Introduction
Plague, a historic deadly disease caused by Yersinia pestis, remains endemic in many regions and is responsible for several thousand annual human cases worldwide1,2. Moreover, the emergence of multiple drug-resistant Y. pestis strains in plague outbreak regions poses severe concerns about the effectiveness of antibiotic treatment3–5. Thus, plague is classified by the World Health Organization (WHO) as a re-emerging disease despite its low incidence rates6,7.
In addition, Y. pestis is categorized as a Tier 1 Select Agent due to its low infectious dose and high case-fatality rate, its ability through aerosol transmission, and its potential use as a biological weapon8,9. There are two main clinical forms of plague: bubonic and pneumonic. The pneumonic form is the most fatal disease and results in nearly 100% mortality of the subjects if the medical treatment is delayed beyond 24 h2,9.
Subunit plague vaccines demonstrated highly effective protection against bubonic plague but not pneumonic plague10–12. Thus, the effectiveness of a vaccine against pneumonic plague is a crucial measure.
Previously, we have developed highly immunogenic Yersinia outer membrane vesicles (OMVs), which are exclusively and intrinsically composed of detoxified monophosphoryl-lipid A (MPLA) and refined a leading Yersinia OMV-based plague vaccine candidate, OMV46-LcrV13,14. The OMV46-LcrV was fully coated with the MPLA, loaded with high amounts of LcrV antigen, and showed reduced TLR4-stimulating activity14. Intramuscular (IM) OMV46-LcrV vaccination elicited robust adaptive responses in lung mucosal and systemic compartments, orchestrated innate immunity in the lung, and provided excellent short- and long-term protection against pneumonic plague.
However, direct OMV46-LcrV vaccination caused certain levels of reactogenicity in mice14. To address the safety concerns, we have adopted the aluminum (Alum)-absorption strategy in the Yersinia OMV vaccine, which had been used in licensed Neisseria OMV vaccines (MenBVAC, MeNZB, and Bexsero)15,16 to minimize reactogenicity and retain OMV46-LcrV immunogenicity. However, the parenteral [IM or subcutaneous (SC)] immunization is generally not efficient in inducing robust mucosal immunity17,18.
A recent study showed that IM priming with BNT162b2 (Pfizer/BioNTech) and intranasal (IN) boosting with unadjuvanted intranasal SARS-CoV-2 spike protein (termed “prime and spike”) elicited potent antibody (Ab) and T-cell responses in the lung and enhanced protection against a lethal SARS-CoV-2 challenge in comparison to IM priming/boosting with BNT162b219. The IN booster with unadjuvanted antigen(s) would minimize unwanted side effects in the lungs caused by an adjuvant20,21. Given those, we sought to use the “prime-spike” strategy to minimize potential vaccine safety concerns related to the lung spike but also to enhance lung immunity against pneumonic plague.
In this study, we revealed that the Alum-absorbed OMV46-LcrV (designated as OMV46-LcrV/Alum) demonstrated great safety and tolerance in mice after IM injection and retained high immunogenicity. Furthermore, an IN booster with the purified recombinant Y. pestis LcrV (LcrV) protein following the IM prime-boost OMV46-LcrV/Alum vaccination displayed no safety issues and augmented lung immunity against pneumonic plague compared to IM OMV46-LcrV/Alum vaccination alone.
Results
Alum absorption improves the safety of OMV46-LcrV but slightly compromises its protection
To further reduce reactogenicity and enhance the safety of the OMV46-LcrV vaccine14, we formulated 20 μg of OMV46-LcrV with 2% of Alhydrogel [1:1 (v/v), corresponding to 50 µg aluminum hydroxide] and designated the formulation as OMV46-LcrV/Alum. In contrast to OMV46-LcrV alone, IM prime-boost OMV46-LcrV/Alum vaccination completely abrogated animal weight loss and exhibited similar weight-gain profiles to IM 5 µg LcrV/Alum, IM 5 µg F1-V/Alum, and PBS control (Figs. 1A, B and S1A).
Also, IM OMV46-LcrV/Alum vaccination elicited significantly lower serum IL-6 and IL-1β levels than IM OMV46-LcrV vaccination at 1-, 2-, and 3-days post-vaccination (DPV), but higher than those induced by IM LcrV/Alum, IM F1-V/Alum, and PBS injection (Figs. 1C and S1B). The C-reactive protein (CRP) is considered a biomarker of vaccine safety22.
Analysis showed that the IM OMV46-LcrV vaccination induced significantly higher levels of serum CRP than other immunization groups at 1 and 2 DPV, indicating higher reactogenicity. Serum CRP levels were comparable among mice immunized with IM OMV46-LcrV/Alum, LcrV/Alum, and F1-V/Alum, but significantly elevated compared to PBS (sham) controls at 1 DPV, returning to baseline levels at 2 DPV. Serum CRP levels (2.13 µg/ml) in IM OMV46-LcrV/Alum-immunized mice were within a range of normal mice ( < 2.5 µg/ml)23.
At 3 DPV, the CRP levels in all groups returned to the baseline (Figs. 1D and S1C). In addition, the complete blood count (CBC) analysis showed no significant alterations in the hematological parameters, such as white blood cells, red blood cells, hemoglobin, hematocrit test, mean corpuscular volume, and mean corpuscular hemoglobin in the blood from sham, LcrV/alum, F1-V/Alum-, OMV46-LcrV/Alum-, and OMV46-LcrV-immunized mice (SI Table 4).
Among the leukocyte population, the percentages of neutrophil and monocyte were similar in sham-, LcrV/alum, F1-V/Alum-, and OMV46-LcrV/Alum-immunized mice but substantially lower than those in OMV46-LcrV-immunized mice at 1 and 2 DPV. The percentages of basophils were comparable among IM LcrV/Alum, F1-V/Alum-, OMV46-LcrV/Alum-, and OMV46-LcrV-immunized mice but significantly higher than those in sham mice. On 3 DPV, the cell percentages in all groups of mice were comparable except neutrophils, which were higher in OMV46-LcrV immunized mice (Figs.
1E–G and S1D–F). No significant lung histological changes were observed after immunization (Fig. S1G).Fig.
1Evaluation of Alum-absorbed OMV46-LcrV safety and protective efficacy against pneumonic plague.Swiss Webster mice (n = 10, equal male and female, 7 weeks old) were IM immunized with 20 µg of OMV46-LcrV/Alum, 20 µg of OMV46-LcrV, or 5 µg of LcrV/Alum in 50 µL PBS, and PBS alone as a negative control, then boosted on day 21. A Immunization schema. B The percentages (%) of mouse weight changes after IM vaccination.
Parallelly, blood samples were collected from mice on 0, 1, 2, and 3 DPV for analysis. C The levels of IL-6 and IL-1β, and D the levels of CRP in the sera were analyzed using commercial ELISA Kits, respectively. E The percentages of neutrophil, F monocyte, and G basophil in blood were analyzed using a CBC analyzer.
H On 45 DPV, mice were intranasally challenged with 4 × 104 CFU (400 LD50), 4 × 105 CFU (4000 LD50), and 2 × 106 CFU (20,000 LD50) of Y. pestis KIM6 + (pCD1Ap), respectively. Animal survival was recorded for 15 days after the challenge. Each symbol was obtained from an individual mouse, and data were represented as the mean ± SD.
The statistical significance of the difference among groups was analyzed by two-way ANOVA with the Turkey post hoc test. Statistical significance was analyzed by the log-rank (Mantel–Cox) test for survival analysis. ns no significance; *, p < 0.05; **, p < 0.01; ****, p < 0.0001. Source data are provided as a Source Data file.
Challenge studies showed that IM OMV46-LcrV/Alum vaccination provided similar protection against 400 [4 × 104 colony-forming units (CFU)] and 4000 LD₅₀ intranasal Y. pestis challenges, with slightly reduced protection (70%) at 20,000 LD₅₀, compared to IM OMV46-LcrV alone (Fig. 1H). In contrast, mice immunized with PBS, IM LcrV/Alum, or IM F1-V/Alum all succumbed to 4 days post-infection (DPI) (Fig.
