An intranasal vaccine durably protects against SARS-CoV-2 variants in mice

Summary SARS-CoV-2 variants that attenuate antibody neutralization could jeopardize vaccine efficacy. We recently reported the protective activity of an intranasally administered spike protein-based chimpanzee adenovirus-vectored vaccine (ChAd-SARS-CoV-2-S) in animals, which has advanced to human trials. Here, we assessed its durability, dose response, and cross-protective activity in mice. A single intranasal dose of ChAd-SARS-CoV-2-S induced durably high neutralizing and Fc effector antibody responses in serum and S-specific IgG and IgA secreting long-lived plasma cells in the bone marrow. Protection against a historical SARS-CoV-2 strain was observed across a 100-fold vaccine dose range and over a 200-day period. At 6 weeks or 9 months after vaccination, serum antibodies neutralized SARS-CoV-2 strains with B.1.351, B.1.1.28, and B.1.617.1 spike proteins and conferred almost complete protection in the upper and lower respiratory tracts after challenge with variant viruses. Thus, in mice, intranasal immunization with ChAd-SARS-CoV-2-S provides durable protection against historical and emerging SARS-CoV-2 strains.

Correspondence diamond@wusm.wustl.edu In brief Hassan et al. show that immunization with ChAd-SARS-CoV-2-S is durably immunogenic and protects against SARS-CoV-2 challenge in a dosedependent manner. Many months after single-dose intranasal immunization, ChAd-SARS-CoV-2 confers protection against variants of concerns of SARS-CoV-2 in both the upper and lower respiratory tracts of mice.

INTRODUCTION
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the etiologic agent of the coronavirus disease 2019  syndrome, which can rapidly progress to pneumonia, respiratory failure, and systemic inflammatory disease (Cheung et al., 2020;Mao et al., 2020;Wichmann et al., 2020). The elderly, immunocompromised, and those with certain co-morbidities (e.g., obesity, diabetes, and hypertension) are at greatest risk of severe disease, requirement of mechanical ventilation, and death . To date, approximately 184 million infections and 4 million deaths have been recorded worldwide (https://covid19.who.int) since the start of the pandemic. The extensive morbidity and mortality associated with COVID-19 pandemic have made the development and deployment of SARS-CoV-2 vaccines an urgent global health priority.
The spike (S) protein of the SARS-CoV-2 virion is the principal target for antibody-based and vaccine countermeasures. The S protein serves as the primary viral attachment and entry factor and engages the cell-surface receptor angiotensin-converting enzyme 2 (ACE2) to promote SARS-CoV-2 entry into human cells (Letko et al., 2020). SARS-CoV-2 S proteins are cleaved to yield S1 and S2 fragments (Hoffmann et al., 2020), with the S1 protein containing the receptor binding domain (RBD) and the S2 protein promoting membrane fusion and virus penetration into the cytoplasm. The prefusion form of the SARS-CoV-2 S protein  is recognized by potently neutralizing monoclonal antibodies (Barnes et al., 2020;Cao et al., 2020b;Pinto et al., Here, as a further step to evaluating the potential utility of ChAd-SARS-CoV-2-S, we assessed its dose response, durability, and cross-protective activity in mice including effects on upper-and lower-airway infection. At approximately 9 months after IN immunization, neutralizing antibody and anti-S protein IgG and IgA levels in serum of ChAd-SARS-CoV-2-S-vaccinated animals remained high and inhibited infection with SARS-CoV-2 strains with B.1.351, and B.1.1.28 spike proteins. At this time point, susceptible K18-hACE2 transgenic mice were fully protected against upper and lower respiratory tract infection after challenge with a SARS-CoV-2 virus displaying B.1.351 spike proteins.
Long-lived plasma cells (LLPCs) reside in the bone marrow and constitutively secrete high levels of antibody that correlate with serum levels (Amanna and Slifka, 2010). To assess the levels of antigen-specific LLPCs at 200 days after IM or IN immunization with 10 10 vp of ChAd-SARS-CoV-2-S, CD138 + cells were isolated from the bone marrow and assayed for S-specific IgG or IgA production using an ELISPOT assay (Purtha et al., 2011). We observed an 4-fold higher frequency of LLPCs secreting S-specific IgG after IN immunization than IM immunization ( Figure 1N). Additionally, after IN immunization, we de-tected LLPCs producing S-specific IgA, which were absent after IM immunization ( Figure 1N). Together, these data establish the following: (1) single-dose IN immunization promotes superior humoral immunity than IM immunization, (2) 100-fold lower inoculating doses of ChAd-SARS-CoV-2-S induce robust neutralizing antibody responses in mice, (3) IN but not IM immunization induces serum IgA responses and IgA-specific LLPCs against the SARS-CoV-2 S protein, and (4) the humoral immunity induced by ChAd-SARS-CoV-2-S is durable and rises over a 6-month period after vaccination.

