Specific pathogen-free (SPF) 4- to 5-week-old male and female BALB/c mice were purchased from Beijing Charles River Laboratory (Beijing, China). All mice used in this study were treated in accordance with the Guide for the Care and Use of Laboratory Animals of the People’s Republic of China. All protocols were reviewed and approved by the Committee on Ethics of the Institute of Medical Biology, Chinese Academy of Medical Sciences (IMBCAMS; assurance number: DWSP202106004). Animals were bred and maintained under SPF conditions at IMBCAMS at a constant temperature (20–24 °C) and humidity (45–65%), with lighting on a fixed light/dark cycle (12 h/12 h).

The B. pertussis strain B.p-L1 used in this study was recently isolated from a patient in Yunnan Province, China. Total DNA was extracted from the recovered B. pertussis, which was characterized as carrying the ptxP3 genotype by DNA sequencing of the pertussis toxin promoter (ptxP) (30). For B. pertussis infection experiments, bacteria were grown on Bordet-Gengou agar (B-G) plates (Hopebio) containing 20% defibrinated sheep blood (Nanjinglezhen) for 24 to 48 h at 37°C. Colonies from fresh B-G plates were resuspended in phosphate-buffered solution (PBS), diluted to a concentration of 10¹¹ CFU/mL by using a turbidimetric method, and used within 2 h of preparation. For culture of bacteria from tissues, Regan-Lowe plates prepared from Regan-Lowe charcoal agar base (Oxoid) supplemented with 10% defibrinated sheep blood and 40 μg/mL cephalexin (Oxoid) were used.

BMDCs were isolated from mouse bone marrow cells as previously described (31). Briefly, bone marrow cells were isolated and cultured in RPMI 1640 medium containing 20 ng/mL recombinant GM-CSF (Peprotech), 100 U/mL penicillin, 100 µg/mL streptomycin, and 10% FBS. Petri dishes containing 2 × 10⁶ cells in 10 mL were incubated at 37°C in 5% CO₂. At day 3 of incubation, an additional 10 mL of fresh complete medium containing GM-CSF was added, and 10 mL of medium was replaced with fresh medium supplemented with GM-CSF on day 6. Immature BMDCs were collected on day 7 for experiments. The BMDCs were treated with 2′,3′-cGAMP (5 μg/mL, In vivoGen, California, USA) or c-di-GMP (5 μg/mL, In vivoGen, California, USA) in vitro for 24 h. Cells were treated with lipopolysaccharide (LPS, 1 μg/mL, Sigma–Aldrich) as a positive control or sterile PBS as a negative control. Cytokine (IFN-β and TNF-α) levels in the supernatant were quantified by ELISA. The stimulated BMDCs were stained with the specific antibodies described below and analyzed by using FlowJo software.

BALB/c mice (half of which were male and half of which were female, 4–5 weeks old) were immunized via the IN (20 μL volume) or intraperitoneal (IP, 200 μl volume) route three times at three-week intervals with the different tested vaccines (Table 1). Antigens of the aP vaccine were produced by the IMBCAMS under good manufacturing practice conditions (32). All vaccinations were performed under anesthesia (isoflurane). Blood, nasal washes (NWs), and bronchoalveolar lavage fluid (BALF) were collected two weeks after the last immunization. Blood was collected and centrifuged at 860 g for 10 min to obtain plasma. NW and BALF samples were obtained by washing the nasal cavity and lung with 0.2 mL and 1 ml of cold PBS containing protease inhibitor, respectively. Nasal cavity and lung wash fluids were centrifuged at 2400 g for 10 min, and the supernatants were collected. Plasma, NW, and BALF samples were stored frozen at -20°C until the detection of antibodies (Figure 1).

