Rv0309 antiserum development in mice and artificial infection were performed strictly according to the Guidance for the Use and Care of Laboratory Animals, Hubei Province, China. The protocols were approved by the Ethics Committee of Huazhong Agricultural University (protocol no. HZAUMO-2018-027).

M. smegmatis mc²155 (NC_008596.1) and M. bovis BCG-Pasteur (ATCC:35734) were a gift from Professor Luiz Bermudez from Oregon State University. All strains were cultured in Middlebrook 7H9 broth (BD, MD, USA) containing 0.5% glycerol (Sigma, MO, USA), 10% oleic acid-albumin-dextrose-catalase (OADC) and (BD) 0.05% Tween 80 (Sigma) or on Middlebrook 7H11 agar plates (BD, MD, USA) containing 0.5% glycerol (Sigma) and 10% OADC (BD). Before infection, optical densities at 600 nm (OD₆₀₀) of bacterial cultures were adjusted to the required multiplicity of infection (MOI) referring to the standard turbidimetric card. Then, the cultures were centrifuged at 3,000 × g for 10 min. The precipitated bacteria were resuspended in a medium and dispersed by passage through an insulin syringe. Next, 50 μL of 10-fold serially diluted bacterial suspension was plated onto Middlebrook 7H11 agar (BD) to count viable bacteria (colony-forming units, CFUs).

RAW264.7 cells were cultured in Dulbecco’s modified Eagle’s medium (Gibco, NY, USA) containing 10% (v/v) fetal bovine serum (Gibco) and 100 mg/mL streptomycin and 100 IU/mL penicillin at 37°C in an atmosphere of 5% CO₂.

Anti-β-actin antibody (IgG) (60008-1-lg, ProteinTech, IL, USA) was obtained from ProteinTech. Rabbit monoclonal antibodies to JNK2 (56G8) (#9258), phospho-SAPK/JNK (Thr183/Tyr185) (98F2) (#4671), phospho-NF-κB p65 (Ser536) (93H1) (#3033), p38 MAPK (#9212), phospho-p38 MAPK (Thr180/Tyr182) (#9211), p44/42 MAPK (Erk1/2) (137F5) (#4695), and phospho-p44/42 MAPK (Erk1/2) (Thr202/Tyr204) (D13.14.4E) XP^® (#4370) and mouse monoclonal antibodies to NF-κB subunit p65 (L8F6) (#6956), NF-κB IκBα (L35A5) (#4814), and phospho-NF-κB IκBα (Ser32/36) (5A5) (#9246) were purchased from Cell Signaling Technology (Cell Signaling Technology, MA, USA).

The Aliphatic index and grand average of hydropathicity index (GRAVY) value of the Rv0309 were evaluated using ProtParam (https://web.expasy.org/protparam/) (27). Transmembrane structure forecasting was performed at TMHMM web server (http://www.cbs.dtu.dk/services/TMHMM/) (28). Protein homologs to Rv0309 in mycobacterium were identified with Mycobrowser (https://mycobrowser.epfl.ch) (29). Homologous sequence alignment was conducted online with ESPript web server (https://espript.ibcp.fr/ESPript/cgi-bin/ESPript.cgi) and Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/) (30, 31). To validate the presence of a conserved domain in the Rv0309 and its homologs, the resulting six protein sequences were subjected to analysis using the NCBI CDD search (https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) (32). The conserved motifs in these protein sequences were examined by the MEME suite motif search tool (http://memesuite.org/) (33).

