Immunization with virus-like vesicle-based COVID-19 vaccine induces robust systemic and mucosal immunity

Abstract

Coronavirus disease 2019 (COVID-19) has led to significant global morbidity and mortality. Although several vaccines are approved, developing more effective candidates remains essential for long-term prevention. In this study, we present a COVID-19 vaccine candidate using a virus-like vesicle (VLV) platform, an enveloped self-amplifying RNA replicon incorporating an evolved Semliki Forest virus RNA polymerase and VSV glycoprotein. Two constructs were generated: VLV-S-FL (full-length spike protein) and VLV-S-RBD (receptor-binding domain). In C57BL/6J mice, VLV-S-FL elicited robust anti-spike antibody and T cell responses, with antibody levels comparable to those induced by the BNT162b2 mRNA vaccine. Prime-boost immunization with VLV-S-FL provided in vivo protection against SARS-CoV-2. Notably, intranasal boosting enhanced mucosal immunity, including IgA production and recruitment of CD4+ T, CD8+ T, and B cells in BALF. These findings suggest that VLV-S-FL is a promising COVID-19 vaccine capable of inducing both systemic and mucosal immune responses to prevent infection and reduce disease severity.

Introduction

The 2019 coronavirus disease (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has become an increasingly problematic global pandemic that has caused significant morbidity and mortality¹. SARS-CoV-2 has resulted in over 484 million documented cases of COVID-19 and nearly 7 million deaths worldwide, since 2024. The rapid and uncontrolled growth of COVID-19 has posed continuous, sizeable public health crisis. Persistent virus circulation raises concerns about the emergency of new variants with increased transmissibility and disease severity². Therefore, exploring a more effective long-term strategic approach is necessary. Currently, vaccination remains the most effective method for preventing SARS-CoV-2 infection and reducing disease severity³.

Currently, various types of vaccines against SARS-CoV-2 have been developed, including virus-based vaccines, protein-subunit vaccines, nucleic acid-based vaccines (RNA and DNA), and viral vector vaccines⁴. The former two directly introduce the antigen into the body, while the latter two, utilize the body’s own cells to produce viral antigens⁵. Adenovirus and mRNA-based vaccines have demonstrated effectiveness in protecting against severe symptoms, decreasing the spread of the virus, and reduce infection rates³. The BioNTech/Pfizer BNT162b2 mRNA vaccine, administered in two 30 μg doses, 21 days apart, conferred 95% protective immunity against COVID-19 disease in persons 16 years of age and older⁶, while Johnson & Johnson’s Ad26.COV2-S viral vector vaccine was 85.4% effective at preventing severe-critical disease⁷. Although vaccination offers protection, long-term prevention of global recurrence and mitigation of viral transmission is still limited by continuously emerging viral mutations and the waning of acquired immunity. Consequently, developing innovative vaccines that utilize alternative technologies may aid in efforts to contain the spread of SARS-CoV-2 and the prevention of future outbreaks. Furthermore, the development of additional vaccines also can help broaden and strengthen the immune response because of concerns on emerging variants.

The route of immunization plays an important role in dictating the type and quality of the immune responses, particularly at sites of pathogen entry or initial infection. Most vaccination strategies focus on vaccines administered via intramuscular route that primarily induce systemic immunity. However, respiratory mucosal immunity is crucial for protecting against COVID-19, as the virus infects the upper and lower respiratory tracts via the ACE2 receptor^8,9. Given this respiratory tropism, eliciting mucosal immunity at the virus’s entry port could help prevent transmission^10,11. The mucosal immune system can produce specific IgA antibodies locally, forming the first line of defense against viral invasion. It is capable of eliciting a powerful immune response to external pathogenic antigens while preserving immunological tolerance to non-pathogenic antigens¹². Intranasal vaccines are designed to induce mucosal immunity, offering a first line of defense against respiratory pathogens. Additionally, intranasal administration, being needle-free, improves ease and speed, decreases costs, and reduces the pain associated with vaccination. The importance of mucosal immunity and mucosal vaccines for SARS-CoV-2 has been emphasized, with several studies exploring intranasal (IN) vaccination to meet this need and highlight the significance of generating mucosal immunity to control infection^13–15.

Virus-based platforms have been widely studied and tested in clinical trials for vaccine¹⁶. As the natural ability of viruses is to infect cells, the viral vector mimics the infection process, thus having high gene transduction capability and the ability to generate a more robust humoral and cytotoxic (CD8+) T cell responses¹⁷. One viable option is the virus-like vesicle (VLV), an alphavirus-rhabdovirus hybrid vector that functions as a self-amplifying RNA (saRNA) replicon and is capable of eliciting robust T-cell and antibody responses¹⁸. VLVs contain a Semliki Forest Virus (SFV) derived RNA-dependent RNA polymerase and a Vesicular Stomatitis Virus (VSV) derived glycoprotein as its essential components, therein replicating like SFV but entering and spreading from cells via the VSV glycoprotein¹⁹. Studies found that VLVs can infect DCs, promote CD8+ T cell activation²⁰, and allow for rapid and efficient expression of the desired antigens in the immunized host²¹. The efficacy and safety profiles of VLV vector vaccines were evaluated in several studies using mice²² and non-human primates (as a booster)²³. Immunization with VLVs expressing HBV middle surface antigen (MHBs) induced HBV-specific CD8+ T cells and protected mice against acute HBV infection²⁴. Furthermore, expression of the intact middle S antigen (MHBs) enabled robust and long-lasting antibody responses²⁵. This VLV-based vaccine offers distinct advantages over the existing adenoviral vector vaccine Ad26. COV2-S (John & Johnson)²⁶. One limitation of adenoviral vector vaccines is the presence of pre-existing immunity, as prior exposure can reduce vaccine efficacy by eliciting vector-specific immune responses²⁷. In contrast, VLVs are derived from VSV glycoprotein-enveloped RNA replicons, which do not have widespread pre-existing immunity in the human population. This may provide an advantage in eliciting robust immune responses even in individuals with prior exposure to adenovirus-based vaccines. Additionally, unlike non-replicating adenoviral vectors, VLVs can generate viral particles upon cell entry and undergo intracellular replication, resulting in increased expression of viral genes and elevated levels of the target antigen. This induces a strong and sustained immune response against SARS-CoV-2. Moreover, upon entry into host cells, VLV-S-FL exposes its viral RNA to PRRs in the cytoplasm and endosomes, thereby stimulating innate immune responses and reducing the need for an external adjuvant. Therefore, the VLV-based vector platform has proven to be a suitable option as a COVID-19 vaccine candidate.