Spleen tissues were minced, filtered through a 70-μm nylon mesh (BD Biosciences), and diluted 1:1 in RPMI 1640. The mixture was loaded onto a Ficoll-Paque (GE, USA) layer and centrifuged at 1500 rpm for 30 minutes at room temperature. For monocellular isolation from lungs, tissues were minced and digested with collagenase D (1 mg/mL; Roche) and DNase I (20 U/mL; Roche) for 45 min at 37°C on a shaker. Next, tissues were passed through a 70-μm cell strainer, washed twice in complete RPMI 1640 medium, mixed with 40% Percoll, loaded onto a 70% Percoll layer and centrifuged at 1800 rpm for 30 minutes at room temperature. The cells obtained from organs were washed twice with RPMI 1640 and/or resuspended in complete cell culture medium (RPMI 1640 supplemented with 10% (V/V) fetal calf serum (FCS) and a 1% (v/v) premixed penicillin–streptomycin solution). The cells were used for cytokine detection, ELISpot, and/or flow cytometry.

Microplates (96-well) were coated with PT, FHA, or PRN at 3 μg/mL and incubated at 4°C overnight. Then, the plates were blocked with 3% (w/v) bovine serum albumin (BSA, Abcam) in PBS at 37°C for 2 h. Diluted serum, NW, or BALF sample was added to each microplate and incubated at 37°C for 1 h. After washing, a horseradish peroxidase (HRP)-labeled sheep anti-mouse IgG (Jackson ImmunoResearch, USA), anti-mouse IgG1 (Southern Biotech, USA), anti-IgG2a (Southern Biotech, USA), or anti-mouse IgA (Southern Biotech, USA) antibody was added to the microplate, and the plate was incubated at 37°C for 1 h. All the ELISA plates were developed using tetramethylbenzidine (TMB; Solarbio, CHN) to generate a colorimetric reaction, and the reaction was terminated with 2 mmol/L H₂SO₄. The absorbance of the plates at 450 nm was read. Endpoint titers were determined as the dilution that exhibited an optical density exceeding ≥2.1 times the background level (secondary antibody alone).

T cell detection by IFN-γ, IL-4 and IL-17 ELISpot assays was performed according to the manufacturer’s instructions (Cellular Technology Limited, USA). Briefly, precoated 96-well plates were seeded with 5 μg/mL specific stimulants and 4 × 10⁵ mouse splenocytes or pneumonocytes in a total volume of 100 μL and incubated. Following a 24-h incubation at 37°C with 5% CO₂, the cells were removed, and an anti-IFN-γ, anti-IL-4 or anti-IL-17 detection antibody was added, followed by the addition of streptavidin-ALP. For antigen-specific IgA or IgG plasma cell detection, MultiScreen filter 96-well plates (Millipore, USA) were precoated with PT, FHA, or PRN (each 5 μg) for each well overnight at 4°C. After rinsing with PBS, the plates were blocked with culture medium for 30 min at room temperature. Single-cell suspensions of splenocytes or pneumonocytes in culture medium were added to the coated plates and incubated at 37°C with 5% CO₂ for 24 h. After washing with PBS, the plates were incubated with biotinylated anti-IgA or anti-IgG antibodies (Southern Biotech, USA) followed by incubation with streptavidin-conjugated horseradish peroxidase (Jackson ImmunoResearch, USA), each for 1 h at room temperature. After additional washes with PBS, AEC substrate solution (BD Bioscience, USA) was added for spot development. The reaction was stopped by rinsing with water. Spots were developed using BCIP/NBT and analyzed by a Cellular Technology Limited (CTL) ELISpot reader.

Cytokine levels in the supernatant of cultured cells were measured by an ELISA kit according to the manufacturer’s recommendation. Values were calculated based on a standard curve of recombinant cytokines. The results are expressed as picograms per milliliter (pg/mL). IFN-β (Cat# VAL612, Novus Biologicals) and TNF-α (Cat# VAL609, Novus Biologicals) levels in the supernatant of BMDCs were quantified. For detecting the cytokines level in the supernatant of cultured splenic or pulmonary lymphocytes, cells were cultured at a concentration of 2 × 10⁶/ml and stimulated with the antigens PT (2 μg/ml), FHA (2 μg/ml), and PRN (2 μg/ml). Supernatants were removed after 3 days and stored at -20°C before testing. IFN-γ (Cat# VAL607, Novus Biologicals), TNF-a (Cat# VAL609, Novus Biologicals), and IL-2 (Cat# VAL602, Novus Biologicals) levels were measured for Th1 responses; IL-5 (Cat# KA0253, Novus Biologicals) and IL-6 (Cat# VAL604, Novus Biologicals) levels were detected for Th2 responses; and IL-17A (Cat# VAL610, Novus Biologicals) and IL-22 (Cat# M2200, R&D Systems) levels were tested for Th17 responses (33, 34).

