Preparation and characterization of the AIBP vaccine

Ciprofloxacin is a broad-spectrum antibiotic that inhibits bacterial DNA synthesis by rapidly inhibiting the activity of DNA gyrase or DNA topoisomerase IV, thus disrupting DNA replication and repair processes²⁶. We found that ciprofloxacin at 0.1 mg ml⁻¹, 0.25 mg ml⁻¹ or 0.5 mg ml⁻¹ completely inactivated B. pertussis at 6 × 10⁷ colony forming units (CFU) ml⁻¹ after 2–4 h (Fig. 1a). The ciprofloxacin-treated bacteria had intact membrane morphology, and a proportion were elongated but not lysed (Extended Data Fig. 1a). Treatment of B. pertussis with levofloxacin, another fluoroquinolone antibiotic, had a similar effect on bacterial morphology (Extended Data Fig. 1b). By contrast, B. pertussis treated with the beta-lactam antibiotic chloramphenicol had lost their membrane structure, consistent with lysis of the bacteria (Extended Data Fig. 1b).

a, Survival of B. pertussis cultured in S&S medium at a starting concentration of 6 × 10⁷ CFU ml⁻¹ with 0.1–0.5 mg ml⁻¹ of ciprofloxacin or medium only; live bacteria were quantified by performing CFU counts on BG agar plates at 2 h, 4 h and 6 h. Data are the mean ± s.d. of biological replicates (n = 3). b, Female 6–8-week-old C57BL/6 mice were exposed to an aerosol of live B. pertussis or AIBP vaccine, and CFU were assessed in lungs and nasal tissue after 2 h, 24 h and 72 h. Data are the mean ± s.d. of biological replicates (n = 3 for each time point). c–e, Female 6–8-week-old C57BL/6 mice were immunized by aerosol administration of the AIBP vaccine (equivalent to a culture of 1 × 10⁹ CFU ml⁻¹) or were immunized intramuscularly with a wP vaccine (1/50 human dose), and the concentrations of TNF, IL-1β and IL-6 were quantified in serum after 4 h and 24 h (n = 4 for each time point) (c), the concentration of CRP was quantified in serum after 48 h (n = 5) (d) and the number of neutrophils in the spleen was quantified by flow cytometry after 24 h and 48 h (n = 5 for each time point) (e). f, Female 6–8-week-old C57BL/6 mice were exposed to an aerosol of live B. pertussis or AIBP vaccine, and after 24 h, the concentration of PT was quantified in the lung by ELISA (n = 5). Data in c–f are presented as the mean ± s.e.m. of biological replicates (n = 4 or 5) shown as individual symbols. Data were analysed by two-way ANOVA followed by Tukey’s test for multiple comparisons. P values are shown above data compared.

Source data

We prepared the AIBP vaccine by treating B. pertussis from an overnight liquid culture (6 × 10⁷ CFU ml⁻¹) with ciprofloxacin at 0.25 mg ml⁻¹ for 3 h, followed by 2 washes in 1% casein solution. Inactivation of the bacteria was confirmed by lack of growth on Bordet–Gengou (BG) agar for 3–5 days. Furthermore, no live B. pertussis could be detected in the lungs or nose 24 h and 72 h after aerosol delivery of the AIBP vaccine to mice, showing that AIBP does not colonize the respiratory tract, whereas significant CFU were detected in mice given an aerosol of live B. pertussis (Fig. 1b).