As SARS-CoV-2 infection is primarily initiated through viral binding of the spike (S) glycoprotein to the human angiotensin-converting enzyme 2 (ACE2) receptor^28,29, the S protein has been the primary protein target of COVID-19 vaccines. The S protein consists of two subunits, the S1 subunit which contains a receptor-binding domain (RBD) that recognizes and binds to the host receptor, and the S2 subunit which mediates viral-cell membrane fusion³⁰. These functions point to the S protein as a critical aspect of the viral life cycle^31,32. Several studies have found that the S protein effectively induces the production of neutralizing antibodies (nAbs), wherein these nAbs then target viral S proteins to prevent further infection at the stage of viral entry³³. Therefore, using the S protein as a vaccine antigen has proven to be effective at conferring protective immunity to the host. Particularly, the RBD of the full-length S protein is the primary target of nAbs in interfering with the viral receptor-binding process³⁴.

In our study, we constructed two kinds of VLVs, VLV-S-FL and VLV-S-RBD, and found that VLV-S-FL, but not VLV-S-RBD, could elicit anti-spike protein-specific antibody responses in C57BL/6J mice and induce effective neutralizing antibody response with protective immunity against the SARS-CoV-2. The intranasal administration of VLV-S-FL induced substantial mucosal immunity, including IgA production and immune cell infiltration in the BALF. This suggests that the VLV platform could be used to develop a COVID-19 vaccine that provides strong systemic and mucosal immunity against SARS-CoV-2.

Results

Design and construction of VLV-S-FL and VLV-S-RBD

Given that SARS-CoV-2 utilizes the S protein to enter host cells and that the S protein’s RBD region is the primary binding region for ACE2, we designed VLV vector vaccines that encode full-length and RBD sequences of the B.1.1.7 strain of spike protein, herein named VLV-S-FL and VLV-S-RBD, respectively (Fig. 1A). Our VLV vector utilized the ribosome-skipping 2A peptide from Thosea asigna virus (T2A), a self-processing viral peptide which separates different protein-encoding sequences contained within a single ORF transcription unit²⁴. Therefore, VLVs may express multiple exogenous antigens, full-length or partial S protein, and VSV-G from the same sub-genomic RNA. To produce our VLVs, we transfected BHK-21 cells with these two kinds of plasmids and subsequently collected supernatants containing VLV-S-FL or VLV-S-RBD. After further propagation, we concentrated the VLVs by ultrafiltration and then determined viral titers using the TCID50 method. To evaluate the expression profile of the full-length or partial S protein, we co-cultured the BHK-21 cells with 10³ or 10⁴ TCID50 VLV-S-FL or VLV-S-RBD for 24 h and harvested for the expression of spike protein. We found that the spike protein showed higher expression levels in VLV-S-FL infected BHK-21 cells, rather that VLV-S-RBD infected cell (Fig. 1B). This difference might be attributed to variants in the structural recognition of the antigens, differences in sequence length, and the antibody recognition site. To further validate RBD expression, we employed a specific anti-RBD antibody and increase the viral dose used for infection. At a higher titer of 10⁷ TCID50 VLV-S-RBD, RBD expression can be detected in BHK-21 cells (Supplementary Fig. 1). Furthermore, we stained the cells for VSV-G with polyclonal anti-VSV-G antibody to detect the expression in the cells. As shown in Fig. 1C, both VLV-S-FL and VLV-S-RBD-infected BHK-21 cells displayed intense green fluorescence.

Fig. 1. Virus-like vesicle platform for COVID-19 vaccine expressing RBD or full length of SARS-CoV-2 spike protein.

A Design of replicating VLV for expression of RBD or full length of SARS-CoV-2 spike protein. The plasmid construction contains T2A self-cleaving peptide that allow it to express both antigen and VSV-G protein. B Expression of spike protein in BHK-21 cells infected with VLV-S-RBD and VLV-S-FL. C Validation of VSV-G expression in BHK-21 cells after infection with VLV-S-RBD or VLV-S-FL (MOI = 1) by immunofluorescence at 24 h post infection.

Antibody responses elicited by VLV-S-FL and VLV-S-RBD in C57BL/6J mice

To determine if VLV-S-FL or VLV-S-RBD could effectively elicit an antibody response, we immunized different groups of C57BL/6J mice with a single i.p. injection of either VLV-S-FL or VLV-S-RBD (10⁸ PFU) on day 0. Antibody levels in the sera were analyzed on day 28 after immunization (Fig. 2A). The results showed that C57BL/6J mice immunized with VLV-S-FL had a significantly greater antibody responses against the spike protein than those immunized with VLV-S-RBD or PBS (Fig. 2B), indicating higher immunogenicity from VLV-S-FL vaccination. Although unclear why, VLV-S-RBD vaccination was less immunogenic in C57BL/6J mice; possibly as a result of its shorter antigen sequence or lower expression levels, which may reduce its effectiveness in stimulating a sufficient immune response. Safe and efficient administration of vaccines is essential for successful immunization. To investigate the impact of different injection methods, we administered a single dose of 10⁸ PFU VLV-S-FL to C57BL/6J mice either intramuscularly (i.m.) or intraperitoneally (i.p.) on day 0 (Fig. 2C). On day 28, we collected blood from the immunized mice to measure anti-spike-specific IgG antibodies using ELISA. Both injection methods resulted in high antibody levels. While i.p. injections elicited a slightly stronger immune response compared to i.m. injections, the overall antibody levels were comparable (Fig. 2D). Additionally, serial dilution of antibody titers showed similar results for both methods, suggesting that i.m. and i.p. injections are equally effective for eliciting antibody responses and are safe and effective methods for vaccine administration (Fig. 2E).

Fig. 2. Anti-spike protein antibody responses in the serum of C57BL/6J mice immunized with PBS, VLV-S-RBD or VLV-S-FL.

A Experimental procedure. B Antibody levels, expressed as OD value, in the serum of C57BL/6J mice (n = 5) on day 26 after the immunization with PBS, VLV-S-RBD (1 × 10⁸ PFU/mice) or VLV-S-FL (1 × 10⁸ PFU/mice). C Experiment procedure of different injection methods on anti-spike protein immune responses. D Antibody levels, expressed as OD value, in the serum of C57BL/6J mice (n = 5) on day 26 after the immunization with VLV-S-FL (1 × 10⁸ PFU/mice) with different injection methods, including intramuscular injection (i.m.) and intraperitoneal injection (i.p.). E Antibody series dilution in the serum of C57BL/6J mice (n = 5) on day 26 after the immunization with VLV-S-FL (1 × 10⁸ PFU/mice) by using different injection methods. **, P < 0.01; ***, P < 0.005; ****, P < 0.001.