Single-cell suspensions were obtained from BMDCs or lung tissues. To detect BMDC maturation, cells were blocked with anti-mouse CD16/32 antibodies and stained for surface markers with anti-CD11c (Biolegend, clone: N418, Cat# 117338), anti-CD80 (Biolegend, clone: 16-10A1, Cat# 104729), anti-CD86 (eBioscience, clone: GL1, Cat# 25-0862-82), anti-CD40 (BD Bioscience, clone: 3/23, Cat# 562846), and anti- major histocompatibility complex molecule class II (MHC II) antibodies (I-A/I-E, eBioscience, clone: M5/114.15.2, Cat# 12-5321-8282). To discriminate circulating cells from lung-resident cells, intravascular staining was performed as previously described (31). In brief, mice were intravenously (i.v.) delivered with 3 μg of PE-labeled anti-CD45 antibody (eBioscience, clone: 30-F11, Cat# 12-0451-82) and sacrificed 10 min after i.v. injection, and lungs were isolated immediately to obtain a single-cell suspension as described. For the detection of lung tissue-resident memory T (TRM) cells, lung cells were stained for cell surface markers with anti-CD3 (Biolegend, clone: 17A2, Cat# 100216), anti-CD4 (Biolegend, clone: GK1.5, Cat# 100406), anti-CD8 (Biolegend, clone: 53-6.7, Cat# 100752), anti-CD44 (Biolegend, IM7, Cat# 103040), anti-CD69 (eBioscience, clone: H1.2F3, Cat# 25-0691-82), and anti-CD62L antibodies (Biolegend, clone: MEL-14, Cat# 104412). Dead cells were excluded by 7-AAD staining (BD Bioscience). All samples were assessed on a flow cytometer (Beckman, USA), and data were analyzed with FlowJo software (TreeStar).

The nasal mucosa was collected on day 14 after the third immunization and homogenized with TRIzol reagent (Invitrogen, CA), and total RNA was isolated with chloroform/isopropanol, followed by purification with a RNeasy Mini Kit (Qiagen). For library and sequencing, the mRNA was isolated and purified from total RNA via Oligo(dT)-attached magnetic beads. Subsequently, purified mRNA was fragmented into small pieces with fragment buffer. Then first-strand cDNA was generated using random hexamer-primed reverse transcription, followed by a second-strand cDNA synthesis. Afterwards, A-Tailing Mix and RNA Index Adapters were added by incubating to end repair. The obtained cDNA fragments were amplified by PCR, and the products were purified by Ampure XP Beads and validated on the Agilent Technologies 2100 bioanalyzer for quality control. The double stranded PCR products were heated, denatured, and circularized by the splint oligo sequence to obtain the final library. The single strand circle DNA was formatted as the final library. The final library was amplified with phi29 to make DNA nanoballs (DNBs) which had more than 300 copies of one molecular, DNBs were loaded into the patterned nanoarray and pair end 100 bases reads were generated on BGIseq500 platform. The raw data were filtered by SOAPnuke software (35). The clean data were mapped on the reference Mus-musculus_GRCm38.p6 with hierarchical indexing for spliced alignment of transcripts (HISAT) software (36). Differentially expressed genes (DEGs) were identified and log2 transformed with DEseq2 (37). The resulting p values were adjusted to the Q value (adjusted p value) using the method of multiple testing adjustment with the R package (https://bioconductor.org/packages/release/bioc/html/qvalue.html). The DEGs identified according to the absolute value of log2(fold change) ≥1 and Q value (adjusted P value) ≤ 0.05. The enriched Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways of the DEGs were analyzed by ClusterProfiler in RStudio. Heatmaps were drawn with the pheatmap R package.