IN inoculation of ChAd-SARS-CoV-2-S induces broad antibody responses with Fc effector function capacity
To characterize the humoral response further, we analyzed antibody binding to SARS-CoV-2 variant proteins and Fc effector functions using serum derived from BALB/c mice at 90 days after IN or IM vaccination. Our panel of SARS-CoV-2 proteins included spike (D614G, E484K, N501Y, D69-70, K417N) and RBD (E484K) antigens corresponding to WA1/2020, B.1.1.7, B.1.351, and B.1.1.28 strains. We first measured the anti-SARS-CoV-2 specific antibody response for several isotypes (IgG1, IgG2a, IgG2b, IgG3, IgM, and IgA) and their ability to bind Fcg receptors (mouse FcgRIIB, FcgRIII, and FcgRIV) using a luminex platform. Consistent with data obtained by ELISA (Figures 1B and 1E), IN vaccination of ChAd-SARS-CoV-2-S induced higher levels of IgG1 to D614G spike and WA1/2020 RBD proteins than IM immunization, and as expected, decreasing doses of the vaccine elicited lower antibody titers ( Figure 2A). Anti-SARS-CoV-2 IgG1 titers after IN immunization also were higher against all spike and RBD variants than after IM immunization, and titers decreased with vaccine dose (Figure 2B). As shown in a heatmap, this trend was observed for all anti-SARS-CoV-2 specific antibody isotypes and correlated with FcgR binding patterns ( Figure 2C). These data suggest that IN vaccination induces a higher magnitude and broader antibody subclass response to SARS-CoV-2 than IM vaccination.
Antibody effector functions, such as opsonization, are mediated in part by Fcg receptor engagement (Bruhns and Jö nsson, 2015). To determine whether the observed differences in antibody titers and FcgR binding titers resulted in differences in effector functions, we performed antibody-dependent neutrophil (ADNP) and cellular phagocytosis (ADCP) assays (Figures 2D and 2E). Sera from IN-vaccinated mice stimulated substantially more ADNP than those obtained from IM-vaccinated mice. However, minimal differences in ADCP were apparent from (B) Serum was analyzed by Luminex to quantify the amount of anti-SARS-CoV-2 IgG1 to different SARS-CoV-2 protein variants. Polar plots represent the IgG1 median percentile rank for each SARS-CoV-2 protein and variant. (C) Heatmap shows the IgG titer and FcgR binding titer of each vaccine regimen to SARS-CoV-2 Spike or RBD proteins. Each square represents the average Z score within a group for the condition. (D) Serum was incubated with primary mouse neutrophils (mADNP) or J774A.1 cells (mADCP) and SARS-CoV-2 spike-coated beads, and phagocytosis was measured after 1 h. Bars represent the mean and the error bars indicate standard deviations. (E) Serum was incubated with primary mouse neutrophils (mADNP) or J774A.1 cells (mADCP) and WA1/2020 D614G, B.1.1.7, or B1.351 spike-coated beads, and phagocytosis was measured after 1 h. Polar plots represent the mADNP or mADCP median percentile rank for each SARS-CoV-2 protein and variant. For (A and D), one-way ANOVA with a Dunnett's post-test comparing vaccine to control groups: **p < 0.01; ***p < 0.001; ****p < 0.0001). In (A) 2E). These data demonstrate that IN vaccination with ChAd-SARS-CoV-2-S induces a greater and more functional antibody response than after IM vaccination.
Intranasally administered ChAd-SARS-CoV-2-S induces durable protection against SARS-CoV-2 challenge in BALB/c mice To assess the efficacy of the ChAd-SARS-CoV-2-S vaccine, immunized BALB/c mice given the dosing regimen described in Figure 1A were challenged with SARS-CoV-2. Virus challenge was preceded by intranasal introduction of Hu-Ad5-hACE2, which enables ectopic expression of hACE2 and productive infection of SARS-CoV-2 in BALB/c mice by historical SARS-CoV-2 strains (Hassan et al., 2020a; Sun et al., 2020). Animals were immunized once via IN or IM routes with 10 10 vp of ChAd-Control or 10 8 , 10 9 , or 10 10 vp of ChAd-SARS-CoV-2-S. At day 95 or 195 post-vaccination, mice were given 10 8 plaque-forming units (PFUs) of Hu-Ad5-hACE2 and anti-Ifnar1 mAb; the latter attenuates innate immunity and enhances pathogenesis in this model . Five days later, BALB/c mice were challenged with 5 3 10 4 focus-forming units (FFUs) of SARS-CoV-2 (strain WA1/2020) via IN route. At 4 days post-infection (dpi), lungs, spleen, and heart were harvested from mice challenged at 100 days post-immunization, and lungs, nasal turbinates, and nasal washes were collected from a second cohort challenged at 200 days post-immunization. Tissues were assessed for viral burden by quantitative reverse transcription PCR (qRT-PCR) using primers for the subgenomic RNA (N gene). IN immunization with all three doses induced remarkable protection at 100 days post-vaccination as evidenced by a virtual absence of viral RNA in lungs, spleen, and heart compared to animals receiving the ChAd-Control vaccine (Figures 3A-3C). At 200 days post-immunization, protection conferred by the IN delivered ChAd-SARS-CoV-2-S remained robust in the upper and lower respiratory tracts compared to ChAd-Control immunized mice. Nevertheless, we observed limited infection breakthrough in the lungs and nasal turbinates in animals immunized with the lowest 10 8 vp dose of ChAd-SARS-CoV-2-S ( Figures 3G and 3I). In comparison, protection at 100 days post-IM immunization was less than after IN immunization at the same challenge time point. Although viral RNA was not detected in the heart and spleen ( Figure 3E-F), at least 1,000-to 30,000-fold (p < 0.0001) higher levels were measured in the lungs of mice immunized with ChAd-SARS-CoV-2-S by the IM compared to IN route ( Figures 3A and 3D). We also observed a greater impact of dosing by the IM route, as the reduction in viral RNA load in the lungs at 10 8 vp dose no longer was different than in the ChAd-control-vaccinated mice (Figure 3D). At 200 days post-IM immunization, we observed less protection against SARS-CoV-2 infection in the lungs, nasal washes, and nasal turbinates than after IN immunization ( Figures  3G-3L).