Comparing the immunogenicity and stability of VLV-S-FL

BNT162b2 (BioNTech and Pfizer) is a lipid nanoparticle-formulated, nucleoside-modified mRNA vaccine that is used to prevent COVID-19³⁵. This mRNA vaccine encodes the SARS-CoV-2 spike protein for endogenous expression to trigger an immunological response. The BNT162b2 vaccine is administered i.m. with two 30 μg doses, at an interval of 21 days. To compare the efficacy of the VLV-S-FL vaccine against BNT162b2, we injected C57BL/6J mice with a single i.m. injection of VLV-S-FL (10⁸ PFU) or the BNT162b2 at the equivalent amount of 10 ng, 100 ng, or 1 μg twice, 14 days apart (Fig. 3A). On day 28, blood was collected to compare anti-S antibody titers. The results showed that antibody levels from mice that received VLV-S-FL vaccination were comparable those immunized with 100 ng and 1 μg of the BNT162b2, but higher than in mice vaccinated with the equivalent amount 10 ng (Fig. 3B). In addition, serial dilution results revealed a similar tendency in that antibody responses induced by VLV-S-FL were stronger than those generated by 10 ng vaccination, but lower than those immunized with 100 ng or 1 μg (Fig. 3C).

Fig. 3. Comparison of antibody responses in C57BL/6J mice immunized with VLV-S-FL or BNT162b2.

A Experimental procedure. B Antibody levels, expressed as OD value, in the serum of C57BL/6J mice (n = 5) on day 28 after the immunization with VLV-S-FL (1 × 10⁸ PFU/mice) or BNT162b2 with different amount (10 ng, 100 ng and 1 μg). C Antibody series dilution in the serum of C57BL/6J mice (n = 5) on day 26 after the immunization with VLV-S-FL (1 × 10⁸ PFU/mice) or BNT162b2 with different amount (10 ng, 100 ng and 1 μg). D Neutralization assay. VSVΔG-GFP pseudotyped with spike protein was incubated with series dilutions of serum and VSV-G antibody for 1 h at 37 °C. The mixture was then transferred to the corresponding wells with Huh7.5 cells and incubated for 48 h at 37 °C. Apparent number of GFP+ spots were counted and showed in the graph. Bar represents mean ± standard deviation (SD). *, P < 0.05; **, P < 0.01; ***, P < 0.005; ****, P < 0.001.

NAbs are generated within weeks of infection or immunization and can protect cells from viral invasion, thereby conferring protective immunity to the host³⁶. Accordingly, the production of NAbs is considered to be the main goal of SARS-CoV-2 vaccination. To determine the protective effects against SARS-CoV-2, we detected neutralizing antibody titers against SARS-CoV2, using a pseudotyped VSV virus-based assay³⁷. Serial dilutions of sera were mixed with 75 i.u. of pseudotyped virus to allow for visualization and quantification of GFP signals. The results showed that the GFP signal spots from i.m. groups were lower than that of the control group (Fig. 3D), indicating the i.m. route could effectively induce NAbs after immunization. Additionally, the GFP signals for VLV-S-FL was lower than the equivalent amount of 10 ng, but higher than the other amounts at 100 ng and 1 μg of BNT162b2. These data suggest that VLV-S-FL immunization elicits neutralizing antibody responses and is effective at protecting against infection from SARS-CoV-2 pseudotyped virus.

The mRNA vaccines typically require storage at ultra-low temperatures to avoid degradation and ensuring vaccine stability is crucial for increasing the availability and affordability of vaccines, thereby making vaccination more accessible worldwide. To assess the stability of VLV-S-FL under different conditions, we stored it at room temperature (RT), 4 °C, −20 °C and −80 °C. We measured the TCID50 of VLV-S-FL over varying time periods. No significant differences in viral titers were observed across all storage temperatures on day 3. However, a noticeable decline in stability was detected at RT by day 7 and day 14. In contrast, storage at 4 °C, −20 °C, and −80 °C showed no significant changes in stability over the same period (Supplementary Fig. 2).

Optimization of VLV-S-FL immunization strategies

To further optimize the immunization procedure, we compared three vaccination strategies: a single dose intramuscular injection, intramuscular priming and boosting (i.m./i.m.), and intramuscular priming followed by intranasal boosting (i.m./i.n.). Initially, we administered VLV-S-FL i.m. and then boosted with VLV-S-FL either i.m. or intranasally (i.n.) (Fig. 4A). Blood samples were collected on day 28 for antibody analysis. Our results showed that the booster immunization with an intramuscular injection of VLV-S-FL led to increased antibody titers at various dilutions (Figs. 4B, C). This indicates that the intramuscular priming and boosting enhances vaccine efficacy more effectively.

Fig. 4. Comparison of antibody response in C57BL/6J mice immunized with one or two doses of VLV-S-FL.

A Experimental procedure. B Antibody levels, expressed as OD value, in the serum of C57BL/6J mice (n = 5) on day 28 after the immunization with VLV-S-FL (1 × 10⁸ PFU/mice). C Antibody series dilution in the serum of C57BL/6J mice (n = 5) on day 28 after the immunization with VLV-S-FL (1 × 10⁸ PFU/mice). ****, P < 0.001.

Effects of VLV-S-FL on the type I IFN induction and CD8+ T cell response stimulation

Natural viral infection can initiate adaptive immunity as well as type I IFN induction²⁰. To study the host immune responses to VLV-S-FL, we isolated splenocytes from the naïve C57BL/6J mice and co-cultured in vitro with the addition of 10³ TCID50 VLV-S-FL or PBS for 3–48 h followed by detection of the kinetic expression of type I IFN-related signaling pathway molecules. The results showed that the mRNA level of Tlr7, Irf3, Irf7 and Ifna showed similar trends that they were elevated at 6 h and reached the peak at nearly 24 h (Supplementary Fig. 3A). The mRNA levels of Ifnb were increased 3 h and reached a peak at 9 h. These dynamic expression changes indicate that VLV-S-FL could effectively induce the activation of type I IFN related anti-viral immune responses. To explore the activation of T cell responses, we detected the levels of CD4+ and CD8+ T cells of the mice administrated with the VLV-S-FL at 10⁸ PFU/mouse. The results showed that the VLV-S-FL administration significantly increase the percentage of CD8+ T cells on day 3 and day 7, whereas no significant changes of CD4+ T cells were observed (Supplementary Fig. 3B).

