Active and Passive Immunization of Syrian Hamsters with An Attenuated SARS-CoV-2 Protects against New Variants of Concern

Abstract Detection of secretory antibodies in the airway is highly desirable when evaluating mucosal protection by a vaccine against a respiratory virus like the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). We show that a single intranasal delivery of an attenuated SARS-CoV-2 (Nsp1-K164A/H165A) induced both mucosal and systemic IgA and IgG in Syrian hamsters. Interestingly, either active or passive immunization of hamsters with Nsp1-K164A/H165A offered protection against heterologous challenge with variants of concern (VOCs) including Delta, Omicron BA.1, and Omicron BA.2.12.1. Among challenged animals, Nsp1-K164A/H165A vaccination specifically reduced viral loads in the respiratory tract and suppressed infection-induced macrophage accumulation and MX1 upregulation in the lung. The absence of variant-specific mucosal and systemic antibodies was associated with breakthrough infections, particularly of the nasal cavity following challenges with Omicron isolates. Together, our study demonstrates that an attenuated nasal vaccine may be developed to boost mucosal immunity against future SARS-CoV-2 VOCs.


Introduction
The early success of the two mRNA vaccines against SARS-CoV-2 boasted up to 95% e cacy against the most severe disease outcomes 1,2 . Nevertheless, the emergence of new virus variants has resulted in reduced vaccine effectiveness against symptomatic COVID-19. Since the emergence of the Omicron variant of concern (VOC) in late 2021, SARS-CoV-2 Omicron sub-lineages (e.g., BA.1, BA.2, BA.3, BA.4, BA.5, etc.) have gradually replaced previously circulating variants worldwide 3 . Owing to increased immune evasion, infection with Omicron has led to large numbers of hospitalizations and deaths [4][5][6] .
Vaccine effectiveness against Omicron after two BNT162b2 doses was modest at 2 to 4 weeks and fell to nearly zero after 6 months or more [7][8][9] . A third and fourth mRNA vaccine booster did improve e cacy against Omicron to 60-70% [10][11][12] , but the durability of protection was less than impressive, with a mean 30-day rate of decay in neutralizing antibody titers of nearly 20% against BA.4/5 13 .
Several studies have suggested that COVID-19 vaccine performance might be improved if mucosal immunity can be enhanced. For example, a recent study demonstrated that combining systemic mRNA vaccination with mucosal adenovirus-S immunization induced strong neutralizing antibody (nAb) responses against both the ancestral virus and the Omicron BA.1.1 variant in mice. By contrast, systemic mRNA vaccination alone induced weak respiratory mucosal neutralizing antibody responses 14 . Earlier in another proof-of-concept study, an intranasal vaccination with nonadjuvanted spike subunit protein following intramuscular mRNA vaccinations in mice elicited protective mucosal immunity via memory T/B cells and IgA that signi cantly lowered viral load in the upper and lower airways and prevented disease and death from a lethal SARS-CoV-2 challenge 15 . These ndings highlight the importance of understanding the mucosal immunogenicity and e cacy of next-generation nasal vaccines 16 .
Of the many vaccines in development against SARS-CoV-2, live attenuated virus (LAV) vaccines are a substantial minority despite the potential for nasal administration and the advantage of presenting all viral antigens to the host immune system [17][18][19] . To facilitate the evaluation of LAVs, we recently developed a genetic approach to attenuate SARS-CoV-2. Our strategy consists of three modi cations to the viral genome: the removal of the furin cleavage site (PRRA) 20 , the deletion of ORFs 6-8 21 , and introduction of a pair of mutations to the Nsp1 gene (Fig. 1a) 22 . The resulted WA1-ΔPRRA-ΔORF6-8-Nsp1 K164A/H165A (abbreviated as Nsp1-K164A/H165A in this paper) is attenuated both in vitro and in vivo compared to wildtype virus and was immunogenic and protective against the ancestral SARS-CoV-2 challenge 23 . In the current study, we assessed the mucosal immunogenicity, e cacy in protecting diseases caused by recent variants of concern, as well as the transmission of a candidate attenuated SARS-CoV-2 vaccine.