The full-length rv0309 gene (gene ID: 886574, NC_000962.3: 377931-378587) was amplified from M. tb H37Rv genomic DNA, using specific primers ( Table 1 ). The target gene was cloned into the pMV261 vector to generate pMV261-Rv0309. pMV261-Rv0309 was electroporated into M. smegmatis mc²155 to generate a recombinant strain, Ms_rv0309. Briefly, mc²155 were electroporated in the presence of 2 μg of pMV261-Rv0309 plasmid DNA with a Gene Pulser (Bio-Rad, USA). The conditions of electroporation were 200 μl volume, 2.5 kV, 25 μF and 1000 Ω, with a 0.2-cm-gap electroporation cuvette. M. smegmatis transformed with empty pMV261, designated as Ms_Vec, was used as a control. The constructs were confirmed by colony PCR and western blot assay. The sequencing of the colony PCR product was outsourced to TSINGKE Biological Technology (Wuhan, China). A 657-bp band was amplified from the Ms_rv0309 strain. The specific primers listed in Table 1 were used for identifying the rv0309 gene. Ms_rv0309 and Ms_Vec were cultured until an OD₆₀₀ of 0.6. The cells were pelleted and resuspended in lysis buffer (0.1 M PBS, 1 mM phenylmethylsulfonyl fluoride) for ultrasonic lysis (250W, 5s on/5s off, lasting for 25 min). The whole- cell lysates were subjected to western blot assay for detecting the expression of Rv0309 using mouse antiserum to rRv0309, which was prepared and stored at our laboratory.

The Ms_rv0309 strain was cultured in Middlebrook 7H9 broth (BD) containing 10% OADC (Sigma) and 0.05% Tween-80 until an OD₆₀₀ of 0.6. Rv0309 expression was induced in a water bath at 45°C for 1 h. Ms_rv0309 bacteria were centrifuged at 3,000 × g for 10 min and subsequently resuspended in phosphate-buffered saline (PBS) and lysed by ultrasonication (300W, 5s on/5s off, lasting for 30 min). Recombinant rRv0309 was purified using Ni-NTA agarose chromatography (Qiagen, Hilden, Germany) and verified using 12% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). The purified rRv0309 protein was stored at –80°C until use (6).

The BCGΔRS01790 mutant was constructed as described previously, with some modification (34). Cosmid p0004s and phAE159 vectors were kindly offered by Professor Jiaoyu Deng from Wuhan Institute of Virology, Chinese Academy of Sciences. The Phasmid harboring allelic exchange substrates (AES) including flanking sequences of BCG_RS01790 was electroporated into M. smegmatis mc²155 to get the specialized transducing phages (STPs) from transduced M. smegmatis and then used STPs to disrupt BCG_RS01790. Briefly, flanking genomic regions of BCG_RS01790, 817 bp upstream and 699 bp downstream, were amplified and inserted into p0004s at the Van91I site to generate p0004s-L+R. The recombinant cosmid phAE159-p0004s-L+R was obtained by ligating p0004s-L+R with phAE159 and was transformed into Escherichia coli HB101 through an in-vitro λ packaging reaction with commercial MaxPlax Lambda Packaging Extract (Epicentre Biotechnologies, WI, USA). Phasmid DNA was extracted from confirmed hygromycin-resistant E. coli transductants using Plasmid Maxi Kit (Omega, United States). The temperature-sensitive shuttle phasmid DNA was transduced into M. smegmatis mc²155, which was cultured at 30°C to generate transgenic phages. As previously reported, the phage transduction rate was optimal at an MOI of 10 (35). BCG was infected with the recombinant STPs at 37°C to generate BCGΔRS01790 mutant by integrated phage-specific DNA into BCG genome, which was verified with PCR by using the primers RS01790LYZ and RS01790RYZ ( Table 1 ) and sequencing. BCGΔRS01790 was distinguished from the wild-type BCG based on hygromycin resistance. In the mutant, the hygromycin resistance cassette replaced the target gene BCG_RS01790 and expected PCR product sizes for WT BCG (~2.5 kb) and BCGΔRS01790 (~5.5 kb) were obtained.

To construct the complement strain cBCGΔRS01790::RS01790, pMV261-rv0309 was electroporated into BCGΔRS01790, and positive colonies were screened on 7H11 plates containing hygromycin (75 μg/mL) and kanamycin (50 μg/mL). Rv0309 expression in the complement strain was confirmed using PCR and immunoblotting. Briefly, BCGΔRS01790, cBCGΔRS01790::RS01790, and WT BCG were cultured until an OD₆₀₀ of 0.6. The cells were pelleted and resuspended in lysis buffer (0.1 M PBS, 1 mM phenylmethylsulfonyl fluoride) for ultrasonic lysis (250W, 5s on/5s off, lasting for 30 min). Then whole cell lysate was subjected to western blot assay for detecting the absence of Rv0309 using mouse antiserum to rRv0309. An anti-GroEL2 antibody prepared in our laboratory was used to detect the cytoplasmic marker protein GroEL2 of BCG.