Three weeks after the third immunization, all immunized mice were challenged with strain B.p-L1 via aerosol exposure using an aerosolization apparatus (38). Animals were infected via the challenge chamber for 30 min. Within the 30 min period, the air sample was removed from the sampling port at 5, 10, 20, and 30 min for assessment of the concentration of B. pertussis inside the chamber. The mock-infected animals received PBS aerosol exposure. At the indicated timepoints (2, 4, 7, 11, 16, and 23 days post infection), the lung, trachea, and nasal turbinate were harvested from mice to measure the bacterial burden (Figure 1).

Total RNA from lung tissues was extracted using TRIzol reagent (Invitrogen, USA) according to the manufacturer’s recommendations. Total RNA concentration and quality were measured by using a Thermo Scientific Varioskan Flash (Thermo Fisher Scientific, USA). cDNA was synthesized from 1 µg of RNA with oligo-dT primers and a PrimeScriptTM RT kit (Accurate Biotechnology, CHN). Cytokine mRNA levels were determined by RT–PCR performed on a LightCycler 96 system (Applied Biosystems, USA) using gene-specific primers (Supplementary Table 1) and a SYBR Green Premix Pro Taq HS qPCR Kit (Accurate Biotechnology, CHN). The expression of the housekeeping gene GAPDH was quantified in parallel for RNA normalization. The relative expression of the target genes was calculated by the ΔΔCt method.

For histopathologic analysis, lung tissues from necropsied mice were fixed in 10% neutral buffered formalin, embedded in paraffin, and sectioned at 3-5 μm. Then, the sections were stained with hematoxylin and eosin (H&E) after dehydration. The pathological sections were observed and photographed under a microscope (Leica, Germany).

The results are presented as the means ± SEMs or GMTs and their 95% confidence intervals (CIs). For the data in accordance with normal distribution, one-way ANOVA followed by Tukey’s multiple comparisons test was used to compare the difference, while for the data not in accordance with normal distribution, Kruskal-Wallis followed by Dunn’s multiple comparisons test was used to compare the difference. A value of P < 0.05 was considered significant. Statistical analysis and plots were performed with the Prism version 8.0 (GraphPad Software, Inc.).

We studied the maturation efficacy of BMDCs by evaluating the BMDC maturation markers MHC II and the costimulatory molecules CD86, CD80, and CD40. In comparison, c-di-GMP treatment resulted in significantly higher expression of both MHC II and costimulatory molecules (CD86, CD80, and CD40) than 2’,3’-cGAMP, as revealed by the enhanced mean fluorescence intensity (MFI) (Figures 2A, B). Additionally, we measured the cytokine (IFN-β and TNF-α) levels secreted from BMDCs treated with LPS, PBS, 2’,3’-cGAMP, or c-di-GMP. The PBS-treated BMDCs showed undetectable amounts of IFN-β in the culture supernatant. Interestingly, compared with 2’,3’-cGAMP treatment, c-di-GMP treatment enhanced IFN-β secretion (Figure 2C). Moreover, compared with 2’,3’-cGAMP treatment, c-di-GMP treatment caused a slight increase in TNF-α production; however, the difference did not reach statistical significance (Figure 2D). Overall, these results indicated that c-di-GMP could better promote the production of the cytokines IFN-β and TNF-α and stimulate DC maturation than 2’,3’-cGAMP.

To evaluate the systemic and mucosal humoral immune responses induced by different vaccines (Table 1), PT-, FHA-, and PRN-specific antibodies in serum, nasal washes (NW), and bronchoalveolar lavage fluid (BALF) were detected using ELISA method, and splenocytes and pneumonocytes producing antigen-specific IgA or IgG were detected using the ELISpot method at 14 days after the third immunization.