To further examine the presence of viral antigens and host innate immune activation, lung sections from uninfected (mock) or challenged hamsters were stained with hematoxylin and eosin (H&E) and also immunostained for viral nucleocapsid protein (NP) and myxovirus resistance 1 (MX1), an interferoninduced antiviral host response marker 25,26 (Fig. 4). Lungs from Delta-challenged unvaccinated hamsters at 4 DPC showed widespread immune in ltrates and regions of viral NP deposition characterized by prominent staining of the epithelial lining of infected bronchioles accompanied by intense staining of surrounding alveolar epithelium (Fig. 4a&b). Two of the four BA.1-challenged unvaccinated animals at 4 DPC showed NP deposition with a similar staining pattern (Fig. 4b&c). Delta-and BA.1-challenged unvaccinated groups also showed increased MX1 immunoreactivity in these NP-positive lung regions particularly evident in the bronchiolar epithelium. Vaccination with Nsp1-K164A/H165A or infection with WA1/2020 blocked NP deposition and MX1 upregulation in the Delta-and BA.1-challenged groups (Fig. 4b&c). High resolution imaging of representative lung sections from Delta-challenged unvaccinated hamsters highlighted the pronounced upregulation of MX1 in nuclear and cytoplasmic compartments of infected bronchiolar epithelial cells and the attenuation in Nsp1-K164A/H165A-vaccinated hamsters ( Fig. 4d). Altogether, the absence of NP staining and MX1 upregulation implies that the challenge virus, whether it is the Delta or Omicron BA.1 variant, failed to establish infection in the lungs of Nsp1-K164A/H165A-vaccinated and WA1/2020-convalescent hamsters.
To further characterize these histopathological changes using speci c markers of in ammation and epithelial damage, serial lung sections from uninfected hamsters (mock), unvaccinated hamsters, and WA1/2020-convalescent and Nsp1-K164A/H165A-vaccinated hamsters were immunostained for Iba1 (a marker of macrophages), prosurfactant protein C (ProSPC, a marker of AT2 cells), RAGE (a marker of AT1 cells), and E-cadherin (a marker of intercellular epithelial junctions). At 7 DPC, lungs from all four Deltachallenged unvaccinated hamsters and two of the four BA.1-challenged unvaccinated hamsters showed regions of consolidation by H&E that corresponded with areas containing extensive accumulation of Iba1-expressing macrophages (Fig. 6a&b) and increased terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) (Fig. S1). These consolidated regions also showed marked reduction of ProSPCexpressing AT2 cells and pronounced loss of alveolar RAGE expression along the borders of AT1 cells.
Distinct areas of epithelial cell loss and Iba1-positive macrophage consolidation formed around Ecadherin-labeled bronchioles that were presumably subject to virus attack (Fig. 6c). Increased E-cadherin staining also identi ed hyperplastic epithelium in consolidated spaces that was particularly prominent in Delta-challenged unvaccinated lungs. Vaccination with Nsp1-K164A/H165A attenuated Iba1-positive macrophage consolidation and protected against the loss of ProSPC and RAGE expression in Delta-and BA.1-challenge groups ( Fig. 6a-c). Prior infection with WA1/2020 also protected against epithelial damage in the Delta-challenge group but, interestingly, appeared less effective in protecting against BA.1induced in ammation and damage.
We then performed a correlational analysis between serum binding IgG, serum nAB, and nasal SIgA titers and tissue viral loads and severity of lung pathology. In general, antibody titers inversely correlated with tissue viral loads and severity of lung pathology, i.e., lower antibody titers correlate with increased frequencies of breakthrough infections (i.e., detectable viral loads) (Fig. S2). The exception is found in nasal turbinates following challenge with Omicron BA.1, where no clear correlation between antibody titers and viral loads was noticed. A plausible explanation of this nding is that the very low level or absence of BA.1-speci c mucosal and systemic neutralizing antibodies failed to offer any protection in nasal turbinates (Fig. S2).

Transmission of Nsp1-K164A/H165A in Syrian Hamsters
To characterize the transmissibility of Nsp1-K164A/H165A, we performed an additional study in an airborne transmission model (Fig. 7a) in which two Syrian hamsters in the same cage are separated from each other by a customized, perforated metal divider that prevents physical contact while permitting air exchange (Movie. S1) 27 .

