ copies μL^–1 detected in the assay]) were used in the RT-qPCR analysis to validate the assay. The median tissue-culture infectious dose (TCID50) represents the dilution of a virus-containing sample at which half of the inoculated cells show signs of infection. To perform the assay, lung and brain tissues were weighed and homogenized using ceramic beads in a tissue homogenizer (Precellys Bertin Instruments) in 0.5 mL of RPMI media supplemented with 2% FCS and 25 mM HEPES. Tissue homogenates were then clarified by centrifugation and stored at 80 °C until use. Forty thousand cells per well were seeded in 96-well plates containing 200 μL of DMEM + 4% FCS and incubated for 24 h at 37 °C. Tissue homogenates were serially diluted (1:10) in RPMI media, and 50 μL of each dilution was transferred to the plate in six replicates for titration at 5 days post inoculation. Plates were read for the CPE (cytopathology effect) using a microscopy reader, and the data were recorded. Viral titers were then calculated using the Spearman-Karber formula and expressed as TCID50/mg of tissue.

Histology Analysis

At day 5 post infection, three mice per group were euthanized and lung and brain tissues were harvested and fixed in Sterilin tubes with 4% paraformaldehyde. Fixed left lungs (4% paraformaldehyde) were paraffin-embedded (Leica Pearl). Three 3 μm transversal sections were cut, mounted on Superfrost glass slides (Fischer Scientific), and stained with hematoxylin–eosin saffron (HES). To determine the severity of tissue remodeling observed on histological sections, the area of stained tissue was measured and reported to the total section area. Measurement was realized with the Cell Dimension software coupled to a DP72 camera (Olympus) on three sections for each animal.

Cytokine and Chemokine Protein Measurements

K18-hACE2 transgenic mice were infected (intranasally challenged with 2.5 × 10⁴ PFU of SARS-CoV-2 Wuhan) and treated according to administration regimen by intranasal inoculation of P2712 or P2570 (6 h before infection and days 1 and 2 post infection). At day 5 post infection, blood was collected and plasma were isolated after centrifugation. Lung and brain were harvested in Precellys tubes containing RPMI medium completed with Pen/Strep, Hepes, and FCS and then homogenized with the Precellys. Then, the samples were inactivated by a mix of Triton 10X and RPMI medium at a final concentration of 0.5% Triton. A kit, created by Merck Millipore, in order to cover the overall cytokine and chemokine responses was used here. This kit coupled with the Luminex platform in a magnetic bead format provides the advantage of ideal speed and sensitivity, allowing quantitative multiplex detection of dozens of analytes simultaneously, which can dramatically improve productivity. It contained cytokine standards to make a standard curve and to ensure lot-to-lot consistency, two quality controls to qualify assay performance, detection antibody cocktails designed to yield consistent analyte profiles within the panel, streptavidin-PE, and a panel of magnetic beads that recognize each one of the following analytes: IL-1a, IL-1b, IL-2, IL-6, IL-10, IL12p40, IL-12p70, IL-13, M-CSF, G-CSF, GM-CSF, IP-10 (CXCL10), IFN-g, MCP-1 (CCL2), MIP 1-a (CCL3), TNF-a, VEGF, CXCL9 (MIG), CXCL1 (KC), and MIP-2. In a 96-well plate previously washed with wash buffer, we incubated overnight (at 4 °C and on a shaker), according to a template, the standards, the quality controls, and the samples (plasma and organs supernatants) with assay buffer and the magnetic bead panel. The following day, the plate was washed twice with wash buffer and then we added the detection antibody cocktail and incubated it for 1 h at room temperature on a shaker before we added the streptavidin-PE and incubated it for 30 min. Then, the plate was washed twice and the samples were suspended in the assay buffer for 5 min on the shaker before reading on the MAGPIX Instrument.

Data Availability Statement

All data are available on request

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsptsci.3c00306.

• (Figure S1) Size exclusion chromatography profiles of purified nanoCLAMPs used in this study; (Figure S2) Binding affinity of nanoCLAMP 2632 to Wuhan-RBD; (Figure S3) SDS PAGE analysis of nanoCLAMPs used in this study; (Figure S4) Differential scanning fluorimetry (DSF) determination of melting temperature of nanoCLAMPs; (Figure S5) Cell viability of different nanoCLAMP binders; and (Figure S6) Neutralization potency of nanoCLAMP P2715, formed via a combination of two nanoCLAMP P2710 (PDF)

The authors declare the following competing financial interest(s): RJS is a significant shareholder of Nectagen, Inc.