In serum, the aP+c-di-GMP group vaccinated via IN administration (aP+c-di-GMP/IN) showed comparable levels of PT-, FHA-, and PRN-specific IgG as the aP+Al(OH)₃ group vaccinated via IP administration (aP+Al(OH)₃/IP), but both groups showed higher PT-specific IgG than the group vaccinated with aP vaccine without any adjuvant (aP/IN) or aP+2’,3’-cGAMP via IN immunization (aP+2’,3’-cGAMP/IN) (Figure 3A). In addition, the FHA- and PRN-specific IgG of the aP+c-di-GMP/IN group was not significantly different from that of the aP-2’,3’-cGAMP group but was significantly higher than that of the aP/IN group (Figure 3A). The aP+c-di-GMP/IN treatment also elicited higher serum PT-, FHA-, and PRN-specific IgA antibody levels than the aP+Al(OH)₃/IP treatment and higher PRN-specific IgA antibody levels than the aP/IN treatment (Supplementary Figure 1A). Regarding antibody subclasses (IgG1 and IgG2a), aP+c-di-GMP/IN-immunized mice showed higher serum PT-specific IgG1 than aP/IN-immunized mice but comparable FHA- and PRN- specific IgG1 than aP+Al(OH)₃/IP-immunized mice (Supplementary Figures 1B–D). In contrast, compared to aP+Al(OH)₃/IP and aP/IN treatment, aP+c-di-GMP/IN treatment induced a significant increase in the levels of IgG2a against PT and PRN, and there was no significant difference in the levels of IgG2a against PT, FHA, and PRN between the two CDN-adjuvanted aP vaccine groups (Supplementary Figures 1B–D). These data suggested that the c-di-GMP-adjuvanted aP vaccine delivered via IN administration induced a balanced Th1 and Th2 immune response.

Regarding nasal washes (NW), aP+c-di-GMP/IN-vaccinated mice showed higher PT, FHA, and PRN-specific IgA levels than aP/IN-vaccinated mice and higher FHA and PRN-specific IgA levels than aP+2’,3’-cGAMP/IN-vaccinated mice, while there were no differences between the aP/IN and aP+2’,3’-cGAMP/IN groups (Figure 3B). Regarding BALF, aP+c-di-GMP/IN-vaccinated mice also showed the higher levels of IgA against PT, FHA, and PRN than aP/IN-vaccinated mice, but there was no significant difference in anti-PT, FHA, and PRN IgA levels in BALF between the two CDN-adjuvanted groups (Figure 3C).

For the antibody-secreting splenocytes and pneumonocytes producing IgA or IgG, the number of B cells secreting PT-, FHA-, and PRN-specific IgG in splenocytes from the alum-adjuvanted aP group was higher than that from the aP/IN and aP+2’,3’-cGAMP/IN groups, but it showed no difference from the aP+c-di-GMP/IN group (Figure 3D). However, in the pneumonocytes, both CDNs adjuvanted with aP vaccines produced a slight increase in the number of IgG-producing B cells (Supplementary Figure 1E). Of note, a higher frequency of B cells secreting anti-PT, anti-FHA, or anti-PRN IgA was observed in splenocytes from the aP+c-di-GMP/IN group than in those from all other groups, with a 4.5-fold higher frequency than in the aP+2’,3’-cGAMP/IN group (Figure 3E). And a higher frequency of B cells secreting anti-PT, anti-FHA, or anti-PRN IgA was also observed in pneumonocytes from the aP+c-di-GMP/IN group than in those from aP/IN group (Figure 3F). Although the frequency of B cells secreting anti-PT, anti-FHA, or anti-PRN IgA in pneumonocytes did not show statistical difference between the two CDN-adjuvanted groups, aP+c-di-GMP/IN-immunized mice caused a slight increase with 3.7-fold higher frequency than in the aP+2’,3’-cGAMP/IN group (Figure 3F). Taken together, these results suggest that aP+c-di-GMP treatment can elicit stronger systemic and mucosal humoral responses than 2’,3’-cGAMP.