To examine growth patterns, OD₆₀₀ values of triplicate cultures of BCG strains (WT BCG, BCGΔRS01790, and cBCGΔRS01790::RS01790) and M. smegmatis strains (Ms_rv0309 and Ms_Vec) were adjusted to 0.2. Then, OD₆₀₀ values were determined every 3 days for 60 days for BCG and every 4 h for 5 days for M. smegmatis. Growth curves were plotted based on the average OD₆₀₀ values.

The pictures of colonies for each strain were taken with VHX-5000 microscope with a super-wide depth of field and the circularity and diameter of 10 independent colonies of each strain were determined using ImageJ software, according to the developer’s instructions (36, 37). Circularity was calculated as 4π × area/perimeter²; a value of 1.0 indicates a perfect circle, and a decrease in the circularity value indicates a less circular colony. For scanning electron microscopy (SEM) observation of the bacteria, Ms_rv0309, Ms_Vec, WT BCG, BCGΔRS01790, and cBCGΔRS01790::RS01790 were cultured until an OD₆₀₀ of 0.6. The bacterial pellets were resuspended in 0.1 M phosphate buffer (pH 7.2) after centrifugation, and then put on poly-L-lysine-coated slides (Thermo Fisher Scientific, Rochester, NY). The bacteria were fixed with 2.5% glutaraldehyde (Solarbio, Beijing, China) at 4°C for 2 hours, washed 3 times with 0.1 M phosphate buffer (pH 7.2), and then postfixed with 1% osmium tetroxide (Sigma-Aldrich, MO, USA) for 1 hour at room temperature. After being dehydrated in a graded ethanol series and permuted with 3-methyl butyl acetate, samples were placed in the critical point dryer for drying (Leica EM CPD300, IL, United States). SEM was performed on a VEGA3 TESCAN instrument (Brno-Kohoutovice, Czech Republic) using an accelerating voltage of 20 kV. The bacterial cell length and width of 20 bacteria of each strain were determined using ImageJ software.

The subcellular localization of Rv0309 was determined using previously reported methods (38). Briefly, Ms_rv0309, Ms_Vec, and WT BCG were cultured until an OD₆₀₀ of 0.6. The cells were pelleted and resuspended in lysis buffer (0.1 M PBS, 1 mM phenylmethylsulfonyl fluoride) for ultrasonic lysis (250W, 5s on/5s off, lasting for 30 min). The lysates were centrifuged at 3,000 × g, 4°C for 5 min, and the supernatants were ultracentrifuged at 30,000 × g, 4°C for 30 min. After ultracentrifugation, the supernatants (cell membrane and cytoplasmic fractions) and pellets (cell wall fraction) were collected separately, and the pellets were resuspended in PBS. Equal amounts of pellets and supernatants were subjected to western blotting to determine Rv0309 expression. An anti-GroEL2 antibody prepared in our laboratory was used to detect the cytoplasmic marker protein GroEL2 of M. smegmatis and BCG. An anti-Ag85A antibody, a gift from Professor Xiang Chen from Yangzhou University, was used to detect the cell wall-associated protein Ag85A of BCG (39).