DISCUSSION
The durability of vaccine-induced immune responses is a key for providing sustained protection against SARS-CoV-2 infection and curtailing the current pandemic. Here, we show that a single IN immunization with ChAd-SARS-CoV-2-S induced S-and RBD-specific binding and neutralizing antibodies that continued to rise for several months, suggestive of sustained germinal center reactions. LLPCs in the bone marrow were detected 6 months after IN vaccination, secreting SARS-CoV-2specifc IgG and IgA that likely contributed to the durably high antiviral antibody levels in circulation (Amanna and Slifka, 2010). In comparison, IM immunization with ChAd-SARS-CoV-2-S induced lower levels of serum neutralizing antibodies, fewer spike-specific IgG secreting LLPCs, and virtually no serum or cellular IgA response. At least in mice, a single IN dose immunization with ChAd-SARS-CoV-2-S produced durable humoral immunity that was observed across a 100-fold dose range. These pre-clinical immunogenicity results compare favorably with studies in humans with mRNA vaccines against SARS-CoV-2, which show humoral immune responses lasting at least several months (Doria-Rose et al., 2021;Widge et al., 2021). In comparison, the durability of antibody responses after natural SARS-CoV-2 infection can vary considerably (Dan et al., 2021;Gudbjartsson et al., 2020).
A single immunization of ChAd-SARS-CoV-2-S conferred durable protection against SARS-CoV-2 (WA1/2020 strain) challenge in hACE2-tranduced BALB/c mice or K18-hACE2 transgenic C57BL/6 mice at multiple time points. IN immunization in particular provided virtually complete virological protection against upper and lower respiratory tract infection, with only a limited infection breakthrough seen at the 100-fold lower vaccine dose. The abrogation of infection in the upper respiratory tract suggests that IN vaccination could prevent transmission, although corroborating studies are needed in other rodent (e.g., hamsters or ferret) models better suited to studying this question (Muñ oz-Fontela et al., 2020). In comparison, IM immunization reduced the viral RNA levels in the lungs but showed substantially less protection against the homologous WA1/ 2020 strain in samples from the upper respiratory tract. While many SARS-CoV-2 vaccine candidates from different platforms have demonstrated immunogenicity and protective efficacy in animals models (García-Arriaza et al., 2021;Hennrich et al., 2021;Tostanoski et al., 2020;van Doremalen et al., 2020;Vogel et al., 2021;Yao et al., 2021;Yu et al., 2020), to our knowledge, none have established durability or protection against variant viruses. The long-term protection conferred by IN immunization even at 100-fold lower inoculating doses is promising but remains to be validated in human clinical trials with ChAd-SARS-CoV-2-S. If results in mice were recapitulated, dose sparing strategies could enable production of a large number of vaccine doses that could curtail infection and transmission of SARS-CoV-2.
The emergence of SARS-CoV-2 S variants with mutations of amino acids in the receptor binding motif (e.g., B.1.351,  B.1.1.28, and B.1.617) is of concern because of their resistance to the inhibitory activity of many neutralizing antibodies Wang et al., 2021aWang et al., , 2021b. Indeed, human sera from subjects vaccinated with BNT162b2 mRNA or ChAdOx1 nCoV-19 (AZD1222) vaccines showed reduced neutralization against B.1.351 Madhi et al., 2021;Zhou et al., 2021). Concerningly, IM-administered ChAdOx1 nCoV-19 (AZD1222) showed reduced protective efficacy against mild to moderate B.1.351 infection in humans (Madhi et al., 2021). In K18-hACE2 transgenic mice, when we compared the immunogenicity of IN-delivered ChAd-SARS-CoV-2-S against WA1/2020 and chimeric SARS-CoV-2 strains expressing B.1.1.28 or B.1.351 spike proteins or a B.1.617.1 isolate, we also observed reduced (3-to 8-fold) neutralization of the variant viruses although the titers remained >1,000. At 6 weeks post-IN immunization of ChAd-SARS-CoV-2-S, K18-hACE2 mice were fully protected against weight loss and infection in the upper and lower respiratory tracts and brain by WA1/2020, Wash-B.1.351, and Wash-B.1.1.28. Remarkably, in a separate cohort of K18-hACE2 mice challenged 9 months after single IN immunization, animals maintained high neutralizing titers against all of the variant strains and were fully protected against Wash-B.1.351 challenge. Although correlates of protection are not fully established for SARS-CoV-2 vaccines, the high levels of cross-neutralizing antibodies against the variant viruses combined with robust virus-specific systemic and mucosal CD8 + T cell responses described previously  likely contribute to protection. Beyond this, antibody effector functions also might contribute to prevent SARS-CoV-2 infection and disease (Bartsch et al., 2021;Schä fer et al., 2021;Winkler et al., 2021). Indeed, we observed enhanced Fc effector functions against SARS-CoV-2 variant proteins in serum derived from IN-delivered ChAd-SARS-CoV-2-S including robust induction of ADNP and ADCP responses.