The B. pertussis-specific cellular immune responses were assessed in vitro by exposing splenic and pulmonary lymphocytes from immunized mice to the specific antigens PT, FHA, and PRN. Regarding frequencies of IFN-γ secreting T cells, aP+c-di-GMP/IN-treated mice showed increased frequency, but there was no significant difference compared to mice in other groups in splenic lymphocytes (Figure 4A), while the frequency was significantly higher than aP/IN-immunized mice, reaching a mean response of 542 spot-forming cells (SFC) per 4×10⁵ input splenocytes (Figure 4D). For frequencies of IL-17-secreting T cells, aP+c-di-GMP/IN-treated mice showed higher frequencies than mice in aP/IN and aP+Al(OH)₃/IP groups and comparable frequencies to aP+2’,3’-cGAMP/IN-treated mice in splenic lymphocytes and comparable frequencies to aP+2’,3’-cGAMP/IN- and aP/IN-treated mice but higher than aP+Al(OH)₃/IP- treated mice in pulmonary lymphocytes (Figures 4B, E). For IL-4-secreting T cells, the aP+c-di-GMP/IN group showed levels comparable to those in the aP+Al(OH)₃/IP group but higher levels than the aP/IN and aP+2’,3’-cGAMP/IN groups in the spleen (Figure 4C), while the level increased but did not show significant differences from that in all other vaccine groups in pulmonary lymphocytes (Figure 4F).

To further examine whether TRM cells were induced by CDN-adjuvanted aP vaccines, flow cytometry was performed on lung tissues from mice (the TRM cell gating strategy is shown in Supplementary Figure 2). Not surprisingly, IP administration of the aP+Al(OH)₃/IP vaccine did not generate an increased population of lung TRM cells (Figures 4G, H). IN immunization with aP vaccine alone also failed to promote the population of lung TRM cells (Figures 4G, H), indicating that mucosal immunization alone is not sufficient to induce lung TRM cells. In contrast, mice immunized with aP+2’,3’-cGAMP/IN or aP+c-di-GMP/IN showed significantly promotion of the population of lung TRMs (Figures 4G, H). Together, these data suggested that IN immunization with both CDN-adjuvanted aP vaccines promoted the generation of CD4+ TRM populations.

After incubation with PT, FHA, and PRN for 3 days, cultures of splenic and pulmonary lymphocytes isolated after the third immunization were tested for cytokines in the supernatant. In culture supernatants of splenic lymphocytes, the levels of cytokines associated with Th1 responses (IFN-γ, TNF-α, and IL-2) and Th17 responses (IL-22) were significantly increased in the aP+c-di-GMP/IN group compared with those in the aP+Al(OH)₃/IP or aP/IN group (Figures 5A, B). The aP+c-di-GMP/IN treatment also induced a slight increase in Th1 (IFN-γ, TNF-α, and IL-2) and Th17 (IL-17A and IL-22) related cytokine levels than aP+2′, 3′-cGAMP/IN treatment, but there was no significant difference (Figures 5A, B). Compared with aP/IN and aP+c-di-GMP/IN groups, the aP+Al(OH)₃/IP group produced the higher IL-5, which is a Th2-related cytokine, in the supernatant of splenic lymphocyte cultures, while there was no significant difference in IL-6 levels between the aP+Al(OH)₃/IP and aP+c-di-GMP/IN groups (Figure 5C).

In culture supernatants of pulmonary lymphocytes, similar to the results for splenic lymphocytes, compared with aP+Al(OH)₃/IP or aP/IN treatment, aP+c-di-GMP/IN treatment greatly promoted the secretion of Th1 and Th17 response-related cytokines by pulmonary lymphocytes (Figures 5D, E). And aP+c-di-GMP/IN treatment also induced a slight increase in Th1 (IFN-γ, TNF-α, and IL-2) and Th17 (IL-17A and IL-22) related cytokine levels than aP+2′, 3′-cGAMP treatment, but there was no significant difference (Figures 5D, E). Similarly, the alum-adjuvanted aP vaccine induced higher IL-5 but not IL-6 levels than all other treatments, and the difference between the 2′, 3′-cGAMP/IN and aP+c-di-GMP/IN treatments was not significant (Figure 5F). Overall, these results indicated that IN delivering c-di-GMP-adjuvanted aP vaccine generates balanced Th1, Th2, and Th17 responses in both the spleen and lungs and that this effect is better than that with the 2′, 3′-cGAMP-adjuvanted aP vaccine.

