Animal studies were conducted under the approval of the Institutional Ethics Committee of Second Affiliated Hospital of Air Force Medical University, using the recommendations from the Guide for the Care and Use of Laboratory Animals of the Institute (Approval No. TDLL-2016325).

M. tuberculosis H37Ra was obtained from the National Food and Drug Administration (China). M. tuberculosis was grown in Middlebrook 7H9 medium (BD, USA) supplemented with 10% oleic acid-albumin-dextrose-catalase (OADC) (BD, USA) and 0.05% Tween 80, or on 7H10 agar plates (BD, USA) supplemented with 10% OADC. Murine alveolar macrophage cell line MH-S was purchased from Procell Life Science & Technology Co., Ltd. (China). Female BALB/c and C57BL/6 mice were purchased from the Animal Center of Air Force Medical University.

The amino acid sequences of CnpB in M. tuberculosis (NP_217353) were acquired from GenBank of NCBI. IEDB (https://www.iedb.org) and BCPREDS (http://ailab-projects1.ist.psu.edu:8080/bcpred/) were used to predict the B-cell epitopes. The NetCTL (http://www.cbs.dtu.dk/services/NetCTL/) and NetMHCIIpan (http://www.cbs.dtu.dk/services/NetMHCIIpan) were used to predict the T-cell epitopes.

Open reading frames (ORFs) of full-length CnpB (1-336 aa), DHH (29-179 aa), and DHHA1 (275-333 aa) domains of CnpB were amplified using M. tuberculosis genomic DNA as a template and using primers listed in Table S1 . The DNA fragments were cloned into plasmid pET28a(+) as previously described, respectively (Pozzi et al., 2012; Yang et al., 2014; McCarthy et al., 2015). E. coli BL21 (DE3) strains harboring each recombinant plasmid were inoculated in LB broth, and the recombinant proteins of CnpB, DHH, and DHHA1 domains were purified using affinity chromatography according to previous work (Pozzi et al., 2012; Yang et al., 2014; McCarthy et al., 2015). Ag85B proteins were expressed and purified as previously described (Lu et al., 2018).

The enzyme activity was detected by high-performance liquid chromatography (HPLC) as previously described (Yang et al., 2014). Beforehand, reaction contained 50 mM Tris-HCl (pH 7.5), 1 mM MnCl₂, 10 mM NaCl, 1 mM c-di-AMP (Invivogen, France), and 3 μM purified protein CnpB. The mixtures were incubated for 1 h at 37 °C, terminated by adding 1 μL 0.5 M EDTA and diluted with 40 μL ddH₂O (Yang et al., 2014). Subsequently, 50 μL methanol was added and 20 μL of the mixture was uploaded into reverse-phase HPLC with a C18 column (250 × 4.6 mm, Vydac) as previously reported (Ryjenkov et al., 2005). Finally, different components were separated, in which nucleotides were monitored at a wavelength of 254 nm.

Mice sera and bronchoalveolar lavage fluids (BALFs) were collected for antibody detection using ELISA as previously reported (Ning et al., 2021). M. tuberculosis H37Rv infected sera samples from intravenously challenged mice and guinea pigs were obtained from our previous work (not published). TB patients sera were collected by the Tuberculosis Institute of Shannxi Province under patients’ informed consent (not published). For TB patients sera assay, the accuracy of detection was assessed by calculating the area under the receiver operating characteristic (ROC) curve. Diluted sera samples were added into CnpB, CnpB domain or Ag85B proteins coated microplates, and HRP-conjugated goat anti-mouse IgG (1:2 000, Zhongshan Co., Beijing, China) were used as the detection antibody. For immunoglobulin subclasses detection, HRP-conjugated goat anti-mouse IgG, IgG1, IgG2a, IgG2b, IgG3, IgM (1:5 000, InCellGenE LLC., Germany), or sIgA (1:2 000, Zhongshan Co., Beijing, China) were used as detection antibodies.