Discussion
Current SARS-CoV-2 vaccines, particularly the mRNA-based ones, induce robust systemic humoral and cellular immunity and prevent severe disease caused by SARS-CoV-2 28 . However, protection against infection and transmission of SARS-CoV-2 variant viruses, in particular the Omicron sub-lineage viruses, by mRNA vaccines may be limited [29][30][31] . Results from recent studies suggest that intramuscular vaccination tends to induce potent systemic but not local humoral response at the mucosa. By contrast, nasal vaccines induce more robust mucosal immune responses, characterized by secretion of IgA and IgG at mucosal surfaces and detection of resident memory T/B cells in respiratory tract 16,24,32−34 . Here we show that intranasal immunization of Syrian hamsters with an attenuated SARS-CoV-2 (Nsp1-K164A/H165A) induced SIgA and IgG at respiratory mucosa, albeit the levels of mucosal IgA/IgG appeared to be at least two orders of magnitude lower than their serum counterparts. Vaccinated hamsters were largely protected in the lungs from subsequent challenge with Delta, Omicron BA.1 and BA.2.12.1, even when circulating nAb titers against the latter two variants were below limit of quanti cation. Together, our study demonstrates the great potential of attenuated Nsp1-K164A/H165A as a nasal vaccine capable of inducing SIgA/IgG in the respiratory mucosa.
Many of the pathologies found in human cases of COVID-19 are recapitulated in mature (> 4-6-month-old) Syrian hamsters, making them an excellent animal model for studying SARS-CoV-2 pathogenesis 35 . For example, the damage to AT1 and AT2 cells with the prominent in ltration of macrophages observed in the lungs of Delta-and BA.1-challenged unvaccinated hamsters partially mimics the histopathology of COVID-19 in humans 36,37 . One of the challenging aspects of this model, however, is the relative lack of species-speci c reagents that has made immunological analyses in Syrian hamsters di cult. In this study we established methods to reliably measure SIgA/IgG in nasal wash and BALF samples. These techniques have now enabled us to assess the mucosal immunogenicity of SARS-CoV-2 vaccine candidates in Syrian hamsters. We detected signi cant amounts of SIgA/IgG in nasal wash and BALF samples from hamsters after intranasal vaccination of Nsp1-K164A/H165A. One signi cant and noteworthy observation is that SIgA/IgG titers, like circulating nAb titers, inversely correlate with viral loads in respiratory tract and the severity of lung pathology (Fig. S2). When SIgA/IgG is speci cally raised against the ancestral virus (WA1/2020) and hence does not e ciently neutralize Omicron subvariants, vaccinated hamsters were not free of virus in the upper respiratory tract upon challenge with Omicron sub-lineage viruses. This nding is in line with the recent observation that anti-Spike mucosal IgA protects against SARS-CoV-2 Omicron infection in human population 38 . Interestingly, Nsp1-K164A/H165Avaccinated animals were protected in the lung against both Delta and Omicron BA.1 challenge. It is possible that cellular immunity, induced by Nsp1-K164A/H165A, contributes to the protection against heterologous virus challenge, as we detected splenocytes reacting to the nucleocapsid protein of the virus. Future research is warranted to decipher the mechanism of live attenuated viral vaccines in conferring cross-protection in hamsters as more reagents become available.
An early and robust activation of interferon (IFN) signaling pathways contributes to the protective mucosal immune response against viral infections in the respiratory tract 39 . Viral recognition in infected immune and/or epithelial cells triggers the production of IFNs (types I, II, and III) that subsequently limit viral replication and dissemination by activating the transcription of numerous antiviral IFN-stimulated genes (ISGs) including IFIT1-3, OAS1, IRF7 and MX1 in infected cells and bystander cells. While IFNs are critical to the antiviral host defenses, excessive or persistent IFN signaling may also aggravate lung pathologies of viral infections including SARS-CoV-2 40 . Activation of IFN signaling in the lungs of Deltaand BA.1 unvaccinated animals was clearly evidenced by the robust upregulation of MX1 protein that localized speci cally to infected bronchioles and neighboring alveolar parenchyma stained by viral NP. In this regard, we surmise that the lack of MX1 upregulation in the lungs of Nsp1-K164A/H165A-vaccinated hamsters supports that the vaccine exerts a strong and speci c protective response along the upper airways that blocks viral dissemination in the lungs. The lack of macrophage accumulation in the lungs of Nsp1-K164A/H165A-vaccinated hamsters suggests that robust viral infection did not occur, hence vaccinated animals did not develop pneumonia.

