Inhalable SARS-CoV-2 vaccine for single-dose dry powder aerosol immunization

doi:10.21203/rs.3.rs-2301923/v2

Download PDF

Research Article

Inhalable SARS-CoV-2 vaccine for single-dose dry powder aerosol immunization

https://doi.org/10.21203/rs.3.rs-2301923/v2

This work is licensed under a CC BY 4.0 License

Version 2

posted

You are reading this latest preprint version

The coronavirus disease pandemic has fostered major advances in vaccination technologies; however, there are urgent needs for vaccines that induce mucosal immune responses and single-dose, noninvasive administration. Here, we develop an inhalable, single-dose, dry powder aerosol SARS-CoV-2 vaccine that induces potent systemic and mucosal immune responses. Our vaccine encapsulates proteinaceous cholera toxin B subunit-assembled nanoparticles displaying the SARS-CoV-2 RBD antigen within microcapsules of optimal aerodynamic size, and this unique nano-micro coupled structure supports efficient alveoli delivery, sustained antigen release, and antigen-presenting cell uptake, which are favourable features for induction of immune responses. Moreover, our vaccine successfully induces strong production of IgG and IgA, as well as a local T-cell response, collectively conferring effective protection against SARS-CoV-2 in mice, hamsters, and nonhuman primates. Finally, we also demonstrate a “mosaic iteration” of our vaccine that codisplays ancestral and Omicron antigens, extending the breadth of antibody response against cocirculating strains and transmission of the Omicron variant. These findings support our inhalable vaccine as a promising multivalent platform for fighting COVID-19 or other respiratory infectious diseases.

Vaccine Development

Immunology

COVID-19

inhalable vaccine

dry powder vaccine

mucosal immunity

single dose

SARS-CoV-2 challenge

nonhuman primates

Coronavirus disease 2019 (COVID-19), a severe acute respiratory syndrome caused by SARS-CoV-2, was declared a pandemic in 2020 ^1,2. Vaccination has been a successful public health intervention for preventing and controlling the spread of this disease ³. To date, multiple types of SARS-CoV-2 vaccine, such as inactivated vaccines ^4-6, protein subunit vaccines ^7-9, mRNA vaccines ^10,11, and viral-vectored vaccines ^12-15, have been approved. Most of these are administrated by intramuscular injections that induce production of serological IgG, thereby neutralizing the infectivity of the virus and alleviating COVID-19. However, vaccines delivered intramuscularly fail to provide a first line of protection at the respiratory tract, owing to deficiencies for secretory IgA and IgG ^13,16. Another recognized issue is the requirement of two or three invasive administrations, which greatly compromised enthusiasm and reduced immunization rates ¹⁷.

Seeking to overcome the problems with intramuscular injection, several vaccine candidates for intranasal or nebulization administration are under development or have been approved ^18,19, most of which are based on viral vectors, such as adenovirus, attenuated influenza viruses, parainfluenza viruses, or Newcastle disease viruses ^20-24. Notably, human clinical studies have shown that inhaled vaccines, at a much lower dose than an injected vaccine, can induce comparable antibody responses ^25,26. However, the aforementioned vaccines are in liquid form, necessitating cold-chain transportation and storage, and generally require two or three inhaled immunizations or use of a heterologous booster vaccination following primary muscular immunization.

Given this background and the abundant immune cells known to reside in the lung ²⁷, we were motivated to develop a dry powder vaccine suitable for single-dose inhalation, a goal requiring several considerations. Specifically, the vaccine vector should retain its structure and performance after lyophilization. As smaller particles are liable to be exhaled along with airflow and larger particles tend to deposit in the superficial lung, the aerodynamic size must be well controlled in the 1-4 µm range, enabling predominant deposition in the deep lung (in alveoli) ²⁸. Additionally, vaccine particles with a sustained-release profile can supply continuous antigen (Ag) stimulation, which may induce a long-lasting immune response.

We initially used the pentameric cholera toxin B subunit (CTB), a classic mucosal adjuvant ^29,30, and fused trimer-forming peptides to assemble a self-adjuvanting nanoparticle (CNP) chassis, which supported the display of SARS-CoV-2 receptor-binding domain (RBD) (R-CNP) based on a bioorthogonal system and enhanced antigen-presenting cells (APCs) uptake owing to the nano size. Next, we encapsulated R-CNPs into porous poly(lactic-co-glycolic acid) (PLGA) microcapsules, and the resulting R-CNP@M, with an optimized aerodynamic diameter, facilitated dry powder aerosol delivery to alveoli. Along with gradual degradation of microcapsules, R-CNPs were released in a sustained manner and were internalized by APCs, leading to potent T-cell and B-cell responses. Upon a single dose, our vaccine induced high IgG and IgA production in mice, hamsters, and nonhuman primates, and conferred efficient protection against SARS-CoV-2. Further, an engineered “mosaic iteration” of this vaccine showed promise for readily responding to future cocirculation of multiple strains and for preventing transmission of the Omicron variant.

Fabrication and characterization of R-CNP

Mucosal adjuvanticity and the pentamer domains of CTB initially prompted us to construct a self-adjuvanting nano chassis, which was achieved by the fusion of a trimer peptide at the C-terminus of CTB and subsequently expressed in the Escherichia coli BL21 strain³¹ (Fig. 1a). Via a SpyTag/SpyCatcher bioorthogonal system, the resulting CNP chassis displayed RBD antigens, which are favourable for improving antigen stability, APC uptake, and activation ^32-34. As evident from SDS-PAGE and immunoblotting analyses (Fig. 1b and Extended Data Fig. 1a), a clear shift in the molecular weight of the SC-RBD band was observed after mixing with ST-CNP, indicating covalent bond formation. Transmission electron microscopy revealed that the R-CNPs displayed well-dispersed nanoparticulate features (Fig. 1c) and a dynamic light scattering (DLS) analysis showed a narrow size distribution (Extended Data Fig. 1b), with a peak at 24 nm.

We next characterized R-CNP antigenicity by detecting the binding affinity and kinetics with the hACE2 receptor. Enzyme-linked immunosorbent assays (ELISA) and surface plasmon resonance (SPR) assays showed that the RBD monomer and the R-CNP bound to hACE2 in a similar dose-dependent manner (Fig. 1d, Extended Data Fig. 1c, d), indicating that the epitopes on the R-CNP were properly exposed and correctly folded. Moreover, we verified the APC activation capacity of R-CNPs in vitro (Extended Data Fig. 1e-g), finding increased levels of major histocompatibility complex (MHC) and costimulatory molecules, as well elevated secretion of multiple cytokines. Thus, our SpyTag/SpyCatcher-mediated attachment of the RBD antigen to the CNP worked as anticipated.

Design and preparation of R-CNP@M vaccine

Given that direct inhalation of nanosized formulations suffers from poor deposition at alveoli and previous studies reported efficient alveolar delivery from particles with an aerodynamic diameter ranging from 1-4 μm ^35,36, we pursued an “R-CNP@M” design by placing the R-CNPs inside microcapsules of suitable aerodynamic size. This should permit efficient delivery to alveoli and enable a sustained release of R-CNPs to support continuous antigen stimulation. We used a membrane emulsification method to prepare porous PLGA microspheres, which encapsulated R-CNPs through simple mixing and a mild heating-mediated sealing (Fig. 1e) ^37,38. By fine-tuning the membrane pore size, osmotic gradient, and pore evolution time, the microspheres can be controlled for particle size, porosity, and cavity volume, which facilitated optimization of R-CNP@M to a suitable aerodynamic size and high encapsulation efficiency (Extended Data Fig. 2a-c).

We used confocal laser scanning microscopy (CLSM) to characterize the resulting R-CNP@M , and the 3D reconstruction provided strong evidence that the inner cavities of Nile red-labelled microcapsules were filled with fluoresceine isothiocyanate (FITC)-labelled R-CNPs (Fig. 1e (Ⅰ-Ⅲ)). After lyophilization, R-CNP@M appeared as a white powder with excellent dispersity, which could be used to create a uniform fog upon spraying with a dry powder aerosol generator (DPAG) (Fig. 1f). These lyophilized microcapsules were almost entirely within our desired aerodynamic size range (2-4 μm; Fig. 1g). Moreover, they showed sustained release of R-CNP over 5 weeks (Extended Data Fig. 2d), and the binding affinity for hACE2 did not differ between the released vs. fresh R-CNPs (Fig. 1h). Testing after one-month of storage showed only negligible changes in the aerodynamic diameter of R-CNP@M and in the hydrodynamic diameter of the encapsulated R-CNP (Extended Data Fig. 2e, f), supporting good storage stability.

In vivo biodistribution and release

After characterizing the acceptable endotoxin level and verifying the in vivo safety of the inhalable microcapsule vaccine (Extended Data Fig. 3a-e), we investigated its retention, deposition, release, and uptake profiles in the lung after administration with a dry powder inhaler (DPI) (Fig. 1i). To measure retention, we delivered a microcapsule vaccine containing Cy7-labelled R-CNPs, and a group of mice receiving aerosol administration of equivalent Cy7-labelled R-CNPs was included for comparison (Fig. 1j, Extended Data Fig. 3f). Compared to the absence of a fluorescence signal in lungs of the R-CNP group mice after 5 days, 50% of the initial signal intensity was still evident in lungs of R-CNP@M group mice on Day 5, and a signal was evident until Day 42. The area-under-curve values of R-CNP@M were improved by 3.5-fold compared to those of R-CNP (Extended Data Fig. 3g), suggesting that R-CNP@M can induce continuous antigen stimulation in lungs.

We also labelled the trachea and microcapsule with streptavidin-AF647 and Nile red, respectively, and examined the distribution by light-sheet microscopy. Taking the upper right pulmonary lobe on Day 5 post-inhalation as an example, microcapsules were uniformly spread throughout the lobe (Fig. 1k). Whole lung tissues sections (Day 5) showed few microcapsules in the trachea, primary bronchi, or secondary bronchi, yet most deposited in the distant lung, including in terminal bronchi and alveoli (Fig. 1l, Extended Data Fig. 3h). Sections from Day 10 revealed noncolocalized signals from the R-CNPs and microspheres, supporting successful release of R-CNPs from the degraded microcapsule (Fig. 1m, Extended Data Fig. 3i). The favourable uptake of these R-CNPs by APCs in lungs was verified by flow cytometry (FCM), with nearly half of APCs present in R-CNP-positive cells (Fig. 1n). These results established that our R-CNP@M vaccine was efficiently delivered to alveoli and that R-CNPs released from microcapsules were internalized by APCs.

Potent responses of immune cells

To investigate the impact of the inhaled vaccine on pulmonary immune responses, we performed single-cell RNA sequencing (scRNA-seq) of sorted immune cells (CD45⁺) from the lungs on Day 21. The mice immunized with free R-CNP (5 μg RBD, twice) received an equivalent dose of RBD antigen as those mice immunized with R-CNP@M (10 μg RBD, once); mice receiving PBS or CNP at the same regimen as R-CNP were also included. The scRNA-seq data clustered into nine cell types (Fig. 2a and Extended Data Fig. 4a). Focusing on APCs (DCs and macrophages), the proportions and numbers in the vaccination groups both increased (Extended Data Fig. 4b). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis revealed involvement of several immune activation-associated pathways, such as the “antigen processing and presentation” and “chemokine signalling” pathways (Extended Data Fig. 4c-e). Moreover, both vaccination groups exhibited upregulation of a panel of DC- and macrophage-activation-related genes, such as Cd40 and Cd86, most of which showed the highest extent of upregulation in the R-CNP@M group (Fig. 2b). Notably, several homing-related genes were upregulated in activated APCs (such as H2-Ab1 and Cd74), which drew our attention to their homing to mediastinal lymph nodes (mLNs). Correspondingly, R-CNP@M outperformed the other groups for increasing the intra-LN proportion of activated APCs (Fig. 2c).