RNA-seq analysis of the nasal mucosa in all immunized mice revealed 44 (7 upregulated, 37 downregulated), 42 (37 upregulated, 5 downregulated), 561 (407 upregulated, 154 downregulated), and 2076 (860 upregulated, 1216 downregulated) DEGs (|fold change| ≥2) in aP+Al(OH)₃/IP-, aP/IN-, aP+2’,3’-cGAMP/IN-, and aP+c-di-GMP/IN-immunized mice, respectively, compared with PBS (IN)-immunized mice (Figure 6A).

The KEGG pathway functional enrichment results showed 20 top significant pathways (Figure 6B), among which Th1 and Th2 cell differentiation, Th17 cell differentiation, and IgA production signaling pathways were observed. Using a heatmap, 59 genes involved in these three signaling pathways were identified. Th1 and Th2 cell differentiation-, Th17 cell differentiation-, and IgA production signaling pathway-related genes were noticeably upregulated in aP+c-di-GMP-immunized mice (Figure 6C).

To assess protective efficacy against pertussis, immunized mice were challenged with aerosolized B. pertussis, and the bacterial colony-forming units (CFU) in nasal, tracheal, and lung homogenates were counted at the indicated times. aP+c-di-GMP/IN immunization provided a high level of protection against lung infection with B. pertussis, resulting undetectable bacterial colonization at 4 days post infection (dpi) with the lowest level of the areas under the bacterial clearance curve (AUC, 0.7195) among all groups (Figures 7A, B). When compared with those in PBS-immunized mice, the CFU counts in the lungs after B. pertussis aerosol challenge were significantly reduced in mice IP immunized with aP+Al(OH)₃, resulting in undetectable CFU counts at 7 dpi and an AUC of 1.385, but the reduction was not as rapid as that seen in mice IN immunized with the aP+c-di-GMP vaccine (Figures 7A, B). Additionally, there was a significant reduction in the number of CFU in the lungs from aP (IN)- and aP+2’,3’-cGAMP (IN)-immunized mice at 4 dpi; however, these changes were not observed at 2 dpi, and the AUC of each was very similar (aP/IN: 6.209; aP+2’,3’-cGAMP/IN: 5.739) (Figures 7A, B).

In the trachea, B. pertussis colonization was not detected at any indicated time in the aP+di-GMP/IN group (Figures 7C, D). aP/IN- and aP+2’,3’-cGAMP/IN-immunized mice showed very low bacterial burdens at 2 dpi and 4 dpi, with AUCs of 0.30 and 0.24, respectively (Figures 7C, D). Surprisingly, aP+Al(OH)₃/IP group mice did not exhibit detectable CFU counts in the trachea at 2 dpi; however, rebound was observed at 4 dpi, the numbers peaked at 7 dpi, and the bacteria were completely cleared at 23 dpi (Figures 7C, D).

Regarding nasal homogenate, PBS-immunized mice were heavily colonized at 2 days post B. pertussis challenge and reached the highest level, 1.0 × 10⁶ CFU/mL, at 7 dpi; the number of colonies gradually decreased until 23 dpi, with an AUC of 16.36 (Figures 7E, F). In contrast, mice immunized with aP+c-di-GMP/IN had very low CFU counts at 2 dpi and 4 dpi, and bacteria were completely cleared by day 7, with an AUC of 1.20 (Figures 7E, F). aP+2’,3’-cGAMP/IN treatment also resulted in protection in the nose, although not as effectively as that seen with aP+c-di-GMP/IN treatment, with an AUC of 4.191 (Figures 7E, F). Intriguingly, aP+Al(OH)₃/IP treatment did not confer any protection against B. pertussis colonization of the mouse nose, resulting in a steady level of nasal colonization at 10³ CFU/mL through at least 23 dpi (Figures 7E, F). In contrast, the PBS-immunized mice completely cleared the nasal infection by that time. Together, these results indicated that IN immunization with the aP+c-di-GMP/IN vaccine enabled accelerated bacterial clearance in the respiratory tract following B. pertussis infection compared to that in aP+Al(OH)₃/IP- or aP-2’,3’-cGAMP/IN-immunized animals.

