ESAT6 gene of M. tuberculosis strain H37Rv with human optimized codon was synthesized and cloned into vaccinia virus delivery vector PVVT1-C7L (PVVT1-C7L-Tpa-esat6) which contains tPA gene for secretion of intracellular signal peptide. Sfi1 restriction enzyme was used for cloning. PVVT1-C7L-Tpa-esat6 was transformed to E. coli DH5 competent cells for amplification. The expression of ESAT6 gene was confirmed by PCR using the following primers; 5′-TTT GAA GCA TTG GAA GCA ACT-3′ (VVTK-F) and 5′-ACGTTGAAATGTCCCATCGACT-3′ (VVTK-R).

Vero cells in 12-well plates were infected with vaccinia virus (KCCM11574P) at a multiplicity of infection (MOI) of 0.02 for 2 h, and the infected Vero cells were transfected with PVVT1-C7L-Tpa-esat6 plasmid using Lipofectamine 2000 (Thermo Fisher Scientific, Waltham, MA, USA) transfection reagent for 4 h. Vero cells were incubated for 3–4 days to observe the cytopathic effects, and recombinant viruses were obtained by plaque isolation. For high efficacy and purity, recombinant vaccinia virus expressing ESAT6 (vacESAT6) was concentrated by ultracentrifugation. The expression of ESAT6 protein by Vero cells and isolated B cells after transduction with vacESAT6 was confirmed by confocal microscopy (Figures 1A,B, Supplementary Figure 1).

B220⁺ cells were magnetically purified from splenocytes of naïve C57BL/6 mice using CD45R/B220 biotin (BD biosciences, California, USA) and anti-biotin microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer's instructions. Labeled cells were purified through LS column (Miltenyi Biotec, Bergisch Gladbach, Germany). Isolated B220⁺ cells (2 × 10⁷ cells seeded) were transduced with vacESAT6 (MOI of 1) for 2 h and then loaded with αGC (Enzo life sciences, New York, USA) (1 μg/ml) for 22 h and incubated in a CO₂ incubator. After washing three times with PBS, mice were immunized with cultured cells (B cell-based vaccine) by tail vein injection. As for the comparison group, mice were intramuscularly immunized with BCG (10⁵ CFU/mouse).

To confirm preventive effect of Bvac, mice were either immunized with BCG by intramuscular injection at a 10⁵ CFU/mouse or administered with Bvac by tail vein injection at day 0 for the priming and day 7 for the boost. At day 14, mice were challenged with 10⁷ CFU/mouse of M. kansasii. To confirm the therapeutic effect of Bvac, mice were infected with 10⁷ CFU of M. kansasii per mouse at Day 0, and were administrated with BCG or Bvac at 3 days post-infection. The mice were boosted with Bvac at 7 days post-infection. We analyzed histology, protein levels and bacterial loads from lung and liver after 14 days following M. kansasii infection.

C57BL/6 mice were purchased at 6–7 weeks of age from Charles River Laboratories (Orient Bio Inc., Seongnam, Korea). All animal experiments, including the M. kansasii challenge experiment, were approved by the Institutional Animal Care and Use Committee of Kangwon National University (Permit Number: KW-160201-4). A hypervirulent M. kansasii SM#1 clinical isolate was used for challenge in vivo (14). To induce infection, mice were intravenously injected with M. kansasii (10⁷ CFU/mouse). We checked the bodyweight and survival rate of mice every day following M. kansasii infection.

The lungs and liver of infected mice were isolated for determining the bacterial count in these organs at 2 weeks post M. kansasii infection. Lungs were homogenized in 1X PBS containing 0.04% tween 80 and liver was homogenized in 1X PBS containing 1 mM EDTA (125 mg of tissues/ml). The homogenized supernatants were 10-fold serial diluted in Difco™ Middlebrook 7H9 Broth (BD biosciences, California, USA) containing ADC [Sodium Chloride (Duchefa, BH Haarlem, The Nederlands), Dextrose (SHOWA, Gyoda, Japan), Bovine Albumin Fraction V (MPBio, Santa Ana, USA), Catalase (Sigma-Aldrich, St. Louis, USA)] and each diluent was drop cultured in Difco™ Middlebrook 7H10 Agar (BD biosciences, California, USA) containing OADC [Sodium Chloride, Dextrose, Bovine Albumin Fraction V, Catalase, Oleic acid (Sigma-Aldrich, St. Louis, USA)]. Smear plates were cultured in a 37°C incubator and colonies were counted after 2–3 weeks.