We also evaluated T-cell responses in the respiratory tract, which are known as beneficial for broadly neutralizing antibody-evading SARS-COV-2 variants^39-41. In the lower respiratory tract, the proportion of CD8⁺ memory T cells at 21- and 70-days post-immunization increased in the order of PBS, CNP, R-CNP, and R-CNP@M (Fig. 2d, Extended Data Fig. 5a). By Day 70, the proportion of CD8⁺ tissue-resident memory T (T_RM) cells in the R-CNP group fell to 11.06±2.81% from the 30.63±1.24% observed on Day 21, while the proportion remained at approximately 21% on both sampling days for the R-CNP@M group, indicating that sustained release was favourable for long-term maintenance of CD8⁺ T_RM cells (Fig. 2e, Extended Data Fig. 5b).

Upon identifying RBD epitopes (CGPKKSTNL and VGGNYNYLYRLFRKS for CD8⁺ and CD4⁺ T cells, respectively, Extended Data Fig. 5c-f), we were able to investigate epitope-specific T cell responses. In the lower respiratory tract, the R-CNP@M group obviously outperformed the other groups in both magnitude and duration of interferon (IFN)-γ-secreting T_RM cell response (Fig. 2f, Extended Data Fig. 5g, h). We also assessed intracellular cytokines (IFN-γ, tumour necrosis factor (TNF)-α, and interleukin (IL)-2) to evaluate the multifunctionality of CD8⁺ T cells in the lung, and the R-CNP@M group showed substantially increased proportions of bi- and tri-positive CD8⁺ T cells upon stimulus with the epitope (Fig. 2g). In the upper respiratory tract, we found that the numbers of both CD4⁺and CD8⁺T cells increased in the vaccinated group, and IFN-γ-secreting cells in the R-CNP@M group were much more durable than those in the other groups (Extended Data Fig. 5i-l). The superior extent and duration of T-cell responses induced by R-CNP@M, also supported by the upregulated expression of T-cell activation-associated genes (Extended Data Fig. 4f-h), were favourable to prevent infections by antibody-evading SARS-CoV-2 variants.

We next assessed B-cell responses, which largely mediated humoral immunity ⁴². Light-sheet imaging revealed significant enlargement of B-cell follicles (increasing in the order of PBS, CNP, R-CNP, and R-CNP@M), suggesting induction of secondary follicles containing germinal centres (GCs) in the latter two groups (Fig. 2h). Moreover, the potent humoral response induced by R-CNP@M was supported by the highest proportion of GC B cells (IgD^-GL7⁺B220⁺) detected by FCM (Fig. 2i), as well as the significantly upregulated gene expression of memory B-cell differentiation (Extended Data Fig. 4i–k).

Antibody responses and protective effects

The above results encouraged us to evaluate the antibody production of mice (Fig. 3a). We collected blood samples at multiple time points over a 12-month period and measured serological anti-RBD antibody titers by ELISA. The antibody titers of RBD-specific IgM, IgG, IgG1, and IgG2a in serum were more rapidly and more robustly increased by a single dose of R-CNP@M than by two doses of free R-CNP (10 μg equivalent RBD), while RBD-specific antibodies were undetectable in the CNP group (Fig. 3b and Extended Data Fig. 6a). Notably, the anti-RBD antibody titers of R-CNP@M remained at a relatively high level over the 12-month period, which might be related to the continuous supply of immunogenic R-CNPs from the degraded microcapsules (Extended Data Fig. 6b, c). Our BCR sequencing data for pulmonary lymphocytes showing a high level of germline divergence further indicated that this long-term antibody response seemed promising for improving antibody quality by accelerating affinity maturation (Extended Data Fig. 6d-j).

To assess the mucosal immune response, we vaccinated another batch of mice and sampled on Day 70 post-inhalation. Consistent with the data for nasal lavage fluid (Extended Data Fig. 6k), bronchoalveolar lavage fluid (BALF) from R-CNP@M-vaccinated mice showed higher levels of RBD-specific IgA and IgG than those from R-CNP-vaccinated mice, whereas no RBD-specific IgA or IgG was detected in the BALF from the CNP group (Fig. 3c). We further conducted neutralization assays with a panel of pseudotyped viruses (displaying the SARS-CoV-2 S protein) and detected the pseudovirus neutralizing antibody titers at 50% inhibition (pVNT₅₀) (Fig. 3d, e). Both serum and BALF from the R-CNP@M group showed the highest pVNT₅₀against all variants. Specifically, for the D614G variant, the pVNT₅₀ of mice immunized with R-CNP@M approached ∼1259 in serum and ∼120 in BALF, while the corresponding values in the R-CNP group were lower, at ∼181 and ~85 in serum and BALF, respectively. The superior performance of R-CNP@M in the neutralizing antibody (NAbs) response supported that the increased NAbs benefits markedly from continuous antigen stimulation.

We next evaluated in vivo protection of R-CNP@M against wild-type SARS-CoV-2 in K18-hACE2 mice on Day 28 post-inhalation and assessed potential protection benefits of cellular or humoral responses after additional depletion of CD8⁺ T cells (CD8⁺ T^dep) or B cells (B^dep) (Extended Data Fig. 7). R-CNP@M-vaccinated mice without T/B depletion (naïve) showed extremely low viral loads, with no sign of lesions, body weight loss, or death during the observation window. We also observed that the protection against SARS-CoV-2 was significantly reduced in mice with either B or CD8⁺ T-cell depletion. Compared to the B^dep group, higher viral loads in the lung and nose, yet less pronounced body weight loss and fewer deaths, were observed for CD8⁺ T^dep mice. These findings suggest that strong cellular and humoral immunity each contributed to the observed in vivo protection against SARS-CoV-2 infection.

We also evaluated the performance of the R-CNP@M vaccine in hamsters (a gold standard animal model for SARS-CoV-2 studies). Briefly, hamsters were immunized with R-CNP@M (30 μg equivalent RBD, once) via DPI, and PBS was used as control (Fig. 3f). The high anti-RBD IgG antibody titers (Fig. 3g) and the high SARS-CoV-2 neutralizing titers in serum and BALF (Fig. 3h) together supported establishment of systemic and mucosal immunities. Accordingly, viral RNA (ORF1a/b region) copies and infectious virions in the lung and nose were much lower in the R-CNP@M group than in PBS group (Fig. 3i, j). Supporting this distinct viral burden, immunohistochemical and haematoxylin-eosin staining of lung sections revealed abundant SARS-CoV-2 nucleocapsid protein, substantial thickening of alveolar septa and prominent infiltration of inflammatory cells in the PBS group but not in the R-CNP@M group (Fig. 3k). These exciting protection effects of R-CNP@M could be attributed to the orchestration of systemic and mucosal responses, whereas the licenced intramuscular vaccine only elicited a systemic response, compromising protection efficacy (Extended Data Fig. 8a-g).

Immune responses in nonhuman primates

Aiming to bring our vaccine closer to clinical relevance, we experimented with nonhuman primates (Fig. 4a). One monkey treated with a Nile red-labelled vaccine was sacrificed on Day 14, with the lung collected for frozen tissue sectioning and imaging. R-CNP@M were mostly deposited in alveoli; only a few microcapsules were evident in trachea and bronchi (Fig. 4b, c). The mLNs were also collected for immunofluorescence staining, and the GC B cells were observed as clearly proliferative (Extended Data Fig. 9a).

After verifying the safety in nonhuman primates (Extended Data Fig. 9b, c), we investigated systemic and mucosal responses against SARS-CoV-2. Five monkeys received R-CNP@M (50 μg equivalent RBD, once) via DPI; five other monkeys received PBS. In peripheral blood, the proportion of granzyme B⁺ CD8⁺ T cells was significantly increased in the R-CNP@M group compared to the PBS group (Extended Data Fig. 9d). In terms of the serological antibody responses, the anti-RBD IgG titers remained greater than 10⁴ on Day 84, with higher antibody quality than that on Day 14 (Fig. 4d, e). Regarding the antibody responses in BALF, high titers of RBD-specific IgA and IgG were also detected in the R-CNP@M group (Fig. 4f). Note that both serum and BALF samples showed good performance against pseudotyped viruses and live viruses (Fig. 4g, h, Extended Data Fig. 9e), which reinforced the value of the R-CNP@M vaccine for effective protection against SARS-CoV-2 challenge.

Mosaic iteration for extended immune responses

Given that our CNP chassis can display user-selected antigens (using SpyCatcher/SpyTag), we were interested in codisplaying both wild-type RBD (R_W) and Omicron’s RBD (R_O), seeking to meet the challenge of a two strain cocirculation scenario. The resulting mosaic R_WR_O-CNPs were loaded in our self-healing microcapsules (R_WR_O-CNP@M); R_W-CNP@M and R_O-CNP@M were prepared for comparison (Fig. 5a). With these three vaccines, we evaluated the prophylactic antibodies in mice on Day 28 after a single dose of dry powder aerosol administration (10 µg equivalent RBD).

The R_WR_O-CNP@M group exhibited high levels of both anti-RBD_W and anti-RBD_O titers, whereas the titers for the unmatched RBD were highly compromised in the R_W-CNP@M and R_O-CNP@M groups (Fig. 5b). The R_WR_O-CNP@M vaccine performed well in pseudovirus neutralization for both the wild-type strain and the Omicron variant, whereas serum from mice immunized with R_W-CNP@M or R_O-CNP@M failed to neutralize the pseudotyped viruses of the unmatched strain (Fig. 5c), even though the serological IgG antibody titers were found to be greater than 10⁴. For BALF samples, similar results were observed for IgA and IgG titers, enabling efficient pseudovirus neutralization for both strains in the R_WR_O-CNP@M group but not the R_W-CNP@M or R_O-CNP@M groups (Fig. 5d, e, Extended Data Fig. 10a). Note that a correlation analysis of pVNT₅₀ and the corresponding antibody titers indicated that the IgA antibody response in BALF was a predictive immune correlate with the pseudovirus neutralizing results (Extended Data Fig. 10b-d).

To provide a clearer visualization of the above data, we performed principal component analysis. After a dimension reduction process of ten immunological indices derived from the data features of antibody reactivities, sample sources, detection method, and antibody types, the data from the three vaccine groups were clearly separated as distinct clusters (Fig. 5f). A heatmap visualization revealed that both the R_W-CNP@M and R_O-CNP@M vaccines showed very low levels of strain-unmatched immunological indices, whereas the Z scores of most immunological indices for our mosaic R_WR_O-CNP@M vaccine were high for both the wild-type strain and the Omicron variant (Fig. 5g). Note that primary immunization with R_W-CNP@M and booster immunization with R_WR_O-CNP@M also resulted in higher RBD binding titers for both wild-type and Omicron (Extended Data Fig. 10e-f). Thus, our mosaic iteration of the Ag-CNP@M design could extend the breadth of Ag-induced antibody responses, which should help in managing future scenarios with cocirculation of multiple variants.

Benefits against host-host transmission

As our inhalable vaccine can induce a robust immune response in the respiratory tract, we were finally interested in assessing potential benefits against host-host transmission. We treated hamsters with R_WR_O-CNP@M (30 μg dose, single inhalation), R_WR_O+Al (15 μg dose, two intramuscular injections), or PBS. Upon confirming that the inhaled vaccination group had superior antibody and NAb responses in both serum and BALF (Extended Data Fig. 10g-i), we tested performance in intervening against Omicron’s spread based on multiple hamster transmission models (Fig. 5h).