1H). Thus, Alum absorption improved the safety of OMV46-LcrV but slight offset protective efficacy.
IN adjuvant-free LcrV booster following IM OMV46-LcrV/Alum immunization augments protection against pneumonic plague
Given that Alum-absorbed OMV46-LcrV resulted in attenuated protective efficacy, we attempted an additional boost with adjuvant-free LcrV, following a “prime-spike” regimen described previously19, to enhance protection without compromising vaccine safety. Thus, mice received an IM prime-boost vaccination with 20 μg OMV46-LcrV/Alum, followed by an IN booster with 3 μg adjuvant-free LcrV (designated as IM OMV46-LcrV/Alum + IN LcrV), as outlined in Fig. 2A.
Control groups received either 5 μg IM LcrV/Alum + 3 μg IN LcrV or 5 μg IM F1-V/Alum + 3 μg IN F1-V. The IN LcrV or F1-V booster did not affect weight gain (Fig. S2A), serum CRP levels (Fig.
S2B), hematological parameters (Table S5), or lung histology (Fig. S2C) compared to IM OMV46-LcrV/Alum alone. However, blood neutrophil, monocyte, and basophil levels (Fig.
S2D–F) transiently increased on days 1 and 2 post-IN booster and returned to baseline by day 3.Fig. 2Protection against pneumonic plague in mice by IM OMV46-LcrV/Alum and intranasal adjuvant-free LcrV boosting.Groups of Swiss Webster mice (n = 10, equal male and female, 7 weeks old) were IM primed with 20 µg of OMV46-LcrV/Alum, 5 µg of LcrV/Alum in 50 µL PBS, and boosted on day 21 after the initial vaccination. Then, the IM LcrV/Alum immunization group and another group of IM OMV46-LcrV/Alum-immunized mice were subjected to intranasal (IN) boosting with 3 µg of adjuvant-free LcrV in 20 µL PBS on day 37 post initial immunization, respectively.
A group of PBS-injection mice (Sham) was used as a negative control. A Immunization schema. B On 45 DPV, mice were intranasally challenged with 4 × 104 CFU (400 LD50), 4 × 105 CFU (4000 LD50), and 2 × 106 CFU (20,000 LD50) of Y. pestis KIM6 + (pCD1Ap), respectively.
C On 2 DPI and D 6 DPI, the bacterial burden was determined in the lung, liver, and spleen of mice challenged with 20,000 LD50 of Y. pestis KIM6 + (pCD1Ap). E Lung representative histopathological analysis from each group of mice (n = 3) on 2 DPI. The lungs were microscopically examined and imaged using Nanozoomer 2.0 RS Hamamatsu slide scanner (scale bar, 100 nm).
Lung damage was assessed using a histology scoring system as described in the “Methods”. F Groups of Swiss Webster mice (n = 10, equal male and female, 7 weeks old) were immunized as shown in figure (A). On 45 DPV, mice were intranasally challenged with 4 × 104 CFU and 4 × 105 CFU of ∆caf1 Y. pestis KIM6 + (pCD1Ap), respectively.
G Nine-month-old C57BL/c and Hfe−/− mice (n = 8, mixed males and females) were immunized with IM OMV46-LcrV/Alum and IN LcrV shown above. A group of PBS-injection C57BL/c was used as a negative control. On 45 DPV, mice were intranasally challenged with 200 LD50 of Y. pestis KIM6 + (pCD1Ap).
Animal survival was recorded for 15 days after the challenge. Each symbol was obtained from an individual mouse, and data were represented as the mean ± SD. The statistical significance of difference among groups was analyzed by two-way ANOVA with Turkey post hoc test.
Statistical significance was analyzed by the log-rank (Mantel–Cox) test for survival analysis. ns no significance; *, p < 0.05; **, p < 0.01; ****, p < 0.0001. Source data are provided as a Source Data file.
The challenge results showed that the IM OMV46-LcrV/Alum + IN LcrV immunization conferred complete protection against 400, 4000, and 20,000 LD50 of IN Y. pestis challenge, respectively. However, PBS, IM LcrV/Alum + IN LcrV, or IM F1-V/Alum + IN F1-V-immunized mice completely died within 5 DPI (Figs. 2B and S2G).
Two days post-infection (DPI) with 20,000 LD50 of Y. pestis, both sham and IM LcrV/Alum + IN LcrV-immunized mice exhibited an enormous bacterial load (mean 10.0 log10 CFU/g tissue) in the lungs, spleen (mean 6.0 log10 CFU/g tissue), and liver (7.0 log10 CFU/g tissue) (Fig. 2C). The OMV46-LcrV/alum-immunized mice showed a moderate bacterial burden in the lung (mean 6.6 log10 CFU/g tissue), liver (mean 3.0 log10 CFU/g), and spleen (mean 3.0 log10 CFU/g), while the IM OMV46-LcrV/Alum + IN LcrV immunized mice had a very low bacterial load in the lung (mean 1.4 log10 CFU/g tissue), liver (mean 1.0 log10 CFU/g tissue), and spleen (mean 1.0 log10 CFU/g tissue) (Fig.
2C). Mice immunized with IM F1-V/Alum + IN F1-V showed similar bacterial burden profiles across organs those immunized with IM LcrV/Alum followed by IN LcrV (Fig. S2H).
At 6 DPI, OMV46-LcrV/Alum-immunized mice exhibited an approximately 3–4 log₁₀-fold reduction in bacterial load in the lungs (mean 2.13 log₁₀ CFU/g), liver (mean 1.16 log₁₀ CFU/g), and spleen (mean 0.7 log₁₀ CFU/g) compared to 2 DPI. In contrast, IM OMV46-LcrV/Alum + IN LcrV immunized mice showed minimal or no detectable bacterial growth in these organs (Fig. 2D).
At 15 DPI, no detectable bacteria were found in the lungs, liver, or spleen in either immunized group (Fig. S2I), indicating that the IM OMV46-LcrV/Alum + IN LcrV vaccination induced sterilizing immunity. Histopathological analysis of lung tissue at 2 DPI revealed pronounced inflammation, edema, disruption of alveolar macrophages, and neutrophil infiltration in IM OMV46-LcrV/Alum-immunized mice.
In contrast, IM OMV46-LcrV/Alum + IN LcrV-immunized mice showed only mild inflammation and minimal immune cell infiltration. Severe immune cell infiltration, hemorrhage, edema, and cellular disruption were observed in both sham and IM LcrV/Alum + IN LcrV-immunized mice (Fig. 2E).
Naturally isolated F1-negative strains have been shown to cause human disease similar to wild-type (WT) Y. pestis24. Thus, we sought to evaluate vaccine efficacy against pulmonary infection with a Δcaf1 Y. pestis KIM6 + (pCD1Ap) mutant, a laboratory-generated mimic of F1-negative strains. Results showed that IM OMV46-LcrV/Alum + IN LcrV immunization provided 100% and 90% protection against pulmonary challenges with 4 × 104 CFU (400 LD50) and 4 × 105 CFU (4000 LD50) of Δcaf1 Y. pestis, respectively.
In contrast, sham, IM LcrV/Alum + IN LcrV, and IM F1-V/Alum + IN F1-V immunized mice completely died within 5 DPI (Figs. 2F and S2J). In addition, IM 10 μg of OMV46-LcrV/Alum immunization then IN LcrV boosting (Fig.
S2K) afforded similar levels of protection against pulmonary challenges with either WT Y. pestis or ∆caf1 Y. pestis to that of 20 μg of IM OMV46-LcrV/Alum then IN LcrV boosting (Figs. 2B and F and S2L, M). It has been reported that individuals with hereditary hemochromatosis, manifesting iron overload in tissues, are susceptible to Yersinia infection25,26.
The aged Hfe−/− mice exhibited an iron-overload phenotype resembling hereditary hemochromatosis in humans27. Therefore, we sought to evaluate the vaccine efficacy in aged Hfe−/− mice using the “prime-spike” immunization. Nine-month-old Hfe−/− and C57/BL6 mice were IM immunized with 20 μg of OMV46-LcrV/Alum and then boosted with IN LcrV as shown in Fig.
