Recombinant BCG vaccine expressing multistage antigens of Mycobacterium tuberculosis provides long-term immunity against tuberculosis in BALB/c mice

ABSTRACT

Tuberculosis (TB) caused by Mycobacterium tuberculosis (Mtb) persistently kills nearly 1.5 million lives per year in the world, whereas the only licensed TB vaccine BCG exhibits unsatisfactory efficacy in adults. Taking BCG as a vehicle to express Mtb antigens is a promising way to enhance its efficacy against Mtb infection. In this study, the immune efficacy of recombination BCG (rBCG-ECD003) expressing specific antigens ESAT-6, CFP-10, and nDnaK was evaluated at different time points after immunizing BALB/c mice. The results revealed that rBCG-ECD003 induced multiple Th1 cytokine secretion including IFN-γ, TNF-α, IL-2, and IL-12 when compared to the parental BCG. Under the action of PPD or ECD003, rBCG-ECD003 immunization resulted in a significant increase in the proportion of IL-2⁺ and IFN-γ⁺IL-2⁺ CD4⁺T cells. Importantly, rBCG-ECD003 induced a stronger long-term humoral immune response without compromising the safety of the parental BCG vaccine. By means of the protective efficacy assay in vitro, rBCG-ECD003 showed a greater capacity to inhibit Mtb growth in the long term. Collectively, these features of rBCG-ECD003 indicate long-term protection and the promising effect of controlling Mtb infection.

KEYWORDS:

• Tuberculosis
• Mycobacterium tuberculosis
• recombinant BCG vaccine
• multicomponent fusion antigen
• cytokine

Introduction

TB has threatened humanity for thousands of years in the world. Despite the only TB vaccine BCG approved for use has been widely available worldwide for decades, the TB pandemic continues to pose challenges. In 2021, TB accounted for approximately 10 million new cases and 1.6 million deaths in the world.¹ It is estimated that nearly a quarter of the world’s population has been latently infected with Mtb, and approximately 10% of these individuals will develop active TB during their lifetime.² While BCG vaccination offers some protection against severe TB disease forms in children, its effectiveness in preventing the reactivation of latent tuberculosis infection (LTBI) and reinfection in adults is variable and even incapable.³ Clearly, a novel effective TB vaccine used for the prevention of TB reactivation in adults is urgently needed.

Several approaches have been used to design new TB vaccines, including subunit vaccines, live attenuated vaccines, and recombinant BCG vaccines. However, none of them could replace the BCG vaccine now. In the Phase IIb controlled trial of the most hopeful subunit vaccine M72/AS01E, it provided 49.7% protection against progression to active pulmonary tuberculosis in Mtb-infected, HIV-negative adults for at least 3 years.⁴ The result of M72/AS01E is inspiring, but further testing is required to assess its efficacy over longer periods and in a larger population. Recombinant BCG would be a promising way to achieve the goals of the End TB strategy. Different from the subunit vaccines, recombinant BCG has the ability to duplicate, which may represent robust and long-standing immunity. The biggest obstacle in developing recombinant BCG lies in understanding what constitutes protective immunity and which antigens can elicit such immunity.

Identifying essential mycobacterial antigens and determining the ideal antigen composition is a breakthrough for achieving the development of TB vaccines. Mtb primarily stimulates CD4⁺ T cells of which the critical role in TB control has been demonstrated in both animal models⁵ and human.⁶ The Th1 cytokine IFN-γ is considered an indispensable factor in protective immunity. However, the results of MVA85A vaccine trials indicated that robust IFN-γ response alone failed to offer sufficient protection in adults, raising the doubt that the potent IFN-γ simply means better protection.⁷ Thus, the role of multiple immune factors in protection against TB has been further investigated. Studies have highlighted the potential activities of TNF-α, IL-2, IL-12, IL-17, GM-CSF, and so on in contributing to protective immunity.^8–12 Meanwhile, there has been evidence emphasizing the importance of antibody-mediated response, which has aroused great enthusiasm for developing a TB vaccine that can elicit specific antibodies against Mtb.¹³

Here we constructed the rBCG-ECD003, which expresses the fusion protein consisting of the ESAT-6, CFP-10, and the epitopes-rich area of the latent antigen DnaK. Its immunogenicity was evaluated by means of measuring multiple cytokines-secreting abilities, polyfunctional CD4⁺T cell subsets and humoral immune responses in BALB/c mice. Furthermore, we verified the efficacy of rBCG-ECD003 against Mtb growth in mice in vitro and its safety. Overall, the results suggest that rBCG-ECD003 has the potential to enhance the effectiveness of the BCG vaccine in the long term.