In an airborne transmission protection model, we designed a cage divider, which allowed airflow but no direct contact or fomite (including diet and bedding) transmission (Fig. 5i). One donor was inoculated with the live Omicron strain and was introduced into one side of the cage at 24 h after infection. Three recipients, which were administered PBS, inhaled vaccine, or intramuscular vaccine, were placed on the other side (downflow from one infected hamster). After 24 h of exposure to the donor, the inhaled vaccine group had the lowest RNA copy numbers and lowest TCID₅₀ values in lung homogenates at 3-days post exposure (dpe) (Fig. 5j, k) and had the lowest pathological scores (with few if any pulmonary histopathologic lesions) (Fig. 5l). In a contact/airborne transmission protection model, one donor and three recipients were placed in a cage without any barrier (Extended Data Fig. 10j). The inhaled vaccine group again outperformed the intramuscular vaccine group in terms of RNA copy numbers, TCID₅₀ values, and pathological scores (Extended Data Fig. 10k-m).

In an airborne transmission blocking model, the cages we used were similar to the cages from the above airborne transmission protection experiment (Fig. 5m). Each donor hamster prior to inoculation with the live Omicron strain was treated with inhaled vaccine, intramuscular vaccine, or PBS. After 24 h of infection, these donor hamsters were introduced into one side of the cage divider, while two recipients (naïve hamsters) were placed downflow from the infected hamster. After 24 h of exposure, the recipient hamsters were fed separately and sacrificed in 3 days. The recipient hamsters exposed to the vaccinated donors had significantly reduced viral loads in nose homogenates (Fig. 5n). Notably, the infection was detected for 11 recipients (11/12) exposed to donors treated with PBS and 8 recipients (8/12) exposed to the intramuscular vaccine-immunized donors, yet only 3 recipients (3/12) exposed to the inhaled vaccine donors (Fig. 5o, p). These results supported that the robust immune response in the respiratory tract induced by our inhaled vaccine conferred benefit against host-host transmission of the Omicron variant, a capacity that should be useful in managing highly contagious viruses.

Mucosal vaccines are developed rapidly, and several have entered clinical trials or have been approved for use based on urgent need protocols. Intranasal immunization of these vaccines in mice and hamsters has shown superior performance over intramuscular immunization in inducing systemic and mucosal responses ^21,24. However, some viral-vector vaccines showed weak immunogenicity in clinical trials ^23,43. This may be attributable to the poor infectivity to the human nasal epithelium, leading to weak expression of the antigen. Moreover, these vaccines may trigger anti-viral vector backbone immunity, hence potentially hindering the efficacy of repeated vaccinations ⁴⁴. In addition, nasal delivery of these vaccines is largely limited to the nasal passage, whereas aerosol-inhaled vaccines can penetrate deeper and wider (to major and small airways). Beyond supporting the benefits of licensed Ad5-nCoV vaccine upon delivery to the lower respiratory tract ^25,45, our aerosol-inhaled vaccine, combining a recombinant protein antigen with a proteinaceous adjuvant, presents an alternative approach for inducing a robust immune response without concerns of inherent immunogenicity and differential antigen expression among species.

Some promising application prospects of our vaccine warrant discussion. The inhalable vaccination matches a known public health issue in that there is more enthusiasm than for traditional injection, and a single-dose regimen is favourable for substantially increasing the proportion of total completed vaccination recipients ^46,47. Further, our vaccine’s dry powder form can greatly save costs (storage, transportation), potentially supporting increased immunization coverage to remote areas. Additionally, our vaccine’s clinical translation prospects are booted by its use of a proteinaceous nanochassis and a Food and Drug Administration-approved material-based microcapsule.

In summary, we have developed an inhalable, single-dose, dry powder aerosol SARS-CoV-2 vaccine. The suitable aerodynamic diameter of the microcapsules facilitated deposition in the alveoli, while the released nanosized R-CNPs were readily internalized by APCs. Our vaccine induced long-term production of robust IgG in serum and abundant IgA in BALF, and showed efficient protection in mice, hamsters, and nonhuman primates. With the mosaic iteration, we succeeded in extending the breadth of RBD-induced antibody responses in both serum and mucus. Considering the flexibility of displaying antigens on the CNP chassis, we envisioned that our inhalable vaccine could serve as a promising multivalent platform for fighting COVID-19 and other respiratory infectious diseases.

1 Petersen, E. et al. Comparing SARS-CoV-2 with SARS-CoV and influenza pandemics. Lancet Infect Dis 20, e238-e244, doi:10.1016/S1473-3099(20)30484-9 (2020).

2 Wu, F. et al. A new coronavirus associated with human respiratory disease in China. Nature 579, 265-269, doi:10.1038/s41586-020-2008-3 (2020).

3 Lambert, H. et al. COVID-19 as a global challenge: towards an inclusive and sustainable future. Lancet Planet Health 4, E312-E314 (2020).

4 Gao, Q. et al. Development of an inactivated vaccine candidate for SARS-CoV-2. Science 369, 77-81, doi:10.1126/science.abc1932 (2020).

5 Wang, H. et al. Development of an Inactivated Vaccine Candidate, BBIBP-CorV, with Potent Protection against SARS-CoV-2. Cell 182, 713-721 e719, doi:10.1016/j.cell.2020.06.008 (2020).

6 Xia, S. et al. Effect of an Inactivated Vaccine Against SARS-CoV-2 on Safety and Immunogenicity Outcomes: Interim Analysis of 2 Randomized Clinical Trials. JAMA 324, 951-960, doi:10.1001/jama.2020.15543 (2020).

7 Dai, L. et al. A Universal Design of Betacoronavirus Vaccines against COVID-19, MERS, and SARS. Cell 182, 722-733 e711, doi:10.1016/j.cell.2020.06.035 (2020).

8 Hong, S. H. et al. Immunization with RBD-P2 and N protects against SARS-CoV-2 in nonhuman primates. Sci Adv 7, doi:10.1126/sciadv.abg7156 (2021).

9 Yang, J. et al. A vaccine targeting the RBD of the S protein of SARS-CoV-2 induces protective immunity. Nature 586, 572-577, doi:10.1038/s41586-020-2599-8 (2020).

10 Corbett, K. S. et al. Evaluation of the mRNA-1273 Vaccine against SARS-CoV-2 in Nonhuman Primates. N Engl J Med 383, 1544-1555, doi:10.1056/NEJMoa2024671 (2020).

11 Zhang, N. N. et al. A Thermostable mRNA Vaccine against COVID-19. Cell 182, 1271-1283 e1216, doi:10.1016/j.cell.2020.07.024 (2020).

12 Folegatti, P. M. et al. Safety and immunogenicity of the ChAdOx1 nCoV-19 vaccine against SARS-CoV-2: a preliminary report of a phase 1/2, single-blind, randomised controlled trial. Lancet 396, 467-478, doi:10.1016/S0140-6736(20)31604-4 (2020).

13 Hassan, A. O. et al. A Single-Dose Intranasal ChAd Vaccine Protects Upper and Lower Respiratory Tracts against SARS-CoV-2. Cell 183, 169-184 e113, doi:10.1016/j.cell.2020.08.026 (2020).

14 Logunov, D. Y. et al. Safety and immunogenicity of an rAd26 and rAd5 vector-based heterologous prime-boost COVID-19 vaccine in two formulations: two open, non-randomised phase 1/2 studies from Russia. Lancet 396, 887-897, doi:10.1016/S0140-6736(20)31866-3 (2020).

15 Mercado, N. B. et al. Single-shot Ad26 vaccine protects against SARS-CoV-2 in rhesus macaques. Nature 586, 583-588, doi:10.1038/s41586-020-2607-z (2020).

16 Amorij, J. P. et al. Pulmonary delivery of an inulin-stabilized influenza subunit vaccine prepared by spray-freeze drying induces systemic, mucosal humoral as well as cell-mediated immune responses in BALB/c mice. Vaccine 25, 8707-8717, doi:10.1016/j.vaccine.2007.10.035 (2007).

17 Moss, W. J. Measles. Lancet 390, 2490-2502, doi:10.1016/S0140-6736(17)31463-0 (2017).

18 Waltz, E. How nasal-spray vaccines could change the pandemic. Nature 609, 240-242, doi:10.1038/d41586-022-02824-3 (2022).

19 Warfel, J. M. & Merkel, T. J. Bordetella pertussis infection induces a mucosal IL-17 response and long-lived Th17 and Th1 immune memory cells in nonhuman primates. Mucosal Immunol 6, 787-796, doi:10.1038/mi.2012.117 (2013).

20 Chen, J. et al. A live attenuated virus-based intranasal COVID-19 vaccine provides rapid, prolonged, and broad protection against SARS-CoV-2. Sci Bull (Beijing) 67, 1372-1387, doi:10.1016/j.scib.2022.05.018 (2022).

21 Bricker, T. L. et al. A single intranasal or intramuscular immunization with chimpanzee adenovirus-vectored SARS-CoV-2 vaccine protects against pneumonia in hamsters. Cell Rep 36, 109400, doi:10.1016/j.celrep.2021.109400 (2021).

22 Le Nouen, C. et al. Intranasal pediatric parainfluenza virus-vectored SARS-CoV-2 vaccine is protective in monkeys. Cell 185, 4811-4825 e4817, doi:10.1016/j.cell.2022.11.006 (2022).

23 Ponce-de-Leon, S. et al. Safety and immunogenicity of a live recombinant Newcastle disease virus-based COVID-19 vaccine (Patria) administered via the intramuscular or intranasal route: Interim results of a non-randomized open label phase I trial in Mexico. medRxiv, doi:10.1101/2022.02.08.22270676 (2022).

24 Afkhami, S. et al. Respiratory mucosal delivery of next-generation COVID-19 vaccine provides robust protection against both ancestral and variant strains of SARS-CoV-2. Cell 185, 896-915 e819, doi:10.1016/j.cell.2022.02.005 (2022).

25 Wu, S. et al. Safety, tolerability, and immunogenicity of an aerosolised adenovirus type-5 vector-based COVID-19 vaccine (Ad5-nCoV) in adults: preliminary report of an open-label and randomised phase 1 clinical trial. Lancet Infect Dis 21, 1654-1664, doi:10.1016/S1473-3099(21)00396-0 (2021).

26 Jeyanathan, M. et al. Aerosol delivery, but not intramuscular injection, of adenovirus-vectored tuberculosis vaccine induces respiratory-mucosal immunity in humans. JCI Insight 7, doi:10.1172/jci.insight.155655 (2022).

27 Kunda, N. K., Somavarapu, S., Gordon, S. B., Hutcheon, G. A. & Saleem, I. Y. Nanocarriers targeting dendritic cells for pulmonary vaccine delivery. Pharm Res 30, 325-341, doi:10.1007/s11095-012-0891-5 (2013).

28 Patton, J. S. & Byron, P. R. Inhaling medicines: delivering drugs to the body through the lungs. Nat Rev Drug Discov 6, 67-74, doi:10.1038/nrd2153 (2007).

29 Smits, H. H. et al. Cholera toxin B suppresses allergic inflammation through induction of secretory IgA. Mucosal Immunol 2, 331-339, doi:10.1038/mi.2009.16 (2009).

30 Wakabayashi, A., Shimizu, M., Shinya, E. & Takahashi, H. HMGB1 released from intestinal epithelia damaged by cholera toxin adjuvant contributes to activation of mucosal dendritic cells and induction of intestinal cytotoxic T lymphocytes and IgA. Cell Death Dis 9, 631, doi:10.1038/s41419-018-0665-z (2018).

31 Pan, C. et al. Biosynthesis of Self-Assembled Proteinaceous Nanoparticles for Vaccination. Adv Mater 32, e2002940, doi:10.1002/adma.202002940 (2020).

32 Shin, M. D. et al. COVID-19 vaccine development and a potential nanomaterial path forward. Nat Nanotechnol 15, 646-655, doi:10.1038/s41565-020-0737-y (2020).