We assessed the potential safety of the AIBP vaccine. Aerosol administration of the AIBP vaccine did not significantly increase inflammatory cytokine concentrations in the serum over those detected in mice immunized with phosphate buffered saline (PBS (Fig. 1c). By contrast, immunization of mice with the wP vaccine led to significant increases in the concentrations of IL-1β and tumour necrosis factor (TNF) in serum of mice at 4 h and IL-6 at 4 h and 24 h (Fig. 1c). Furthermore, immunization with the wP vaccine resulted in significantly elevated concentrations of C-reactive protein (CRP) in the serum, whereas serum CRP concentrations were not significantly increased in mice immunized with the AIBP vaccine (Fig. 1d). We also showed that the number of neutrophils in the spleen was not enhanced 24 h or 48 h after immunization with the AIBP vaccine, whereas neutrophils were significantly elevated 24 h and 48 h after immunization with the wP vaccine (Fig. 1e). Furthermore, the number of T cells in the spleen was not enhanced 24 h or 48 h after immunization with the AIBP vaccine (Extended Data Fig. 2a). PT is a major contributor to pertussis disease. We detected high concentrations of PT in the lungs of mice following aerosol administration of live B. pertussis, but not following aerosol administration of the AIBP vaccine (Fig. 1f). To address the longer-term safety of the vaccine, we examined body weight changes and showed that mice immunized with the AIBP vaccine gained weight to the same extent as mice immunized with PBS (Extended Data Fig. 2b). These findings show that unlike the wP vaccine, the AIBP vaccine does not have any features associated with potential toxicity in humans.

The AIBP vaccine activates antigen-presenting cells

Dendritic cells (DCs) have a central role in priming naive T cells. We assessed the effect of the AIBP vaccine on DC maturation. The AIBP vaccine enhanced expression of major histocompatibility complex class II (MHCII), CD40 and CD80 on bone marrow-derived DCs, and this was significantly greater than that induced with the wP vaccine (Fig. 2a). The AIBP vaccine promoted production of the T cell-polarizing cytokines IL-1β, IL-12p70 and IL-23, which was significantly greater than that induced with the wP vaccine (Fig. 2b). We next assessed production of T cell-polarizing cytokines in vivo at the site of immunization. Aerosol immunization with the AIBP vaccine significantly enhanced the concentrations of IL-1β and IL-23 in the lungs 4 h after administration (Fig. 2c), similar to or greater than that induced by parenteral immunization with the wP vaccine (Fig. 2c). Aerosol delivery of the AIBP vaccine, but not the wP vaccine, induced significant IL-6 production in the nasal tissue 4 h after administration (Supplementary Fig. 1). These findings show that aerosol delivery of the AIBP vaccine induces production of T_H1- and T_H17-polarizing cytokines at the site of immunization in the respiratory tract, but not inflammatory cytokines in the circulation, which can cause side effects.

a,b, Bone marrow-derived DCs were stimulated for 24 h with AIBP or wP vaccine (bacterium-to-cell ratio of 10:1) or medium (Med.). a, Surface expression of MHCII, CD40 and CD80 was evaluated using flow cytometric analysis gated on live single cells. Results are expressed as mean fluorescence intensity (MFI) and show individual values for four biological replicates. b, Concentrations of IL-1β, IL-12p70 and IL-23 in supernatants were quantified by ELISA (n = 3). c, Groups of eight female 6–8-week-old C57BL/6 mice were immunized by exposure to an aerosol of the AIBP vaccine or intramuscularly with a wP vaccine or PBS. After 4 h and 24 h, concentrations of IL-1β and IL-23 in lung homogenates were quantified by ELISA. d, CD4 T cells enriched from the spleen of B. pertussis convalescent mice were cultured with AIBP or wP vaccines at concentrations equivalent to 10 bacteria to one DC, and after 3 days, IL-17 was quantified in supernatants by ELISA. Data are presented as the mean ± s.e.m. of four biological replicates shown as individual symbols. Data were analysed by one-way ANOVA followed by Tukey’s test for multiple comparisons. P values are shown above relevant datasets.

Source data

Finally, we assessed the ability of the AIBP vaccine to stimulate B. pertussis-specific T cell responses in vitro. B. pertussis-specific CD4 T cells enriched from spleens of convalescent mice, with residual antigen-presenting cells (APCs), secreted IL-17 when cultured with AIBP, and this was significantly greater than observed following culture with the wP vaccine (Fig. 2d). We also compared the AIBP vaccine with heat-killed B. pertussis (HKBP) over a range of antigen concentrations. B. pertussis-specific CD4 T cells, purified from spleens of convalescent mice, secreted IL-17 and interferon-γ (IFNγ) when cultured with AIBP in the presence of APCs (irradiated spleen cells), and this was significantly greater than that induced with HKBP, especially at low antigen concentrations (Extended Data Fig. 3). Collectively, our findings show that the AIBP vaccine induces DC maturation and production of T cell-polarizing cytokines and activates APC to drive T_H1 and T_H17 responses.