Materials and methods

Ethics approval

The animal experiments in this study were carried out in strict accordance with the recommendations in the Guide of the Ethics Committee of the National Institute for Communicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention (No. 2019-CCDC-IACUC-011). The suffering of the mice was minimized during animal immunization and surgery procedures.

Bacterial strains and culture conditions

Mycobacterium bovis (M. bovis) BCG-China and recombinant BCG(rBCG-ECD003)were cultured at 37°C in Middlebrook 7H9 broth supplemented with 0.2% glycerol, 0.05% Tween 80, and 10% oleic acid-albumin-dextrose-catalase (OADC) or on Middlebrook 7H10 agar supplemented with 10% OADC. Kanamycin was added to the media at the final concentration of 25 μg/ml for recombinant BCG strain. Mtb H37Rv was cultured on Roche’s medium.

MHC-II binding epitopes retrieval and prediction of DnaK

For enhancing the immunogenic, the protein sequence of the DnaK in Mtb H37Rv was used for major histocompatibility complex class II (MHC-II) binding T cell epitope retrieval and prediction. The Immune Epitope Database (IEDB) was used for MHC-II binding epitope retrieval, and eight identified MHC-II epitopes of DnaK in human studies were collected. For CD4 epitope prediction, the IEDB MHC-II Binding Predictions (http://tools.iedb.org/main/) was used. The IEDB recommended 2.22 prediction method and recommended the 27 HLA-II alleles were selected. The threshold was set as percentile rank < 1 for prediction reliability. On the basis of the results of epitopes retrieval and prediction, we selected a high epitope density region in DnaK named new DnaK (nDnaK), which contains six identified HLA-II epitopes and three candidate epitopes, ranging from DNA sequence 778 to 1143 (amino acid sequence 260 to 381)in the sequence of DnaK (Figure 1). The higher immunogenicity of nDnaK had been proven by ELISPOT assay in our previous work.

Figure 1. Schematic diagram of T-cell epitope distribution in nDnak and DnaK. The scale unit is amino acid (AA).

Construction of recombinant BCG strain rBCG-ECD003

The Mycobacteria-E. coli shuttle vector pMV361 was used to generate the recombinant BCG which can integrate the heterologous gene into the chromosome of Mycobacteria. The DNA sequence of nRv0350 was derived from the T-cell epitope-rich region of Rv0350. The gene fragment coding for fusion protein ECD003 consisted of three sequentially linked in tandem genes, Rv3875, Rv3874, and nRv0350 with peptide linker sequence coding “GGSGG” added between each two genes. The fusion gene was then synthesized by gene synthesis technology with restriction EcoRIand HindIII digest site at 5’ and 3’ end, respectively. Then, the sequence was inserted into the pMV361 expression vector. The recombinant plasmids were transformed into parental BCG-China by electroporation. The transformed BCG cells were plated on Middlebrook 7H10 agar with kanamycin (25 μg/ml). After 4–5 weeks of growth at 37°C, individual colonies were picked. To identify the rBCG-ECD003 strain, the ECD003-coding gene was confirmed by PCR and DNA sequencing.

Immunization and sample collection

Specific pathogen-free, female BALB/C mice (6–8 weeks of age) were purchased (Beijing HFK Bioscience Co. Ltd.).The mice were kept in specific pathogen-free conditions. At the 0^th week, the groups of mice (12 mice in each group) were immunized subcutaneously (s.c.) on the scruff of the neck with 1 × 10⁶ CFU parental BCG and rBCG-ECD003 in 200 μl of PBS respectively, with 200 μl PBS as the negative controls. At each time point (4^th and 12^th week), 6 mice in each group were euthanized to collect spleen, liver, and lung samples sterilely to analyze. The spleen lymphocytes were isolated and then the cell concentration of lymphocyte suspension was adjusted to 1 × 10⁶ cells/ml. The serum were collected at the 4^th, 8^th, and 12^th week, and stored at −80°C

Luminex multiplex cytokine assay

For multiplex cytokine assay, spleen lymphocyte suspensions were seeded at 1 × 10⁵ cells/well in 96-well U-bottomed plates and stimulated with 10 μg/well of PPD, ECD003, and medium alone as the negative control, respectively. After incubation at 37°C for 24 h, supernatants were harvested by centrifuge at 400 g for 5 min. The cytokine productions including IFN-γ, TNF-α, IL-2, IL-12, IL-4, IL-10, IL-6, IL-17, and GM-CSF were detected by Luminex multiplex cytokine assay according to manufacturer’s instructions (LabEx).

