Specific pathogen-free female C57BL/6 mice at 5-6 weeks of age were purchased from Japan SLC, Inc. (Shijuoka, Japan), maintained under barrier conditions in a biohazard animal room at Yonsei University Medical Research Center, and fed a sterile commercial mouse diet and water. All animal experiments were done according to the regulation of Institutional Animal Care and Use Committee, Yonsei University Health System.

Mycobacterium tuberculosis H37Rv (ATCC 27294) and M. bovis BCG (Pasteur strain 1173P2) used in this study were prepared as previously described.21 Briefly, M. tuberculosis and M. bovis BCG were grown for about 10 days at 37℃ as a surface pellicle on Sauton medium enriched with 0.4% sodium glutamate and 3.0% glycerine. The surface pellicles were collected and disrupted with 6 mm glass beads by gentle vortexing. After clumps were settled down, the upper suspension was collected and aliquots were stored at -70℃ until use. After thawing, viable organisms were then counted by plating serial dilutions on the Middlebrook 7H11 agar (Difco, Detroit, MI, USA). For inoculation into mice, bacterial suspensions were sonicated briefly in a sonic bath and diluted with phosphate-buffered saline (PBS) to reach desired bacilli numbers.

Culture filtrate protein antigens (CFP) were produced from the culture of M. bovis BCG as previously described.21 Briefly, M. bovis BCG was grown in Sauton medium on an orbital shaker for 7 days. The supernatants were sterile filtered and concentrated via ultrafiltration over an Amicon ultrafiltration stirred cell (Amicon, Danvers, MA, USA) fitted with a PM10 membrane (Millipore, Bedford, MA, USA). Samples were further concentrated to a final volume of about 5 mL by centrifugation in a Savant speed-Vac (Savant Instruments, Holbrook, NY, USA). These preparations were aliquoted in 1 mL samples and stored at -20℃. Residual endotoxin levels were <50 IU/mg as determined by a Limulus amebocyte lysate assay (Cambrex, East Rutherford, NJ, USA).

The vector pGX10 expressing IL-12 or IL-18 was prepared as described previously.16,22 Briefly, pGX10, plasmid DNA vector, was generated by replacing the ampicillin resistance cassette of pTV2 with the kanamycin resistance cassette. pGX10 is composed of a cytomegalovirus (CMV) immediate-early gene promoter, tripartite leader (TPL) sequences, simian virus 40 (SV40) late polyA tail, and an SV40 enhancer, linked in tandem. For generation of pGX10-IL-12, murine IL-12p35 (p35), the internal ribosomal entry site (IRES) of encephalomyocarditis virus, and murine IL-12p40 (p40) were also linked in a tandem, unidirectionally arranged and inserted into pGX10 vector, resulting in pGX10-p35/IRES/p40 (pGX10-IL-12) (Fig. 1). For generation of pGX10-IL-18, murine IL-18 cDNA was provided kindly from Bank for Cytokine Research at Chonbuk University. IL-18 cDNA was inserted into pGX10 vector (Fig. 1). pGX10, pGX10-IL-12, and pGX10-IL-18 plasmid DNA for immunization were prepared by cesium chloride density gradient centrifugation which was performed at 50,000 rpm for 15 h at 20℃ using a Beckman LE80 ultracentrifuge.22

Mice were immunized with experimental vaccines as previously described.22 Mice were given subcutaneously a single dose (2×10⁵ CFU/mouse) of the BCG vaccine in a volume of 200 µL on the back. For DNA vaccination, mice were injected twice at 3-weeks intervals intramuscularly into the quadriceps muscle with 50 µg of pGX10, pGX10-IL-12 or pGX10-IL-18 DNA vaccine constructs in a volume of 200 µL. For the co-immunization of plasmid DNA vaccine constructs and the BCG vaccine, mice were injected together subcutaneously with the BCG vaccine and simultaneously with the plasmid DNA intramuscularly at the first immunization, and then were injected intramuscularly with plasmid DNA on the third week after the first immunization. For negative controls, mice were injected subcutaneously with 200 µL of sterile pyrogen-free saline or, intramuscularly with 50 µg of pGX10, pGX10-IL-12, and pGX10-IL-18 DNA vaccine constructs.