Aerosol AIBP vaccine potently induces respiratory T_RM cells and confers sterilizing immunity in the lungs and nasal tract

We have reported that induction of local T cell responses in the respiratory tract is key to protective immunity against B. pertussis²²; therefore, we first examined the ability of AIBP vaccines, prepared with B. pertussis treated with ciprofloxacin or levofloxacin, compared with chloramphenicol, to induce respiratory T_RM cells. We also compared the immunogenicity of bacteria treated with ciprofloxacin for 3 h or 24 h. Aerosol immunization of mice with B. pertussis inactivated with ciprofloxacin (3 h or 24 h) or levofloxacin resulted in expansion of CD4 T cells and CD4 T_RM cells in the lungs and nasal tissue of immunized mice (Extended Data Fig. 4a–d). By contrast, the number of CD4 T cells and CD4 T_RM cells in mice immunized with chloramphenicol-inactivated B. pertussis was at background levels, like that observed in mice immunized with PBS (Extended Data Fig. 4a–d). Furthermore, aerosol immunization of mice with fluoroquinolone antibiotics generated B. pertussis-specific CD4 T_RM cells in the lungs that secreted IFNγ and IL-17, whereas immunization with chloramphenicol-inactivated B. pertussis did not induce B. pertussis-specific T_RM cells (Extended Data Fig. 4e,f). The T cell responses were not significantly different in mice immunized with B. pertussis treated with ciprofloxacin for 3 h compared with 24 h (Extended Data Fig. 4a–f).

We next examined the immunogenicity and protective efficacy of aerosol-delivered AIBP vaccine, compared with a current licensed parenterally delivered aP vaccine. CD4 T_RM cells were detected in the lungs and nasal tissue after a single aerosol immunization with the AIBP vaccine, and this was significantly enhanced following a booster immunization (Fig. 3a). By contrast, respiratory CD4 T_RM cells were not enhanced in mice immunized with the aP vaccines. Furthermore, B. pertussis-specific IL-17- and IFNγ-producing CD4 T_RM cells were induced in lung and nasal tissues of mice immunized with the AIBP vaccine, but not with the aP vaccine (Fig. 3b). These responses were strongest after two immunizations with the AIBP vaccine. Representative flow cytometry plots are shown in Extended Data Fig. 5a–d.

Female 6–8-week-old C57BL/6 mice were immunized by aerosol administration of the AIBP vaccine (once or twice at 0 week and 4 weeks), an aP vaccine (i.m., twice, 0 week and 4 weeks; 1/50 human dose) or PBS. The mice were aerosol challenged from a culture at 1 × 10⁹ CFU ml⁻¹ of live B. pertussis at week 6. On the day of but before the challenge with live B. pertussis, mice were injected intravenously with anti-CD45 antibody 10 min before euthanasia (to identify tissue-resident cells), and lung or nasal tissue cells were stained with antibodies specific for T_RM cells, or cells were stimulated with HKBP, anti-CD28 and anti-CD49d (both 1 μg ml⁻¹) for 16 h, followed by Brefeldin A (5 μg ml⁻¹) for the final 4 h of culture before intracellular cytokine staining (ICS) and flow cytometric analysis. a, Mean absolute number of CD4 T_RM cells (CD45iv⁻ CD4⁺CD44⁺CD62L⁻CD69⁺CD103^+/−) quantified in lung and nasal tissue by flow cytometric analysis. Data are mean ± s.e.m. (n = 5). b, Number of B. pertussis-specific IFNγ- or IL-17-secreting CD45iv⁻CD4 T_RM cells in lung and nasal tissues determined using ICS and flow cytometry. Data in a and b are presented as the mean ± s.e.m. of biological replicates shown as individual symbols (n = 5). c, Before the B. pertussis challenge, concentrations of FHA-specific IgA in nasal tissue homogenates and FHA-specific IgG1 and IgG2c in serum were quantified by ELISA. Data are presented as the mean ± s.e.m. of biological replicates (n = 5). d, Live bacterial loads in lung and nasal tissue were quantified by CFU counts at 2 h and 7 days, 14 days and 21 days after the live B. pertussis challenge. Data are presented as the mean ± s.e.m. of biological replicates (n = 5). Data were analysed by one-way ANOVA followed by Tukey’s test for multiple comparisons. P values are shown above relevant datasets.