Intracellular flow cytometry assay

To detect specific cytokine-secreting T cell responses, intracellular cytokine staining was performed. At the 12^th week post-immunization, isolated spleen lymphocyte suspensions were stimulated with 10 μg PPD or ECD003 and added with Brefeldin A (8 h，37°C, 5% CO₂). The cells were stained with LIVE/DEAD Fixable Dead Cell Stain and fluorophore-conjugated antibodies against surface markers anti-CD3, anti-CD4, and anti-CD8. Then, intracellular staining was performed with fluorophore-conjugated antibodies against IFN-γ, TNF-α, and IL-2, respectively. Finally, the cells were washed with PBS, and fixed with 4% paraformaldehyde. Each sample was analyzed by flow cytometer, using the gating strategy illustrated in Supplementary Figure S1.

Enzyme-linked immunosorbent assay for IgG antibody

The antigen-specific antibody titer of the immunized mice sera in each group was measured by enzyme-linked immunosorbent assay (ELISA) at the 4^th, 8^th, and 12^th week after immunization. In brief, the plates were coated with PPD or purified ECD003 (2 µg/mL) at 4°C overnight. The free-binding sites were blocked by 200 μl/well PBS containing 5% defatted milk powder. Serum samples were added at serial 2-fold dilutions. Following 100 μl/ml horseradish peroxidase-conjugated rabbit anti-mouse IgG antibodies (1:5000) were added. After washings, 100 μl of tetramethylbenzidine substrate was added to each well. Reactions were terminated with 50 μl/well of 0.3 M H₂SO₄. The PBS group was treated as the negative group and the titer of the experimental groups was determined by the value (OD450 [experimental group]/OD450 [PBS group]) > 2.1.

Mycobacterial growth inhibition assay in vitro

For assessing the protective efficacy of the rBCG-ECD003, Mycobacterial Growth Inhibition Assay (MGIA) in vitro was performed. Briefly, spleen lymphocytes (1 × 10⁵/well) were co-cultured with ultrasonic dispersed Mtb H37Rv (500 CFU/well) for 72 h in 24-well plates. The co-cultures were harvested at 12,000 rpm for 10 minutes, then the supernatants were carefully removed by pipetting. The adherent cells were lysed with the addition of 500 μl sterile water after incubating at room temperature for 10 minutes. After mixing with repetitive pipetting, the lysates were transferred to the corresponding tubes containing the co-cultures. The tubes were pulse vortexed and 50 μl bacterial suspensions were spread-plate on Middlebrook 7H10 agar plate supplemented with 5 μg/ml TCH was used specifically to inhibit BCG growth and eliminate potential influence from residual BCG resulting from the immunization.¹⁴ Three weeks later, the colony-forming units on the Middlebrook 7H10 agar plate were counted.

Recombinant BCG safety assay

For assessing the persistence of BCG in the visceral organs of the immunized mice, the colony-forming unit of BCG was counted after vaccination. Right lung lobes, half of the spleens, and livers were homogenized in 1 ml of PBS after weighing up, following plated 100 μl each sample onto Middlebrook 7H10 agar plate respectively. Following 4 weeks of culturing, colonies were counted.

Histological analysis was also determined by using H&E staining. The immunized mice’s visceral organs including left lung lobes, spleens, and livers were fixed with 4% paraformaldehyde and embedded into paraffin blocks. After cutting into serial sections, processed organs were stained with H&E and performed microscopic examination.

Statistical analysis

Statistical analysis was performed by means of SPSS 26.0 software. Two group analyses were carried out with Student’s t-test. For multiple groups comparison, One-Way ANOVA with Turkey’s multiple comparison tests was used. The difference was considered significant when the P value was <  .05.