2A. Immunization did not affect animal weight gain (Fig. S2N).
Serum anti-LcrV IgG titters showed no differences between immunized Hfe−/− and C57/BL6 mice on 21 and 39 DPV (Fig. S2O). All immunized aged mice survived pulmonary challenges with 200 LD50 of Y. pestis (Fig.
2G). Our results demonstrated that the IN LcrV booster following IM OMV46-LcrV/Alum vaccination effectively addressed vaccine safety concerns and provided excellent protection against pneumonic plague in both immunocompetent and immunocompromised mice. Since no difference in protection was observed between the LcrV and F1-V groups, subsequent immune analyses focused on the LcrV-immunized group.
IN adjuvant-free LcrV boosting elicits enhanced lung humoral immunity
Given that the “prime-spike” immunization substantially augmented protection against pneumonic plague, we sought to determine humoral responses in vaccinated mice based on the experimental schema illustrated in Fig. S3A. Ab analysis showed that high and comparable levels of serum anti-LcrV IgG titers were elicited in IM LcrV/Alum-, OMV46-LcrV-, and OMV46-LcrV/Alum-immunized mice at 21 DPV and boosted at 36 DPV, respectively (Fig.
S3B). Serum Ab isotyping demonstrated that OMV46-LcrV vaccination induced a balanced Th1/Th2 response. In contrast, IM OMV46-LcrV/Alum vaccination significantly switched the response to Th2 bias, like IM LcrV/alum (Fig.
S3C). Also, we compared anti-LcrV titers in the sera and bronchoalveolar lavage fluid (BALF) collected from immunized mice with or without IN LcrV booster on 42 DPV. Results showed that IN LcrV booster failed to elevate serum anti-LcrV IgG titers (Fig.
3A). In our previous study, only IM OMV46-LcrV vaccination induced moderate anti-LcrV IgG titers in the BALF, and no anti-LcrV secretory IgA (sIgA) was detected in either IM OMV46-LcrV- or LcrV-immunized mice14. However, the IN LcrV boosting significantly elevated anti-LcrV IgG (Fig.
3B) and sIgA titers (Fig. 3C) in the BALF compared to no IN LcrV booster. In addition, immune sera from mice immunized with OMV46-LcrV/Alum + IN LcrV, OMV46-LcrV/Alum, LcrV/Alum + IN LcrV, or PBS were adoptively transferred intraperitoneally into naïve Swiss Webster mice (Fig.
S3D, upper). Mice receiving 100 µL sera from OMV46-LcrV/Alum + IN LcrV and OMV46-LcrV/Alum showed 60% and 50% survival, respectively, after 2 × 103 CFU pulmonary Y. pestis challenge. while those given sera from LcrV/Alum + IN LcrV completely died within 3 days of infection (Fig. S3D, down).Fig.
3Antibody responses and lung BRM cell responses.Swiss Webster mice (n = 10 females, 7 weeks old) were intramuscularly primed and boosted with 20 µg of OMV46-LcrV/Alum, 5 µg of LcrV/Alum in 50 µL PBS on day 0 and 21, respectively, then IN boosting with and without adjuvant-free LcrV on day 37. Sham (PBS) mice were used as negative controls. On 42 DPV, the lung single cells were isolated from euthanized mice for analysis of lung BRM cells as described in the “Methods”.
A Serum anti-LcrV IgG titers at 36 and 42 DPV. B BALF anti-LcrV IgG titers at 42 DPV. C BALF anti-LcrV IgA titers at 42 DPV.
D The absolute counts of lung BRM (IV−CD19b+) cells in immunized mice. E Lung ASC BRM (IV−CD19b+CD138+) cells. Representative flow plots (Left) and absolute counts (Right) of lung F IgG+ and G IgA+ ASC BRM cells H Lung matured BRM (IV−CD19b+CD38+GL7−) cells.
Representative flow plots (Left) and absolute count of lung (Right) I IgG+ and J IgA+ matured BRM cells. The unit of cell numbers in mice was per lung. Each symbol was obtained from an individual mouse, and data were represented as the mean ± SD.
The statistical significance of difference among groups was analyzed by two-way ANOVA with Turkey post hoc test: ns, no significance; *, p < 0.05; **, p < 0.01; *** p < 0.001; ****, p < 0.0001. Source data are provided as a Source Data file.
Resident memory B (BRM) cells in the lungs assist in the rapid recall response of Ab-secreting plasma cells upon secondary infection28 or IN administration with SARS-CoV-2 spike protein19. Given those, we sought to determine BRM cells via intravenous (IV) labeling with FITC-conjugated mouse anti-CD45 antibodies to differentiate circulating immune cells within lung tissue29. Then, single cells from the lungs of immunized mice at 42 DPV were isolated, stained with specific markers, and analyzed by flow cytometry using a representative gating strategy (Fig.
S3E). Compared to IM OMV46-LcrV/Alum vaccination alone, the IN LcrV booster following IM OMV46-LcrV/Alum vaccination resulted in a 1.4-fold increase in IV−CD19b+ (depicted as IV−CD19b+) B cells (BRM) within lung tissue (Fig. 3D).
The lung BRM-cell counts in the IM LcrV/Alum + IN LcrV immunized mice were higher than those in sham mice but lower than those in IM OMV46-LcrV/Alum-immunized mice (Fig. 3D). Further, we examined the polyclonal tissue response, representing the B-cell response within the lung28.
We found that IN rLcrV boosting after IM OMV46-LcrV/Alum vaccination significantly increased the number of lung Ab-secreting cells (ASCs) (CD138+ BRM) and both IgG- and IgA-producing ASCs compared to the other three groups (Fig. 3E–G). Similar patterns were also shown in the number of mature (GL7−CD38+) BRM cells and their IgG- and IgA-producing cells.
(Fig. 3H–J). Thus, results implied that IN LcrV booster following IM OMV46-LcrV/Alum vaccination remarkably enhanced responses of lung BRM cells.
IN adjuvant-free LcrV boosting augments lung cellular immune responses
Studies also have shown that lung tissue-resident memory T (TRM) cells are pivotal for protection against respiratory infections30,31. To probe the induction of lung TRM cells, we parallelly stained single lung cells isolated above with canonical TRM markers, such as CD62L−, CD69+, and/or CD103+30,32. Cells were collected via flow cytometry and analyzed using a representative gating strategy (Fig.
S4A). Quantitative plots showed that the IN LcrV booster following IM OMV46-LcrV/Alum vaccination induced the highest lung IV−CD3+CD4+ T-cell counts, while IM OMV46-LcrV/Alum vaccination alone elicited moderately higher cell counts compared to sham and IM LcrV/Alum + IN LcrV vaccination (Fig. 4A).
Lung IV−CD3+CD8+ T-cell counts were similar between groups of IM OMV46-LcrV/Alum and IM OMV46-LcrV/Alum + IN LcrV vaccination but were higher than those in sham and IM LcrV/Alum + IN LcrV vaccination groups. The IN LcrV booster following IM LcrV/Alum vaccination failed to increase both T-cell counts compared to PBS (Fig. 4A).
Further, flow and quantitative plots showed that patterns of lung IV−CD3+CD4+CD44+CD62L−CD69+ T cells (designated to CD69+CD4+ TRM cells) (Fig. 4B) and IV−CD3+CD8+CD44+CD62L−CD69+ T cells (designated to CD69+CD8+ TRM cells) were similar to those of IV−CD3+CD4+ and IV−CD3+CD8+ T cells, respectively. In addition, the patterns of CD103+CD4+ TRM cells (Fig.
4C) were the same as those of lung CD69+CD4+ TRM cells in those immunized mice, whereas lung CD103+CD8+ TRM cell counts were no different across all groups.Fig. 4Lung TRM cells and their responses in immunized mice.Swiss Webster mice (n = 10 females, 7 weeks old) were intramuscularly primed and boosted with 20 µg of OMV46-LcrV/Alum, 5 µg of LcrV/Alum in 50 µL PBS on day 0 and 21, respectively, then IN boosting with and without adjuvant-free LcrV on day 37. Sham (PBS) mice were used as negative controls.