Source data

Immunization of mice with two doses of AIBP vaccine induced potent FHA-specific IgA in the nasal tissue, whereas the aP vaccine failed to induce FHA-specific IgA (Fig. 3c). Two doses of AIBP or aP vaccine induced FHA-specific IgG1 and IgG2c in the serum. By contrast, FHA-specific antibodies were undetectable after a single dose of the AIBP vaccine (Fig. 3c).

Mice immunized with 2 doses of the AIBP vaccine completely cleared bacteria from the lungs 14 days after the live B. pertussis challenge (Fig. 3d). A single immunization with the AIBP vaccine conferred similar protection against B. pertussis infection of the lung to two immunizations with an aP vaccine (Fig. 3d). A single dose of the AIBP vaccine also conferred a high level of protection against infection of the nose, and mice that received two doses had completely cleared the infection from the nose 14 days after the challenge (Fig. 3d). By contrast, immunization with the aP vaccine did not protect against nasal infection with B. pertussis.

We also assessed intranasal (i.n.) delivery of the AIBP vaccine by dropping it onto the nose of mice and found that doses equivalent to 3 × 10⁵ CFU, 3 × 10⁶ CFU or 3 × 10⁷ CFU conferred high levels of protection against lung and nasal infection with B. pertussis, with the most complete protection achieved with the highest dose (Extended Data Fig. 6).

Collectively, our data show that the AIBP vaccine delivered by the aerosol or i.n. route confers protective immunity against lung and nasal infection, even after a single dose, and this is associated with the induction of potent T cell and antibody responses in the respiratory tract.

Finally, we assessed the durability of immunity induced with the AIBP vaccine. CD4 T_RM cells were still elevated in lung and nasal tissues of mice 7 days after the B. pertussis challenge of mice immunized 6 months earlier with the AIBP vaccine, and the numbers were significantly greater than in mice immunized with PBS (Extended Data Fig. 7a,b). Furthermore, substantial numbers of B. pertussis-specific IL-17- and IFNγ-producing CD4 T_RM cells were detectable in lung and nasal tissues. Lower numbers of IL-5-secreting CD4 T_RM cells were detected in the lungs, and very low numbers were detected in the nasal tissue (Extended Data Fig. 7c,d). The B. pertussis challenge of mice 6 months post-immunization showed a high level of protection against infection of the lungs and nose, with complete bacterial clearance by day 21 (Extended Data Fig. 7e). These findings show that the AIBP vaccine confers durable immunity against B. pertussis.

Efficacy of the AIBP vaccine not affected by priming with an aP vaccine

Existing aP vaccines promote B. pertussis-specific T_H2, but do not generate strong T_H1 or T_H17 responses in mice or humans^25,27. Immunization with the aP vaccine can also suppress the induction of CD4 T_RM cells and thereby bacterial clearance after the B. pertussis challenge²⁸. Here we assessed the induction of T cell subtypes and protective efficacy of two doses of the AIBP vaccine in mice previously immunized with two doses of an aP vaccine (Fig. 4a). The number of CD4 T_RM cells in the respiratory tract was at background levels in mice immunized with the aP vaccine (Fig. 4b). By contrast, there was significant accumulation of CD4 T_RM cells in the lungs and nasal tissue of mice immunized with the AIBP, and this was not significantly different in mice immunized with the aP vaccine and boosted with the AIBP vaccine. The aP vaccine failed to generate B. pertussis-specific CD4 T_RM cells in respiratory tissue (Fig. 4c). By contrast, IL-17- and/or IFNγ-secreting B. pertussis-specific CD4 T_RM cells were substantially augmented in the lungs and nasal tissue of mice immunized with the AIBP vaccine and these responses were not significantly different in mice immunized with the aP vaccine and boosted with the AIBP vaccine (Fig. 4c).