For the isolation of T cells, splenocytes were labeled with CD8α-PE and anti-PE microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer's instructions. The labeled cells were then purified through LS column (Miltenyi Biotec, Bergisch Gladbach, Germany). CD4⁺ T cells were isolated by using the mouse CD4⁺ T cell isolation kit (Miltenyi Biotec, Bergisch Gladbach, Germany). CD11c⁺ DCs were purified by using CD11c⁺ microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany). B220⁺ cells were purified by using CD45R/B220 biotin (BD biosciences, California, USA) and anti-Biotin microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany) from splenocytes of naïve C57BL/6 mice. Purified B cells were transduced with vacESAT6 at a MOI of 1 and/or loaded with 1 μg/ml of αGC and then co-cultured with naïve splenocytes for 24 h. Incubated cells were stained to examine the expression of B220, CD40, and CD86 using antibodies such as APC-conjugated anti-B220 (BD biosciences, California, USA), PE-conjugated isotype control (eBioscience, San Diego, USA), anti-CD40 (Biolegend, San Diego, USA), and anti-CD86 Ab (BD biosciences, California, USA). Cells were analyzed by flow cytometry.

For measurement of intracellular cytokines, dendritic cells and CD4⁺ T cells were co-cultured and stimulated for 3 days with heat-killed H37Rv at 0.1 MOI or overnight with anti-CD3 and anti-CD28 antibody in culture media. H37Rv strain was generously provided by Sang-Nae Cho (Yonsei University). Brefeldin A Solution (1000 x) Thermo Fisher Scientific, Waltham, MA, USA) was added for 4 h before harvest and then harvested cells were stained with PerCP-Cy™5.5 rat anti-mouse CD4 (BD biosciences, California, USA) or PE rat anti-mouse CD8α (BD biosciences, California, USA). Stained cells were permeabilized with IC Fixation Buffer (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer's recommendations. Next, permeabilized cells were stained with TNF-α mAb, APC (Thermo Fisher Scientific, Waltham, MA, USA) and IFN-γ mAb (XMG1.2) PE (Thermo Fisher Scientific, Waltham, MA, USA).

The supernatants of homogenized tissues were analyzed for cytokine production using BD™ Cytometric Bead Array (CBA) Mouse Inflammation kit (BD biosciences, California, USA) according to the manufacturer's instructions.

Total protein lysates of M. kansasii were sonicated with PRO-PREP™ protein extraction solution (iNtRON Biotechnology, Daejeon, Korea). Lysates were boiled at 100°C and proteins were separated by performing SDS-PAGE. Proteins were transferred onto PVDF membranes (Millipore, Burlington, USA) and then blocked with 5% skim milk in TBS with tween 20. Next, the membranes were incubated with anti-ESAT6 primary antibody [11G4] (Abcam, Cambridge, USA) and proteins were detected using HRP conjugated goat anti-mouse polyclonal antibody (Enzo life sciences, New York, USA). Membranes were developed using femtoLUCENT™ PLUS-HRP chemiluminescence detection system (G-Biosciences, St. Louis, USA).

Cells were collected from spleen and stained with markers such as APC-conjugated anti-B220 (BD biosciences, California, USA) and anti-Ly6C (BD biosciences, California, USA) Ab, FITC-conjugated anti-CD11b Ab (BD biosciences, California, USA), PE-conjugated isotype control (eBioscience, San Diego, USA), anti-CD40 Ab (Biolegend, San Diego, USA), and anti-CD86 (BD biosciences, California, USA) Ab. Flow cytometry was performed on a FACSVerse instrument (BD biosciences, California, USA) and data were analyzed using FlowJo software (Flowjo, San Carlos, USA).