Bone marrow-derived macrophages (BMDMs) from female C57BL/6 mice were collected and cultured in RPMI 1640 supplemented with 15% fetal bovine serum (FBS), 100 U/mL Penicillin, 100 μg/mL Streptomycin, and 25% L929 culture supernatant. Murine alveolar macrophage MH-S and BMDM were seeded in 6-well plates at 1×10⁶ cells/well in RPMI 1640 medium supplemented with 10% FBS and incubated overnight at 37°C with 5% CO₂. Cells were then stimulated with different concentrations of endotoxin-removed CnpB protein for the time indicated in figure legends, and medium alone was used as control. Cells were collected at time points, and total RNA was extracted using Trizol and then quantified for qRT-PCR. For Western-blot analysis, cells were lysed by RIPA buffer (Solarbio, China) supplemented with protease inhibitor cocktail (Roche, Switzerland) and phosphatase inhibitor cocktail (EpiZyme, China) for total proteins extraction at indicated time points. Anti-LC3 antibody (Sigma, USA) and anti-NF-κB (Abcam, UK) were incubated as primary antibodies, and β-actin was used as the loading control.

MH-S cells were seeded at 1×10⁵ cells per well in a 24-well microplate. RFP-GFP-LC3 plasmids were transfected with lipofectamine 2000 (Invitrogen, USA) in a mass ratio of 2.5:1 without FBS for 4 h. Then the medium was replaced with a fresh medium containing 10% FBS for an additional 20 h. Cells were treated with serum-free medium or 5 μg/mL CnpB protein and examined by fluorescence microscopy. Cells were fixed with 4% paraformaldehyde, permeabilized with 0.5% Triton X-100 and blocked with 3% BSA. Cells were stained with Hoechst 33342 and observed under an Olympus fluorescence microscope.

1×10⁵ MH-S cells per well in 24-well microplate and 1×10⁵ BMDM cells per well in 96-well microplate were treated with CnpB for 12 h as described above. M. tuberculosis H37Ra was added to the cells with a multiplicity of infection (MOI) at 2:1 or 1:1 for MH-S or BMDM for 4 h. Then the extracellular bacteria were removed by washing the infected cells with sterile PBS three times, and this time point was marked as “0 h” post-infection. Macrophage cells were lysed with 0.025% SDS at indicated time points. Cell lysates were diluted and spread on 7H10 Middlebrook agar plates supplemented with OADC enrichment for bacteria colony forming units (CFU) counting, and results were presented as log₁₀ CFU.

Mice were immunized subcutaneously with 50 μg CnpB protein in PBS mixed with or without an equal volume of incomplete Freund’s adjuvant (IFA, Sigma) for three times at 2-week intervals. The antigen dose was halved during the third immunization. An equal volume of PBS was injected into mice as naïve control. As for Ag85B, DHH, and DHHA1 proteins vaccination, the inoculation doses, and the immunization schedule were the same as that of CnpB.

Four weeks after the third vaccination, mice were challenged intranasally (i.n.) with 2.5×10⁵ CFU of M. tuberculosis H37Ra in 50 μL PBS. The unimmunized mice (UN group) were only infected with the same dose of M. tuberculosis without immunization. The naïve group of mice was treated with an equal volume of PBS.

For quantitative determination of complements C3 and C5, mouse blood samples were collected, of which sera were separated by centrifugation at 4 °C. Sera were diluted to 1:2 500 and 1:200 for complement C3 and C5 assays respectively. Lung and spleen organ homogenates and BALF were assayed without dilution. Complement C3 and C5 were measured by the C3 ELISA kit (Alpha Diagnostic International, USA) and the C5 ELISA kit (Cloud-Clone Corp, USA) according to the manufacturer’s instructions.

For cytokines measurement, 1×10⁶ splenocytes were inoculated in 96-well plates with proteins of 5μg/mL CnpB and incubated at 37 °C with 5% CO₂ for 72 h, then supernatants were collected for Interferon-γ (IFN-γ) and Interleukin-10 (IL-10) cytokines detection with commercial ELISA kits (Mouse ELISA, eBioscience, USA).