33 Reddy, S. T. et al. Exploiting lymphatic transport and complement activation in nanoparticle vaccines. Nat Biotechnol 25, 1159-1164, doi:10.1038/nbt1332 (2007).

34 Li, X. et al. Orthogonal modular biosynthesis of nanoscale conjugate vaccines for vaccination against infection. Nano Res 15, 1645-1653, doi:10.1007/s12274-021-3713-4 (2022).

35 Heida, R., Hinrichs, W. L. & Frijlink, H. W. Inhaled vaccine delivery in the combat against respiratory viruses: a 2021 overview of recent developments and implications for COVID-19. Expert Rev Vaccines, 1-18, doi:10.1080/14760584.2021.1903878 (2021).

36 Pulliam, B., Sung, J. C. & Edwards, D. A. Design of nanoparticle-based dry powder pulmonary vaccines. Expert Opin Drug Deliv 4, 651-663, doi:10.1517/17425247.4.6.651 (2007).

37 Xi, X. B. et al. Self-healing microcapsules synergetically modulate immunization microenvironments for potent cancer vaccination. Science Advances 6, doi:ARTN eaay7735, doi:10.1126/sciadv.aay7735 (2020).

38 Xie, X. L. et al. Therapeutic vaccination against leukaemia via the sustained release of co-encapsulated anti-PD-1 and a leukaemia-associated antigen. Nat Biomed Eng 5, 414-+, doi:10.1038/s41551-020-00624-6 (2021).

39 Liao, M. et al. Single-cell landscape of bronchoalveolar immune cells in patients with COVID-19. Nat Med 26, 842-844, doi:10.1038/s41591-020-0901-9 (2020).

40 Toor, S. M., Saleh, R., Sasidharan Nair, V., Taha, R. Z. & Elkord, E. T-cell responses and therapies against SARS-CoV-2 infection. Immunology 162, 30-43, doi:10.1111/imm.13262 (2021).

41 Grifoni, A. et al. Targets of T Cell Responses to SARS-CoV-2 Coronavirus in Humans with COVID-19 Disease and Unexposed Individuals. Cell 181, 1489-1501 e1415, doi:10.1016/j.cell.2020.05.015 (2020).

42 Yu, D., Walker, L. S. K., Liu, Z., Linterman, M. A. & Li, Z. Targeting T(FH) cells in human diseases and vaccination: rationale and practice. Nat Immunol 23, 1157-1168, doi:10.1038/s41590-022-01253-8 (2022).

43 Madhavan, M. et al. Tolerability and immunogenicity of an intranasally-administered adenovirus-vectored COVID-19 vaccine: An open-label partially-randomised ascending dose phase I trial. EBioMedicine 85, 104298, doi:10.1016/j.ebiom.2022.104298 (2022).

44 Zhu, F. C. et al. Immunogenicity and safety of a recombinant adenovirus type-5-vectored COVID-19 vaccine in healthy adults aged 18 years or older: a randomised, double-blind, placebo-controlled, phase 2 trial. Lancet 396, 479-488, doi:10.1016/S0140-6736(20)31605-6 (2020).

45 Li, J. X. et al. Safety and immunogenicity of heterologous boost immunisation with an orally administered aerosolised Ad5-nCoV after two-dose priming with an inactivated SARS-CoV-2 vaccine in Chinese adults: a randomised, open-label, single-centre trial. Lancet Respir Med 10, 739-748, doi:10.1016/S2213-2600(22)00087-X (2022).

46 Zhou, Q. T., Tang, P., Leung, S. S., Chan, J. G. & Chan, H. K. Emerging inhalation aerosol devices and strategies: where are we headed? Adv Drug Deliv Rev 75, 3-17, doi:10.1016/j.addr.2014.03.006 (2014).

47 Walters, A. A., Krastev, C., Hill, A. V. S. & Milicic, A. Next generation vaccines: single-dose encapsulated vaccines for improved global immunisation coverage and efficacy. J Pharm Pharmacol 67, 400-408, doi:10.1111/jphp.12367 (2015).

Cell lines

Vero, HEK293T and HEK293T-hACE2 cells were cultured in Dulbecco’s modified Eagle medium (DMEM, high glucose; BI) supplemented with 100 U/mL of penicillin, as well as 100 μg/mL streptomycin solution (VivaCell), and 10% fetal bovine serum (FBS, GIBCO) in a 5% CO₂ environment at 37°C. Trypsin-EDTA (0.25%, BI) was used to detach cells for subculturing every 2-3 days. Bone marrow-derived cells (BMDCs) and peritoneal macrophages were derived from female BALB/c mice (6-8 weeks old) using standard techniques.

Animals

Female BALB/c mice (6-8-weeks old) and female K18-hACE2 mice were purchased from Beijing Vital River Laboratories and GemPharmatech respectively, and were randomly divided into various groups for subsequent experiments. This study was performed in strict accordance with the Regulations for the Care and Use of Laboratory Animals and Guideline for Ethical Review of Animal (China, GB/T 35892-2018). BALB/c mice were reviewed and approved by the Animal Ethics Committee of the Institute of Process Engineering (approval ID: IPEAECA 2020061). K18-hACE2 mice experiments were reviewed and approved by the Peking Union Medical College Animal Care and Use Committee (approval ID: DWSP202302 001).

Female Syrian Golden hamsters were purchased from Beijing Vital River Laboratories, and were randomly divided into various groups for subsequent experiments. All animal experiments were reviewed and approved by the Peking Union Medical College Animal Care and Use Committee (approval ID: DWSP202207 001, DWSP202302 001).

A total of 10 cynomolgus monkeys were purchased from the Animal Center of the Academy of Military Medical Sciences and were also housed in the Center. This study was performed in strict accordance with the Regulations for the Care and Use of Laboratory Animals and Guideline for Ethical Review of Animal (China, GB/T 35892-2018). All animal experiments were approved and conducted in accordance with the institutional guidelines of the Academy of Military Medical Sciences Institutional Animal Care and Use Committee (approval ID: IACUC-DWZX-2021-058)

Human sera

Sera from 5 convalescent patients (admission date from 2020/1/21 to 2020/1/26, discharge date from 2020/2/1 to 2020/2/13, follow-up after two weeks; severity grading: mild or moderate) were provided by Beijing You An Hospital, Capital Medical University. Written informed consent was obtained from each individual for serum collection. This study was reviewed and approved by the Clinical Ethics Committee of Beijing You An Hospital, Capital Medical University (approval ID: LL-2022-063-K).

METHOD DETAILS

Protein expression and purification

ST-CNP was expressed and purified from Escherichia coli. Briefly, DNA sequences of ST-CNP were cloned to pET28a vector in the following order: SpyTag, GSGTAGGGSGS linker, CTB, GGSG linker, trimer-forming peptide, and 6×His tag. The construct was transformed into E. coli BL21. A Single colony was amplified in 10 mL Luria-Bertani (LB) with kanamycin (50 μg/mL) for 12 h at 37℃ while shaking. The preculture was diluted 1:100 in 1L LB medium (50 µg/mL kanamycin) and cultured at 200 rpm at 37°C, until OD₆₀₀~0.6. Protein expression was induced with isopropyl β-D-1-thiogalactopyranoside (IPTG) and grown at 30 °C, 200 rpm for 12 h. The culture was centrifuged and the pellet was resuspended in 20 mM Tris-HCl (pH8.0), 300 mM NaCl, 8 M Urea, and 20 mM Imidazole. The cell suspension was homogenized and centrifuged at 10,000 × g for 20 min at 4 °C. The precipitate was incubated with His-Tag purification resin column to enrich ST-CNP, followed by protein elution with wash buffer (20 mM Tris–HCl, pH 8.0, containing 300 mM NaCl, 8 M Urea, and 300 mM Imidazole). The purified protein was dialyzed and finally the buffer was replaced with conventional PBS buffer. The concentration of ST-CNP was determined by BCA assay.

SC-RBD was expressed and purified from HEK293T. Briefly, DNA sequences of SC-RBD were cloned to pcDNA3.4 vector in the following order: SpyCatcher, GGSG linker, RBD and 6×His tag. The construct was transformed into HEK293T cells. Five days later, supernatants were collected and centrifuged at 500 g for 10 min at 4 °C. The cleared supernatants were incubated through His-Tag purification resin column to enrich SC-RBD proteins, followed by elution with Imidazole-containing Tris buffer. The purified proteins were concentrated and buffer-replaced with PBS buffer. The concentrations of SC-RBD were determined by BCA assay.

For vaccine generation, SC-RBD was incubated with ST-CNP in PBS buffer and reacted overnight at 4℃. The ST/SC conjugated proteins were proceeded to size-exclusion chromatography (SEC). The elution of nanoparticles was concentrated, and the concentration was measured by BCA assay. Coomassie blue staining, western blotting against His tag and CTB, transmission electron microscopy, and dynamic light scattering, were executed to confirm the purity and homogeneity.

Surface Plasmon Resonance (SPR) assay

Interactions between the hACE2 protein and R-CNP were analyzed using the Biacore 8K system (Cytiva) at 25 °C. RBD or R-CNP were immobilized in pairs onto CM5 chips. Subsequently, gradient concentrations of hACE2 (from 7.813 nM to 500 nM for SARS-CoV-2 RBD and R-CNP) were then used to flow over the chip surface at 30 μL/min with a contact time of 120 s and a dissociation time of 300 s. After each cycle, the sensor surface was regenerated using 3 M magnesium chloride. Data were analyzed with the Biacore Insight Evaluation Software 3.0.12 by curve fitting using a 1:1 binding model.

Fabrication and characterization of R-CNP@M

Membrane emulsification

Briefly, given concentration (from 0.05% to 0.5%, w/v) NaCl solution was added into 12 mL of ethyl acetate containing 100 mg/ml compound (90 mg poly(DL-lactic-co-glycolic acid) and 10 mg poly(ethylene glycol)-b-poly-dl-lactide) to prepare the primary emulsion by ultrasonic equipment (120W, Digital Sonifier 450, Branson Ultrasonics Corp, USA). Then the emulsion was mixed with 90 ml 3% (w/v) PVA solution for membrane emulsification. Microporous tubular membranes with pore sizes of 5 μm, 8 μm, 12 μm, and 15 μm were used in the experiments. The membrane was wetted in the continuous phase under ultrasonic field before the installation. Membrane emulsifier (Senhui Microsphere Tech Co., Ltd., China) controls the emulsification pressure precisely and stably through the pressure sensor. The O/W emulsion was extruded through microporous membrane three times with a certain pressure under 250 rpm magnetic stirring, following pre-solidification by vertical suspension at 45 rpm for 30 min and solidification in 500 mL deionized water by magnetic stirring at 200 rpm for 10 min. The obtained microspheres were collected by centrifugation, following three washes with deionized water. The R-CNP were mixed with microspheres and incubated at 4 ^oC overnight, followed by 2 h of healing at approximately 40 ^oC using incubator (DGX-9143B-1, Shanghai Fuma Test Equipment Co., Ltd, China).

Characterization of microcapsules

The filled microcapsules came freeze-dried by freeze dryers (Alpha 2-4 LDplus, Martin Christ Gefriertrocknungsanlagen GmbH, Germany). The particle sizes of microsphere and microcapsule were analyzed by laser diffraction using Mastersizer (Mastersizer 2000, Malvern Instruments Ltd, UK). The aerodynamic particle diameter of lyophilized microcapsule was determined by the Aerodynamic Particle Sizer (TSI 3321, TSI Incorporated, USA). The shape and surface morphology of microsphere, microcapsule, and freeze-dried powder was observed using scanning electron microscopy (JSM-6700, JEOL, Japan). To obtain the distribution of nanoparticles in the cavities of the microcapsule, nanoparticles and microcapsules were labelled with FITC-SE and Nile red and observed by using confocal laser scanning microscopy (A1/SIM/STORM, Nikon, Japan). Protein concentrations of supernatant after encapsulation were determined by Micro BCA Protein Assay Kit (PI23235, Thermo Scientific) following the corresponding protocols for evaluation of loading capacity (mass of R-CNP in microcapsule/mass of microcapsules × 100%) and encapsulating capacity (mass of R-CNP in microcapsule/mass of total R-CNP × 100%).