Since Rv0309/BCG_RS01790 contains a YkuD_like superfamily domain with a L,D-transpeptidase catalytic site which has been shown to be related to the hinge of bacterial cell wall peptidoglycans (26), it was speculated that expression of Rv0309/BCG_RS01790 should change the permeability of the bacterial cell wall. To confirm the effect of Rv0309/BCG_RS01790 on cell wall permeability, the permeability of the recombinant M. smegmatis and BCG was determined using a reported method (40). Briefly, Ms_rv0309, Ms_Vec, WT BCG, BCGΔRS01790, and cBCGΔRS01790::RS01790 were cultured with and without the antibiotics hygromycin until an OD₆₀₀ of 0.6 and then, the cultures were centrifuged at 3,000 × g for 10 min. The bacterial pellets were washed with PBS containing 0.05% Tween-80 three times and then, the cells were resuspended in uptake buffer (5 mM MgSO₄, 50 mM KH₂PO₄) and diluted to an OD₆₀₀ of 0.5. Glucose solution (25 mM) was added into the suspension, and the mixture was incubated to pre-energize the strains at room temperature (RT) for 5 min. Then, 200 μL of the bacterial solutions were added to a black, clear-bottomed 96-well microplate, and ethidium bromide (EB) was added to each well at a final concentration of 20 μM. The fluorescence intensity of EB was determined at 5-min intervals for 1 h, using a microplate reader (BMG-Labtech, Offenburg, Germany), with excitation and emission wavelengths of 530 nm and 590 nm, respectively.

RAW264.7 cells were seeded in 12-well plates (1 × 10⁶ cells/well) for 12 h before infection. The cells were infected with Ms_rv0309, Ms_Vec, WT BCG, BCGΔRS01790, and cBCGΔRS01790::RS01790 at an MOI of 10:1 for 4 h (defined as –4 h). Then, the infected cells were washed with PBS three times to remove extracellular bacteria (referred to as 0 h). The infected cells were further cultured in complete medium supplemented with 100 μg/mL gentamicin and lysed with 0.025% (v/v) SDS at 0, 2, 4, 8, and 24 h post-infection (hpi) for M. smegmatis and at 0, 2, 4, 8, 24, and 48 hpi for BCG strains. After 10-fold serial dilution, the lysates were plated onto 7H11 agar plates containing 10% OADC. Colonies were counted 7 days after plating for M. smegmatis strains and 21 days after plating for BCG strains. The CFU/mL was calculated for each strain.

RAW264.7 cells were infected with Ms_rv0309, Ms_Vec, BCGΔRS01790, cBCGΔRS01790::RS01790, and WT BCG at an MOI of 10 for quantitative reverse-transcription PCR (RT-qPCR). At indicated time points, total cellular RNA was extracted using TRIzol reagent (Invitrogen, CA, USA) and was reverse-transcribed into cDNA using HiScript Reverse Transcriptase (Vazyme, Nanjing, China). mRNA levels of IL-1β, IL-6, and TNF-α were detected on a ViiA7 Real-time PCR System (Applied Biosystems, CA, USA) using SYBR Green Master Mix (Vazyme) and were quantified using the 2^–ΔΔCt method. The primers used are listed in Table 1 .

To detect the production of IL-6, TNF-α, and IL-1β, culture supernatants of infected macrophages were collected and analyzed using ELISA kits (Neobioscience, Shenzhen, China) according to the manufacturer’s protocol. To investigate the effect of recombinant protein rRv0309 derived from Ms_rv0309 on cytokine production, RAW264.7 cells were stimulated with rRv0309 protein at various concentrations (5, 10, and 50 μg/mL) and LPS (1 μg/mL) for 18 h (13). Then, cell culture supernatants were harvested to detect IL-6 levels using an ELISA kit (Neobioscience). Each sample was analyzed in triplicate.

RAW264.7 cells were infected with Ms_rv0309, Ms_Vec, BCGΔRS01790, and WT BCG (MOI = 10) and lysed at indicated time points. The lysates were centrifuged at 12,000 × g, 4°C for 10 min. Proteins were separated by SDS-PAGE and transferred to polyvinyl difluoride membranes, which were blocked with 5% bovine serum albumin in TBST. The membranes were incubated overnight with antibodies against non-phosphorylated and phosphorylated JNK, ERK1/2, p38, IκBα, and p65. β-Actin was used as a control. Then, the membranes were washed and incubated with horseradish peroxidase (HRP)-conjugated goat anti-rabbit (1:2000) or anti-mouse IgG (1:2000) (SouthernBiotech, AL, USA) at 25°C for 1 h. Protein signals were detected using WesternBright ECL HRP substrate (Advansta, CA, USA) per the manufacturer’s instructions.