Mice were sacrificed under anesthesia at 4 weeks post last vaccination or 8 weeks post-infection. Spleens were separated and single-cell splenocyte suspensions were prepared as our previous work (Ning et al., 2019). Splenocytes totalling 1×10⁶ were seeded in 96-well plates with stimulation of 5 μg/mL CnpB proteins or Ag85B proteins and incubated at 37 °C with 5% CO₂ for 72 h. Then 20 μL MTS (CellTiter 96^® AQueous One Solution Cell Proliferation Assay, Promega, USA) was added and incubated for 4 h. Finally, wells were measured at the absorbance of 490 nm (A₄₉₀) to calculate the stimulation index (SI). SI = (A₄₉₀ of the stimulated group – A₄₉₀ of blank control)/(A₄₉₀ of the negative group – A₄₉₀ of blank control).

For the CFSE assay (Lu et al., 2018), 10⁷ splenocytes were stained with 5 μM CFSE (CellTrace™ CFSE, Invitrogen, USA) at 37 °C for 20 minutes, and then stopped with 4 volumes of cold PRIM 1640 containing 10% FBS. After washing and resuspending, CFSE-labeled cells were seeded in 12-well microplates with 5 μg/well CnpB protein and incubated for 6 days. Cells were collected for flow cytometry analysis using a BD FACS Calibur cytometer. The flow cytometry data were analyzed using Modfit 5.0 software. The percentage of proliferated cells was analyzed statistically.

After mice were sacrificed, lungs were separated aseptically and stored in RNAlater^® Solution (Ambion, USA). Total RNA was extracted using Trizol reagent (Ambion, USA) and then quantified. For cell assay, MH-S cells were seeded at 1 × 10⁶ cells per well in 6-well plates and cultured at 37 °C with 5% CO₂ overnight. CnpB protein was added to the cell culture at different concentrations and cells were collected at time points. Total RNA was extracted using Trizol and then quantified as described above. qRT-PCR was carried out using primers listed in Table S1 for detecting the transcriptional levels of cytokines by Quantitative PCR Kit (Takara, JPN) according to the instructions of the manufacturer.

Four weeks post-vaccination or eight weeks post-infection, lungs were aseptically removed, cut into small pieces, and then digested in 3 mL digestion media (RPMI 1640 containing 50 μg/mL DNase I (Sigma, USA), 1 mg/mL Collagenase V (Sigma, USA), 5% fetal bovine serum, 100 U/mL Penicillin and 100 μg/mL Streptomycin (Solarbio, China) for 1 h at 37°C with 5% CO₂. Suspensions were passed through a 70 μm cell strainer to obtain a single cell suspension. Cells were pelleted by centrifugation and lysed with red blood cell lysis buffer. After being washed with RPMI 1640 medium, cells were resuspended and adjusted to a suitable density in complete RPMI 1640 medium.

Four weeks post-vaccination or eight weeks post-infection, single-cell suspensions of the spleen and lung were prepared according to our previous work (Ning et al., 2019). For different immune cell subsets analysis, 1.5×10⁶ cells were stained with Live/Dead Zombie NIR dye (BioLegend, USA). Fc receptors were blocked by anti-mouse CD16/32 (BioLegend, USA). Lung cells were stained with antibodies of BV510-anti-CD3, PE/Cy7-anti-CD4, PE-anti-CD8, FITC-anti-Ly6G, BV421-anti-CD19, PerCP-anti-CD49b, Alexa Fluor 700-anti-CD11b, and Alexa Fluor 647-anti-F4/80 (BioLegend, USA) for 30 min on ice and in darkness. Spleen cells were stained with BV510-anti-CD3, Alexa Fluor 700-anti-CD11b, Alexa Fluor 647-anti-F4/80, PE/Cy7-anti-CD69, PerCP-anti-CD49b, FITC-anti-CD19, PE/Cy7-anti-MHC II, PerCP-anti-CD86, and FITC-anti-CD80 (Abcam, UK).