Protective immunity induced by VLV-S-FL immunization in K18-hACE2 mice

To determine the protective immunity induced by VLV-S-FL immunization, the K18-hACE2 mice were immunized i.m. with VLV-S-FL or BNT162b2 on day 0 and Day 14, followed by infection with SARS-CoV2 on day 28 (Fig. 5A). The mice were monitored for body weight and survival. The results showed that the antibody levels induced by VLV-S-FL were comparable to those induced by the BNT162b2 (Fig. 5B). Compared to the VLV-S-FL and BNT162b2 immunization groups, PBS and VLV-GFP groups exhibited significant decreases in body weights (Fig. 5C). Notably, all mice vaccinated with VLV-S-FL were completely protected from death (Fig. 5D). Importantly, VLV-S-FL demonstrated increased protective effects against infection compared to BNT162b2. At the experiment endpoint, lung tissues were harvested from all groups for histopathological analysis. Compared to the PBS and VLV-Ctrl groups, mice immunized with either BNT162b2 or VLV-S-FL displayed significantly reduced lung pathology (Fig. 5E), as evidenced by decreased alveolar wall thickening, reduced immune cell infiltration, and preserved alveolar structures. These findings indicate that both vaccination strategies showed protective effects against SARS-CoV-2 infection. Furthermore, we assessed the inflammatory responses by measuring the levels of pro-inflammatory cytokines in lung tissues. We found that the expression levels of Ccl5 and Tnfa, rather than Ccl4 and Il6, were significantly decreased following immunization, supporting the protective effect of the vaccines (Fig. 5F). To further evaluate the protective efficacy against SARS-CoV-2 variants, we performed a neutralizing assay using serum collected from the immunized mice. Huh7.5 cells were used as target cells, and a pseudotyped virus expressing the Omicron (B1.1.529) spike protein was used to mimic variant-specific infection. As shown in Fig. 5G, serum from the VLV-S-FL immunized group significantly inhibited Omicron pseudovirus infection, demonstrating the potential of this platform to elicit broadly NAbs against emerging SARS-CoV-2 variants.

Fig. 5. Protective effects of VLV-S-FL immunization in vivo study.

A Schematic of experimental plans. The immunized K18-hACE2 mice were infected with SARS-CoV-2, followed by recording the survivals and weights of the mice. After immunization, the serum was separated from the blood to analyze the antibody responses (B). Body weight (C) and survival (D) analysis were observed and recorded. E The lungs were isolated and sectioned for HE staining. The stained sections were observed under the microscope. F The mRNA analysis of Tnfa, Il6, Ccl5 and Ccl4 in lung tissues. RNA was extracted from FFPE mouse lung samples, and each data point represents the average of three independent assays. G Neutralization assay. VSVΔG-GFP pseudotyped with Omicron spike protein was incubated with serum and VSV-G antibody for 1 h at 37 °C. The mixture was then transferred to the corresponding wells with Huh7.5 cells and incubated for 48 h at 37 °C. Apparent number of GFP+ spots were counted and showed in the graph. Bar represents mean ± standard deviation (SD). *, P < 0.05; **, P < 0.01; ****, P < 0.001.

Enhanced mucosal immunity induced by intranasal immunization in mice

SARS-CoV-2 primarily infects the upper respiratory tract through airborne transmission and can cause severe acute respiratory disease. Most COVID-19 vaccines in development are administered i.m., focusing on inducing systemic immunity. However, intranasal vaccines have added advantage of inducing mucosal immunity, targeting the primary route of viral entry, and potentially blocking transmission. To investigate the potential mucosal immunity induced by VLV-S-FL, we immunized the mice i.n., i.m. combined with i.n., and i.n. combined with i.n., as shown in Fig. 6A. On day 28, blood samples were collected and analyzed, revealing that the i.m. combined with i.n. group showed increased levels of anti-S IgG in serum and anti-IgA in BALF, indicating enhanced mucosal immunity with the combination of nasal inoculation. Further analysis of immune cells in BALF revealed an increased in CD45+ cell in the nasal inoculation group, suggesting that nasal inoculation enhances mucosal immune responses (Fig. 6B). Additionally, mRNA levels of various indicators showed an increase in CD4, CD8a, and CD19. In comparison, the i.n. combined with i.n. group showed an increase in CD8a and CD4 T cells (Fig. 6C). Furthermore, the mRNA levels of Ifnb, Ifnr, and Tnfa showed an increase in lung tissues of the nasal inoculation combination groups (Fig. 6D). To assess whether VLV-S-FL can induce spike-specific T cell activation via intranasal administration, we administered 10⁸ PFU of VLV-S-FL, VLV-Ctrl or PBS into C57BL/ mice on day 0 and day 14 via intramuscular injection and intranasal inoculation. On day 28, BALF was collected to isolate cells and evaluate cytokine secretion following in vitro restimulation with spike protein for 36 h. Compared to the PBS and VLV-Ctrl, VLV-S-FL immunized mice produced significantly higher levels of IFN-γ (Fig. 6E). These data indicate that intranasal boosting with VLV-S-FL results in substantial induction of mucosal immunity, which may help block infection and onward transmission.

Fig. 6. VLV-S-FL elicited mucosal immune responses after intranasal vaccination.

A Experimental procedure and the IgG and IgA levels in the serum and BALF of C57BL/6J (n = 5) on day 28 after the immunization with VLV-S-FL (1 × 10⁸ PFU/mice), respectively. B CD45+ immune cell analysis in BALF of C57BL/6J mice (n = 5) on day 28 after the immunization with VLV-S-FL (1 × 10⁸ PFU/mice). C, D The mRNA analysis of immune cell markers and inflammation cytokines in lung tissues of C57BL/6J (n = 5) immunized with VLV-S-FL. E The cells in BALF were collected and co-cultured with purified spike protein. IFN-γ production was analyzed using ELISPOT assay. *, P < 0.05; **, P < 0.01; ***, P < 0.005; ****, P < 0.001.