In vivo antigen release

To evaluate the duration of nanoparticles with/without microcapsules in vivo, nanoparticle was labeled with near-IR water-soluble fluorescence probe Cy7-NHS ester (Cy7-SE), and microcapsule was labeled with hydrophobic Cy5 by mixing into the framework during primary emulsion. The inhaled vaccines containing Cy7-nanoparticles and Cy5-microcapsules were delivered into the lung using a customized dry powder inhaler for mice (DPI, Beijing Huironghe Technology). At different time intervals, the mice were captured using an In Vivo Imaging System (IVIS) optical imaging system (Caliper Life Sciences Inc., PerkinElmer, USA), and euthanized to extract organs following to capture again at the end of this experiment.

Immunofluorescence analysis

The distribution of R-CNP@M in lung

The animals were vaccinated with Nile Red-labeled R-CNP@M. The lungs were fixed with 4% paraformaldehyde and then dehydrated with 30% sucrose solution for subsequent sectioning on day 5 post-inhalation. Then 10 µm lung slides were prepared by freezing microtome (CM1950, Leica, Germany). Slides were mounted for imaging by the Vectra platform (v 3.0.5, PerkinElmer, USA) and counted by inform software.

GC production of monkey mLNs

The mLNs of the cynomolgus monkeys were harvested on day 14 after immunization, the mLNs were fixed with 4% paraformaldehyde and then embedded in paraffin blocks for serial sectioning. Then 5 µm mLN slides were prepared for immunofluorescence staining. Microwave treatment was conducted with ethylenediaminetetraacetic acid (EDTA) antigen retrieval buffer to retrieve antigen, followed by blocking with 5% (w/v) bovine serum albumin for 30 min. The sections were stained with anti-B220 (mouse), and anti-Ki67 (rat) at 4℃ overnight. The corresponding fluorescent secondary antibodies were incubated for 1 hour at room temperature. Counterstain was done using DAPI, and slides were mounted for imaging by the Vectra platform.

Light-sheet imaging

To verify whether microcapsules uniformly distribute over the whole lung, lung trachea and microcapsules were labeled with streptavidin-AF647 and Nile red, respectively. The whole lung was processed with Tissue-Clearing Reagent and captured by light-sheet microscopy (Z.1, Zeiss, Germany). Image reconstructions were generated using Imaris v9.0 software.

To get a visualization of germinal center in mLNs, B cells were labelled with B220-AF647. The mLNs were profile by CD31-FITC antibody. The mLNs were processed with Tissue-Clearing Reagent and captured in CUBIC-R⁺ reagent by light-sheet microscopy (Light Innovation Technology Limited.). Laser power and gain were adjusted consistently for each group. Image reconstructions were generated using LitScan v3.0 software.

Immunization

Dry powder inhaler

To simulate the process of human dry powder inhalation, we developed a dry powder inhaler (DPI, Supplementary Figure 2), which comprised the dry powder aerosol generator (DPAG) (Beijing Huironghe Technology Co., Ltd.), a cushion chamber and an endotracheal tube. To ensure that the dry powder aerosol is inhaled into the lungs through the mouth and trachea, the cushion chamber was designed with one end connected to an endotracheal tube and the other end connected to the DPAG. The volumes of the cushion chamber are distinguished for different animal species (about 35 mL, 80 mL, and 700 mL for mouse, hamster, and monkey, respectively). During immunization, the endotracheal tube of DPI is inserted into the animal's trachea, and a needle free syringe is used as the gas supply source to inject the dry powder vaccine in DPAG into the cushion chamber with the form of dry powder aerosol. The animal is separated from the DPI device after continuously inhaling the dry powder vaccine for 2 minutes. Note that mouse and hamster can be assisted with a small animal ventilator to assist when using the DPI device.

Immunization for mice

For long-term immunological evaluation, six female BALB/c mice (6-8-weeks old) were immunized twice with R-CNP, which contain 5 μg equivalent SARS-CoV-2 RBD protein intratracheally through a customized liquid aerosol delivery device for mice (Beijing Huironghe Technology); the second dose was the same as the first dose and given on day 14 post-prime vaccination. Six female BALB/c mice (6-8-weeks old) were immunized once with R-CNP@M, which contain 10 μg equivalent SARS-CoV-2 RBD protein intratracheally through DPI for mice. Six female BALB/c mice (6-8-weeks old) were immunized twice with CNP as control.

For the evaluation of mucosal response, another batch of mice were immunized with CNP, R-CNP and R-CNP@M as above described. Six female BALB/c mice (6-8-weeks old) were immunized with 5 μg SARS-CoV-2 RBD protein twice, combined with 100 μg aluminum adjuvant (RBD+Al) through intramuscular route as control, and the second dose was the same as the first dose and given on day 14 post-prime vaccination. On day 70 post-immunization, these mice were sacrificed for the harvest of BALF and serum. The obtained samples were subsequently inactivated at 56 ℃ for pseudovirus neutralization assay.

For the live SARS-CoV-2 challenge, another batch of mice were vaccinated with PBS, RBD+Al and R-CNP@M as above described.

Immunization for Syrian golden hamsters

Six-to-seven weeks old female Syrian Golden hamsters were immunized once with R-CNP@M, which contain 30 μg equivalent SARS-CoV-2 RBD protein intratracheally through DPI. Hamsters were intramuscularly immunized three times with licensed recombinant protein subunit vaccine Zifivax, which has been approved for use in China, Colombia, Indonesia, and Uzbekistan. The formulation contained 10 μg equivalent of SARS-CoV-2 dimer RBD protein. Blood samples were taken every 14 days, and BALF were collected on the 42nd day. The obtained samples were then subjected to ELISA and live virus neutralization tests. For the live SARS-CoV-2 challenge, another batch of hamsters were vaccinated with PBS, Zifivax and R-CNP@M as above described.

To evaluate the effectiveness of iterative vaccines, six-to-seven weeks old female Syrian Golden hamsters were immunized once with iterative vaccine (R_WR_O-CNP@M), which contain 30 μg equivalent SARS-CoV-2 RBD protein intratracheally through DPI. Hamsters were immunized twice with R_WR_O+Al, which contains 15 μg equivalent SARS-CoV-2 RBD protein intramuscularly. Blood samples and BALF were collected on the 42nd day. For the live SARS-CoV-2 challenge, another batch of hamsters were vaccinated with PBS, R_WR_O+Al and R_WR_O-CNP@M as above described.

Immunization for cynomolgus monkey

A total of 10 cynomolgus monkeys (female, 4-5-years old) were divided into PBS group and R-CNP@M group. All five animals were immunized once with R-CNP@M, which contain 50 μg equivalent SARS-CoV-2 RBD protein through DPI. Blood was collected every 2 weeks. The BALF was collected with assistance of bronchoscope at day 56.

Pulmonary single-cell RNA sequencing

The female BALB/c mice (6-8-weeks old) were immunized with CNP, R-CNP or R-CNP@M as described above, and mice were sacrificed on day 21 post-prime vaccination. The lungs were harvested and minced with scissors, following treated with collagenase I (1 mg/mL) and DNase I (0.05 mg/mL) and dissociated to a single-cell suspension. After being treated with red blood cell lysing reagent, the single-cell suspension was passed through a 40 μm strainers and resuspended in PBS. The single-cell suspension of four mice in each group was pooled. The CD45⁺ cells were isolated from single-cell suspension in each group by CD45 MicroBeads, and then these CD45⁺ single-cell suspensions containing about 10000 cells in each group were loaded on a 10X Genomics GemCode Single-cell instrument to generate single-cell Gel Bead-In-EMlusion (GEMs). Libraries were generated and sequenced from the cDNAs with Chromium Next GEM Single Cell 3’ Reagent Kits v3.1. The Gene Denovo Biotechnology Co. were entrusted to perform GEMs generation, cDNA amplification, library preparation, and sequencing in this study. Overall, 10156 cells from the PBS group, 6400 cells from the CNP group, 8400 cells from the R-CNP group, and 11753 cells from the R-CNP@M group passed the quality control threshold of more than 200 genes were identified in each cell.

BCR repertoires sequencing

RNA samples were obtained from pulmonary lymphocytes by cell sorting and subsequently analyzed by High-throughput sequencing of IGH using the ImmuHub® BCR profiling system at a deep level (ImmuQuad Biotech, Hangzhou China). Briefly, a 5’ RACE unbiased amplification protocol was used. This protocol uses unique molecular barcodes (UMBs) introduced in the course of cDNA synthesis to control bottlenecks and to eliminate PCR and sequencing errors. Sequencing was performed on an Illumina NovSeq® system with PE150 mode (Illumina). One common adaptor with UMB was added on the 5’ of cDNA during the first-strand cDNA synthesize and one reverse primer corresponding to the constant (C) regions of each of the IGH were designed to facilitate PCR amplification of cDNA sequences in a less biased manner. The UMB attached to each raw sequence reads were applied to correct PCR and sequencing errors correction and PCR duplicates removing. Map V, D, J and C segments with IMGT® and then extract CDR3 regions and assemble clonotype for all clones. The resulting nucleotide and amino acid sequences of CDR3 of IGH were determined and those with out-of-frame and stop codon sequences were removed from the identified IGH repertoire. We further defined amounts of each IGH clonotype by adding numbers of IGH clones sharing the same nucleotide sequence of CDR3.

Flow cytometry analysis

In vitro Activation of APC

To evaluate the activation of primary DCs and macrophages resulting from RBD, the mixture of RBD and CTB, and R-CNP, BMDCs and peritoneal macrophages were plated at 1.0 × 10⁵ cells/well in a 24-well plate respectively, and then stimulated with RBD, CTB+RBD, and R-CNP for 24 h. Non-stimulated BMDCs and peritoneal macrophages were used as a negative control. Cells were collected and stained with anti-CD11c-FITC, anti-CD40-PE, anti-CD86-PE-Cy7, anti-MHC I-Percp-cy 5.5, and anti-MHC II-Alexa Fluor 700 at 4 °C for 30 min. After washing, the cells were resuspended in 300 μL of staining buffer and analyzed using a CytoFLEX LX flow cytometer (Beckman Coulter, USA). The cell culture supernatants of BMDCs and peritoneal macrophages were collected to assess the levels of IL-6, TNF-α, IL-12, IFN-γ, and MCP-1 by a CBA Mouse Inflammation Kit.

Analysis of antigen uptake by APC in lung

To investigate antigen uptake by APC in lung, flow cytometry was executed. After 10 days post-vaccination with microcapsule vaccine (Cy5-labeled R-CNP), single-cell suspensions from the lungs were prepared by mechanical disruption and isolated from red cells using red blood cell (RBC) lysis buffer, following centrifugation and resuspension with staining buffer. A part of the collected cells was stained with anti-CD11c-AF700, anti-CD64-BV421, anti-MERTK-PE and anti-CD11b-BV510 to define DCs/macrophages, and other cells were stained with anti-CD3-BV605 and anti-CD19-APC-Cy7 to define T cells/B cells. The proportion of APC that internalized antigen (Cy5⁺F4/80⁺ or Cy5⁺CD11c⁺) was measured using CytoFLEX LX flow cytometer (Beckman Coulter, USA), and analyzed using CytExpert software (version 2.3).