For cytokines staining of T cells, 3×10⁶ splenocytes were seeded in 24-well microplates with 5 μg/mL CnpB proteins and incubated at 37 °C for 48 h (Lu et al., 2018). Protein transport inhibitor of Brefeldin A Solution (BioLegend, USA) was added for 12 h. Surface markers of PerCP-anti-CD4 and PE-anti-CD8 were stained as described above. Then, cells were permeabilized using Fixation/Permeabilization Kit (BD, USA), followed by cytokines staining of FITC-anti-IFN-γ, and APC-anti-IL-10 (BioLegend, USA). Cells were analyzed by flow cytometer (BD FACS Canto), and data were analyzed with FlowJo software (Treestar, USA).

Lung tissues of infected mice were fixed in formalin, then embedded in paraffin, and sectioned. For immunohistochemistry, anti-CD4 and CD8 antibodies (Abcam, UK) were used for the immunohistochemistry assay. The integrated optical density of CD4 or CD8 in each immunohistochemistry slide was calculated by Image-Pro Plus 6.0 software. For pathological analysis, lung tissue sections were stained with hematoxylin-eosin (HE). All the sections were observed under an optical microscope. The pathological changes (e.g., peribronchiolitis, perivasculitis, alveolitis, and granuloma formation) of the lung were scored as 0, 1, 2, 3, 4, or 5 for absent, minimal, slight, moderate, marked, or strong changes respectively, according to Dormans’ report (Dormans et al., 2004).

For CFU counting, lungs and spleens were homogenized, and homogenates were diluted and plated on 7H10 Middlebrook agar plates. Plates were incubated at 37 °C until colonies were visible. CFUs were counted and the results were expressed as log₁₀CFU per lung or per spleen.

Statistical analysis was performed using the software GraphPad Prism 5.0. One-Way ANOVA was performed for comparison of groups over two. Two-Way ANOVA was performed for comparison of groups in each treatment. T-test was performed for two groups comparison. P < 0.05 was considered a statistically significant difference. The data of the ROC curve were generated by SPSS Statistics 25.0.

Since the deletion of CnpB led to reduced virulence of M. tuberculosis (Yang et al., 2014; Dey et al., 2017), CnpB could be strongly taken considered as a prophylactic candidate vaccine antigen against M. tuberculosis. The prediction of allergen and B-cell and T-cell epitopes demonstrated that CnpB possesses antigenic epitopes ( Figure 2A and Tables S2 , S3 ), implying CnpB could be a potential antigen for TB vaccines like Ag85B and ESAT-6.

As is known, Ag85B is one of the strongest antigens of M. tuberculosis, and novel fusion proteins with Ag85B have entered clinical trials as TB vaccine candidates (Suliman et al., 2019). In this study, recombinant CnpB was purified by affinity chromatography as we previously reported (Yang et al., 2014) ( Figure 1A ). In addition, the phosphodiesterase activity of CnpB hydrolyzing c-di-AMP was confirmed by HPLC ( Figure 1B ), which was consistent with our previous report (Yang et al., 2014). We found that the levels of anti-CnpB antibodies in sera of M. tuberculosis infected mice were as high as those of anti-Ag85B, and gradually increased with the infection time and reached the highest level at 12 weeks post-infection ( Figure 1C ). Similarly, the levels of anti-CnpB antibodies in M. tuberculosis infected guinea pigs were even higher than those of anti-Ag85B antibodies ( Figure 1D ). In sera of TB patients, the levels of anti-CnpB antibodies were also elevated. The area under the receiver operating characteristic (ROC) curve (AUC) of CnpB was 0.58 ( Figure 1E ) which was not comparable with two classic antigens of M. tuberculosis CFP-10 and ESAT-6 (Luo et al., 2017). These data suggested that CnpB exhibited strong immunogenicity, but showed weak diagnostic potential.