We evaluated the pathological changes in the different vaccination groups after B. pertussis challenge. We observed lung histopathological changes by hematoxylin & eosin (H&E) staining at 4 and 7 dpi. Large amounts of inflammatory cell infiltration, especially neutrophils (blue arrows), lung interstitial thickening, and severe bronchial obstruction, were observed in the lungs of PBS-immunized mice after challenge (Figure 8A). A small quantity of inflammatory cells and relatively slight bronchial obstruction were also observed in lungs from aP/IN- or AP+2’,3’-cGAMP/IN-immunized mice (Figure 8A). Both aP+Al(OH)₃/IP- and aP+c-di-GMP/IN-immunized mice showed less evidence of histopathological lesions in the lungs, especially no bronchial obstruction (Figure 8A). Interestingly, eosinophil infiltration was observed in mice IP immunized with the alum-adjuvanted aP vaccine, which was not found in other groups. And bronchial-associated lymphoid tissue (BALT) hyperplasia was observed in three groups of mice IN immunized vaccines, and with the c-di-GMP-adjuvanted aP vaccine most obviously, which was not found in control mice or mice IP immunized with the alum-adjuvanted aP vaccine (Figure 8A and Supplementary Figure 3).

Moreover, mRNA levels of cytokines in the lung tissues on 2 and 4 dpi were detected by RT-qPCR. The production of proinflammatory cytokines (IL-1β, TNF-α, and IL-6) in the lung tissues of B. pertussis-infected mice was expressed as a fold change compared to that in lung tissues from mock-infected mice (Figure 8B). At 2 dpi, PBS-immunized mice showed induction of only IL-1β levels in the lung tissues. However, the mRNA levels of IL-1β, TNF-α, and IL-6 were significantly increased at 4 dpi and were 14-, 5- and 28-fold higher than the levels in mock-infected lung tissues, respectively (Figure 8C). As expected, the levels of IL-1β, TNF-α, and IL-6 from mice immunized with aP+Al(OH)₃/IP did not show obvious changes at 2 dpi and 4 dpi compared with those from mock-infected mice (Figures 8B, C). Interestingly, both CDN-adjuvanted groups showed higher TNF-α production than the mock-infected group at the two detected time points (Figures 8B, C). The difference was that the aP/IN and aP + 2’,3’-cGAMP/IN groups showed a slight increase and indistinguishable IL-6 production at 2 dpi compared with the mock-infected group, while no increase was observed in the aP+c-di-GMP/IN group (Figures 8B, C). We also analyzed the Th cell bias in the lungs of immunized mice at 2 dpi and 4 dpi by detecting the mRNA levels of IFN-γ, IL-17A, and IL-5, which are indicators of Th1, Th17, and Th2 activation, respectively. Mice immunized with aP+c-di-GMP/IN showed the higher levels of IL-17A and IFN-γ gene expression at 2 dpi than PBS- and aP+Al(OH)₃/IP- immunized mice (Figure 8D). The aP/IN and aP+2’,3’-cGAMP/IN groups showed moderate upregulation of IL-17A gene expression at 2 dpi compared with that of the mock-infected group (Figure 8D). However, mice immunized with aP+Al(OH)₃/IP did not show obvious changes in IL-17A and IFN-γ gene expression at 2 dpi and 4 dpi compared with that of mock-infected mice (Figures 8D, E), while they showed a significant increase in IL-5 expression at 2 dpi (Figure 8D).