Systemic and mucosal immune responses induced by VLV-S-FL

VLV-S-FL immunization induced effective systemic and mucosal immune responses. To further compare these responses, we immunized the mice i.m., i.m. combined with i.m., and i.m. combined with i.n. administration, as shown in Fig. 7A. Blood samples were collected on day 28 and analyzed for antibody responses at various dilutions (Fig. 7B). The results showed that the levels of antibody responses followed the same trend across the groups. We then collected BALF from the different immunization groups and found that the i.m. combined with i.n. showed increased anti-S IgA levels (Fig. 7C) and enhanced cell counting numbers (Fig. 7D), indicating the mucosal immunity induced by VLV-S-FL. Additionally, the cells in BALF from different groups was collected and analyzed. Analysis of cells in BALF revealed a significant increase in the percentages of CD4, CD8 and CD19 cells (Fig. 7E).

Fig. 7. Mucosal immune response comparisons after VLV-S-FL immunization.

A Experimental procedure and B the antibody levels in the serum of C57BL/6J with different procedures. C The anti-S IgA levels in the BALF of C57BL/6J (n = 5) on day 28 after the immunization with VLV-S-FL (1 × 10⁸ PFU/mice). D Total cell counts in BALF of C57BL/6J. E Immune cell population analysis in BALF of C57BL/6J mice (n = 5) on day 28 after the immunization with VLV-S-FL (1 × 10⁸ PFU/mice). *, P < 0.05; **, P < 0.01; ****, P < 0.001.

Discussion

The exploration of novel vaccines for COVID-19 remains crucial in combatting the threat of emerging viral mutations and the long-term public health challenges posed by the pandemic⁴. Viral vector vaccines utilize recombinant viruses to deliver antigenic sequences, enabling host cells to produce high levels of recombinant proteins that stimulate immune responses³⁸. VLV-based vaccines offer several distinct advantages, including efficient antigen delivery, sustained antigen expression, and the potential for intranasal administration, positioning this platform as a promising alternative to current vaccine technologies. In this study, we present a novel COVID-19 vaccine candidate based on a VLV platform engineered to additionally express the full-length spike protein (VLV-S-FL). A prime-boost immunization strategy using VLV-S-FL demonstrated effective protective effects against SARS-CoV-2 infection in vivo, and intranasal boosting with VLV-S-FL led to a robust induction of mucosal immunity.

The selection of a potent immunogenic target is an essential factor in developing safe and effective COVID-19 vaccines⁴. The spike protein of SARS-CoV-2 plays a critical role in viral entry by facilitating receptor binding and membrane fusion and is therefore considered a key antigenic target for vaccine design³³. Within the spike protein, the RBD region of the S1 subunit is particularly important, as it directly mediates attachment to host cell receptors and is a primary target of NAbs. However, in our study, we were initially unable to detect RBD expression in the VLV-S-RBD infected BHK-21 cells (Fig. 1B). This observation is consistent with our immunogenicity results, in which mice immunized with VLV-S-RBD failed to generate a significant antibody response (Fig. 2B). To further validate RBD expression, we employed a specific anti-RBD antibody and increase the viral dose used for infection. At a higher titer of 10⁷ TCID50 VLV-S-RBD, RBD expression can be detected in BHK-21 cells (Supplementary Fig. 1). These findings suggest that the RBD expression level may have been too low for effective antigen presentation, resulting in weak immunogenicity that was insufficient to activate a robust immune response in C57BL/6J mice. To improve immunogenicity, further optimization of the antigen sequence is necessary. One potential strategy is to extend the RBD coding region by repeating the sequence multiple times, either with or without linker sequences, which could enhance antigen presentation by increasing overall protein size and epitope availability. Notably, the presence and design of linkers also impact the folding, stability, biological activity, and immunogenicity of proteins³⁹, which need be carefully considered, as well.

Currently, mRNA and adenoviral vector vaccines have been the primary strategies for COVID-19 immunization. The BNT162b2 mRNA vaccine delivers messenger RNA encoding the full-length SARS-CoV-2 spike protein into host cells, thereby triggering an adaptive immune response. Despite their promising, these vaccines face limitations, including the requirement for cold storage, relatively short duration of protection, and limited accessibility in low-resource regions⁴⁰. Adenoviral-based vaccines vector such as Ad26.COV2.S are widely used non-replicating DNA vaccines⁴¹. However, pre-existing immunity to the viral vector can reduce vaccine efficacy in some individuals. In contrast, saRNA vaccines have emerged as a promising alternative, capable of enhancing antigen expression and immune activation while requiring lower RNA doses⁴⁰. The Arcturus COVID-19 vaccine (ARCT-154), based on self-amplifying mRNA, has been approved in Japan after demonstrating non-inferiority to the Pfizer-BioNTech BNT162b2 vaccine for the Wuhan-Hu-01 strain and superiority against the Omicron BA.4/5 variant⁴². Similar to this sa-mRNA vaccine, the VLV platform employs a self-replicating RNA genome to drive sustained antigen production in vivo, leading to enhanced immunogenicity. Unlike sa-mRNA vaccine (ARCT-154) that require nanostructured lipid carriers for delivery⁴⁰, VLVs are enveloped RNA vectors that can directly infect host cells, eliminating the need for additional delivery vehicles. This unique feature may allow more efficient in vivo transduction, prolonged antigen presentation, and enhanced immune responses^43,44. In this study, we selected the mRNA vaccine as a comparator due to its well-documented immune response profile and widespread clinical use, making it a suitable benchmark for assessing the immune responses induced by VLV-S-FL. As both platforms rely on intracellular antigen expression, they share key mechanistic similarities. Notably, VLV-S-FL demonstrated protective effects comparable to those of the mRNA vaccine, as evidenced by survival rates and body weights following SARS-CoV-2 infection (Fig. 5). To further assess the protective potential of VLV-S-FL, future studies should include comparisons with other self-amplifying RNA platforms and adenovirus-based vaccines, providing a more comprehensive evaluation of its advantages.