Cells and Viruses
Vero E6 cell line (Cat # CRL-1586) was purchased from American Type Culture Collection (ATCC) and cultured in Dulbecco's minimal essential medium (MEM) supplemented with 10% fetal bovine serum An intrinsic safety concern for a LAV against a pandemic virus is the potential transmission of the vaccine virus. In principle, transmission of a vaccine virus may contribute to the establishment of herd immunity, but there is a possibility that an attenuated virus may directly or through reversion to a more virulent form cause illness among some at-risk individuals. We have previously demonstrated the genome stability of Nsp1-K164A/H165A 23 . Here we found that the airborne transmission of Nsp1-K164A/H165A was less e cient compared to wild-type SARS-CoV-2 in hamsters, but most (except for one animal) sentinel hamsters became seroconverted after 14 days. Airborne transmission of Nsp1-K164A/H165A did not cause weight loss in sentinel hamsters. Another intriguing nding is that four and half months after the initial exposure, these seroconverted hamsters were largely protected from a BA.2.12.1 challenge. Thus, passive immunization via transmission of Nsp1-K164A/H165A was achieved. Nonetheless, the risks and bene ts of administering an infectious attenuated SARS-CoV-2 vaccine will need to be further evaluated.

Study Limitations
Typically, SARS-CoV-2 infection of Syrian hamsters induces pronounced consolidation and pathologies in the lung at 7 DPC 35 . However, lung pathology was limited to a subset of BA.1 and BA.2.12.1-infected hamsters rather than being widespread as in Delta challenged animals. For this reason, determination of vaccine e cacy against Omicron sub-lineage viruses using Syrian hamsters was problematic due to the low pathogenicity of BA.1 and BA.2.12.1 in this species. Low pathology scores in unvaccinated controls led to di culties in determining e cacy of vaccination in preventing lung damage post-challenge.
Additionally, persistent damage after WA1/2020 infection likely led to increases in bronchiole mucosal hyperplasia in hamsters observed at 4 DPC in convalescent animals challenged with Delta as well as BA.1 and BA.2.12.1 Omicron. In separate experiments being prepared for publication, we have observed long-term lung damage (including hyperplasia in bronchioles) in hamsters up to 4 weeks after WA1/2020 infection by the intranasal route. This would provide an explanation for the observed lung pathology in WA1/2020 convalescent hamsters after Omicron challenge which was not observed in control or Nsp1-K164A/H165A vaccinated animals.

Concluding Remarks
Nsp1-K164A/H165A vaccination induces humoral immunity at mucosa as well as cellular-immunity targeting nucleocapsid protein. Additionally, transmission of Nsp1-K164A/H165A leads to passive immunization. Thus, Nsp1-K164A/H165A may be further developed into a nasal vaccine for primary series or as a booster.
Production of Nsp1-K164A/H165A was described elsewhere 23 . The SARS-CoV-2 isolate WA1/2020 (NR-52281, lot 70033175) was obtained from BEI Resources, NIAID, NIH, and had been passed three times on Adult male (5-6 months old) Syrian hamsters (Mesocricetus auratus) were anesthetized with (3-4% v/v) iso urane and oxygen following procedures as described previously 35,41,42 . Intranasal inoculation was done by pipetting 10 2 PFU or 10 4 PFU SARS-CoV-2 in 50 µl volume dropwise into the nostrils of the hamster under anesthesia. Following infection, hamsters were monitored daily for clinical signs and weight loss. Nasal washes were collected by pipetting ~ 200 µl sterile phosphate buffered saline into one nostril when hamsters were anesthetized by 3-5% iso urane. Nasal swabs were done as described previously 32 .
For airborne transmission, a subset (n = 7) of hamsters inoculated with 10 2 PFU WA1/2020 or Nsp1-K164A/H165A were paired in divided cages to prevent direct contact to measure transmission to naive sentinels 27 . One hamster (WH363), paired with an actively shedding Nsp1-K164A/H165A vaccinated animal, did not show evidence of productive infection or seroconvert at 14 DPE and remained seronegative until just prior to BA.2.12.1 challenge 4.5 months later. For these reasons, WH363 was removed from the challenge datasets.
For tissue collection, a subset of hamsters was humanely euthanized by intraperitoneal injection of pentobarbital at 200mg/kg at 4 and 7 DPC. Lungs, trachea, and nasal turbinates were dissected for histopathology or homogenized for RNA extraction or titration in cell culture. Blood collection was performed under anesthesia (3-5% iso urane) through gingival vein puncture or cardiac puncture when animals were euthanized. The left lobes of hamster lungs (~ 0.2 gram) were diced, divided, and resuspended in 1 milliliter MEM or TriZol reagent (RNA extraction) and homogenized on a Precellys Evolution tissue homogenizer with a Cooling Unit (Bertin). Trachea and nasal turbinates were homogenized the same way in TriZol Reagent. Splenocytes were extracted at 14 DPI from vaccinated and naive hamsters and IFNγ-secreting cells were identi ed after stimulation with spike and nucleocapsid antigen pools (Genscript) by ELISpot (MABTECH, 3102-2H).