CnpB of M. tuberculosis is a DHH-DHHA1 domain protein as Pde2, a c-di-AMP phosphodiesterase in Streptococcus pneumoniae (Bai et al., 2013). The N-terminal DHH and C-terminal DHHA1 domains consist of residues 29-179 aa and 275–333 aa, respectively (Yang et al., 2014). The DHH domain, having a five-parallel strand β-sheet that is sandwiched by 10 α-helices, has the phosphodiesterase activity catalytic core (Srivastav et al., 2014). The DHHA1 domain contributes to both the recognition and stabilization of substrates (He et al., 2016). By immunoinformatics analysis, we found that most B-cell and T-cell epitopes were distributed in the DHH domain and spacer sequence instead of the DHHA1 domain ( Figure 2A ). Then, two truncated proteins of DHH and DHHA1 were expressed and purified for immunogenicity evaluation in mice. It was shown that the DHH domain dominated the major immunogenicity of CnpB rather than the DHHA1 domain ( Figure 2B ), which was consistent with the epitopes prediction of CnpB ( Figure 2A ). These data suggested the immune response was mainly derived from the DHH domain and the spacer sequence.

Next, we evaluated the humoral immune response induced by full-length CnpB subcutaneous vaccination. The levels of anti-CnpB antibody in sera rose sharply and reach a plateau 4 weeks post 3rd immunization ( Figure 2C ). In sera of immunized mice, anti-CnpB and anti-Ag85B antibodies had comparable OD₄₅₀ detected by the same dose of antigens (10 μg/mL) at 1 : 6 400 dilution in ELISA assay. Meanwhile, the antibody titers of anti-CnpB and anti-Ag85B were both 1 : 102 400 (the ratio of OD_sample/OD_{negative control} was over 2.1) by serial dilutions. Therefore, we believe that CnpB is able to induce comparable antibody titers with Ag85B. IgG1 was the main antibody subclass ( Figure 2D ), and the ratio of IgG2a/IgG1 was significantly decreased in the CnpB group ( Figure 2E ), which represents a reduced Th1 balance (Paydarnia et al., 2020). Meanwhile, IgG levels in bronchoalveolar lavage fluid (BALF) were significantly elevated in CnpB subcutaneously immunized mice ( Figure 2F ). These data suggested that CnpB could trigger an antigen-specific humoral response at both systemic and local levels by subcutaneous inoculation.

Cellular immunity is important for protection against intracellular bacterium such as M. tuberculosis. CD69 is the early leukocyte activation antigen and is mainly expressed by activated T cells, B cells, and natural killer (NK) cells (Lauzurica et al., 2000). For splenocytes, CnpB stimulation in vitro up-regulated CD69 expression in NK cells, but not in T cells, B cells, and macrophages ( Figure 3A ). Besides, the expressions of the costimulatory molecules CD80/CD86 and MHC-II in macrophages were comparable between CnpB immunized and naïve mice ( Figure S2A ), implying that CnpB immunization had little effect on the antigen presentation of macrophages.

Splenocytes’ proliferative response is one of the indicators of cellular immune response in the host. Through MTS and CFSE assay, we found that CnpB immunization stimulated significant splenocyte proliferation ( Figures 3B, C ). Th1/Th2 immune responses play important roles in the protection against M. tuberculosis in the host (Abebe, 2019). T cells from PPD-positive TB patients could produce both IFN-γ and IL-10, and the proliferation levels of Th1 and Th2 types T cells were elevated (Harling et al., 2019). However, CnpB immunized mice failed to induce IFN-γ and IL-10 release from splenocytes ( Figure S2B ). Neither, CnpB vaccination didn’t alter the proportions of CD4⁺ and CD8⁺ T cells in the spleen ( Figure S2C ), and the percentages of CD4⁺ T cells secreting Th1/Th2 cytokines were not affected by CnpB immunization ( Figure S2D ). These indicated that CnpB immunization could induce splenocyte proliferation, but did not stimulate T cell activation, corresponding to the ratio of IgG2a/IgG1 ( Figure 2E ).