On 42 DPV, the lung single cells were isolated from euthanized mice for analysis of lung TRM cells and in vitro stimulated with 10 µg/mL LcrV for 48 h to determine cytokine-producing lung antigen-specific TRM cells as described in the “Methods”. A The absolute counts of IV−CD3+CD4+ and CD8+ T cells. B Representative flow plots (Left) of lung CD69+ TRM (IV−CD3+ CD4+CD44+CD62L−CD69+) cells and the absolute counts (Right) of those cells.
C Representative flow plots (Left) of lung CD103+ TRM (IV−CD3+ CD4+CD44+CD62L−CD103+) cells and the absolute counts (Right) of those cells. D Representative flow plots (Left) of IFN-γ-producing lung CD69+ TRM cells and the absolute counts (Right) of those cells. E Representative flow plots (Left) of IL-17A-producing lung CD69+ TRM cells and the absolute counts (Right) of those cells.
F Representative flow plots (Left) of IL-4-producing lung CD69+ TRM cells and the absolute counts (Right) of those cells. The unit of cell numbers in mice was per lung. Each symbol was obtained from an individual mouse, and data were represented as the mean ± SD.
The statistical significance of difference among groups was analyzed by two-way ANOVA with Turkey post hoc test: ns no significance; *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001. Source data are provided as a Source Data file.
In vitro LcrV stimulation showed the number of lung CD69+CD4+ TRM cells producing interferon-gamma (IFN-γ) (Fig. 4D), interleukin 17 A (IL-17A) (Fig. 4E), or IL-4 (Fig.
4F) was the highest in IM OMV46-LcrV/Alum + IN LcrV immunized mice. The number of lung CD69+CD4+ TRM cells producing corresponding cytokines was moderate in IM OMV46-LcrV/Alum-immunized mice but substantially higher than those in IM LcrV/Alum + IN LcrV-immunized and sham mice (Fig. 4D–F).
The number of lung CD69+CD4+ TRM cells producing tumor necrosis factor (TNF-α) was comparable between IM OMV46-LcrV/Alum and IM OMV46-LcrV/Alum + IN LcrV immunized mice but was higher than that in IM LcrV/Alum + IN LcrV-immunized and sham mice (Fig. S4B). There were no differences in the number of lung CD69+CD8+ TRM cells producing IFN-γ, IL-4, IL-17A, or TNF-α across all immunized groups (Fig.
4D–F). In addition, only IN LcrV boosting after OMV46-LcrV/Alum vaccination significantly heightened Ki67 expression in both IV−CD3+CD4+ and IV−CD3+CD8+ T cells (Fig. S4C).
Alveolar macrophages (AMs) and lung neutrophils provide a first-line defense against respiratory pathogens33. Therefore, we examined AMs and neutrophils using a gating strategy (Fig. S4D) in immunized mice on 42 DPV.
Results showed that IN LcrV boosting after IM OMV46-LcrV/Alum vaccination substantially increased the number of AMs (Ly6G−CD64+Siglec F+) (Fig. S4E, F), MHCII+ AMs, and MerTK+ AMs (Fig. S4G) in the lung and BALF compared to other groups.
However, no changes in neutrophil counts were observed in the lung or BALF of mice after IN LcrV boosting compared to other groups (Fig. S4H). Thus, results indicated that the IN LcrV booster following IM OMV46-LcrV/Alum vaccination enhanced lung cellular immunity and provided excellent protection against pneumonic plague.
Given that IM OMV46-LcrV/Alum + IN LcrV immunization elevated lung CD4⁺ TRM cells producing IL-17A and IFN-γ and provided complete protection against a huge dose of pulmonary Y. pestis challenge (Figs. 2B and 4D, E), we further evaluated the roles of these cytokines. IM OMV46-LcrV/Alum + IN LcrV immunized mice were intranasally administered anti-mouse IL-17A, anti-mouse IFN-γ, or both Abs on days 39, 40, and 41 following the initial immunization, then challenged with 2000 LD₅₀ of Y. pestis on day 42 (Fig.
S5A). Neutralizing IL-17A or IFN-γ reduced survival to 50% and 40%, respectively; combined neutralization further decreased survival to 20%, while all isotype control mice survived (Fig. S5B).
Both cytokines are known to enhance macrophage-mediated immunity against pneumonic plague34. Thus, we examined the impact of cytokine neutralization on lung TRM cells, AMs, and neutrophils at 2 DPI. Compared to isotype controls, neutralization of IFN-γ, IL-17A, or both markedly reduced the number of lung-resident CD4⁺ and CD8⁺ TRM cells (Fig.
S5C, D) and AMs in the lungs and BALF (Fig. S5E). Consistent with our previous study30, blockade of both cytokines led to a maximal increase in neutrophil numbers in the lungs and BALF during Y. pestis infection (Fig.
S5F). Our results indicate that both IL-17A and IFN-γ are critical for protection against pneumonic plague, with their combined neutralization showing an additive effect. Both cytokines support AM proliferation, as well as humoral and cellular memory responses, aiding in Y. pestis clearance and limiting excessive neutrophil infiltration.
The long-interval IN adjuvant-free LcrV boosting enhances mucosal immunity
Next, we asked whether IN LcrV boosting at a long interval would impact lung immune responses and protection against pneumonic plague. Mice were subjected to IM prime-boost OMV46-LcrV/Alum vaccination on days 0 and 21, and IN LcrV boosting on day 171 (Fig. 5A).
At 190 DPV, assessment of humoral immune responses revealed that long-interval IN LcrV (LI LcrV) boosting following IM OMV46-LcrV/Alum vaccination significantly enhanced serum anti-LcrV IgG titers (Fig. 5B), as well as BALF anti-LcrV IgG and IgA titers (Fig. 5C, D) compared to OMV46-LcrV/Alum immunization alone.
Similarly, increased numbers of lung-resident IgG- and IgA-producing ASCs and matured BRM cells were observed (Fig. 5E, F). However, these humoral immune responses were absent in sham-treated mice (Fig.
5B–F).Fig. 5Immune responses and protection of the long interval IN LcrV boosting.Swiss Webster mice (n = 10 females, 7 weeks old) were intramuscularly primed and boosted with 20 µg of OMV46-LcrV/Alum, 5 µg of LcrV/Alum in 50 µL PBS on day 0 and 21, respectively, then IN boosting with and without adjuvant-free LcrV on day 171. Sham (PBS) mice were used as negative controls.
On 190 DPV, blood, BALF, and lung tissues were collected from euthanized mice for the following analysis. A Experimental schema for long-term study. B Serum anti-LcrV IgG titers.
C BALF anti-LcrV IgG titers. D BALF anti-LcrV IgA titers. E The absolute counts of lung IgG+ and IgA+ ASC BRM cells.
F The absolute counts of lung IgG+ and IgA+ mature BRM cells. G The absolute counts of IFN-γ-producing lung CD69+ CD4+ and CD8+ TRM cells. H The absolute counts of IL-17A-producing lung CD69+ CD4+ and CD8+ TRM cells.
I The absolute counts of IL-4-producing lung CD69+ CD4+ and CD8+ TRM cells. Immunofluorescence staining of BRM cells (J), and TRM cells (K) in the lung sections from sham (PBS), IM OMV46-LcrV/Alum, and IM OMV46-LcrV/Alum + IN LcrV immunized mice (n = 3). Cell surface markers were stained with the indicated antibodies, and cell nuclei were stained with DAPI.
A representative image from each group was shown. L Experimental schema for long-term survival study (left) and survival of a pulmonary Y. pestis challenge (Right). On 190 DPV, mice (n = 10) were intranasally challenged with 4 × 105 CFU (4000 LD50) of Y. pestis KIM6 + (pCD1Ap).
Mouse mortality and morbidity were monitored for 15 days. Each symbol was obtained from an individual mouse, and data were represented as the mean ± SD. The statistical significance of difference among groups was analyzed by two-way ANOVA with Turkey post hoc test.
Statistical significance was analyzed by the log-rank (Mantel–Cox) test for survival analysis: ns no significance; *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001. Source data are provided as a Source Data file.