RNA isolation and qRT-PCR
Procedures as described previously 35,41 . In brief, RNA was extracted from 0.1-gram tissue homogenates using QIAamp vRNA mini kit or the RNeasy 96 kit (QIAGEN) and eluted with 60 µl of water. 5 µL RNA was used for each reaction in real-time RT-PCR. When graphing the results in Prism 9, values below the limit of quanti cation (LoD) were arbitrarily set to half of the LoD values. Unless otherwise speci ed, the unit for RNA copies are as presented as Log 10 RNA copes/µg tissue RNA.

Histopathology Analyses
Procedures as described previously 35,41 . Tissues (lungs, trachea, and nasal turbinates) were xed in 10% neutral buffered formalin overnight and then processed for para n embedding. The 5-µm sections were stained with hematoxylin and eosin for histopathological examinations. Images were scanned using an Aperio ImageScope. Blinded samples were graded by a licensed pathologist for the following twelve categories: consolidation, alveolar wall thickening, alveolar airway in ltrates, perivascular in ltrates, perivascular edema, type II pneumocyte hyperplasia, atypical pneumocyte hyperplasia, bronchiole mucosal hyperplasia, bronchiole airway in ltrates, proteinaceous uid, hemorrhage, and vasculitis.

Virus titration
Tissue culture infectious dose 50% (TCID 50 ) assays were done described previously 35,41 for initial nasal wash titrations post-inoculation. In brief, Vero E6 cells were plated the day before infection into 96 well plates at 1.5 × 10 4 cells/well. On the day of the experiment, serial dilutions of 20 µl nasal wash samples were made in media and a total of six to eight wells were infected with each serial dilution of the virus.
After 48 h incubation, cells were xed in 4% PFA followed by staining with 0.1% crystal violet. The TCID50 was then calculated using the formula: log (TCID50) = log(do) + log (R) (f + 1). Where do represents the dilution giving a positive well, f is a number derived from the number of positive wells calculated by a moving average, and R is the dilution factor.
For focus-forming assay, nasal wash, BALF, and lung homogenate samples were 10-fold serially diluted in 96-well plates and dilutions added to 96-well black-well plates for uorescent focus forming assays in H1299-hACE2 cells 43 . After 1 h the Tragacanth gum overlay ( nal concentration 0.3%) was added. Cells were incubated at 37°C and 5% CO 2 for 1 day, then xed with 4% paraformaldehyde, followed by staining of cells with primary rabbit anti-N Wuhan-1 antibody (Genscript) overnight followed by secondary antirabbit Alexa-488 conjugated antibody and DAPI staining. The infectious titers were then counted using Gen5 software on a Cytation7 machine and calculated and plotted as focus forming units per milliliter (FFU/ml).