Mice were sacrificed and lungs were isolated from each group. They were fixed with 4% formalin overnight. Lung tissues were processed using a tissue processor (Leica, Wetzlar, Germany) and then embedded in paraffin. Paraffin-embedded tissue blocks were cut into 5 μm thick sections and stained with hematoxylin and eosin.

Uninfected or vaccinia virus-infected Vero cells and B cells were cultured in a 37°C incubator for 24 h. Cells were fixed with 4% paraformaldehyde and stained with anti-ESAT6 Ab (Abcam, Cambridge, USA), and further stained with anti-mouse IgG (H+L), F(ab')₂ Fragment (Alexa Fluor® 488-conjugated) (Cell signaling, Danvers, USA) or DyLight™ 405 affinipure donkey anti-mouse IgG (H+L) (DyLight™ 405-conjugated). Cells were visualized by confocal microscopy (Carl Zeiss, LSM880 with Airyscan, Zena, Germany).

Statistical analysis was conducted with GraphPad Prism 5.0 (GraphPad Software, La Jolla, USA). Differences between groups were assessed by the Student's t-test. Comparisons between multiple-groups were carried out by one-way ANOVA analysis of variance followed by the Bonferroni's multiple comparison test. P values < 0.05 were considered as significant at a 95% confidence interval for all analyses.

It has been previously reported that αGC-loaded B cell-based vaccines expressing tumor antigens showed significant antitumor effects in vivo (15, 16). Thus, we decided to adopt this vaccine strategy for the development of preventive and therapeutic anti-mycobacterial vaccine. B cells were transduced with recombinant vaccinia virus expressing ESAT6 (vacESAT6), and the transduced B cells were loaded with αGC. We found that B cells loaded with αGC/vacESAT6 (B/αGC/vacESAT6) (Bvac) increased the expression of co-stimulatory molecules including CD40 and CD86 when they were co-cultured with splenocytes from naïve C57BL/6 mice for 24 h (Figures 1C,D). B cells loaded with αGC or B cells transduced with vacESAT6 increased the expression of CD86, which further increased in B/αGC/vacESAT6 (Figures 1C,D). B cells transduced with vacGFP control virus also increased the expression of CD86 (Supplementary Figure 2) suggesting that infection of vaccinia virus alone could be stimulatory for B cells. These results suggested that activated NKT cells by αGC on B cells as well as vaccinia virus infection activated the B cells to increase the expression of co-stimulatory molecules including CD40 and CD86, which help B cells to function as professional antigen presenting cells to induce effective T cells.

We next determined whether B/αGC/vacESAT6 induced CD4⁺ T cell response in vivo. Groups of mice were immunized with either saline, BCG or B/αGC/vacESAT6, and mice were sacrificed to obtain splenocytes. We analyzed TNF-α- and IFN-γ-producing CD4⁺ T cells after 3 days of co-culturing the splenocytes with dendritic cells pulsed with heat-killed H37Rv. As a result, we found that the percentage of CD4⁺ T cells producing TNF-α⁺ and IFN-γ⁺ was higher in mice vaccinated with B/αGC/vacESAT6 as compared to control and BCG-immunized mice (Figures 2A,B). In addition, IFN-γ production was also increased in the culture supernatant of splenocytes from B/αGC/vacESAT6 group compared to control and BCG group (Figure 2C). These results show that B/αGC/vacESAT6 induced the H37Rv-specific CD4⁺ T cell-mediated cellular immunity which might be critical for the regulation of mycobacterial infection.