a, Schematic of the experiment; mice were immunized intramuscularly with the aP vaccine (twice at 0 week and 4 weeks; 1/50 of the human dose) followed by aerosol administration of the AIBP vaccine (twice, 8 weeks and 12 weeks) or given two doses of the aP or AIBP only or PBS. Female 6–8-week-old C57BL/6 mice were aerosol challenged with live B. pertussis at week 14. b,c, On the day of but before the challenge with live B. pertussis, CD4 T_RM cells (b) and B. pertussis-specific IL-17- and IFNγ-producing CD4 T_RM cells (c) were analysed by ICS and flow cytometry as described in the legend of Fig. 3. Data in b and c are presented as the mean ± s.e.m. of biological replicates shown as individual symbols (n = 5). d, On the day before the B. pertussis challenge, concentrations of B. pertussis-specific IgA, IgG1 and IgG2c in nasal tissue homogenates and serum were quantified by ELISA. Data are presented as the mean ± s.e.m. of biological replicates (n = 5). e, CFU counts on lung and nasal tissue 2 h, 7 days, 14 days and 21 days post-challenge. Data are presented as the mean ± s.e.m. of biological replicates (n = 5). Data were analysed by two-way ANOVA followed by Tukey’s test for multiple comparisons. P values are shown above relevant datasets. Panel a created with BioRender.com.

Source data

Immunization with the AIBP vaccine induced nasal IgA, and this response was not affected by previous immunization with the aP vaccine (Fig. 4d). Two doses of the aP vaccine induced B. pertussis-specific IgG1 in the serum, and this was not enhanced by boosting with the AIBP vaccine (Fig. 4d). Substantial concentrations of B. pertussis-specific IgG2c were induced with two doses of the AIBP vaccine. By contrast, IgG2c responses were close to the background in mice immunized with two doses of the aP vaccine (Fig. 4d). However, IgG2c was induced in mice primed with two doses of the aP vaccine and boosted with the AIBP.

Consistent with the data in Fig. 3, two immunizations with the AIBP vaccine conferred a high level of protection against lung and nasal infection with B. pertussis and this was similar to that observed in mice immunized with two doses of the aP vaccine and boosted with two doses of the AIBP vaccine (Fig. 4e). By contrast, immunization with two doses of the aP vaccine alone conferred only modest protection in the lungs, but not in the nose.

We also assessed the effect of previous immunization with two doses of an aP vaccine on the immune responses and protective efficacy of a single dose of the AIBP vaccine. A single dose of the AIBP vaccine induced CD4 T_RM cells and IL-17- and/or IFNγ-secreting B. pertussis-specific CD4 T_RM cells, and this was not significantly different in mice that had previously been immunized with two doses of the aP vaccine (Extended Data Fig. 8a,b). Furthermore, boosting with a single dose of the AIBP vaccine enhanced protection induced with the aP vaccine in the lungs (Extended Data Fig. 8c), whereas the protection against nasal infection induced with a single dose of the AIBP vaccine was similar in mice previously immunized (twice) with the aP vaccine versus PBS (Extended Data Fig. 8c).

Our findings show that previous immunization of mice with two doses of the aP vaccine, which is known to selectively induce T_H2 responses⁷, does not compromise the ability of one or two doses of the AIBP vaccine to induce CD4 T_RM cells and B. pertussis-specific IL-17- and IFNγ-secreting CD4 T_RM cells or to protect against lung or nasal infection with B. pertussis.