All cell signaling pathway inhibitors were purchased from Sigma-Aldrich (Shanghai, China), dissolved in sterile dimethyl sulphoxide (DMSO) (Sigma-Aldrich, Shanghai, China) and utilized at varying concentrations. The cells were pretreated with 20 mM PDTC (NF-κB inhibitor), 10 mM U0126 (ERK1/2 inhibitor), 20 mM SP600125 (JNK inhibitor), and 20 mM SB202190 (p38 inhibitor) for 1 hour before being infected with BCGΔRS01790. The 0.1% (v/v) DMSO was used as the vehicle control (41). After a 12 infection, the cell supernatants were harvested and subjected to cytokine measurements with the commercial ELISA kits.

In total, 120 6-week-old female C57BL/6 mice weighing 20 ± 2 g were purchased from the Laboratory Animal Center of Huazhong Agricultural University and were randomly divided into four groups of 30 mice each. The mice were infected intratracheally with WT BCG, BCGΔRS01790, or cBCGΔRS01790::RS01790 at a dose of 2 × 10⁶ CFU/mouse in 25 μL of PBS, or mock-infected with an equal amount of PBS. At 0, 2, 4, 8, 16, and 21days post-infection (dpi), five mice were weighed, euthanized, and sampled, and sera were collected to detect cytokine levels.

The lungs were collected for bacterial load enumeration and histopathological and immunohistochemical examinations. For enumeration of the lung bacteria, left lung tissues were homogenized. The homogenates were serially diluted and plated onto 7H11 agar plates. Colonies were counted after incubation at 37°C for 14–21 days. For histopathological and immunohistochemical examinations, right lung lobes were fixed in 10% neutralized formalin, embedded in paraffin, and cut to 4-mm sections. Histopathological analysis was performed through conventional hematoxylin and eosin staining (6). Inflammatory cells, including neutrophils, lymphocytes, and monocytes, were counted manually by an experienced pathologist based on morphological criteria in 10 random fields in one random slide for each lung (37). These were assessed in a blinded fashion by the pathologist. The total numbers of the different cell types were compared between the groups. The levels of IFN-γ, TNF-α, IL-4, and IL-17 in the lung tissues were determined by immunohistochemical staining using relevant antibodies. Positive brown signals were quantified using Image-Pro Plus 6.0 (IPP6) software (Media Cybernetics) in five random fields in one random slide for each lung and were expressed as integrated optical density (IOD) and were compared between the groups (42).

In addition, total RNA was extracted from the spleens using TRIzol reagent (Invitrogen) per the manufacturer’s instructions and used for RT-qPCR.

All assays were conducted in triplicate, and data are expressed as the mean ± standard error of the mean. All experiments were repeated three times independently. GraphPad Prism 7.0 (La Jolla) was used for statistical analysis. A two-tailed unpaired t-test with Welch’s correction was used for comparison of two groups, and one-way or two-way ANOVA followed by the LSD test was used for comparison of multiple groups. Statistical significance is expressed at four levels: not significant (ns, p > 0.05), *p < 0.05, **p < 0.01, and ***p < 0.001. Gray-scale values for western blot bands were analyzed using ImageJ (National Institutes of Health).