Limitations of study
Although a single intranasal administration of ChAd-SARS-CoV-2-S durably protected against SARS-CoV-2 variant replication in the upper and lower respiratory tracts even 9 months after immunization, we note several limitations in our study. (1) We performed challenge studies in BALB/c mice transduced with hACE2 or C57BL/6 mice expressing an hACE2 transgene. Durability and protection studies will need to be corroborated in hamsters, non-human primates, and ultimately in humans. (2) Although our studies suggest that the mucosal immunity induced by intranasal vaccination could limit SARS-CoV-2 transmission, the use of mice precluded formal respiratory transmission analysis, which is better studied in hamsters and ferrets (Muñ oz-Cell Reports 36, 109452, July 27, 2021 9 Article ll OPEN ACCESS Fontela et al., 2020). (3) We observed robust protection in vivo against viruses displaying B.1.351 and B.1.1.28 spike proteins likely due to the high serum neutralizing antibody titers. Even though neutralizing antibody levels were lower with the variant strains due to mutations at sites in the receptor binding motif, the high starting level against the historical SARS-CoV-2 likely provided a sufficient cushion to overcome this loss in potency. Studies in other animals or with even lower doses of vaccine where neutralizing titers might be lower are needed to determine whether the protective phenotype against variants of concern is maintained. (4) Finally, we did not establish the correlates of protection in these studies, as passive antibody transfer or T cell depletions were not performed. Such studies could be performed in follow-up experiments.
In summary, our study shows that IN immunization with ChAd-SARS-CoV-2-S induces robust and durable binding IgG and IgA antibody, neutralizing antibody, Fc effector functions, and LLPC responses against SARS-CoV-2. In mice, a single IN immunization with ChAd-SARS-CoV-2-S confers cross-protection against SARS-CoV-2 strains displaying spike proteins corresponding to B.1.351, B.1.1.28, and B.1.617.1 variants, even 9 months after vaccination. Given the efficacy of preclinical evaluation in multiple animal models (Bricker et al., 2021;Hassan et al., 2021;Hassan et al., 2020b) and the durable protective immunity against variants of concern, IN delivery of ChAd-SARS-CoV-2-S may be a promising platform for preventing SARS-CoV-2 infection, curtailing transmission, and, thus, warrants further clinical evaluation in humans.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:

ACKNOWLEDGMENTS
This study was supported by NIH contracts and grants (R01 AI157155, R01 EB026468-02S1, 75N93019C00062, 75N93021C00017, HHSN27220140001 8C, and U19 142790) and INV-00613 from the Bill and Melinda Gates Foundation. This work also was supported by Woodruff Health Sciences Center 2020 COVID-19 CURE Award.

RESOURCE AVAILABILITY
Lead contact Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Michael S. Diamond (diamond@wusm.wustl.edu).

Materials availability
All requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact author. This includes mice, antibodies, viruses, vaccines, and proteins. All reagents will be made available on request after completion of a Materials Transfer Agreement.

Data and code availability
All data supporting the findings of this study are available within the paper or from the corresponding author upon request. This paper does not report original code. Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.