SARS-CoV-2 neutralization assay
Samples were serially diluted 2-fold in 5% FBS DMEM and mixed with 100 PFU of SARS-CoV-2 in a 96well plate at 37°C for 1 hour. Sample:virus mixtures were then added to con uent H1299-hACE2 cells in 96-well plates. Cells were infected for 1 hour before the inoculum was removed and washed three times with DPBS. A second overlay containing 1.2% Tragacanth gum, 2X MEM, 5% FBS, and DMEM was added to the plate. Cells were incubated at 37°C for 1 day, then xed with 4% paraformaldehyde, followed by staining of cells with primary rabbit anti-SARS-CoV-2 N antibody (Genscript U739BGB150-5) overnight followed by secondary anti-rabbit Alexa-488 conjugated antibody and 4′,6-diamidino-2-phenylindole (DAPI) staining. Plates were imaged on a Cytation7 (Agilent), and foci were counted using Gen5 software.
For the neutralization assays, recombinant LY-CoV555 (Bamlanivimab) mixed with WA1/2020 44 was included as a positive control. The 50% endpoint neutralization titers were determined as the reciprocal of the highest dilution providing ≤ half of the number of foci obtained from the negative control well (plain DMEM mixed with 100 PFU virus).

Measurement of antibody by ELISA
The preparation of SARS-CoV-2 RBD antigen in a baculovirus expression system and its use in ELISA were previously described 45 . ELISAs were performed with slight modi cations. Brie y, Immulon 2 HB plates were coated with recombinant RBD protein at 1 µg/mL overnight at 4°C. Test serum samples were prediluted in assay diluent (PBS containing 0.05% Tween-20 [PBST] and 10% fetal bovine serum), followed by serial two-fold dilutions of each sample in duplicates across the plate. A starting dilution of 1:160, 1:80, and 1:20 was used for serum (IgA and IgG), BALF (IgG) and nasal wash and BALF (IgA) samples, respectively. Plates were incubated with the test serum samples for 2 h at 37°C. After rigorous plate washes in a microplate washer, plates were incubated with anti-hamster antibodies. For IgG ELISA, a 1:4000 dilution of an HRP-conjugated goat anti-hamster IgG (6060-05, Southern Biotech, Birmingham, Alabama) was added to assay wells. For IgA ELISA, a rabbit anti-hamster IgA antibody [sandwich antibody; (cat. #sab 3001a) Brookwood Biomedical, Jemison, Alabama] was added to assay wells at 1:4000 dilution and plates incubated for 1 hour at 37°C. Unbound sandwich antibody was washed off and a 1:4000 dilution of an HRP-conjugated goat anti-rabbit IgG (4030-05, Southern Biotech, Birmingham, Alabama) was added to assay plates. In both IgG and IgA ELISAs, incubation with HRP-conjugated secondary antibodies lasted 1 hour after which plates were rigorously washed to remove unbound antibodies. The ABTS/H 2 O 2 peroxidase substrate (SeraCare, Gaithersburg, Maryland) was added to assay wells and plates left at room temperature for 20 to 30 minutes. Color development was stopped by adding 1% SDS and OD 405 values were captured on the VersaMax microplate reader with Softmax Pro 7 software (Molecular Devices). In the IgG ELISA, the mean OD 405 values of PBS treatment groups were subtracted from the mean OD 405 values from other treatment groups and the assay endpoint was a mean OD 405 value 0.05 (i.e., after background subtraction). In the IgA ELISA, the assay endpoint was a mean OD 405 value 0.02 of duplicate wells. Antibody titer was de ned as the reciprocal of the highest dilution of a sample at which the mean OD 405 value for duplicate wells was 0.02 (IgA) or 0.05 after background subtraction (IgG). were mixed and incubated overnight at 4°C. For negative controls, sections were incubated without the primary antibody or mouse and rabbit isotype antibody controls. Sections stained with conjugated secondary antibodies alone showed no speci c staining. Whole slide uorescence imaging was performed using a Hamamatsu NanoZoomer 2.0-RS whole-slide digital scanner equipped with a 20x objective and a uorescence module #L11600. Analysis software NDP.view2 was used for image processing (Hamamatsu Photonics, Japan). Immuno uorescence and differential interference images were also captured using an Axio Observer Z1 inverted microscope (Carl Zeiss, Thornwood, NY) equipped with an Axiocam 506 monochrome camera, an ApoTome.2 optical sectioning system, and a Plan-Apochromat 63x/1.4NA oil immersion with WD = 0.19 and Plan-Apochromat 20x/0.8 objective lens. Digital image post-processing and analysis were performed using the ZEN 2 ver. 2.0 imaging software.