The AIBP vaccine induces stronger T_RM responses and confers superior protection against nasal infection than a wP vaccine

Good-quality wP vaccines are the gold standard for pertussis vaccine efficacy, whereas immunity generated by previous infection is effective and long lived^22,29,30. Here we compared the AIBP vaccine with previous infection or immunization with the wP vaccine given by the conventional intramuscular (i.m.) route. CD4 T_RM cells accumulated in the lungs and nasal tissue of mice immunized with the AIBP vaccine, and this was significantly stronger than that induced by immunization with the wP vaccine (Extended Data Fig. 9a). IL-17- or IFNγ-secreting B. pertussis-specific CD4 T_RM cells were significantly higher in the lungs of mice immunized with the AIBP vaccine than in the lungs of mice immunized with the wP vaccine or previously infected (Extended Data Fig. 9b). Furthermore, IFNγ-secreting B. pertussis-specific CD4 T_RM cells were significantly higher in the nasal tissue of mice immunized with the AIBP vaccine than in the nasal tissue of mice immunized with the wP vaccine (Extended Data Fig. 9b).

Assessment of systemic B. pertussis-specific T cell responses revealed that lymph node and spleen cells from mice immunized with two doses of the AIBP vaccine produced substantial quantities of B. pertussis-specific IL-17 and IFNγ. The IL-17 production was similar to that induced by previous infection and significantly greater than that induced by two doses of the wP vaccine (Supplementary Fig. 2).

The AIBP vaccine and previous infection induced B. pertussis-specific IgA in the nasal mucosa, whereas the wP vaccine failed to induce IgA (Extended Data Fig. 9c). However, i.m. immunization with the wP vaccine generated higher concentrations of B. pertussis-specific serum IgG1 and IgG2c than immunization with the aerosol-delivered AIBP vaccine or previous infection (Extended Data Fig. 9c).

The AIBP vaccine, the wP vaccine and previous infection all conferred complete protection against B. pertussis infection of the lungs (Extended Data Fig. 9d). However, the best protection against nasal infection was observed with the aerosol-delivered AIBP vaccine. AIBP-immunized mice and previously infected mice had completely cleared the infection from the nose by day 14 and 21, respectively, whereas bacteria were still detectable in the nose 14 days and 21 days after the live B. pertussis challenge of mice intramuscularly immunized with the wP vaccine (Extended Data Fig. 9d).

Collectively, our findings show that while the AIBP induced weaker serum IgG responses, it generated more potent IgA, systemic T_H1 and T_H17 responses, and IL-17- and IFNγ-secreting CD4 T_RM cells. Importantly, the aerosol-delivered AIBP vaccine conferred greater protection against infection of the nose than the parenterally delivered wP vaccine or previous infection.

As the respiratory route of immunization may have contributed to the superior efficacy of the AIBP vaccine, we examined the immunogenicity and protective efficacy of the AIBP vaccine compared with those of the wP or aP vaccines, when all vaccines were delivered by the i.n. route. Intranasal immunization with AIBP induced CD4 T_RM cells in the lungs and nasal tissue (Fig. 5a). By contrast, respiratory CD4 T_RM was at background levels in mice immunized with the wP or aP vaccine by the i.n. route. Furthermore, B. pertussis-specific IL-17- and IFNγ-producing CD4 T_RM cells were induced in lung and nasal tissues of mice immunized with the AIBP vaccine but were close to background in mice immunized intranasally with the wP or aP vaccine (Fig. 5b).

Female 6–8-week-old C57BL/6 mice were immunized by i.n. administration of AIBP vaccine (equivalent to 1 × 10⁸ bacteria per mouse), wP vaccine (1/160 human dose, equivalent to 1 × 10⁸ bacteria per mouse), aP vaccine (1/160 human dose) or PBS twice at 0 week and 4 weeks and were challenged with live B. pertussis 2 weeks later. a,b, On the day of but before the challenge with live B. pertussis, CD4 T_RM cells (a) and B. pertussis-specific IL-17- and IFNγ-producing CD4 T_RM cells (b) were analysed by ICS and flow cytometry as described in the legend of Fig. 3. Data in a and b are presented as the mean ± s.e.m. of biological replicates shown as individual symbols (n = 5). c, On the day before the B. pertussis challenge, concentrations of B. pertussis-specific IgA in nasal tissue homogenates and IgG1 and IgG2c in serum were quantified by ELISA. Data are presented as the mean ± s.d. of biological replicates (n = 5). d, CFU counts on lung and nasal tissue 2 h and 7 days, 14 days and 21 days post-challenge. Data are the mean ± s.d. of biological replicates (n = 5). Data were analysed by one-way ANOVA followed by Tukey’s test for multiple comparisons. P values are shown above relevant datasets.