Analysis of activated APC in mLNs

To discover the activation of homed APC in mLNs, flow cytometry was executed. After 21 days post-vaccination, single-cell suspensions from the mLN were prepared by mechanical disruption and isolated from red cells using red blood cell (RBC) lysis buffer, following centrifugation and resuspension with staining buffer. The collected cells were stained with anti-CD11c-BV650 and anti-F4/80-FITC to define DCs/macrophages, and anti-CD86-PE-Cy7 as activation markers. The proportion of activated APC was measured using CytoFLEX LX flow cytometer (Beckman Coulter, USA), and analyzed using CytExpert software (version 2.3).

Analysis of T cells and B cells in mice

To evaluate T cell responses in lung, mLNs, NALT and nose of mice, flow cytometry was executed (Supplementary Figure 3). Mice were sacrificed and the tissues were harvested. Then tissues were mechanically ground to prepare single-cell suspensions and divided extra several parts. The following antibodies were used to detect GC B cells and CD4⁺ T cells in mLNs: anti-CD3-FITC, anti-CD4-PE-Cy7, anti-CD19-APC-Cy7, anti-GL7-AF647, anti-IgD-BV421. To evaluate memory T cell responses, tissue resident T cell responses and multifunctional T cell responses, cells were stained as follows. (1) The following antibodies were used to detect T_RM CD8⁺ T cell in lung: anti-CD3-BV605, anti-CD4-PE-Cy7, anti-CD8-APC-Cy7, anti-CD44-PE, anti-CD69-APC, anti-CD103-AF700, anti-CD11a-FITC, and anti-IFN-γ-PE antibodies (2) The following antibodies were used to detect IFN-γ⁺ cells in NALT and nose: to anti-CD3-FITC, anti-CD4-PE-Cy7, anti-CD8-BV605, anti-IFN-γ-PE antibodies. (3) The following antibodies were used to detect multi-functionality T cell in spleen and lung: anti-CD3-FITC, anti-CD8-APC-Cy7, anti-IFN-γ-PE, anti-TNF-α-BV510, anti-IL2-PE-Cy5, (4) The following antibodies were used to detect T_EM in spleen: anti-CD3-BV605, anti-CD4-PE-Cy7. anti-CD8-APC-Cy7, anti-CD44-FITC, anti-CD62L-APC antibodies. The stained cells were washed with PBS and measured using CytoFLEX LX flow cytometer, and analyzed using CytExpert software.

Analysis of T cells in monkey

To evaluate T cell responses in monkeys, flow cytometry was executed. First, about 1 mL of peripheral blood was taken from each monkey on day 21, and the lymphocytes were separated using Monkey blood lymphocyte separation Kit. Then the cells were stained with anti-CD8-Alexa Fluor 700 for 30 min at 4°C. Then the cells were washed twice and fixated with a fixative solution for 30 min, centrifugation to collect cells. Permeabilization reagent was then added to suspend the cells and incubated with anti-Granzyme B-Pacific Blue for 30 min at 4°C. The stained cells were washed with PBS and analyzed by flow cytometry (CytoFLEX, Beckman Coulter, USA).

IFNγ ELIspot assay

The assay was performed according to the manufacturer’s instructions (Mabtech, Sweden). Briefly, standard 96-well plates were incubated with anti-mouse IFN-γ antibody with a final concentration of 15 μg/mL overnight at 4 ℃. On day 21 post-vaccination, mice were sacrificed and the spleen was collected to prepare a single-cell suspension. After red-cell lysis and cell counts, these splenocytes were plated at 5 × 10⁵ per well, in triplicate, with peptide at 5 μM. After overnight restimulation at 37°C in a cell incubator, plates were washed with cold ultrapure water and incubated with biotinylated anti-mouse IFN-γ antibody for 2 hours at room temperature. Then the plates were washed with PBST five times and incubated with the streptavidin-ALP conjugated antibody for 1 hour. After extensive washing, 100 μL of substrate (bromochloroindolyl phosphate-1-nitro blue tetrazolium step solution, Pierce, USA) was added into each well of plates for the coloration of IFN-γ spots. The plates were thoroughly washed with plenty of ultrapure water to stop the reaction. After complete air-dry of these plates, spots were counted using an automated ELISpot reader (AT-Spot 2100, Antai Yongxin Medical Technology, China).

Enzyme-linked immunosorbent assay (ELISA)

Binding properties of serum and BALF from BALB/c mice, hamsters and monkey to SARS-CoV-2 RBD protein were determined by ELISA. 96-well plates (Corning) were coated with 2 μg/mL of RBD protein (without his tag, 40592-VNAH, from Sino Biological, China) overnight at 4 ℃ in coating buffer, and blocked in 5% bovine serum albumin in PBS for 2 hours at 37℃. Serially diluted serum or BALF samples were added to the ELISA plates and incubated for 1 hour at 37℃. The plates were washed five times with 1× PBS containing 0.05% Tween-20 (PBST), followed by addition of 100 µL of HRP-conjugated goat anti-mouse IgM, goat anti-mouse IgG, goat anti-mouse IgG1, goat anti-mouse IgG2a, goat anti-mouse IgA, goat anti-hamster IgG, goat anti-monkey IgG secondary antibody with a 1:10000 dilution and goat anti-monkey IgA (4A Biotech) secondary antibody with a 1:5000 dilution. After being incubated at 37℃ for 0.5 hours, plates were again washed five times with PBST. Then the plates were developed with 3,3′,5,5′-tetramethytlbenzidine for 10 min at room temperature, and the reaction was stopped with 2 M H₂SO₄. The absorbance at 450 nm was measured by a microplate reader. The endpoint titer was defined as the highest reciprocal dilution of detected sample to give an absorbance greater than 2.1-fold of the OD₄₅₀value from the PBS group.

Pseudovirus neutralization assay

The SARS-CoV-2 pseudotyped viruses (for Figure 4), which were expressing a luciferase reporter gene, were gifted from Professor Youchun Wang of Institute for Biological Product Control, National Institutes for Food and Drug Control (NIFDC), and WHO Collaborating Center for Standardization and Evaluation of Biologicals, Beijing, China. The SARS-CoV-2 pseudotyped viruses based neutralization assay was performed as described previously^48,49. The pseudotyped virus used in this study was developed based on vesicular stomatitis virus (VSV). All samples were heat-inactivated to eliminate any complement activity. In brief, 293T-hACE2 cells were seeded in 96-well plates (20,000 cells/well) and incubated with 200 TCID₅₀ (50% tissue culture infective dose) SARS-CoV-2 pseudotyped viruses, which premixed with serial dilutions of the serum (starting at 16×, 3-fold serial dilutions) or BALF (starting at 3×, 3-fold serial dilutions) for 1 hour. After 24 hours of incubation, the luminescence of each well was detected. The sample pVNT₅₀ was calculated using the Reed-Muench method.

The SARS-CoV-2 pseudotyped viruses (for Figure 6 and 7), which were expressing green fluorescent protein (GFP) gene, were gifted from Professor Xin Zhao of Institute of Microbiology, Chinese Academy of Sciences, Beijing, China. All samples were heat-inactivated to eliminate any complement activity. In brief, Vero cells were seeded in 96-well plates (20,000 cells/well) and incubated with SARS-CoV-2 pseudovirus-GFP for 1000 IU/well, which premixed with serial dilutions of the serum (starting at 16×, 3-fold serial dilutions) or BALF (starting at 3×, 3-fold serial dilutions) for 1 hour. After 24 hours of incubation, the counts of positive cells in each well were detected by high content analysis (Operetta CLS, Perkin Elmer). The sample pVNT₅₀ was calculated using the Reed-Muench method.

Live virus neutralization assay

This study was performed as described previously in a biosafety level 3 (BSL3) laboratory following institutional biosafety guidelines by the Sinovac Life Sciences Co., Ltd. All samples were heat-inactivated to eliminate any complement activity. In brief, the serial dilutions of the serum (starting at 10×, 2-fold serial dilutions) or BALF (starting at 2×, 2-fold serial dilutions) were mixed with a SARS-CoV-2 virus suspension (SARS-CoV-2/human/CHN/CN1/2020, GenBank number MT407649.1 and SARS-CoV-2 Delta variants, GISAID number Delta EPI_ISL_1911197) of 100 CCID₅₀ in 96-well plates, followed by 2h incubation at 36.5 °C in a 5% CO₂ incubator. Then the Vero cell were added into the 96-well plates at 36.5°C in 5% CO₂ for 5 days. Cytopathic effect (CPE) of each well was recorded under microscopes, and the neutralizing titer was the reciprocal of the highest serum dilution with 50% cytopathic effect. Geometric mean titer (GMT) was calculated based on these data.

In vivo cell depletion

Depletion of CD8⁺ T cells was achieved by intraperitoneal (i.p.) injection of 300 µg/mouse YST-169.4 (anti-CD8a) monoclonal antibodies (YTS 169.4, BE0117, BioXcell, West Lebanon, NH), at day 6 before vaccination with R-CNP@M and every 3 days until SARS-COV-2 challenging. For B cell depletion, 200 µg/mouse anti-CD19 (1D3, BE0150, BioXcell) and 200 µg/mouse anti-B220 (RA3.3A1/6.1, BE0067, BioXcell) were injected i.p. at day 6 before vaccination with R-CNP@M and every 3 days until SARS-COV-2 challenging. A mixture of isotype antibodies (2A3, BE0089; LTF-2, BE0090; BE0094.) were injected i.p. at day 6 before administration with PBS and every 3 days until SARS-COV-2 challenging.

Live SARS-CoV-2 virus challenge

The preimmunized K18-hACE2 mice were challenged with 1×10³ TCID₅₀ of wild-type strain (SARS-CoV-2-KMS1/2020 /GenBank accession number: MT226610.1) via the intranasal route at day 28. The body weight and number of deaths of mice was recorded every day post infection (dpi), and after being euthanized on 4 dpi, the lung and nose tissues were harvested for virus RNA levels (about 0.1-gram lung or nose tissues) determination and pathological examination (the left lung).

The Syrian Golden hamsters were challenged with 2×10⁴ TCID₅₀ of wild-type strain (SARS-CoV-2-KMS1/2020 /GenBank accession number: MT226610.1) via the intranasal route at day 42. Hamsters were euthanized and necropsied at day 3 post-challenge. The lung and nose tissues were harvested for virus RNA levels (about 0.2-gram lung) determination, viral titer assessment and pathological examination.

Airborne transmission protection was examined by housing hamsters in specially designed cages with a cage divider dividing the living space in half. The donor animal was inoculated with 1×10⁵ TCID₅₀of the live Omicron strain (CCPM-B-V-049-2112-18) via the intranasal route and was introduced into one side of the cages at 24 h after infection. Three recipients, which were respectively vaccinated with PBS (mock group), mixed antigen via intramuscular injection and mosaic vaccine via inhalation, were placed on another side of the divider downflow from one infected hamster. Hamsters were sacrificed 1- and 3-days post-exposure, and lung tissues were obtained for viral load and pathological sections.

As to contact/airborne transmission protection, one donor and three recipients were placed in a cage without any barrier, sharing the diet, bedding, and airiness. Donor animals were inoculated with 1×10⁵ TCID₅₀ of the live Omicron strain (CCPM-B-V-049-2112-18) by intranasal route, and 24 h after infection, they were raised in the same cage with three recipients for 24 h, and then sacrificed on the 1st and 3rd day to collect lung tissues.

In airborne transmission blocking model, the cages used were the same as those used in the airborne transmission protection model. One of the donor hamsters (given PBS, inhaled vaccine, or intramuscular vaccine) was inoculated with 2×10⁴ TCID₅₀of live Omicron strain (CCPM-B-V-049-2112-18) via the intranasal route, and was introduced into one side of the cage divider at 24 h after infection. Two recipients (naïve hamsters) were placed on the other side of the divider (downflow from the infected hamster). After 24 h of exposure to the donor, the recipient hamsters were taken out and raised in a clean individually ventilated cage (IVC). The recipient hamsters were sacrificed at 72 h post exposure (not including the initial 24 hours of exposure), and the viral loads of their nose homogenates were used for further analysis. Note that all of hamsters were survived until were sacrificed post challenged with SARS-CoV-2.