Bioinformatics analysis was used to identify the conservation and function of Rv0309. As shown in Figure 1A , Rv0309 protein contained rich aliphatic amino acids (aliphatic index = 78.30) and the grand average of hydropathicity (GRAVY) value of Rv0309 was 0.146, suggesting that Rv0309 had strong lipophilicity, which might be with transmembrane domain residues and/or with the membrane lipids ( Figure 1A ). Furthermore, a transmembrane structure was predicted in Rv0309 (TMHs = 1) ( Figure 1B ), indicating that this protein was a transmembrane protein, this result also suggested that Rv0309 might interact with extracellular substances. Homology analysis showed that Rv0309 was variably conserved in mycobacteria. Compared to Rv0309 in M. tb, the Mb_0317 in M. bovis, or BCG_RS01790 in BCG have an identity of 100%. However, the homolog MSMEG_0635 in M. smegmatis, MMAR_0559 in M. marinum, ML2522c in M. leprae have the identity of 72%, 78%, 88%, respectively ( Figure 1C ). Conserved domain analysis revealed that these proteins contain a YkuD_like superfamily domain related to mycobacterial cell wall synthesis ( Figure 1D ). Furthermore, Rv0309 and its homologs belonging to YkuD_like superfamily shared multiple conserved motifs with high homology in Mycobacteria ( Figure 1D ), indicating that this protein is vital to the physiological process of Mycobacteria and its potential function is worthy of deep investigation.

Ms_rv0309 and BCGΔRS01790 strains were confirmed with PCR ( Supplementary Figure 1 ). The expression of Rv0309 was identified in Ms_rv0309, WT BCG, and the complement strain cBCGΔRS01790::RS01790, with a band size of 22 kDa using Western blot assay, whereas Rv0309/BCG_ RS01790 was absent in Ms_Vec and BCGΔRS01790 mutant ( Figure 2A, B ). The effect of Rv0309 on colony morphology was further investigated. For M. smegmatis strains, Ms_rv0309 generated more rugose and thicker colonies than the Ms_Vec strain ( Figure 2C , upper panel). For BCG strains, BCGΔRS01790 had an obviously thinner and less rugose colonies than WT BCG and cBCGΔRS01790::RS01790 strains ( Figure 2C , lower panel) which were significantly larger and more like a circle compared with BCGΔRS01790 ( Figure 2D ). To further explore the alterations of colony size, we conducted an SEM image analysis on the bacterial size and the results showed that the proportion of bacteria with shorter length (3.5μm, the median of bacterial length of BCG strains) in the images of BCGΔRS01790 strain is 60%, while that of WT BCG only 10% showing significant difference between them (p<0.01) ( Figure 2E ). The widths did not differ among the three BCG strains. In addition, no difference in bacterial length and width was observed between Ms_rv0309 and Ms_Vec ( Figure 2E ). Furtherly, the growth curve was also determined, but showed no significant difference in neither M. smegmatis strains nor BCG strains ( Figure 2F ).

These results suggested that Rv0309 can increase the formation of Mycobacterium wrinkle and help BCG to grow bigger but didn’t influence the growth rate of Mycobacterium.

To determine the subcellular location of Rv0309, Ms_rv0309, and WT BCG cells were disrupted by sonication, and cell wall and cell membrane/cytoplasmic fractions were extracted and analyzed by western blotting. Results showed that Rv0309/BCG_RS01790 protein was solely present in the cell wall of Ms_rv0309 and WT BCG, not in the cytoplasm or cell membrane ( Figures 3A, B ). As expected, the cytoplasmic protein GroEL2 of M. smegmatis and BCG, evaluated as a positive control, was detected in cytoplasm of M. smegmatis and BCG. As a positive control for cell wall fraction, Ag85A was detected only in the cell wall fraction of BCG.

To explore possible Rv0309-induced alterations in the cell wall architecture, a permeability assay was conducted. At 1 h after EB treatment, EB accumulation was significantly lower in Ms_rv0309 cells than in Ms_Vec cells ( Figure 3C ), and the intracellular EB accumulation of WT BCG and cBCGΔRS01790::RS01790 was significantly lower than BCGΔRS01790 ( Figure 3D ). These results indicated that Rv0309 can reduce cell wall permeability.