Source data

Immunization of mice with the AIBP vaccine induced potent B. pertussis-specific IgA in the nasal tissue and IgG1 and IgG2c in serum. By contrast, antibody responses were weak or undetectable in mice immunized by the i.n. route with the wP or aP vaccine (Fig. 5c).

Intranasal immunizations with the AIBP vaccine conferred a high level of protection against lung and nasal infection with B. pertussis (Fig. 5d). By contrast, i.n. immunization with the wP vaccine induced modest protection against lung or nasal infection and this was substantially poorer than that observed with the AIBP vaccine. The aP vaccine failed to protect against lung or nasal infection when administered by the i.n. route (Fig. 5d). These findings show that the superior efficacy of the AIBP vaccine is not solely due to its delivery via the respiratory tract.

Mechanism of protective immunity induced with the AIBP vaccine

Although antibodies have a role in preventing infection of the lungs in mice^7,31 and maternal antibodies protect against pertussis disease in human infants³², studies in mice have shown that IL-17-secreting T cells are required for clearance of B. pertussis from the nasal tract²². Studies in mice, baboons and humans have shown that current aP vaccines selectively induce T_H2 cells and antibody responses, but not respiratory T_RM cells, and consequently fail to prevent nasal colonization with B. pertussis^7,18,19,25. We examined the possible role of IL-17-secreting CD4 T cells by depleting CD4 cells or neutralizing IL-17 before and after challenge with B. pertussis in mice immunized with one dose of the AIBP vaccine.

Flow cytometry analysis showed that recruitment of CD4 T cells to the lungs and nose in mice immunized with the AIBP vaccine was abrogated in mice treated with the anti-CD4 antibody and was also reduced in mice treated with the anti-IL-17 antibody (Fig. 6a). IL-17 is known to recruit neutrophils, especially Siglec-F⁺ neutrophils, to the respiratory tissue of B. pertussis-infected mice. Here we found enhanced recruitment of Siglec-F⁺ neutrophils to the lungs and nasal tissue after the B. pertussis challenge of mice immunized with the AIBP vaccine, and this was reversed in mice treated with anti-IL-17 and significantly reduced in mice treated with anti-CD4 (Fig. 6b).

Female 6–8-week-old C57BL/6 mice were immunized once by aerosol administration of the AIBP vaccine or PBS and challenged by aerosol with live B. pertussis at week 6. Groups of mice that had been immunized with the AIBP vaccine were treated with anti-CD4, anti-IL-17 or an isotype (Iso.) control antibody the day before and every 3 days after the challenge with B. pertussis. a, On the day of but before the challenge with live B. pertussis, CD4 cells were quantified in lungs and nasal tissue by flow cytometric analysis. b, On the day of but before the B. pertussis challenge, Siglec-F⁺ neutrophils were quantified in lung and nasal tissue cells by flow cytometry. Data in a and b are presented as the mean ± s.e.m. of biological replicates shown as individual symbols (n = 5). c, CFU counts were performed on lung and nasal tissue 2 h and 7 days, 14 days and 21 days post-challenge. Data are presented as the mean ± s.d. of biological replicates (n = 5). Data were analysed by two-way ANOVA followed by Tukey’s test for multiple comparisons. P values are shown above relevant datasets.

Source data

Protection against infection induced with a single dose of the AIBP vaccine was completely abrogated after depletion of CD4 T cells or neutralization of IL-17; the CFU counts in the anti-CD4-depleted mice were similar to those in non-immunized control mice (Fig. 6c).

Our findings show that the AIBP vaccine mediates sterilizing immunity largely via induction of IL-17-secreting CD4 T cells that promote recruitment of Siglec-F⁺ neutrophils to the respiratory tract.