Next, we investigated the immune cell population in the lungs after CnpB inoculation. CnpB inoculation did not change the proportions of NK, macrophages and neutrophils ( Figure 3D ). To our surprise, CnpB immunization decreased the proportions of T and B cells in the lungs ( Figure 3D ). The proportions of CD4⁺ and CD8⁺ T cells, the main cellular immune cells against M. tuberculosis and other pathogens, were also reduced by CnpB immunization ( Figure 3E ). On the contrary, CnpB immunization up-regulated transcription of IFN-γ and IL-10 in the lungs ( Figure 3F ). These results suggested that CnpB may have different effects on cellular immunity in spleens and lungs, as well as innate immunity.

CnpB deleted M. tuberculosis strain induced STING-dependent pathway activation and type I interferon response in macrophages (Dey et al., 2017). Further, we investigated whether CnpB could affect the innate immune response of MH-S, a cell line of lung macrophages. IFN-β mRNA was significantly elevated after 6 h, lasting for 12 h by CnpB stimulation in MH-S, and returned to normal 24 h after treatment ( Figure 4A ). However, a high dose of CnpB (10 μg/mL) did not interfere with IFN-β transcription ( Figure 4A ). M. tuberculosis-derived c-di-AMP, a substrate of CnpB, induced significant IFN-β expression through activating STING in macrophages (Yang et al., 2014; Dey et al., 2017; Ning et al., 2019). However, in this study there was no infection in CnpB-treated macrophages, suggesting activation of the STING pathway is not mediated by c-di-AMP. Further, we found that CnpB did not affect the expression of NF-κB ( Figure S3 ), inferring regulation of IFN-β expression was independent of the STING/IKK/NF-kB pathway. In MH-S cells, a low dose of CnpB induced type I interferon response at a short time, but both low and high doses had little effect at a long time, also suggesting that CnpB may act on other innate immune mechanisms of macrophages.

Dey et al. (Dey et al., 2015) showed that overexpression of disA in M. tuberculosis to secrete excessive c-di-AMP enhanced the production of IFN-β and increased macrophages autophagy, and this strain showed reduced virulence in mice. However, it was reported that CnpB inhibited the innate immune cytosolic surveillance such as cGAS-cGAMP pathway in macrophages (Dey et al., 2017), and autophagy is generally believed to help fight against infection (Dey et al., 2015). Western-blot analysis revealed that the LC3-II levels of macrophages was significantly reduced at 12 h post-low-dose CnpB stimulation, and returned to normal 24 h post-stimulation ( Figure 4B ). The transcriptional levels of Atg5 and Atg7, the key molecules in the initial stage of autophagy, were drastically down-regulated 6 h post-CnpB stimulation ( Figures 4C, D ) and back to normal levels 12 h post-stimulation, in line with the reduced levels of downstream LC3-II at 12 h post-stimulation. In addition, MH-S cells were transfected with RFP-GFP-LC3 plasmid, followed by CnpB protein stimulation for 4 h, and cultured for 24 h. The GFP fluorescence showed a trend of inhibition with CnpB pre-treatment, but not the red fluorescent particles ( Figures 4E, F ). These results implied that CnpB negatively affected the autophagy of macrophages in a short time, and the reduced autophagy might negatively activate STING, thereby raising the expression of IFN-β in a short time ( Figure 4A ).

Further, MH-S and BMDM were treated with 5 μg/mL CnpB for 12 h and infected with M. tuberculosis H37Ra (MOI=1). After 4 h infection, M. tuberculosis CFUs showed no significant difference between the negative control and CnpB treatment group in MH-S ( Figure 4G ) and BMDM cells ( Figure 4H ). Besides, M. tuberculosis H37Ra was added to infect MH-S with a multiplicity of infection (MOI) at 2:1 ( Figure S4 ). The untreated cells showed a similar replication of M. tuberculosis as the CnpB group did, and no replication with time, suggesting that higher MOI may be the reason to restrict the replication in MH-S. In addition, CnpB treatment did not change the survival of M. tuberculosis in macrophages, which might be the combined results of IFN-β upregulation ( Figure 4A ) and autophagy inhibition ( Figure 4F ), suggesting innate immune response induced by recombinant CnpB protein is insufficient to restrict bacterial survival in macrophages.