To confirm the role of Rv0309 in intracellular mycobacterial survival, RAW264.7 cells were infected with Ms_rv0309, Ms_Vec, BCGΔRS01790, WT BCG, and cBCGΔRS01790::RS01790 at an MOI of 10:1. A plate counting assay of intracellular mycobacteria showed that the intracellular bacterial amount (CFU/mL) was significantly higher for Ms_rv0309 than for Ms_Vec at 2, 4, 8, and 24 hpi ( Figure 4A ). Similarly, intracellular bacterial survival of BCGΔRS01790 at 4, 8, 24, and 48 hpi were significantly decreased, whereas that of cBCGΔRS01790::RS01790 was recovered to the levels observed for the WT BCG ( Figure 4B ). In addition, compared to 0 hpi, the amount of Ms_Vec (CFU/ml) sharply decreased by 55% at 2 hpi but was not completely cleared, maintained at a stable status through 4 hpi, then continuously decreased by 77% at 8 hpi and 89% at 24 hpi. Although the amount of MS_rv0309 decreased at a similar trend, it kept a slow and constant reduction within the first 8 hpi with the decreasing rate of 20%, 24% and 27% at 2, 4 and 8 hpi respectively, then a higher decreasing rate of 61% at 24 hpi. A similar effect of BCG_RS01790 on BCG survival was observed in BCG strains, but the significant decrease of BCG strains occurred at 4 hpi, 2 h behind of M. smegmatis. These data indicated that Rv0309/BCG_RS01790 significantly enhances the intracellular survival of M. smegmatis and BCG in macrophages.

To reveal the mechanism by which Rv0309 promotes intracellular bacterial survival, cytokines induced by LPS, and bacterially infected models were determined. Data showed that after induced by LPS, 10 μg/mL (p < 0.01) and 50 μg/mL (p < 0.001) rRv0309 can significantly decrease the production of IL-6 ( Figure 4C ). The same trend occurred in both M. smegmatis- and BCG- infected RAW264.7 cells. Cells infected with Ms_rv0309 exhibited significantly lower levels of IL-6 ( Figure 4D ). BCG_ΔRS01790 increased the production of IL-6 compared with WT BCG and cBCGΔRS01790::RS01790 ( Figure 4G ). Besides IL-6, Ms_rv0309 also exhibited the production of IL-1β and TNF-α from 4 hpi (p < 0.05) and 8 hpi (p < 0.001), respectively ( Figures 4E, F ). And BCGΔRS01790 increased the expression level of IL-1β and TNF-α from 4 hpi (p < 0.01) compared with WT BCG and complement strain cBCGΔRS01790::RS01790 ( Figures 4H, I ).

The transcription level of IL-6, IL-1β, TNF-α effected by Rv0309 in RAW264.7 cells was analyzed using RT-qPCR and were in agreement with the ELISA data ( Supplementary Figure 2 ).

Taken together, these results indicated that Rv0309 can inhibit the production of pro-inflammatory cytokines including IL-6, IL-1β, and TNF-α.

To further elucidate the mechanism by which Rv0309 inhibits cytokine production, we examined critical signaling molecules, including MAPK JNK, MAPK ERK, MAPK p38, NF-κB p65, and IκBα in RAW264.7 cells infected with Ms_rv0309, Ms_Vec, WT BCG, and BCGΔRS01790 by western blotting. The phosphorylation of IκBα and NF-κB p65 (from 0 hpi onwards, i.e., the very beginning of the observation) was significantly inhibited in Ms_rv0309-infected cells compared to Ms_Vec-infected cells ( Figures 5A, C ). After Ms_rv0309 infection, in the MAPK signal pathway, the phosphorylation of ERK was decreased, and that of JNK was decreased at 4, 8, and 24 hpi ( Figures 5A, C ). However, p38 phosphorylation showed no significant difference between the Ms_Vec control group and Ms_rv0309 treatment group. Similar results were also obtained from the immunoblotting analysis of RAW264.7 cells infected with different BCG strains including BCGΔRS01790 and WT BCG. The phosphorylation of IκBα (from 4 hpi onwards) and NF-κB p65 (from 0 hpi onwards) were enhanced by BCGΔRS01790 infection compared to WT BCG infected cells ( Figures 5B, D ). In the MAPK signal pathway, the BCGΔRS01790 infection stimulated significantly increased phosphorylation of ERK and JNK ( Figures 5B, D ). However, the phosphorylation of p38 did not differ between BCGΔRS01790 and WT BCG-infected cells. To further confirm the above signaling molecules were engaged in Rv0309 action, RAW264.7 cells were pretreated with the inhibitors PDTC (NF-κB inhibitor), U0126 (ERK1/2 inhibitor), SP600125 (JNK inhibitor), and SB202190 (p38 inhibitor) for 1 hour before being infected with BCGΔRS01790. As a result, the inhibitors PDTC, U0126, and SP600125 treatment in BCGΔRS01790-infected cells exhibited significant inhibition of IL-6, IL-1β, and TNF-α production and release ( Figure 5E ). However, SB202190 did not display any significant inhibitory effect on the cytokine production in the RAW264.7 cells infected by BCGΔRS01790. Thus, Rv0309 regulates cytokine secretion mainly by inhibiting the phosphorylation of NF-κB p65/IκBα and MAPK JNK/ERK signaling pathways.