Concomitantly, assessments of cellular immune responses showed that the LI LcrV booster significantly increased the number of CD69+CD4+ TRM and CD103+CD4+ TRM cells compared to IM OMV46-LcrV/Alum immunization alone and PBS (Fig. S6A, B). Also, the LI LcrV booster increased IFN-γ-, IL-17A-, or IL-4-producing CD69+CD4+ TRM cells but not the CD8+ counterparts (Fig.
5G–I). Compared to PBS, long-term immunized mice exhibited an increasing trend in TNF-α-producing CD69⁺CD4⁺ and CD69⁺CD8⁺ TRM cells, although the differences were not statistically significant (Fig. S6C).
Likewise, the LI LcrV booster significantly increased the number of AMs (Fig. S6D), MHCII+ AMs, and MerTK+ AMs (Fig. S6E) in the lung and BALF compared to PBS and IM OMV46-LcrV/Alum immunization.
The number of AMs and AM subpopulations in the lung and BALF were comparable between sham and IM OMV46-LcrV/Alum-immunized mice (Fig. S6D, E&). Further, immunohistochemical (IHC) stains also showed that the LI LcrV booster significantly increased lung BRM, TRM, and macrophage cells compared to OMV46LcrV/Alum vaccination alone (Figs.
5J, K and S6F).
A following challenge study on 190 DPV revealed that the LI LcrV booster following IM OMV46-LcrV/Alum vaccination offered complete protection against 4000 LD50 of pulmonary Y. pestis challenge, while OMV46-LcrV/Alum immunization alone only provided 40% protection. No sham mice survived the same challenge (Fig. 5L).
Thus, our results underscore that the elevated lung BRM, TRM, and AM cells and their corresponding immune functions activated by the LI LcrV booster confer complete protection against pneumonic plague. We also evaluated whether the short-interval IN LcrV (SI LcrV) boosting could provide comparable protection. Mice received IM prime-boost OMV46-LcrV/Alum on days 0 and 21, followed by IN LcrV boosting on day 37, and were challenged 127 days after initial vaccination with the same dose.
This regimen conferred 80% protection (Fig. S6G). Those results indicate that both immunization strategies are effective, while the LI LcrV booster is superior to SI LcrV.
IN LcrV booster enhances lung tissue-resident lymphocytes derived from circulating lymphocytes primed by IM OMV immunization, conferring protection against pneumonic plague
Given that IM OMV46-LcrV/Alum immunization generated lung BRM and TRM cells, we asked whether the migration of activated lymphocytes from lymph nodes (LNs) after vaccination is associated with the formation of lung BRM and TRM cells. FTY720 (fingolimod) binds the sphingosine-1-phosphate receptor, thereby blocking lymphocytes’ egress from LNs to tissues35. Mice were treated with FTY720 either 10 days before vaccination (termed pre-FTY720) or 10 days after IM OMV46-LcrV/Alum + IN LcrV immunization (termed post-FTY720) until the end of the experiment and without treatment as controls (termed non-FTY720) (Fig.
6A). On day 57 post initial immunization, all animals were challenged with 20,000 LD50 of pulmonary Y. pestis. Results showed that pre-FTY720 mice had 20% survival, while post-FTY720 and non-FTY720 mice had 90% and 100% survival, respectively (Fig.
6B). Ab analysis revealed that pre-FTY720 treatment significantly reduced serum anti-LcrV titers, whereas post-FTY720 treatment had no such effect compared to the non-FTY720 treatment (Fig. 6C).
On 2 DPI, lung single cells isolated from each group of mice were stained with indicated Abs and analyzed using flow cytometry as described above. Quantitative plots showed that the number of CD138+ BRM cells, along with their corresponding IgG+ and IgA+ cells, was comparable between the non-FTY720 and post-FTY720 treatment groups, but significantly higher than in the pre-FTY720 treatment group (Fig. 6D).
Similar patterns were observed for CD138+GL7− BRM cells and their corresponding IgG+ and IgA+ cells (Fig. 6E). Moreover, similar patterns were also observed for lung CD69+CD4+ TRM cells, their IFN-γ-, IL-17A-, or IL-4- producing cells and lung CD103+ CD4+ TRM cells (Fig.
6F-H). While the TNF-α producing lung CD69+CD4+ TRM cell counts were no different among all groups (Fig. 6G).
Also, the number of lung CD69+CD8+ TRM cells, their corresponding IFN-γ-, IL-17A-, TNF-α-, or IL-4-producing cells (Fig. 6F, G), and lung CD103+ CD8+ TRM cells (Fig. 6H) were no different among all groups.
Thus, results suggest that lymphocytes in LNs activated by IM OMV46-LcrV/Alum immunization are a key source of lung resident lymphocytes. Also, similar patterns of humoral and cell-mediated immune responses were observed on 57 DPV (Fig. S7).Fig.
6Impact of FTY720 treatment pre- and post-immunization on lung tissue resident lymphocytes and protection.A Schema of FTY720 treatment and immunization. Up: mice were treated with FTY720 on 10 days prior initial immunization until the end of experiment (termed to pre-FTY720); Bottom: Swiss Webster mice were treated with FTY720 on 10 days of post IN LcrV boosting until the end of experiment (termed to post-FTY720). B Survival of the Y. pestis challenge.
On 57 days post-initial immunization, mice (n = 10 equal males and females) were intranasally challenged with 2 × 106 CFU (20,000 LD50) of Y. pestis KIM6 + (pCD1Ap). Mouse mortality and morbidity were monitored for 15 days. C Serum anti-LcrV Ab levels in FTY720- and non-treated immunized mice at different time points.
On 2 DPI, quantitative analysis of lymphocytes in the lung (per lung) of pre-FTY720-, post-FTY720-, and non-FTY720-immunized mice (n = 5 females). D The absolute counts of lung ASC BRM cells and their IgG+ and IgA+ cells in immunized mice (n = 5 females). E The absolute counts of lung mature BRM cells and their IgG+ and IgA+ cells in immunized mice (n = 5 females).
F The absolute counts of lung CD69+ CD4+ and CD8+ TRM cells in immunized mice (n = 5 females). G The absolute counts of IFN-γ, IL-17A, and IL-4-producing lung CD69+ CD4+ and CD8+ TRM cells in immunized mice (n = 5 females). H The absolute counts of lung CD103+ CD4+ and CD8+ TRM cells in immunized mice (n = 5 females).
Each symbol was obtained from an individual mouse, and data were represented as the mean ± SD. The statistical significance of difference among groups was analyzed by two-way ANOVA with Turkey post hoc test. Statistical significance was analyzed by the log-rank (Mantel–Cox) test for survival analysis: ns no significance; *, p < 0.05; **, p < 0.01; *** p < 0.001; ****, p < 0.0001.
Source data are provided as a Source Data file.
To further confirm the protection provided by lung BRM and TRM cells, the BRM (IV−CD45+CD19+IgG+) and TRM (IV−CD45+CD3+CD4+CD44+) cells were sorted from PBS, IM OMV46-LcrV/Alum-, IM OMV46-LcrV/Alum + IN LcrV-immunized mice, respectively. A total of 2 × 105 sorted lung BRM and/or TRM cells were intratracheally (IT) transferred into naïve mice. After 24 h, the mice were intranasally challenged with 500 CFU of Y. pestis (Fig.
7A). Results showed that naïve mice receiving BRM cells alone, TRM cells alone, and both from IM OMV46-LcrV/Alum + IN LcrV-immunized mice exhibited 60%, 80%, and 80% survival, respectively. In contrast, naïve mice receiving BRM cells alone, TRM cells alone, and both from IM OMV46-LcrV/Alum-immunized mice had 40%, 25%, and 40% survival, respectively.
Naïve mice receiving BRM and TRM cells from sham mice died within 3 days (Fig. 7B).Fig. 7Evaluation of the protective roles of lung BRM and/or TRM cells from different immunized mice by cell transfer.Swiss Webster mice (n = 10 female, 7 weeks old) were IM immunized with 20 µg of OMV46-LcrV/Alum in 50 µL PBS on day 0 and 21.
One group of IM OMV46-LcrV/Alum immunized mice was subjected to IN LcrV boosting on day 37. Sham mice (PBS) were used as negative controls. (A) The schema of the experimental study was Created in BioRender.