RT-qPCR

SARS-CoV-2 viral RNA in lung and nose tissues from challenged mice and hamsters were detected by quantitative reverse transcription PCR (RT-qPCR). Briefly, the lung and nose tissues were weighed and homogenized with Trizol, and virus RNA could be isolated according to the manufacturer’s protocol. SARS-CoV-2 viral RNA levels were measured by One Step PrimeScript RT-PCR Kit (RR064B, Takara) on a CFX96 real-time PCR detection system (Bio-Rad), and one sets of primers and probes were used to detect a region of the ORF1a/b of the viral genome of SARS-CoV-2, with sequences as follows: ORF1a/b-F, CCCTGTGGGTTTTACACTTAA; ORF1a/b-R, ACGATTGTGCATCAGCTGA; ORF1a/b-probe, FAM-CCGTCTGCGGTATGTGGAAAGGTTATGG-BHQ1. Viral RNA levels in lung and nose tissues were expressed as ORF1a/b gene copy numbers per milligram after comparison with a standard curve produced using serial ten-fold dilutions of SARS-CoV-2 RNA.

Viral titration

Lung tissues from hamsters were weighed and homogenized with 1mL of DMEM medium (supplemented with 100 U/mL of penicillin and 100 μg/mL streptomycin solution). Virus titrations were performed by end point titration in Vero cells, which were inoculated with 10-fold serial dilutions of tissue homogenates in 96-well plates (100μL/well). The titrating tissue homogenate were added to 100 μL of Vero cell suspension (DMEM with 4% FBS, 200 U/mL of penicillin and 200 μg/mL streptomycin solution, 2×10⁵ cells/mL) in each well, and then they were incubated at 37℃ with 5% CO₂. Cytopathic effect was read 6 days later.

Immunohistochemical/histopathology analysis

The lung tissues were fixed for more than 48 hours in 4% paraformaldehyde, dehydrated, and embedded in paraffin. 5-μm-thick tissue sections were sectioned and stained with hematoxylin and eosin (H&E). Double-blind evaluations were made for two randomly selected sections of each hamster on the basis of the histopathological changes, including inflammation, structure change and hemorrhage, which were graded according to the pre-defined scoring system.

Briefly, the clear structure of alveolar without inflammatory infiltration was recorded as score 0. The mild inflammation, slightly widened alveolar septum and sparse mononuclear cells (infiltration). The severe inflammation, thickening of alveolar wall and increased infiltration of interstitial monocytes was recorded as score 2. The significantly widened alveolar septum and increased infiltration of inflammatory cells was recorded as score 3-4. The extensive exudation and widened septum, smaller alveolar cavity, septal bleeding, and alveolar cells infiltration was recorded as score 5.

The immunohistochemical reaction in this study was performed on paraffin-embedded lung sections using a rabbit anti-SARS-CoV-2-Nucleocapsid Protein monoclonal antibody followed by an HRP-conjugated goat anti-rabbit secondary antibody. Then the DAB Substrate Kit was used for the chromogenic reaction. After nuclear counterstaining, dehydration, and mounting, the tissue staining was visualized under a microscope to detect virus infection.

Quantification and statistical analysis

Cell type annotation, relative proportion analysis of each cell subsets, differentially expressed genes analysis, Kyoto Encyclopedia of Genes and Genomes pathway and Gene Ontology enrichment analysis were performed on Omicsmart (https://www.omicsmart.com/home.html#/) where the datasets were analyzed and processed based on the R 3.6.1 Seurat package.

To show the multivariate antibody profiles across vaccination groups (R_W-CNP@M, R_O-CNP@M, or R_WR_O-CNP@M), principal component analysis (PCA) were performed and the heatmap were plotted based on corresponding Z score of ten immunological indices (the letters represent different biological meanings as follows: W, RBD_Wild-specific/wild-type strain; O, RBD_Omicron-specific/wild-type stain; B, BALF; S, Serum; P, Pseudovirus neutralization assay; A, IgA titer; G: IgG titer).

For other data, GraphPad Prism 9.0.0 and Origin 2023 were used for plotting and statistical analysis; the values were expressed as means ± SEM. Significance were calculated using unpaired t test and one-way ANOVA with Dunnett’s multiple comparison test or Tukey's test when the data conformed to a normal distribution and exhibited homogeneity of variance. Otherwise, the Mann-Whitney tests and Kruskal‑Wallis test was performed. The p values and R² values in correlation analysis reflect Spearman rank-correlation tests.

Materials & Correspondence

Further information and requests for resources and reagents should be directed to and will be fulfilled by Prof. Wei Wei ([email protected]). All unique reagents generated in this study are available from Prof. W. W. with a completed Materials Transfer Agreement. The published article contains all datasets analyzed during the study can be requested from Wei Wei upon reasonable request. The sequences data reported in this study were archived in the Sequence Read Archive (accession number: PRJNA813749, PRJNA1002859).

REFERENCES

48 Li, Q. et al. The Impact of Mutations in SARS-CoV-2 Spike on Viral Infectivity and Antigenicity. Cell 182, 1284-1294 e1289, doi:10.1016/j.cell.2020.07.012 (2020).

49 Nie, J. et al. Quantification of SARS-CoV-2 neutralizing antibody by a pseudotyped virus-based assay. Nat Protoc 15, 3699-3715, doi:10.1038/s41596-020-0394-5 (2020).

ACKNOWLEDGMENTS

This work was supported by the National Natural Science Foundation of China (grant nos. T2225021, 21821005, 32030062, U2001224, 81930122, U20A20361 and 82202028), Beijing Natural Science Foundation (JQ21027), CAS Project for Young Scientists in Basic Research (YSBR-083), National Key Research and Development Program of China (2021YFC2302600), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB29040303), CAMS Innovation Fund for Medical Science (2021-I2M-1-024), and the Major Science and Technology Special Projects of Yunnan Province (202002AA100009). We gratefully acknowledge Prof. Xin Zhao of Institute of Microbiology, Chinese Academy of Sciences for providing several wildtype and Omicron’s pseudotyped viruses. We gratefully acknowledge Prof. Bin Wang of Fudan University for providing SARS-CoV-2 RBD peptide pool.

AUTHOR CONTRIBUTIONS

W. W., G. M., L. Z., and H. W. conceived, designed, and supervised the experiments. T. Y., Z. J., and X. L. performed the experiments. Y. W. and W. H. provided most pseudotyped viruses and guided pseudovirus neutralization assay. Y. H., Y. Z., and E. W. contributed to live virus neutralization assay. Z. H., Y. L., F. Y. and X. Z. contributed to live virus challenge and transmission experiments. Y. F. provided the convalescent sera and provided professional advices on immunologic evaluation. M. Q. and D. Y. provided constructive suggestions and opinions on immunologic experiment design. W. Y. and L. H. contributed to evaluation of pulmonary delivery. Y. Q. guided immunologic evaluation on non-human primates. T. Y., Z. J., X. L., S. W., H. Y. and Y. W., aided in data analysis. T. Y., Z. J., X. L., and W. W. wrote the manuscript. W. W., G. M., L. Z., and H. W. revised the manuscript. All the authors have read and approved the final manuscript.

DECLARATION OF INTERESTS

The authors declare no competing interests.