The role of Rv0309 in suppressing inflammatory responses was further evaluated in C57BL/6 mice. In general, lung damage aggravated over time. WT BCG and cBCGΔRS01790::RS01790 caused more severe lung damage than BCGΔRS01790. Histopathological changes in lung tissues were observed as of 8 dpi and mainly manifested as immune cell infiltration, alveolar wall thickening, and the fusion of alveolar cavities, which ultimately led to the collapse of the lung tissue structure ( Figures 6A, B ). Most cells in the tissue sections were identified as neutrophils, and lymphocytes and monocytes were also identified. In the early stage (8 dpi), neutrophils and lymphocytes were significantly more abundant in mice infected with BCGΔRS01790 than in mice infected with the other two strains ( Figures 6C, D ) (p < 0.05), whereas in the late stage of the experiment, opposite trends were observed. In addition, the number of monocytes remained low in the early stage but increased in the late stage of the experiment (16 and 21 dpi). Similarly, monocyte numbers were significantly lower in the BCGΔRS01790 infection group than in the other two infection groups at both 16 and 21 dpi ( Figure 6E ).

Next, lung bacterial loads were examined and the bacterial loads of all three strains were significantly decreased at 8 dpi. Notably, in BCGΔRS01790-infected mice, the lung bacterial loads sharply decreased and reached the minimum levels at 8 and 16 dpi compared with that in mice infected with WT BCG or cBCGΔRS01790::RS01790 ( Figure 7A ).

Cytokine production was examined at both the mRNA and protein levels. In general, the BCGΔRS01790 infection group exhibited significantly higher serum levels of IFN-γ, TNF-α, and IL-1β than WT BCG and cBCGΔRS01790::RS01790 groups before 8 dpi. However, after 8 dpi, the cytokine concentrations of IFN-γ, TNF-α, and IL-1β in BCGΔRS01790 infection group were lower or similar compared with the other two infection groups at one or more time points ( Figures 7B–G ).

Immunohistological analysis of lung tissues revealed that the levels of IFN-γ and TNF-α, which are representative Th1 cytokines, were significantly higher in the BCGΔRS01790 infection group than in the WT BCG and cBCGΔRS01790::RS01790 infection groups at 4 and 8 dpi, and they were significantly lower after 16 dpi ( Figures 8A, B ). The level of IL-4, a Th2 hallmark cytokine, was significantly lower in the BCGΔRS01790 infection than in the other two groups throughout the infection period ( Figure 8C ). The IL-17A level was significantly increased in all three infection groups compared to the mock-infected group, without a significant difference among the infection groups ( Figure 8D ).

In addition, we examined the body and organ weights of the mice after infection. Compared with mice infected with WT BCG or cBCGΔRS01790::RS01790, mice infected with BCGΔRS01790 gained significantly more weight in late stage of the experiment. ( Supplementary Figure 3A ) (p < 0.01). After the mice were euthanized, we recorded the weights of the lungs, spleens, and livers and calculated the organ-to-body weight ratios. We observed no differences in the ratios among the bacterial infection and mock infection groups (p > 0.05) ( Supplementary Figures 3B – D ).