Several studies have demonstrated the importance of respiratory mucosal immunity in COVID-19 protection, as the virus infects the upper and lower respiratory tracts via the ACE2 receptor^8,9. While most COVID-19 vaccines are administered i.m. and elicit strong systemic antibody responses, they generally induce limited mucosal immunity. The mRNA vaccines also face challenges in mucosal delivery due to mRNA instability, limited cellular uptake at mucosal sites, and potential local reactogenicity⁴⁵. In our study, intranasal administration of VLV-S-FL effectively stimulated mucosal immune responses, including the production of IgA and significant immune cell infiltration in BALF (Fig. 6). These findings highlight the potent mucosal immune response induced by VLV-S-FL, which is critical for preventing SARS-CoV-2 infection at its primary entry site. The IgA produced in response to VLV-S-FL can neutralize pathogens at the mucosal barrier, thereby blocking infection and reducing transmission risk. Moreover, we found that the intramuscular priming followed by intranasal boosting (i.m./i.n.) regimen elicited stronger mucosal immune responses compared to intranasal only (i.n./i.n.) or intramuscular only (i.m./i.m.) regimens. This enhanced response may be attributed to the synergistic effect of systemic priming and local mucosal restimulation. Intramuscular priming generates a strong systemic immune response and immunological memory, which can be efficiently recalled and directed to mucosal sites upon intranasal boosting. The intranasal boost likely facilitates local immune activation, promoting the recruitment of lymphocytes to mucosal tissues, leading to increased IgA production and immune cell infiltration in the respiratory tract. In contrast, the i.n. + i.n. regimen may not provide sufficient initial immune activation, while the i.m. + i.m. regimen may not effectively induce mucosal-specific responses. Furthermore, the sequential i.m. + i.n. regimen may optimize antigen presentation, ensuring the development of both systemic and mucosal immunity for broader protection. One study exploring the “Prime and Spike” (P&S) strategy, in which animals were primed i.m. with a SARS-CoV-2 mRNA-lipid nanoparticle (LNP) vaccine formulation and boosted i.n. with unadjuvanted spike protein. This approach induced strong antigen-specific mucosal antibodies and tissue-resident memory T and B cells, while also reducing viral shedding compared to mRNA-LNP prime-boost regimens. Intranasal vaccines offer several advantages, including the ability to generate protective mucosal immunity at the site of viral entry⁴⁶, needle-free administration, reduced medical waste, and greater convenience⁴⁷. Therefore, these findings highlight the potential of VLV-S-FL as a promising COVID-19 vaccine candidate capable of eliciting both systemic and mucosal immunity for comprehensive protection against SARS-CoV-2.

Vaccine safety, stability and immunogenicity are critical concerns that should be carefully evaluated. To address safety, viral vectors are modified by removing the virulence genes while maintaining the capacity to enter host cells. This allows for efficient gene delivery without causing disease, mimicking the route of viral infection. Consequently, optimizing dosage and administration remains a key area for future investigation, as it may significantly influence vaccine efficacy and tolerability across individuals. In the study, we administrated a dose of 10⁸ PFU of VLV-S-FL to mice. No significant adverse effects, such as body weight loss or signs of distress, were observed during the experimental period. This dosage was selected based on prior studies using the VLV platform⁴⁸. Further safety assessments of VLV-S-FL are essential to address potential concerns such as systemic toxicity and uncontrolled viral replication. Comprehensive toxicological evaluations, including blood biochemical analysis and long-term monitoring, will be necessary to fully characterize the safety of the VLV platform. Regarding stability, maintaining viral titers is critical for ensuring efficient infection of host cells and robust antigen expression following vaccination. Temperature-induced degradation can impair antigen presentation, potentially diminishing immunogenicity and protective efficacy. For instance, mRNA vaccines require ultra-cold storage to prevent degradation and preserve potency. In contrast, viral vector vaccines generally exhibit greater stability, making them more practical in storage and transportation conditions. In our assessment, VLV-S-FL maintained stable viral titers at we found it the viral titer do not have significant changes at −20 °C for 2 weeks and at 4 °C for 1 week (Supplementary Fig. 2). This suggests that VLV-S-FL retains infectivity and immunogenic potential under moderately cold conditions, enhancing its utility for widespread. However, as our current evaluation focused on viral titer, further studies assessing antigen expression and immunogenicity after storage under various conditions will be essential to determine optimal storage parameters and long-term stability. Scalability and logistical feasibility are also important for the translational potential of VLV-based vaccines. VLV vectors can be produced at high titers using standard cell culture systems, supporting large-scale manufacturing. Additionally, their relatively moderate cold chain requirements may facilitate distribution logistics. Intranasal administration further offers practical advantages by eliminating the need for needles or trained personnel. Thus, these features make the VLV platform a practical, scalable and immunogenically effective strategy for vaccine development, enhancing the practical relevance of our findings and provide a broader perspective on the translational potential of VLV-based vaccines.

To further evaluate the clinical potential of VLV-S-FL, it is essential to assess both the durability of the immune responses it induces and its efficacy against emerging SARS-CoV-2 variants. The evolution of SARS-CoV-2 has led to the emergence of numerous variants, such as Omicron, which have driven successive waves of infection worldwide^49,50. In the study, the serum collected from the VLV-S-FL immunization group showed effective protection against Omicron pseudovirus infection (Fig. 5G). Studies have shown that existing vaccines such as BNT162b2 and Ad5-nCoV exhibit varying degrees of reduced protection against variant strains^51–53, highlighting the need for vaccines capable of eliciting broad-spectrum immunity. Although our current study did not directly assess the longevity of immune responses following VLV-S-FL vaccination, future studies will be essential to evaluate the durability of both mucosal and systemic immunity over time. This investigation will help determine optimal dosing schedules and the potential need for booster immunizations to sustain protective immunity against viral infection^54–59.

Methods

Cell line and plasmid information

Hamster kidney cells (BHK-21) and Huh7.5 cell line cells were cultured and maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% (vol./vol.) fetal bovine serum (FBS), penicillin (100 U/ml) and streptomycin (100 mg/ml) at 37 °C in a 5% CO₂ incubator.

Recombinant plasmids encoding full-length or RBD sequences of the spike protein were constructed. Primers used for PCR amplification of the full-length spike protein sequence using the forward primer: 5′-TTAATTAAAATGACAGAATCTATTGTTAGATTTCCTAATATT-3′ and reverse primer: 5′-CCTGCAGGTTTGTTTTTAACCAAATTAGTAGACTTTTT-3′. Primers used for PCR amplification of the spike protein’s RBD sequence were, forward primer: 5′- GGCGCGGCGCCGGCGCGCCGCGTTAATTAAAATGTTTGTTTTTCTTGTTTTATTGC -3′ and reverse primer: 5′-GCCGCATGCATCCTGCAGGTGTGTAATGTAATTTGACTCC-3′. Then, fragments of full-length or RBD were cloned into pCMV-SFVT2AG plasmids with modified cloning sites (PacI and SbfI) added to the 5′ and 3′ ends. Next, the gene of interest was inserted upstream of a ribosomal T2A skipping site and VSV-G protein. The T2A site allows for the expression of both the protein of interest and VSV G protein from the same promoter and ensures the readthrough of the full-length or RBD spike protein sequences, prior to VLV production.