SIFigure.docx
ExtendeddataFig.1.jpg
Data in c, e, and f (n=3) are shown as the mean ± SEM. Significance was calculated by one-way ANOVA with multiple comparison tests in e and f. Extended Data Fig. 1 Construction and antigenicity characterization of RBD-conjugated CNP (a) Western blotting analysis of covalent complex formation between SC-RBD and ST-CNP. SC-RBD, ST-CNP, and R-CNP were assessed using antibodies against CTB. (b) DLS analysis of CNP and R-CNP. (c) Comparison of hACE2 binding affinity among RBD, SC-RBD, and R-CNP. Affinity was assessed using a competition ELISA. (d) Ka, Kd, and KD of unconjugated RBD and R-CNP determined by surface plasmon resonance assays. Ka: Association rate. Kd: Dissociation rate. KD: Binding affinity constant calculated as Kd/Ka. (e) Flow cytometry analysis of known surface markers to assess the activation of bone marrow dendritic cells (BMDCs) after treatment with RBD, CTB+RBD, or R-CNPs. CTB+RBD: mixture of free CTB and RBD. (f) Flow cytometry analysis of known surface markers to assess the activation of macrophages after treatment with RBD, CTB+RBD, or R-CNP. (g) Heatmap of cytokine secretion by APCs after treatment with RBD, CTB+RBD, or R-CNP.
ExtendeddataFig.2.jpg
Extended Data Fig. 2 Optimization and characterization of the inhalable R-CNP@M vaccine (a) Schematic illustrations of inhaled microcapsule preparation, encapsulation, and characterization. Both the particle size and porosity were adjusted to obtain a suitable aerodynamic diameter to enable efficient alveoli delivery (left panel). The interior structure was optimized for good penetration of R-CNPs and complete sealing of open pores (middle panel). Then, the prepared R-CNP@M was characterized to evaluate the loading performance, morphological features, release profile, and remaining antigen bioactivity (right panel). (b) Fitting surface among aerodynamic size (d_a), hydrodynamic size (d_h) and porosity (po.). The magenta area represents microcapsules with aerodynamic sizes ranging from 1-4 μm. (c) Evaluation of the microencapsulation process. Microspheres with similar hydrodynamic sizes and porosities (po.) but with various pore sizes (red) were prepared and mixed with antigen (blue) followed by self-healing. The different performances on penetration of R-CNPs and self-healing of microcapsules resulted in different encapsulation efficiencies. (d) In vitro release profile of R-CNPs from microcapsules. (e) Aerodynamic size of fresh R-CNP@M and R-CNP@M stored for 1 month. (f) DLS analysis of the harvested R-CNPs from microcapsules stored for 1 month. Data in c and d (n=4) are shown as the mean ± SEM.
ExtendeddataFig.3.jpg
Extended Data Fig. 3 Safety evaluations and in vivocharacterizations of R-CNP@M in mice (a) Endotoxin levels of R-CNPs and R-CNPs@M. The dose of mice was 10 μg. (b) Respiratory function evaluation of vaccinated mice at the indicated time points. (c) Haemogram indices of vaccinated mice at the indicated time points. (d) H&E staining of lung tissues and other main organs at the indicated time points. All images are shown at the same magnification. (e) Changes of cytokine levels in BALF at the indicated time points. (f) Representative images of the Cy7-labelled R-CNP signal in the main organs at the indicated time points. (g) Calculation of the area under the curve (AUC) in Fig 1j. (h) Quantitative distribution of R-CNP@M in various lung structures. (i) Distribution of R-CNP (labelled with Cy5) in healed microcapsules extracted from bronchoalveolar lavage fluid (BALF) at the indicated time points. Data in a, c, e (n=3), g (n=6) and h (n=6, 39, and 46 among the trachea, bronchi, and alveoli, respectively) are shown as the mean ± SEM. Significance was calculated using one-way ANOVA with multiple comparison tests in e and h and unpaired t-test in g.
ExtendeddataFig.4.jpg
(k) GO terms from biological processes related to the activation of memory B cells enriched by DEGs between R-CNP@M and PBS. Extended Data Fig. 4 Analysis of immune cell status in the lungs of immunized mice (a) Bubble chart of canonical cell-type markers of nine major cell types. (b) Proportion of macrophage and DC of mice from the PBS, CNP, R-CNP, and R-CNP@M groups. (c) Heatmap of differentially expressed genes (DEGs) among the PBS, CNP, R-CNP, and R-CNP@M groups from macrophages and DCs in the lungs. (d and e) KEGG (Kyoto Encyclopedia of Genes and Genomes) terms related to the DCs (d) and macrophages (e) activation enriched by DEGs between R-CNP@M and PBS. (f) Annotated T-cell clusters are depicted using t-distributed stochastic neighbour embedding (t-SNE) plots for the merged scRNA-seq data (combined with the PBS, CNP, R-CNP, and R-CNP@M groups). The defined T-cell subsets are naïve CD4⁺ T, naïve CD8⁺ T, effector and/or memory CD4⁺ T (E&M CD4⁺ T), effector and/or memory CD8⁺ T (E&M CD8⁺ T), γδ T, NK T, and regulatory T (Treg) cells. (g) DEGs from E&M CD8⁺ T cells are represented by volcano plots for R-CNP@M vs. PBS. up: up regulation; down: down regulation; ns, no significance. (h) GO (Gene Ontology) terms from biological processes related to the activation of E&M CD8⁺ T cells enriched by DEGs between R-CNP@M and PBS. (i) Annotated B-cell clusters are depicted using t-SNE plots for the merged scRNA-seq data (combined with PBS, CNP, R-CNP, and R-CNP@M groups). The defined B-cell subsets are memory B, plasma, naïve B, germinal centre B and other B cells. (j) DEGs from memory B cells are represented by volcano plots for R-CNP@M vs. PBS.
ExtendeddataFig.5.jpg
Data in a-c and f-l (n=3) are shown as the mean ± SEM. Significance was calculated using a one-way ANOVA with multiple comparison tests in a, b and unpaired t-test in g, h, j and k. Extended Data Fig. 5 Additional evaluations of T-cell responses in the respiratory tract (a) Proportion of memory CD8⁺ T cells (CD44⁺ CD8⁺) in the lung (among CD8⁺ T cells, assessed by flow cytometry) on Day 70. (b) Proportion of tissue-resident memory T cells (T_RM) (CD69⁺CD103⁺) in the lung (among CD44⁺ CD8⁺ T cells, assessed by flow cytometry) on Day 70. (c) ELIspot assay for determination of the immunogenicity of overlapping peptides. SPC: spot-forming cells. (d) Immune Epitope Database analysis (IEDB) prediction scores of the predicted RBD-specific CD8⁺ T-cell epitopes. (e) IEDB prediction scores of the predicted RBD-specific CD4⁺ T-cell epitopes. (f) Verification of the immunogenicity of the candidate CD8⁺ and CD4⁺ T-cell epitopes (CGPKKSTNL and VGGNYNYLYRLFRKS for CD8⁺ and CD4⁺ T cells, respectively, and assessed by flow cytometry). (g) Proportion of IFN-γ⁺ T_RM with/without RBD epitope stimulation in the lung (among CD69⁺ CD103⁺ T cells, assessed by flow cytometry) on Day 70. (h) Proportion of IFN-γ⁺ CD4⁺ T_RM with/without RBD epitope stimulation in the lung (among CD69⁺ CD11a⁺ T cells, assessed by flow cytometry) on Days 21 (left) and 70 (right). (i) Cell numbers of CD4⁺ and CD8⁺ T cells in NALT. (j) Epitope-specific T-cell responses in NALT on Day 21. (k) Epitope-specific T-cell responses in NALT on Day 70. (l) Cell numbers of CD4⁺ and CD8⁺ T cells in nose.
ExtendeddataFig.6.jpg
Data in a-d (n=6) are shown as the mean ± SEM. Significance was calculated using Pearson chi-square test in i, using Kruskal-Wallis test with multiple comparison in k. Extended Data Fig. 6 Additional evaluations of antibody responses in mice (a) Endpoint titers of serum IgG1 and IgG2a determined by ELISA on Days 14, 28, 42, 56, 70, 91 133, 175, 266, and 365. (b) Endpoint titers of serum IgG, IgG1 and IgG2a determined by ELISA on Days 14 and 70. (c) Measurement of pVNT₅₀from serum, with sampling on Days 14 and 70. (d) Serum antibody quality analysis of R-CNP@M (indicated by the ratio of pVNT₅₀/serum IgG titer) on Days 28, 56, 70, 133, 266, and 365. The means of each time point were connected. (e) Relative counts for IGHC/V/D/J gene (frequency > 0.005) usage and pairing in pulmonary BCR repertoires in the R-CNP and R-CNP@M groups on Day 70. The R-CNP@M group shows more high-frequency IGH CDR3 pairing types and more IGHG and IGHA transcripts. (f) Venn diagram among the four indicated groups. There are 1238 convergent clonotypes between the 14 days post vaccination (dpv) and 70 dpv samples for the R-CNP group, 1041 convergent clonotypes between 14 dpv and 70 dpv samples for the R-CNP@M group, and 350 convergent clonotypes among the four samples. (g) Normalized frequency of R-CNP-specific clonotypes among the four samples. The convergent clonotypes among four samples were defined as R-CNP-specific clones. (h) Distribution of CDR3 heavy-chain variable (VH) gene germline divergence of R-CNP-specific clones. (i) Statistical analysis of somatic hypermutation (SHM) in the R-CNP-specific clones. (j) Genealogy tree of Top 5 convergent clone sequences. The Top 5 among R-CNP-specific clonotypes in each sample were selected and analyzed to construct a neighbor-joining tree by MEGA 11.0. (k) Endpoint titers of nasal lavage fluid IgA measured by ELISA on Days 70.
ExtendeddataFig.7.jpg
Data in e-g (n=8) and i (n=4, 6, 7, and 8 among the IsoAb, B^dep, T^dep, and Naïve groups, respectively) are shown as the mean ± SEM. Significance was calculated using Kruskal-Wallis test with multiple comparison in e, f, g, and i. Extended Data Fig. 7 Immunologic performance evaluations of R-CNP@M against SARS-CoV-2 in mice (a) Schematic illustration of immunization, antibody intervention, and sampling in mice. (b) Proportion of CD8⁺ T cells in the lung (among total T cells, assessed by flow cytometry) on Day 28. (c) Proportion of GC B cells (GL7⁺ IgD⁻ B220⁺) in mLNs (assessed by flow cytometry) on Day 28. (d) Proportion of CD8⁺ T cells in mLNs (assessed by flow cytometry) on Day 28. (e) Endpoint titers of serum IgG measured by ELISA on Day 28. (f) Measurement of serum pVNT₅₀ against the D614G strain on Day 28. (g) Percentage of weight loss compared with the initial weight on Day 0 after the wild-type strain challenge. (h) Survival curve of mice post-SARS-CoV-2 challenge. (i) Viral RNA levels measured by qPCR (using ORF1a/b primers) in the lung and nose homogenate from surviving mice at 4 days post-challenge. Different numbers of mice died in the IsoAb, B^dep, and T^dep groups within 4 days post-challenge, and all remaining surviving mice were euthanized on the fourth day post-challenge. (j) Immunohistochemical analysis of the SARS-CoV-2 N protein in lung tissues sections. The arrows indicate the SARS-CoV-2 N protein-positive areas. (k) Histopathology analysis of lung tissues sections.
ExtendeddataFig.8.jpg
Data in b-f (n=6) are shown as the mean ± SEM. Significance was calculated using Mann-Whitney tests in c-f. Extended Data Fig. 8 Comparisons of Zifivax and R-CNP@M on immunologic performance against SARS-CoV-2 in hamsters The hamsters immunized with Zifivax (10 μg equivalent RBD, immunized three times with an interval of 2 weeks, licensed recombinant protein subunit vaccine) received the same equivalent RBD antigen as mice immunized with R-CNP@M (30 μg equivalent RBD, immunized once). (a) Schematic illustration of immunization and sampling in hamsters. (b) Endpoint titers of serum IgG determined by ELISA on Days 14, 28 and 42. (c) Measurement of MN₅₀ from serum, with sampling on Day 42. (d) Measurement of MN₅₀ from BALF, with sampling on Day 42. (e) Viral RNA levels determined in the lung and nose tissues by qPCR (using ORF1a/b primers). (f) Viral titers assessed in lung tissues at 3 dpi. (g) Immunohistochemical analysis and histopathology analysis of the SARS-CoV-2 N protein in lung tissues sections at 3 dpi. The arrows indicate the SARS-CoV-2 N protein-positive areas.
ExtendeddataFig.9.jpg
Data in b-d (n=5) are shown as the mean ± SEM. Significance was calculated using unpaired t tests in b-d. Extended Data Fig. 9 Additional evaluations of R-CNP@M in cynomolgus monkeys (a) Immunofluorescent imaging of germinal centers in mLNs. The nuclei (blue) and B cells (red) were labelled with DAPI and B220 antibodies, respectively. Proliferation was indicated by the expression of Ki-67 (green). (b, c) Biochemical (b) and physiological indices (c) of blood in cynomolgus monkeys. AST: aspartate aminotransferase; LDH: lactate dehydrogenase; ALP: alkaline phosphatase; TP: total protein; BUN: blood urea nitrogen. (d) Proportion of granzyme B⁺ CD8⁺ T cells (Gzmb⁺ CD8⁺) among CD8⁺ T cells in peripheral blood. (e) Table showing binding antibody titers (in WHO BAU/mL) from serum in nonhuman primates on Day 56. We included the WHO International standard for anti-SARS-CoV-2 immunoglobulin (NIBSC code: 21/340) to calibrate binding antibody titers.
ExtendeddataFig.10.jpg
Extended Data Fig. 10 Additional evaluations of mosaic vaccine. (a) Endpoint titers of BALF IgG from mice vaccinated with the mosaic vaccine in Figure 6 determined by ELISA on Day 28. (b) Correlations of IgG antibody titers and pVNT₅₀ in serum. The p values and R² values reflect Pearson’s correlation analysis. (c) Correlations of IgA antibody titers and pVNT₅₀ in BALF. The p values and R² values reflect Pearson’s correlation analysis. (d) Correlations of IgG antibody titer and pVNT₅₀ in BALF. The p values and R² values reflect Pearson’s correlation analysis. (e) Endpoint titers of IgG on Day 14 after booster immunization with mosaic vaccine (R_WR_O-CNP@M) of mice that were prime-immunized with R_W-CNP@M. (f) Endpoint titers of IgG on Day 14 after booster immunization with mosaic vaccine (R_WR_O-CNP@M) of cynomolgus monkeys that were prime-immunized with R_W-CNP@M. (g) Endpoint titers of serum IgG determined by ELISA on Day 42. (h) Measurement of pVNT₅₀ from serum, with sampling on Day 42. (i) Measurement of pVNT₅₀ from BALF, with sampling on Day 42. (j) Schematic illustration of the contact/airborne transmission protection model. (k) Viral RNA levels measured in lung homogenates by qPCR (using ORF1a/b primers). (l) Viral titers assessed in lung homogenates at 3 days post-exposure (dpe). (m) Histopathology analysis and pathological scores of lung tissues sections at 3 dpe. Pathological scoring was carried out for two randomly selected images from each animal and by double-blind scoring on the basis of histopathological changes, including inflammation, structural changes and haemorrhage. Data in c (n=6), e (n=6), f (n=5), and g-l (n=6) are shown as the mean ± SEM. Significance was calculated using Mann-Whitney tests in a, g-i and using Kruskal-Wallis test with multiple comparison tests in k and l.

Download PDF

Version 2

posted

You are reading this latest preprint version

Inhalable SARS-CoV-2 vaccine for single-dose dry powder aerosol immunization

Status:

Version 2

Abstract

Figures

Main

Discussion

References

Methods

Declarations

Supplementary Files

Status:

Version 2