Mycobacterial antigen-specific immune responses were measured as previously described.21 Mice were sacrificed on the fourth week after last immunization, and spleens were removed from three mice in each experimental group. Splenocytes were obtained by preparing single-cell suspension from spleens by dispersion of the tissue with sterilized slide glasses. Erythrocytes were lysed with a solution containing 155 mM ammonium chloride and 10 mM potassium bicarbonate buffer, and cells were thoroughly washed. Isolated cells were cultured in 96-well cell culture plates (Nunc, Roskilde, Denmark), each well containing 2×10⁵ lymphocytes in a 200 µL volume of RPMI 1640 supplemented with 5×10^-5 M 2-mercaptoethanol, 1% penicillin-streptomycin, 1 mM glutamine, and 10% (vol/vol) fetal calf serum. The splenocytes were stimulated in triplicate culture with CFP at final concentration of 10 µg/mL or with medium only. The splenocytes were incubated at 37℃ in the presence of 5% CO₂ for 6 days, the supernatants were taken, and IFN-γ present in culture supernatants was quantified using a mouse IFN-γ OptEIA™ Set (Pharmingen, San Diego, CA, USA) according to the manufacturer's instructions. In brief, enzyme immunoassay (EIA) plates (Costar, Cambridge, MA, USA) were coated overnight with 2 µg/mL monoclonal anti-mouse IFN-γ in polycarbonate buffer (pH9.2) at 4℃. Free binding sites were blocked with PBS containing 0.05% Tween 20 (PBST) and 5% normal goat serum, followed by washing with PBST. Culture superna-tants of day 6 for IFN-γ were added into wells in triplicate, and the plate was then incubated overnight at 4℃. The recombinant mouse IFN-γ was used as a standard. The wells were then washed four times with PBST, and IFN-γ was detected with a biotin-labeled rat anti-mouse IFN-γ monoclonal antibody and phosphate-conjugated streptoavidin. After six times washing with PBST, the enzyme reaction was developed with O-phenylenediamine (Sigma-Aldrich Chemical Co., St. Louis, MO, USA) and hydrogen peroxide. Optical density was measured by an automatic EIA reader (Molecular Devices Co., Sunnyvale, CA, USA) after stopping the reaction by adding 2.5 N H₂SO₄, and the quantity of IFN-γ was calculated by comparing with the results obtained with its standard.

Mice were challenged by aerosol exposure with M. tuberculosis H37Rv using an inhalation device (Glas-Col, Terre Haute, IN, USA) which was calibrated to deliver approximately 200 bacteria into the lungs as previously described.10,22 Mice were challenged on the fourth week after the last immunization, and five mice per group were sacrificed at 30 and 90 days post-challenge and bacterial counts were determined from tissue homogenates of the lung and spleen. The numbers of viable bacteria in the lung and spleen were determined by plating with serial dilution of whole organ homogenates on Middlebrook 7H11 agar (Difco, Detroit, MI, USA). Colonies were counted after 3-4 weeks incubation at 37℃. The resulting values are based on the experimental group of five mice and are presented as means of log₁₀CFU±standard deviation per group.

Lungs were excised from all animals and stored in 10% formalin, and then embedded and stained with haematoxylin and eosin (H&E) for pathological analysis. The level of inflammation in the lungs was evaluated as previously described.10 In brief, H&E stained lung sections were photographed using a microscope (Olympus BX51, Olympus Co., Tokyo, Japan), and the level of inflammation present in the images was analyzed using the ImageJ program (National Institutes of Health, MD, USA). The resulting values are presented as the mean percent of inflamed area from lung sections of five mice per group.

Analyses were performed by the ANOVA test using R statistical software (version 2.6.2). A p-value of less than 0.05 was considered statistically significant.

Production of IFN-γ, a key cytokine of the protective immune response against TB, was examined to investigate whether DNA vaccines encoding IL-12 or IL-18 cytokine were able to augment the protective immune response of the BCG vaccine. Thus, splenocytes from three mice from each group were isolated after four weeks of last immunization and were stimulated with CFP or media for 6 days, and IFN-γ levels in the supernatants were measured. The level of CFP-reactive IFN-γ in the BCG vaccinated mice was approximately 2,000 pg/mL (Fig. 2). The levels of CFP-reactive IFN-γ in the groups co-immunized with the BCG vaccine and IL-12 or IL-18 DNA vaccine were approximately 3,000 pg/mL and 4,600 pg/mL, respectively (Fig. 2), which were significantly higher than that in the BCG vaccine (p<0.01 and p<0.001, IL-12 and IL-18 DNA vaccine constructs, respectively). The levels of CFP-reactive IFN-γ in the pGX10 vector, IL-12 and IL-18 DNA vaccine constructs-immunized mice were below 200 pg/mL as naïve mice (data not shown).