Transfection and recovery

VLV-S-FL and VLV-S-RBD were recovered by transfecting BHK-21 cells with the pCMV-SFVFLT2AG and pCMV-SFVRBDT2AG plasmids, respectively. BHK-21 cells were seeded the day before transfection at a density of 3 × 10⁵ cells per well of a 6-well plate. The next day, Fugene HD transfection reagent (Promega Corporation, WI) and DNA complex were mixed thoroughly and added to the cells. When cytopathic effects were observed, the supernatants containing recombinant VLV-S-FL or VLV-S-RBD were collected by centrifugation at 600 × g for 10 min to remove cell debris. VLV-S-RBD and VLV-S-FL were then concentrated by centrifugation through a 100-kDa Amicon Ultra filter unit (EMD Millipore Corporation, MA). To obtain the 50% tissue culture infective dose (TCID50), the varied dilutions of VLV-S-FL and VLV-S-RBD were incubated with BHK-21 cells to determine the highest dilution producing cytopathic effect in 50% of the incubated cells. The 50% end-point dilution is expressed as TCID50/ml, calculated by the Reed-Muench method. Aliquots were stored at −80 °C.

Immunostaining

BHK-21 cells were seeded on the glass of chamber slides (Corning, NY) one day before infection with VLV-S-FL or VLV-S-RBD at MOI = 1. Cells were washed with PBS twice, fixed with 4% paraformaldehyde for 15 min at 24 h post-infection, and permeabilized with 0.1% Triton-X 100 for 5 min, followed by washing three times with PBS. Cells were blocked with PBS containing 2% FBS and incubated in a 1:200 dilution of mouse monoclonal VSV-G antibodies for 1 h. After extensive wash steps, cells were stained with a 1:500 dilution of secondary antibodies for 1 h at RT. The stained slides were observed using a fluorescent microscope (Keyence, IL).

Protein extraction and western blot analysis

Cell lysates were harvested by RIPA buffer (ThermoFisher Scientific, Waltham, MA) with a cocktail of protease and phosphatase inhibitors. Similar amounts of proteins were separated by 10% SDS-PAGE and transferred to PVDF membranes. After blocking with 5% nonfat dry milk or BSA dissolved in PBST (0.1% Tween in PBS), the membranes were incubated with anti-spike protein primary antibody (NBP3-12854, NOVUS) or anti-RBD primary antibody (A21254, ABclonal) at 4 °C overnight. Subsequently, the membranes were washed three times in PBST, and probed with HRP-conjugated secondary antibodies for 1 h at room temperature. After washing three times in PBST, the target protein bands were visualized by Bio-rad ChemiDoc XRS system (USA).

Animals and immunizations

Six- to eight-week-old C57BL/6J mice and 4- to 5-week-old B6.Cg-Tg(K18-ACE2)2Prlmn/J (K18-hACE2) mice were purchased from the Jackson Laboratory. In this study, mice were anesthetized using isoflurane in oxygen administered via inhalation within a closed chamber, ensuring they were fully unconscious prior to any procedures. Euthanasia was performed using carbon dioxide (CO₂) inhalation. All mice were housed at Yale School of Medicine animal facilities, and all experiments were performed in accordance with Yale Institutional Animal Care and Use Committee-approved procedures.

C57BL/6J mice (n = 5) were injected with PBS or 10⁸ PFU of VLV-S-FL or VLV-S-RBD or VLV-Ctrl on Day 0 and/or Day 14 and bled on day 28 for isolation of sera to detect spike-specific antibody responses. VLV-S-FL was given through different routes of administration, including intramuscular injection (i.m.), intraperitoneal injection (i.p.) and intranasal administration (i.n.). Spike-specific antibodies were detected and compared on day 28 after the injection. To compare the efficacy of this vaccine and the BNT162b2, C57BL/6J mice (n = 5) were i.m. immunized with BNT162b2 at 10 ng (therapeutically equivalent dose), 100 ng, and 1 μg on days 0 and 14, and bled on day 28 for isolating serum to detect spike-specific antibody responses.

Indirect Enzyme-Linked Immunosorbent Assay (ELISA)

96-well plates were coated with 0.2 μg/well of purified spike protein (ACRO Biosystems, DE) in PBS at 4 °C overnight and blocked with 5% non-fat dry milk at 37 °C for 1 h. The serum was diluted at different ratios in PBS containing 0.05% Tween 20 (PBS-T), and then 100 μl of each was added to each well, followed by incubation at 37 °C for 1 h. After washing with PBST, goat anti-mouse IgG/HRP was added and incubated at 37 °C for 1 h. After being washed, the substrate was added to the plate and incubated in the dark for 15 min. Then, the substrate reaction was stopped by adding H2SO4, and the absorbance was read at 492 nm. Alternatively, the antibody levels were also expressed as endpoint titers. The serum obtained from mice was serially diluted, 10-fold, from 1:100 to 1:100000 and then tested by a plate reader to obtain the corresponding values. The IgA levels in BALF were detected as well.

Flow cytometry

Peripheral blood mononuclear cells were collected for analysis due to the minimally invasive nature of blood sampling, which enables consistent and standardized collection across experimental groups. This approach allowed for serial immune monitoring in the same animals without requiring euthanasia. Blood samples were collected from the immunized mice on days 3 and day 7 post-immunization. After red blood cell lysis and counting, the cells were stained with PE-Cy7-conjugated anti-CD4 antibody and APC-conjugated anti-CD8a antibody for 30 min on ice in the dark, followed by two washes with 1X PBS buffer. All stained cells were analyzed by LSRII flow cytometer. Live cells were carefully gated by forward and side scattering.

Bronchoalveolar lavage fluid (BALF), rather than lung tissue leukocytes, was collected to minimize potential confounding factors associated with tissue processing, such as regional variation in immune activation or infiltration. BALF samples were collected on day 28 post-immunization. The BALF was centrifuged at 600 × g for 5 min to obtain the cells, which were stained with PE-Cy7-conjugated anti-CD4 and APC-conjugated anti-CD8a antibodies for flow cytometry analysis. We thank the Yale Flow Cytometry Core Facility for their assistance. The core is supported in part by an NCI Cancer Center Support Grant (NIH P30 CA016359).