Results

Construction of recombinant BCG rBCG-ECD003

The Mycobacteria-E. coli shuttle vector pMV361 was used to take the DNA fragment product which coded the fusion protein ECD003 (Figure 2a). The recombinant BCG (rBCG-ECD003) colony growing in the 7H10 medium containing selective antibiotic Kanamycin was confirmed by specific primers. The PCR analysis confirmed that the recombinant vector pMV361-ECD003 has taken the target sequence integrated into the chromosome of rBCG-ECD003 (Figure 2b).

Figure 2. Construction of rBCG-ECD003. (a) Schematic of the mycobacteria-E. coli shuttle vector pMV361-ECD003. (b) Agarose gel electrophoresis of the PCR products amplified from rBCG-ECD003, transformed with pMV361 shuttle vectors. Lane 1, DNA marker; lane 2, the PCR product amplified from rBCG-pMV361; lane 3, the PCR product amplified from rBCG-ECD003.

Cytokine production profile induced by rBCG-ECD003 in spleen

In order to investigate comprehensive immunity after rBCG-ECD003 vaccination in mice, we compared the production of nine cytokine profiles at the 4^th and 12^th week. As expected, major tested cytokines production in the PBS group was kept at the lowest levels among all groups except for TNF-α. As shown in Figure 3a, after stimulation with ECD003 the rBCG-ECD003 group produced significantly higher Th1 cytokine IFN-γ and IL-2 compared to the parental BCG group at both 4^th (both p < .01)and 12^th week (both P < .01) respectively, while IL-12 enhanced at the 4^th week (P < .05) and the TNF-α production elevated at the 12^th week (P < .05). In terms of Th2 cytokines, the secretion levels of IL-6 in the rBCG-ECD003 group revealed a significant increase compared to that of the parental BCG group at the 4^th week (P < .01), while both IL-4 and IL-10 were not significantly increased. There was no significant difference in the levels of both IL-17 and GM-CSF among the three groups. For stimulation with PPD in Figure 3b, rBCG-ECD003 also elevated the level of multiple Th1 cytokines including IFN-γ, IL-2, and IL-12 in vaccinated mice compared to the parental BCG at the 4^th week (all P < .01) and 12^th week (IFN-γ P < .01, IL-2 P < .05, and IL-12 P < .05), respectively. We observed that TNF-α production was also enhanced in the rBCG-ECD003 group at the 12^th week (P < .01). Higher concentrations of Th2 cytokines IL-4 (p < .01), IL-6 (p < .05), and IL-10 (p < .05) in the rBCG-ECD003 group were only observed at the 4^th week when compared with the BCG group. In regard to the innate immune-related cytokines, rBCG-ECD003 showed a trend to induce higher IL-17 (p < .05) and GM-CSF (p < .01) levels compared to BCG groups at the 4^th week, but not shown at the 12^th week.

Figure 3. Superior cytokines production profile induction in the spleen lymphocytes of the BALB/c mice immunized with rBCG-ECD003. The cytokines production profile was analyzed in isolated spleen lymphocytes at the 4^th and 12^th week after immunization. 1 × 10⁵ cells were restimulated with the fusion protein ECD003 (a) or PPD (b) and the concentration of cytokines (IFN-γ, TNF-α, IL-2, IL-12, IL-4, IL-6, IL-10, IL-17 and GM-CSF) was analyzed with multiplex cytokine assay. Results were shown as the mean ± SEM (n = 3) per group. Statistical significance was determined by ANOVA with Tukey’s multiple comparisons tests (*p < .05, **p < .01).