To evaluate the effect of the plasmid DNA vaccine constructs encoding cytokines on the protective efficacy of the BCG vaccine, ten immunized mice per group were challenged by aerosol exposure with a virulent strain of M. tuberculosis, and the protective efficacy of the vaccine candidates was measured by comparing the bacterial burdens in the organs of vaccinated mice to those of naïve mice (Fig. 3).

The protective efficacy was measured at 30 days post-infection, which is the point when M. tuberculosis bacilli undergoes exponential growth and reaches peak in the lungs. The bacterial CFUs in naïve mice at 30 days post-infection were 6.37±0.13 log and 5.41±0.04 log in the lungs and spleens, respectively. There were significant bacterial reductions in the BCG vaccinated mice [1.1 log (p<0.001) in the lungs and 0.67 log (p<0.01) in the spleens, respectively] compared to naïve control group. There were no significant bacterial reductions in the IL-12 or IL-18 DNA vaccine-immunized mice (data not shown). There was a significant bacterial reduction not only in the lungs but also in the spleens of mice co-immunized with IL-12 DNA vaccine and the BCG vaccine (0.3 log, p<0.05; in the spleen). However, there was no significant bacterial reduction in the mice co-immunized with IL-18 DNA vaccine, and the BCG vaccine.

To examine whether IL-12 or IL-18 DNA vaccine construct helps the BCG vaccine prevent the progression of M. tuberculosis infection, mice were sacrificed and bacterial CFUs were measured at 90 days post-infection (Fig. 3). The CFUs in the naïve mice were 6.13 log and 5.62 log in the lungs and spleens, respectively, at 90 days post-infection. The CFUs in the BCG vaccine-immunized mice were increased at 90 days post-infection (5.76 log and 5.00 log in the lungs and spleens, respectively) compared to CFUs at 30 days post-infection. The protective efficacy of the BCG vaccine was reduced to 0.37 log (p<0.05) in the lungs at 90 days post-infection. There was no significant protective efficacy in the IL-12 or IL-18 DNA vaccine-immunized mice at 90 days post-infection (data not shown).

Although CFUs in the mice co-immunized of with IL-12 DNA vaccine and the BCG vaccine was increased at 90 days post-infection compared to at 30 days post-infection, this vaccine induced high significant protective efficacy in the lungs and spleens at 90 days post-infection (p<0.001 both lungs and spleens). Interestingly, however, there were significant bacterial reductions in the lung (0.30, p<0.05) and spleen (0.47 log, p<0.01) compared to the BCG vaccinated mice. The co-immunization of IL-18 DNA vaccine with the BCG vaccine induced significant protective efficacy at 90 days post-infection and there was a little reduction of CFUs in the spleens compared to the BCG vaccinated mice, but it was not significant.

The level of lung inflammation was assessed to evaluate the effectiveness of IL-12 or IL-18 DNA vaccine constructs on the BCG vaccine. Thus, mice were sacrificed at 30 and 90 days post-infection and five H&E stained lung sections per group were analyzed using the ImageJ software. Representative lung sections and the lung inflammation values are shown in Fig. 4 and Table 1. In the lungs of naïve mice, considerable and consolidated inflammation was seen, and the mean inflammation values were 38% and 43% at 30 and 90 days post-infection, respectively. In the BCG vaccinated mice, less inflammation was observed, and the mean inflammation values were significantly lower compared to naïve group (p<0.001 and p<0.01, at 30 and 90 days post-infection, respectively). In the vector DNA, IL-12 DNA, or IL-18 DNA vaccine constructs-immunized mice, the levels of inflammation in the lungs were similar to that of naïve mice group. In the co-immunized mice of IL-12 DNA or IL-18 DNA vaccine constructs with the BCG vaccine, lung inflammations were not improved at 30 days post-infection, but interestingly, the less lung inflammation was observed in the co-immunized mice of IL-12 DNA and the BCG vaccine at 90 weeks post-infection compared to the BCG vaccinated mice (p<0.05).