Majumder, S. (2025) https://BioRender.com/catuw9m. On 45 DPV, mice were euthanized, and BRM and TRM cells were sorted as described in “Materials and Methods”.
FACS-sorted BRM, TRM, and both cells (a total of 2 × 105 cells) from immunized animals were transferred individually into naïve Swiss Webster mice via the intratracheal (IT) route. B Survival of the Y. pestis challenge. On 24 h post cell transfer, the recipient Swiss Webster mice (n = 10 females, 7-week-old) were intranasally challenged with 500 CFU (5 LD50) of Y. pestis KIM6 + (pCD1Ap).
C The schema of the experimental study was created in BioRender. Majumder, S. (2025) https://BioRender.com/catuw9m.
FACS sorted BRM, TRM, and both cells (a total of 2 × 105 cells) from immunized animals were transferred individually into IM OMV46-LcrV/Alum immunized mice via the intratracheal (IT) route. D Survival of the Y. pestis challenge. On 24 h post cell transfer, the recipient mice (n = 10 females, 7-week-old) were intranasally challenged with 2 × 106 CFU (20,000 LD50) of Y. pestis KIM6 + (pCD1Ap).
Mouse mortality and morbidity were monitored for 15 days. Statistical significance was analyzed by the log-rank (Mantel–Cox) test for survival analysis: ns, no significance; *, p < 0.05; **, p < 0.01; *** p < 0.001; ****, p < 0.0001. Source data are provided as a Source Data file.
Furthermore, we transferred lung BRM and/or TRM cells into IM OMV46-LcrV/Alum-immunized mice and challenged them with an enormous dose (2 × 106 CFU) of Y. pestis (Fig. 7C). The results showed that IM OMV46-LcrV/Alum-immunized mice receiving only lung BRM cells, only TRM cells, or both from IM OMV46-LcrV/Alum + IN LcrV-immunized mice exhibited 100%, 80%, and 100% survival, respectively.
In contrast, IM OMV46-LcrV/Alum-immunized mice that received both BRM and TRM cells from IM OMV46-LcrV/Alum-immunized mice had 60% survival (Fig. 7D). Additionally, sorted TRM cells (IV−CD3+CD4+CD44+) were stimulated with LcrV protein in vitro for 48 h, then stained with the indicated Abs for flow cytometry (Fig.
S8A). Consistent with the previous results (Fig. 4D, E), the number of IFN-γ-, IL-17A-, or Ki67-expressing TRM cells from the IM OMV46-LcrV/Alum + IN LcrV-immunized mice was significantly higher than that from IM OMV46-LcrV/Alum-immunized mice, indicating a heightened recall response mediated by IN LcrV boosting compared with the latter (Fig.
S8B–D). Therefore, our results corroborate that IN LcrV boosting augments immune functions of lung BRM and/or TRM cells imprinted by the initial IM OMV46-LcrV/Alum vaccination, leading to super protection against pneumonic plague.
Discussion
The adjuvant alum can boost Ab response to antigens in vaccines and reduce the incidence and severity of local and systemic reactions to vaccines15,36. Similarly, alum absorbed OMV46-LcrV induced lower levels of proinflammatory cytokines (IL-6 and IL-1β) and drastically abrogated its toxicity and reactogenicity. IM OMV46-LcrV/Alum vaccination demonstrated great safety and negligible side effects in mice (Figs.
1B–G and S1G and Table S4). It is known that the MPLA has been formulated with aluminum salt to generate Adjuvant System 04 (AS04), an adjuvant licensed for human use37. Our OMV46-LcrV vaccine is intrinsically and exclusively composed of detoxified MPLA14.
Thus, the OMV46-LcrV/Alum would likely exhibit similar properties to AS04-adjuvanted vaccines.
Compared to alum alone, AS04 enhances the humoral immune response and promotes a balanced and more Th1-biased immune response in animals38,39. However, IM OMV46-LcrV/Alum immunization elicited comparable anti-LcrV IgG titers in sera to IM LcrV/alum immunization or OMV46-LcrV alone (Fig. S3B) but shifted profiles of IgG subtypes towards a Th2-biased immune response, as observed in LcrV/Alum immunization (Fig.
S3C). Several factors could explain this observation. First, IL-4 induced by alum is an early and potent driver of the Th2 response40,41.
This is supported by the increased presence of IL-4-producing CD4⁺ T cells in the lungs following OMV46-LcrV/Alum immunization (Figs. 4F and 5I). Second, alum has also been shown to promote the release of IL-1042, an anti-inflammatory cytokine that supports Th2 responses and limits inflammation.
Third, the Y. pestis LcrV antigen itself has been reported to induce IL-10 production by immune cells43,44, although it remains debated45. IL-10 has been reported to suppress Th1 and Th17 immune responses46. The Th1 and Th17 responses are required for full protection against pneumonic plague47.
In addition, alum can promote a Th2 immune response via IL-33/ST2 signaling48. Thus, the OMV46-LcrV/alum containing high amounts of alum and LcrV antigen might promote the Th2 shift and diminish vaccine reactogenicity, while slightly compromising its efficacy against a huge dose of pulmonary Y. pestis infection (Fig. 1H).
However, the defined mechanisms warrant further investigation.
Most approved human vaccines are administered via parenteral (IM or SC) routes49. Parenteral vaccination with the alum-conjugated protein antigens generally elicits poor mucosal immune response in the lung17, as evidenced by no antigen-specific IgG and secretory IgA (sIgA) in the BALF and antigen-specific T-cell response in the lung of IM LcrV/Alum-immunized mice14. However, IM OMV46-LcrV/Alum immunization alone induced significant anti-LcrV IgG titers in the BALF and T-cell response in the lung.
Those responses remained long-lasting, although attenuated by alum absorption compared to IM OMV46-LcrV immunization14. Recently, IM immunization with mRNA-N (SARS-CoV-2 nucleocapsid protein)-LNP alone induced modest respiratory T cell responses in the lungs and BALF of mice, while combining mRNA-N with mRNA-S (spike protein) elicited strong respiratory T cell responses50. Thus, we speculate that the immune differences between LcrV/Alum and OMV46-LcrV/Alum may primarily be attributed to the intrinsic properties of the immunogen.
In addition, the alum-absorbed OMV46-LcrV might form a kind of alum nanoparticles51, which rapidly enter the draining LNs and thereby directly interact with antigen-presenting cells (APCs) within the LNs52, in turn triggering both systemic and lung immune responses.
Vaccines after IM injection are restricted to the local LNs mainly via secondary draining LNs (iliac or apical) based on their proximity to the injection site and are not disseminated systemically53,54. In contrast, IN vaccination is more likely to trigger the robust formation of lung BRM and TRM cells, which provide a rapid and localized response against respiratory pathogens55,56. Thus, the IM OMV46-LcrV/Alum vaccination may not directly induce those cells in the lung parenchyma.
In addition, the pre-FTY20 treatment before IM OMV46-LcrV/Alum vaccination, which blocks the egress of circulating lymphocytes, drastically reduced lung BRM- and TRM-cell counts and completely abrogated protection against pulmonary Y. pestis challenge. In contrast, the post-FTY20 treatment did not have this effect (Fig. 6B).
Thus, these results suggest that circulating memory lymphocytes may serve as a source for the generation of lung BRM and TRM cells. Based on the models of TRM cell lineage divergence57, we speculate that the IM OMV46-LcrV/Alum vaccination, instead of IM rLcrV/Alum, induces high complex cues within local LNs for prompting the generation of heterogeneous and circulating B and T cells with high expression of certain adhesion molecules or chemokine receptors. This facilitates the lymphocytes’ commitment to resident memory cell lineages and migration to the lung tissue compartment.
However, how parenteral OMV46-LcrV vaccination shapes lung immunity, specifically BRM and TRM cells, AMs, and their interactions, as well as how important cytokines (IL-17A and IFN-γ) shape the interrelation of these cells, have not been thoroughly investigated. Further studies are needed to pursue the defined mechanisms.