RNA isolation and qRT-PCR analysis

Splenocytes isolated from the naïve C57BL/6J mice were plated in 24-well plates at a concentration of 5 × 10⁶ cells/ml in 1 ml DMEM supplemented with 10% FBS. The cells were co-cultured with VLV-S-FL for 3, 6, 9, 24 or 48 h at 37 °C in a 5% CO₂ incubator. Total RNA was isolated from the splenocytes at different time points and lung tissues by using RNeasy Mini Kit (QIAGEN, 74106). Lung tissues were collected and fixed in 10% NEB buffer. RNA was extracted from FFPE lung samples by using the Allprep DNA/RNA FFPE Kit (QIAGEN, 80234). Isolated RNA was reverse transcribed into cDNA using a cDNA synthesis kit (Thermo Scientific, IL). Quantitative real-time polymerase chain reaction (qRT-PCR) was then performed using two-step SYBR green assays. Target genes were amplified using the following primers. Mouse Tlr7, forward: 5′-ACGCTTTCTTTGCAACTGTG-3′ and reverse: 5′-TTTGTGTGCTCCTGGACCTA-3′; Irf3, forward: 5′-CTACACCCCGGGGAAGGATA-3′ and reverse: 5′-GAGAGGCACCCAGATGTACG-3′; Irf7, forward: 5′-TCCCAGACTGCCTGTGTAGA-3′ and reverse: 5′-ATCCAGATCCCTACGACCGA-3′; Ifna, forward: 5′-TGTCTGATGCAGCAGGTGG-3′ and reverse: 5′-TGTCTGATGCAGCAGGTGG-3′; Ifnb, forward: 5′-TTCCTGCTGTGCTTCTCCAC-3′ and reverse: 5′-AAGGTACCTTTGCACCCTCC-3′; Actb, forward: 5′-GATCAAGATCATTGCTCCTCCTG-3′ and reverse: 5′-AGGGTGTAAAACGCAGCTCA-3′; Il6, forward: 5′-TACCACTTCACAAGTCGGAGGC-3′ and reverse: 5′-CTGCAAGTGCATCATCGTTGTTC-3′; Ccl4, forward: 5′-ACCCTCCCACTTCCTGCTGTTT-3′ and reverse: 5′-CTGTCTGCCTCTTTTGGTCAGG -3′; Ccl5, forward: 5′-CCTGCTGCTTTGCCTACCTCTC-3′ and reverse: 5′-ACACACTTGGCGGTTCCTTCGA-3′; Tnfa, forward: 5′-GGTGCCTATGTCTCAGCCTCTT-3′ and reverse: 5′-GCCATAGAACTGATGAGAGGGAG-3′. The data were acquired using the Step One real-time PCR system (Applied Biosystem, CA). The cycling procedure was as follows: one cycle at 95 °C for 30 s, followed by 40 cycles at 95 °C for 5 s and 64 °C for 31 s. Each assay plate included negative control with no template. The mRNA levels of the gene of interest were normalized to the mRNA levels of Actb and analyzed using the 2^−ΔΔCt method.

Neutralizing assay

1 × 10⁴ Huh7.5 cells were seeded in a 96-well plate. The serum obtained from the mice was serially diluted and mixed with pseudotyped VSVΔG-GFP containing codon-optimized S (Wuhan strain) or omicron variant (B1.1.529) spike protein and anti-VSV G (I1 and I1-4) for 1 h at 37 °C⁶⁰. After that, the mixture was transferred to the corresponding wells and incubated with Huh7.5 cells for 24–48 h. GFP signals were counted using a fluorescent microscope (Keyence, IL).

Generation of SARS-CoV2 stocks

SARS-CoV-2 isolate hCOV-19/USA-WA1/2020 (NR-52281) was obtained from BEI Resources. For generating SARS-CoV2 stocks, Vero E6 cell overexpressing hACE2 and transmembrane serine protease 2 (TMPRSS2) [kindly provided by B. Graham at the National Institutes of Health Vaccine Research Center (NIH-VRC)] were cultures in DMEM supplemented with 1% sodium pyruvate and 5% FBS at 37 °C and 5% CO2 and infected at a multiplicity of infection of 0.01 for 2–3 days to propagate a viral solution. After incubation, the culture supernatant was clarified by centrifugation (500 × g × 5 min), filtered through a 0.45 μm filter, and stored at −80 °C. Viral titers were measured with standard plaque assay by using Vero E6 cells.

SARS-CoV2 infection

Vaccine-immunized mice were anesthetized by using 30% v/v isoflurane diluted in propylene glycol. A pipette delivered 50 μl of SARS-CoV2 virus (Isolate USA-WA1/2020) i.n. at 10⁴ PFU per mouse. Weight and survival were monitored for 2 weeks after infection. Experiments involving SARS-CoV2 infection were performed in a biosafety level 3 facility with approval from the Yale Institutional Animal Care and Use Committee and Yale Environmental Health and Safety.

ELISPOT

T cell response were determined using an IFN-γ enzyme-linked immunospot (ELISPOT) set (BD Biosciences, NJ) according to the manufacturer’s protocol. Briefly, 96-well plates were used for coating overnight with the purified anti-mouse IFN- γ antibody (1:200). The plate was rinsed and then blocked for 2 h using DMEM supplemented with 10% FBS. The cells collected from the BALF in the PBS, VLV-Ctrl and VLV-S-FL immunized mice. The cells were collected and mixed them in pool (5 mice per group). The cells were suspended in DMEM and seeded at 1 × 10⁵ cells/well. Purified spike was used to stimulate cells 48 h at 37 °C. Cells were washed from plates using PBST, and biotinylated anti-mouse IFN-γ antibody (1:250) was added at RT. After washing, HRP (1:100) was added to wells and incubated for 1 h at 25 °C. Following the final washes, AEC substrates was added to the wells and allowed to develop at 25 °C for 20–40 min. Stop substrate reaction by washing wells with DI water. The plates were allowed to air dry before spot-forming cells were counting.

Data analysis

All calculations and statistical analyses were performed using GraphPad Prism 9 (GraphPad Software, CA). Comparisons between groups were conducted using analysis of unpaired t-tests. Data are shown as mean ± SD and differences were considered statistically significant at P < 0.05.

Author contributions

L.Y. and C.L. conceived and designed the experiments. T.O.Y., A.I., and J.K.R. provided methodology assistance. P.L. performed animal challenge studies. L.Y., Y.X., and K.P. analyzed the data and wrote the manuscript. C.L. and J.K.R. supervised the study. All authors contributed to interpretation of the results and development of the manuscript.

Data availability

Data is provided within the manuscript or supplementary files. The data used and/or analyzed during the current study are available from the corresponding author on request.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41541-025-01260-4.