M. kansasii is one of the NTM which expresses ESAT6 homolog as a major antigen. We confirmed the expression of ESAT6 in M. kansasii by western blotting analysis (Figure 3A). We found that the bodyweight of mice significantly decreased with intravenous (i.v) injection of M. kansasii (10⁷ CFU/mouse). For humane reasons, mice were monitored two times every day and sacrificed when they weighed <80% of their initial bodyweight. In addition, there was significant lung injury including infiltration of immune cells around bronchial tubes as well as formation of granuloma-like lesions. We also confirmed the presence of bacteria in lungs of M. kansasii-infected mice at 2 weeks after infection. On the contrary, immunization of mice with Bvac ameliorated loss of bodyweight and increased survival rate following M. kansasii infection (Figures 3B,C). In addition, lungs of mice immunized with Bvac had moderate injury with reduced cell infiltration as compared with non-vaccinated mice after M. kansasii infection (Figures 3D,E). Bvac also decreased the bacterial burden in the lungs of M. kansasii-infected mice (Figure 3F). We also analyzed the proportion of NK cells and NKT cells in splenocytes by flow cytometry. We confirmed that Bvac immunization increased the percentage of NKT cells as compared with that of non-vaccinated mice as assessed after M. kansasii infection (Supplementary Figure 3). We also compared the therapeutic effects of Bvac and BCG vaccine in M. kansasii-infected mice. As a result, mice therapeutically treated with BCG and Bvac showed moderate levels of inflammation with reduced cell infiltration and decreased the bacterial burden in the lungs of M. kansasii-infected mice (Supplementary Figures 4A,B). Intriguingly, however, immunization of mice with Bvac significantly decreased the bacterial burden in the liver than BCG immunization after M. kansasii infection (Supplementary Figures 4C,D). Collectively, we established a murine model of infection of M. kansasii and showed that Bvac had preventive effect against M. kansasii expressing ESAT6.

We speculated whether B/αGC/vacESAT6 (Bvac) had therapeutic effects against mice infected with M. kansasii. To evaluate therapeutic efficacy of Bvac, mice were i.v. challenged with 10⁷ CFU/mouse of M. kansasii, and 14 days later, they were injected with Bvac at day 3 and 7 post-infection. When we checked the bodyweight, mice administered with Bvac showed alleviation in loss of body weight as compared to mice infected with M. kansasii (Figure 4A). Further, survival rate increased in mice treated with Bvac (Figure 4B). Histological analysis of lungs of infected mice confirmed the therapeutic effects of Bvac against M. kansasii infection (Figures 4C,D). Bacterial load in the lungs was significantly reduced in mice administered with Bvac as compared to M. kansasii infected mice (Figure 4E). Also, administered with BCG did not show the therapeutic effect, while Bvac administration reduced bacterial loads in lungs of the infected mice (Supplementary Figure 5). Furthermore, production of TNF and IL-6, which were increased in lungs of infected mice following M. kansasii infection, were significantly decreased when administered with Bvac (Figure 5A). Collectively these results suggested that Bvac had a therapeutic effect in mice infected with M. kansasii.

Finally, we assessed the production of intracellular cytokines in CD4⁺ T cells after therapeutic treatment with Bvac in M. kansasii infected mice. IFN-γ-producing T cells are known to play a key role in resistance against various pathogens including M. tuberculosis (17). Mice were infected with M. kansasii (10⁷ CFU/mouse), and i.v. injected with Bvac at day 3 and 7 post infection. After 14 days following M. kansasii infection, lung homogenates were analyzed for the production of TNF, IFN-γ, and IL-6. Interestingly, although the levels of TNF, IFN-γ and IL-6, were highly increased in mice infected with M. kansasii, the levels of TNF and IL-6 were significantly reduced. When CD4⁺ T cells obtained from the spleen of the treated mice were analyzed for IFN-γ production, we found that splenocytes of mice administered with Bvac showed increased production of IFN-γ in CD4⁺ T cells compared to M. kansasii-infected mice (Figure 5B). Consequently, these results suggest that Bvac could induce IFN-γ-producing CD4⁺ T cell response, resulting in preventive and therapeutic effects against M. kansasii expressing ESAT6 (Supplementary Figure 6).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.