The IT transfer of lung BRM, TRM, or both cells from IN LcrV booster following IM OMV46-LcrV/Alum immunized mice to naïve mice or IM OMV46-LcrV/Alum immunized mice provided enhanced protection against pneumonic plague (Fig. 7). In addition, the IN LcrV booster following IM OMV46-LcrV/Alum vaccination, rather than the IN LcrV booster following IM LcrV/Alum vaccination, conferred excellent protection against pneumonic plague (Fig.
2). These results underscore the importance of preexisting LcrV-specific lung B and T memory cells in enhancing immune protection following the IN LcrV booster. Similar observations were seen in recent studies, like the pre-existing lung immunity induced by prior IM BNT162b2 vaccination19 or sublethal dose of influenza A virus (IAV) infection58 was leveraged to enhance lung immune responses by IN booster with corresponding adjuvant-free protein antigens against SARS-CoV-2 or IAV infection, respectively.
Intriguingly, the short-interval IN LcrV booster following IM OMV46-LcrV/Alum vaccination significantly elevated BALF anti-LcrV IgG titers compared to IM OMV46-LcrV/Alum vaccination alone, but not serum anti-LcrV IgG titers (Fig. 3A, B). In contrast, the long-interval IN LcrV booster following IM OMV46-LcrV/Alum vaccination resulted in an elevation of both serum and BALF anti-LcrV IgG titers compared to IM OMV46-LcrV/Alum alone (Fig.
5B, C). These results suggest that a long-interval booster more effectively recalls memory B cells residing in systemic and lung mucosal compartments than a short-interval booster, consistent with previous findings59. The IN LcrV booster following IM LcrV/Alum vaccination only stimulated very low levels of anti-LcrV IgG titers in the BALF (Fig.
3A, B). In addition, we found that BALF LcrV-specific sIgA titers were similarly elevated in mice given an intranasal LcrV booster following either IM LcrV/Alum or IM OMV46-LcrV immunization, and both were significantly higher than in mice without the IN booster (Fig. 3C).
IM OMV46-LcrV/Alum + IN LcrV provided enhanced protection against a huge dose (20,000 LD50) of Y. pestis challenge compared to IM OMV46-LcrV/Alum alone (Figs. 1H and 2B). In contrast, neither IM LcrV/Alum + IN LcrV nor IM LcrV/Alum alone provided protection, despite the former inducing higher sIgA (Figs.
1H, 2B, and 3C). The findings suggest that pre-existing lung IgG⁺ BRM cells are essential for generating robust antigen-specific IgG antibodies in the BALF following an IN LcrV booster, providing significant protection. In contrast, lung sIgA appears to be dispensable for protection against pneumonic plague, consistent with our previous study60.
In the serum transfer study, sera from either the OMV46-LcrV/Alum + IN LcrV group or the OMV46-LcrV/Alum group conferred partial protection, which was a similar observation to our previous study14. In contrast, sera from the LcrV/Alum + IN LcrV group did not (Fig. S3D), even though all three donor groups exhibited comparable total LcrV-specific IgG titers (Fig.
S3B). We reason that sera from the OMV46-LcrV/Alum groups, with or without IN LcrV boosting, may contain antibodies targeting additional Y. pestis antigens, owing to the presence of other potential Yersinia antigens in the OMV46-LcrV vaccine.
From an immunological perspective, protein antigens alone generally elicit a weak immunogenic response without adjuvants, while the antigens conjugated with proper adjuvants can induce robust immune responses61. However, concerns about the safety of adjuvants in the respiratory system, including the potential risk of triggering Bell’s palsy, pose a challenge for IN vaccination62. An adjuvant-free IN LcrV booster that utilizes preexisting LcrV-specific lung BRM and TRM cells could minimize potential side effects and enhance lung immunity against pneumonic plague.
Moreover, the long-interval IN LcrV booster following IM OMV46-LcrV/Alum vaccination provided sustained protection against pneumonic plague, boosting both systemic and lung mucosal immunity (Fig. 5).
The “prime-spike” immunization strategy, i.e., IM OMV46-LcrV/Alum followed by IN LcrV, demonstrated excellent protection against pneumonic plague in a mouse model. As human challenge studies with pneumonic plague are unethical and unfeasible, vaccine licensure must follow the FDA Animal Rule, requiring efficacy in multiple animal models predictive of human response63. Accordingly, we will evaluate this vaccine and regimen in Brown Norway rats and non-human primates, both established models in plague research64,65.
In addition, the “prime-spike” strategy shows potential advantages in cost-effectiveness and feasibility. IM15,16 and IN66 vaccination routes are already accepted in humans, which may ease coordination and public acceptance. The administration routes of IM OMV46-LcrV/Alum and IN LcrV can fit into existing healthcare systems.
While producing OMV and LcrV separately adds cost and complexity, the enhanced and longer-lasting protection may reduce the need for frequent boosters and help lower the plague burden.
In this study, both male and female mice were included to avoid sex bias in vaccine evaluation. Unlike previous reports of sex-dependent protection against pneumonic plague67,68, no sex-related differences in survival were observed in OMV46-LcrV-immunized Swiss Webster or C57BL/6 mice. The reasons for these discrepant findings remain unclear and will be further interrogated in future studies, including immune responses, bacterial burden, and potential environmental factors.
Altogether, the “prime-spike” regimen shows promise for rapidly recalling lung-resident memory immunity during an emergent pneumonic plague outbreak by timely inhalation of the LcrV antigen in individuals previously vaccinated with IM OMV46-LcrV/Alum and could be applied to combat outbreaks of other respiratory pathogens.
Methods
Mice and animal ethics statement
Seven-week-old male and female Swiss Webster mice were purchased from Charles River Laboratories (Wilmington, MA) and acclimated for one week before experiments. Wild-type C57/BL6 (B6) and Hfe−/− mice on a B6 background were purchased from Jackson Laboratories (Bar Harbor, Maine, USA). B6 and Hfe−/− mice were bred in an animal facility at Albany Medical College.
Age-matched male and female mice were used in the experiments. All animal studies were performed per the NIH “Guide for the Care and Use of Laboratory Animals” and approved by the Institutional Animal Care and Use Committee at Albany Medical College (IACUC protocol# 22-12002).
Strains, plasmids, culture conditions, molecular procedures, and OMV isolation and analysis
All bacterial strains and plasmids used in this study were listed in Supplementary Information (SI) Table 1. Bacterial cultures, molecular procedures, and OMV preparations were described in SI Methods.
Animal immunization and Y. pestis infection
Groups of mice (mixed males and females) were primed with OMV46-LcrV/Alum, LcrV/Alum, F1-V/Alum, or PBS by intramuscular (IM) injection and boosted on day 21 after the initial vaccination. Then, indicated groups of immunized mice were subjected to intranasal boosting with LcrV/PBS on day 37. Blood was collected via the submandibular veins on 21, 36, and 42 DPV to harvest serum for antibody (Ab) analysis.
The BALF was harvested from euthanized mice on 42 DPV for Ab analysis in the lung. Anti-LcrV specific Ab titers in the serum and BALF were measured using an enzyme-linked immunosorbent assay (ELISA)14. Briefly, Polystyrene 96-well flat-bottom microtiter plates (Thermo Fisher Scientific) were coated with 100 ng/well of purified rLcrV.
Antigens suspended in sodium carbonate bicarbonate coating buffer (pH 9.6) were applied in a 100-μl volume in each well. The coated plates were incubated overnight at 4 °C. Then, the LcrV-coated plates were incubated with twofold serial dilutions of serum samples (1:200 to 1:64,000).
The plates were incubated for at 37 °C for 1 h. Post incubation the plates were washed with 1 × TBST (Tris-Buffered Saline with Tween 20). The plates were incubated with anti-mouse IgG (1:1000) (Sigma, USA) at 37 °C for 1 h.
Post incubation, the plates were washed with 1 × TBST. The colorimetric reaction was developed using an alkaline phosphatase-conjugated secondary Ab with nitrophenyl phosphate substrate (Sigma, USA). The Ab titer was defined as the highest serum dilution with an OD405 value at least twice that of sham-immunized controls.