Since that M. tuberculosis attenuated strain H37Ra and virulent strain H37Rv was capable to induce comparable levels of circulating IgG (Ning et al., 2017), we used H37Ra instead of H37Rv. Eight weeks after M. tuberculosis nasal challenge, anti-CnpB IgG, IgG1, IgG2a, and IgG2b increased significantly in sera of immunized mice ( Figure 5A ). The ratio of IgG2a/IgG1 showed an increasing tendency, implying Th1 response was induced (Mansury et al., 2019) ( Figure 5B ), which was different from previous results without infection ( Figure 2E ), suggesting CnpB immunization might reverse the IgG2a/IgG1 balance after M. tuberculosis infection.

Previous studies have reported that patients with active pulmonary TB elicited the elevation of levels of IgG1 and IgG3 subclasses in sera (Correa et al., 2019), which played important roles in complement activation (Jovic and Cymer, 2019). We found that the levels of complement C3 and C5 decreased slightly in sera of CnpB immunized mice ( Figure 5C ), but levels of complement C3 in homogenates of lungs and spleens remained unchanged ( Figure 5D ), suggesting that high levels of IgG subclasses might activate and deplete complement system through the classical pathway. Meanwhile, although CnpB subcutaneous immunization did not induce sIgA production ( Figure 2F ), M. tuberculosis intranasal challenge caused a significant increase of anti-CnpB sIgA secretion in BALF of CnpB immunized mice ( Figure 5E ). Thus, CnpB subcutaneous immunization could induce a significant mucosal immune response when infected with M. tuberculosis by the respiratory tract.

It is known that M. tuberculosis infection could suppress cellular immune response through multiple mechanisms for its survival (Both et al., 2018). The splenocyte proliferation was not changed in CnpB immunized mice compared to M. tuberculosis intranasal infection mice ( Figures 6A, B ). Meantime, the levels of secreted IFN-γ and IL-10 were not significantly changed compared with unimmunized mice ( Figure 6C ). As well, the Th1 (CD4⁺ IFN-γ⁺) and Th2 (CD4⁺ IL-10⁺) cells proportions were not significantly altered compared to naïve mice ( Figure 6D ), indicating that CnpB immunization might not stimulate cellular immune responses in spleens after M. tuberculosis intranasal infection.

However, in the total cells of the lungs, the transcriptional levels of IFN-γ and IL-10 (Th1/Th2 cytokines) of CnpB immunized mice were significantly higher than those of unimmunized mice after M. tuberculosis infection ( Figure 6E ). The immunohistochemistry showed that CnpB immunization could recruit CD4⁺ and CD8⁺ T cells to the interstitium of the lung after M. tuberculosis infection ( Figure 6F ), corresponding to the results of increased IgG2a/IgG1 ratio reflecting Th1 balance ( Figure 5B ). These results implied that CnpB immunization mainly induced the cellular immune response in lungs rather than in spleens after M. tuberculosis infection, and CnpB might perform a different immune mechanism compared to other vaccine antigens such as Ag85B.

Inflammation is a basic pathological process during M. tuberculosis infection. After M. tuberculosis infection, the spleens of mice were enlarged by naked eyes, but not alleviated in CnpB immunized mice. However, in the lungs HE stained sections, inflammatory cell infiltration was significantly reduced in CnpB immunized mice, and the alveolar structure intact was maintained compared to those of unimmunized mice ( Figures 7A, B ). Compared to unimmunized mice, the M. tuberculosis CFUs in lungs decreased significantly in CnpB immunized mice infected with M. tuberculosis ( Figure 7C ), while the CFUs in spleens showed no significant difference ( Figure 7D ) (n = 6). These results were consistent with the results that CnpB induced cellular immune response in lungs rather than in spleens, implying CnpB performed effective protection against M. tuberculosis respiratory tract infection.