IFN-gamma ELISpot
Images were constructed from Z-stack slices collected at 0.48 µm intervals (5 µm thickness in total) and visualized as maximum intensity projections in orthogonal mode. For semiquantitative analysis of NP staining, high resolution whole-slide digital images of each lung section were acquired and the NDP.view2 software was used to measure the NP-stained area as a percentage of total area of the section. For TUNEL staining, sections were depara nized, hydrated, and pretreated with Proteinase K, followed by EDTA, distilled H O wash, and BSA blocking. Sections were then incubated in a reaction mixture (TdT, dUTP, and buffer), washed, and incubated with anti-digoxigenin antibody. Sections were then visualized with alkaline phosphatase-ImmPACT Vector Red and counterstained with hematoxylin.

Statistical analysis
One-way ANOVA or Student t-test was used to calculate statistical signi cance through GraphPad Prism (9.1.2) software for Windows, GraphPad Software, San Diego, California USA, www.graphpad.com.

DATA and MATERIALS AVAILABILITY
All data are available in the main text or the supplementary materials. All unique/stable reagents generated in this study are available from the Lead Contact with a completed Materials Transfer Agreement.  antigen pools for 48 hours. IFNg-secreting splenocytes were enumerated by ELISPOT. Bar graphs in panels a-e represent samples collected at two time points from the same animals in a single experiment with dots representing individual animals. Statistical differences were calculated using ordinary one-way analysis of variance (ANOVA) in GraphPad Prism 9.4.0 with Tukey's multiple comparisons tests. ELISPOT data were compared using two-way ANOVA with Sidak's multiple comparison test. For statistical signi cance, * indicates p<0.05, ** indicates p<0.01, and **** indicates p<0.0001. DPI= days postinfection.   are shown at two levels of magni cation (0.7x and 5x) with white boxes delimiting the regions of magni cation. c Semiquantitative analysis of viral NP staining in hamster lungs at 4 DPC. The plotted values represent the percent NP positive area as a function of the total lung area for each section (n = 3-4 animals per group). d High magni cation immuno uorescence/differential interference contrast images of NP and MX1 in representative bronchioles of lung sections from mock hamsters or Delta-infected nonvaccinated or Nsp1-K164A/H165A-vaccinated hamsters at 4 DPC. Prominent cytoplasmic and nuclear localization of MX1 was detected in NP-positive bronchiolar epithelium in Delta-infected unvaccinated hamsters compared to low cytoplasmic expression of MX1 in mock and vaccinated hamsters. Nuclei were counterstained with Hoechst 33342 dye (blue). Scale bars: 5 mm (0.7x), 500 mm (5x), 20 mm (60x).  Intranasal immunization of Syrian hamsters with Nsp1-K164A/H165A protects against Delta and Omicron challenge. a Weight change was recorded for hamsters (described in Fig. 3a) after challenge by Delta and Omicron BA.1 variants for 7 days. Percentage of consolidation (b) and pathology score (c) in xed lung tissues were compared between WA1/2020 convalescent (n=7), Nsp1-K164A/H165A vaccinated (n=8), and unvaccinated control (n=8) lungs at 4 DPC. Individual pathologies were graded by severity and presented in a heat map (d). Percentage of consolidation (e) and pathology score (f) were also compared at 7 DPC between WA1/2020 convalescent (n=6), Nsp1-K164A/H165A vaccinated (n=6), and unvaccinated control (n=8) lungs. g Heat-map presentation of individual pathologies at 7 DPC. Dot   Airborne transmission of Nsp1-K164A/H165A in Syrian hamsters. a Donor Syrian hamsters (male, 5month-old) were rst inoculated with 100 PFU Nsp1-K164A/H165A (n=14) or WA1/2020 (n=14). One day after inoculation, donor hamsters were paired with recipient hamsters (sentinel, n=7/group) in the specially designed cages with metal dividers for monitoring airborne transmission. During pairing, nasal swabs were collected daily from sentinel hamsters for 4 days. b Weight loss pro le of donor hamsters

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.