On 45 DPV, animals were anesthetized with 100 μl of ketamine/xylazine mixture (25 mg/ml ketamine plus 1 mg/ml xylazine) and intranasally challenged with either Y. pestis KIM6+ (pCD1Ap) or Δcaf1 Y. pestis KIM6+ (pCD1Ap) strain in 40 μl of PBS to recapitulate the development of primary pneumonic plague. Animal morbidity and mortality were monitored for 15 days. To determine the bacterial burden after the challenge.
The lungs, liver, and spleen were collected from euthanized animals at 2, 6, and 15 DPI and homogenized in ice-cold sterilized PBS using Mini Bead Beater (Bio spec Product Inc., USA) at power 7 for 2 min. Serial dilutions of each organ homogenate were spread on HIB Congo Red agar plates to enumerate Y. pestis CFUs.
Histopathology
The lung lobes were collected from groups of euthanized mice at indicated time points and fixed in 10% formalin overnight before being embedded in paraffin. Lung sections were stained with hematoxylin/eosin and then examined using standard procedures described in SI Methods. Each lung section was evaluated for four different phenotypes (A value of 4 for severe, 2 for moderate, 1 for light, and 0 for no damage) indicative of lung damage or injury, which include (a) edema (swelling around the blood vessels or airways), (b) hemorrhage (fluid and or blood leakage into alveolar space), and (c) disruption of alveolar architecture14.
Immunohistochemistry staining followed a standard procedure26. Briefly, lung sections were subjected to antigen retrieval in citrate buffer (pH 6.0), washed with 1 × PBS, blocked in 3% BSA, and incubated with indicated antibodies (1:1000 dilution overnight in 4 °C) in 0.5% BSA for 1 h at 4 °C. Ab details are listed in Table S3.
Then, the tissue sections were washed and probed with a corresponding fluorescent secondary Ab (1:100) in 0.5% BSA and incubated for 1 h at 4 °C. Subsequently, the tissue sections were washed and mounted under a coverslip using FlourshieldTM with DAPI (Sigma, USA) and examined under an Echo Revolve R4 fluorescence microscope (Echo, USA).
Flow cytometry
To determine BRM and TRM cells, mice were injected intravenously with FITC anti-mouse CD45.2 mAb diluted in 200 µL of sterile PBS via the tail vein to discriminate circulating from extravascular immune cells. Five minutes after injection, lung tissues were harvested from euthanized mice using standard procedures29,30. Briefly, the lungs were isolated, chopped, and digested with Liberase TL (1 mg/mL; Roche) and DNase I (10 mg/mL; Sigma) mixture for 30 min at 37 °C for 30 min with gentle shaking.
The digested lung samples were filtered through 70-µm strainers to obtain single cells, followed by RBC lysis. Single-cell suspension was incubated with anti-mouse CD16/32 (Fc block) for 5 min followed by dead live staining with Zombie for 30 min at 4 °C in flow cytometry staining (FACS) buffer. Cells were stained with indicated fluorescence-labeled anti-CD19b, anti-CD138, anti-CD38, anti-GL7, anti-IgG, and anti-IgA antibodies with appropriate dilution in FACS buffer and incubated for 60 min at 4 °C.
Cells were washed in FACS buffer and followed by 4% paraformaldehyde fixation for 15 min at 4 °C. In parallel, cells were stained with indicated fluorescence-labeled CD3, CD4, CD8, CD44, CD62L, CD69, and CD103 antibodies with appropriate dilution in FACS buffer and incubated for 60 min at 4 °C. Cells were washed in FACS buffer and followed by 4% paraformaldehyde fixation for 15 min at 4 °C.
To assess cytokine-producing lung TRM cells, single-cell suspensions (2 × 10⁶ cells) from the lungs of immunized mice were seeded into 24-well plates and stimulated with endotoxin-free LcrV (10 µg/mL) for 48 h. Subsequently, a 1× brefeldin A and monensin cocktail (1:1) was added for 4 h to block Golgi-mediated cytokine secretion. Cells were then harvested, resuspended in FACS buffer containing Fc block (anti-mouse CD16/32, 1:200) for 10 min on ice, and stained with Fixable Viability Dye (Zombie Red) and T cell surface markers (CD3, CD4, CD8, CD44, CD62L, CD69, and/or CD103) for 1 h at 4 °C.
After washing, cells were fixed (30 min, RT), permeabilized (30 min, 4 °C), and stained intracellularly with antibodies against IFN-γ, TNF-α, IL-17A, IL-4, and Ki67 in permeabilization buffer for 45 min at 4 °C. Finally, cells were washed and resuspended in FACS buffer for analysis.
To measure alveolar macrophages and neutrophils in the lung compartment, cells were isolated from the lungs and BALF of immunized mice and stained with Fixable Viability Dye (Zombie Red), followed by fluorochrome-conjugated antibodies against CD11b, Ly6G, CD11c, MHC II, CD64, Siglec F, and MerTK in FACS buffer. Samples were analyzed on a BD FACS Symphony A3 (BD Biosciences), and data were processed using FlowJo v10.7. Ab details are provided in Table S3.
Cell transfer
Lung BRM and TRM cells were distinguished from circulating T cells by intravenous administration of anti-CD45.2 mAb as described above. Single-cell suspensions were prepared from mice, and cell viability was assessed using Trypan blue. Cells were stained with indicated molecular markers, washed with FACS buffer, and sorted using a FACSAria II (BD) three-laser sorter at 25 psi and a flow rate of 2.0, with the “purity” setting.
Electronic compensation was done using single-color controls, and FSC/SSC settings were adjusted with the samples. Antigen specificity of sorted cells was confirmed in Fig. S8A.
Sorted cells (2 × 10⁵) in 40 µL sterile PBS were intratracheally transferred into irradiated naïve or IM OMV46-LcrV/Alum-immunized mice. One day post-transfer, recipient mice were intranasally challenged with Y. pestis and monitored for survival over 15 days.
Serum transfer and cytokine neutralization
Three groups of naïve Swiss Webster mice (n = 10) were intraperitoneally injected with 100 µl of pooled sera from OMV46-LcrV/Alum+ IN rLcrV, OMV46-LcrV/Alum, and IM LcrV/Alum + IN LcrV immunized mice collected on 45 DPV, respectively, then intranasally challenged with Y. pestis KIM6+ (pCD1Ap) (2 × 103 CFU) on 24 h post transfer. To deplete specific cytokines OMV46-LcrV/Alum+ IN rLcrV -immunized mice were intranasally injected with 0.5 µg of mAb specific for IFN-γ (clone XMG 1.2, Bio X Cell), and IL-17A (clone 17F3, Bio X Cell) at 39, 40, and 41 DPV. The control mice received 0.5 µg isotype control rat IgG1 mAb (clone HRPN, Bio X Cell).
Treated mice were challenged with Y. pestis KIM6+ (pCD1Ap) (2 × 106 CFU) on an indicated time point. Challenged mice were monitored for signs of infection, mortality and morbidity over 15 days. On 2 dpi mice were euthanized and lung samples were isolated to analyze the myeloid cell and T cell count as mentioned earlier.
FTY720 treatment
To block the circulating lymphocytes, mice were treated with FTY720 (Santa Cruz Biotechnology) in drinking water at a dose of 0.3 mg/kg/day for 10 days before immunization or 10 days after IN LcrV boosting, as shown in Fig. 6A. Treatment was maintained until the experiment’s termination.
Mice receiving regular drinking water served as controls. All treated and untreated mice were intranasally challenged with Y. pestis and monitored for mortality over 15 days. Additional groups of mice were used to assess changes in lung BRM and TRM cells at 2 DPI.
Statistical analysis
Each experiment included a significant number of biological replicates, to establish reproducibility. Statistical analyses of comparisons of data among groups were performed with one-way ANOVA/univariate or two-way ANOVA with Tukey post hoc tests. The log-rank (Mantel–Cox) test was used for survival analysis.
All data were analyzed using GraphPad PRISM 9.0 software. The data are represented as the mean ± standard deviation (ns, no significance; * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001).
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Supplementary information
Supplemantary InformationReporting SummaryTransparent Peer Review file
Source data
Source data