Polyfunctionality CD4⁺T cells induced by rBCG-ECD003

To verify more details about the long-term T cell immunity response, the polyfunctional CD4⁺T cell subsets were assessed at the 12^th week. After ECD003 stimulation, the spleen lymphocytes showed a preferential accumulation of single-positive IFN-γ⁺ CD4⁺T cell subset in all groups, with the highest population induced by the parental BCG vaccine (Figure 4a). The rBCG-ECD003 induced a significantly higher percentage of ECD003-specific single-positive IL-2⁺ CD4⁺T cells (p < .01) and TNF-α⁺ CD4⁺T cells (p < .05) than the BCG strain. For double-positive T cells, while BCG induced a slightly higher proportion of double-positive IL-2⁺TNF-α⁺ CD4⁺T cells and IFN-γ⁺TNF-α⁺ CD4⁺T cells, rBCG-ECD003 vaccination led to an accumulation of more ECD003-specific IFN-γ⁺IL-2⁺ CD4⁺T cells than BCG group (p < .05). From the pie charts, we also observed an obvious increase in single-positive IL-2⁺ CD4⁺T cell proportion in positive CD4⁺T (Figure 4b). Similar patterns were observed for PPD stimulation, where the IFN-γ⁺ CD4⁺T cell subset constituted the major part of CD4⁺T cells (Figure 4c). We also found a large population of IL-2⁺ CD4⁺T cells in rBCG-ECD003 vaccinated mice, which was significantly higher than the PBS and BCG group, respectively (p < .01). Besides, the mice vaccinated with rBCG-ECD003 exhibited an increased proportion of double-positive IFN-γ⁺IL-2⁺ CD4⁺T cells (p < .01). The pie charts reflect there is a wider variety of combinations of cytokine production in CD4⁺T cells in rBCG-ECD003 groups (Figure 4d). The main increase was observed in the proportion of IL-2⁺ CD4⁺T cells and IFN-γ⁺IL-2⁺ CD4⁺T cells. However, there was no significant difference in the polyfunctional CD8⁺T cells between the BCG group and the rBCG-ECD003 group (Supplementary Figure S2).

Figure 4. Vaccination with rBCG-ECD003 enhanced long-term responses of polyfunctional CD4⁺T cells in the spleen lymphocytes. Intracellular cytokine staining was performed in isolated spleen lymphocytes at the 12^th week post-vaccination. The proportion of cytokine-producing CD4⁺T cells after stimulation with the ECD003 (a) and PPD (b) respectively were shown as the mean ± SEM (n = 3) per group. Pie charts reflected the average proportion of ECD003-responding (c) and PPD-responding (d) CD4⁺T cells with each combination of cytokine production. Statistical significance was determined by ANOVA with Tukey’s multiple comparisons tests (*p < .05, **p < .01).

rBCG-ECD003 confers superior humoral immunity in BALB/c mice

In our study, we measured the serum IgG titer of the vaccinated mice using ELISA at the 4^th, 8^th, and 12^th week after immunization. Of note, the rBCG-ECD003 group produced higher levels of IgG titers at the 8^th (p < .01) and 12^th week (p < .05) compared to the BCG group respectively (Figure 5a). Furthermore, as expected, BCG consistently elicited a modest IgG response against PPD. Compared to the IgG titer in the BCG group, rBCG-ECD003 generated significantly stronger levels of IgG antibodies against PPD at the 8^th week (p < .05) and 12^th week (p < .05; Figure 5b), respectively.

Figure 5. rBCG-ECD003 conferred superior humoral immunity in BALB/c mice. Assay of IgG antibodies against ECD003 (a) and PPD (b) in the sera of the immunized BALB/c mice. Serum samples were collected at the 4^th, 8^th, and 12^th week after immunization, and the titers of IgG were detected by ELISA. The results were shown as the Log₂ of the mean ± SEM (n = 6) per group. Statistical significance was determined by Student’s t-test. (*p < .05, **p < .01).

Mycobacterial growth inhibition assay in vitro

The protective efficacy of rBCG-ECD003 was evaluated with MGIA in vitro. At the 4^th week post-vaccination, the bacterial numbers were similar and remained high in all groups. Both the parental BCG and rBCG-ECD003 did not exhibit significant protection compared to the PBS group (Figure 6a). However, at the 12^th week, there was a noticeable improvement in inhibiting Mycobacterial growth in both the parental BCG group (mean 765 CFU) and rBCG-ECD003 group (mean 290 CFU) compared to the PBS group (mean 1223 CFU). The mean CFU reduction for BCG was 458 CFU and for rBCG-ECD003 was 933 CFU. Notably, the rBCG-ECD003 vaccinated mice exhibited significantly better control compared to the parental BCG group (p < .05; Figure 6b).

Figure 6. The protective efficacy of rBCG-ECD003 in vitro. The BALB/c mice (n = 6) were s.C. immunized with PBS, BCG, and rBCG-ECD003 respectively. The spleen lymphocytes were isolated 4th and 12th week post-immunization and co-cultured with ~ 500 CFU of mtb H37Rv. The bacterial load results of the 4th week (a) and 12th week (b) were presented as mean ± SEM. Statistical significance was determined by ANOVA with Tukey’s multiple comparisons tests (*p < .05, **p < .01).

Safety analysis of rBCG-ECD003 vaccine

Following vaccination, we investigated the persistence of BCG strains in the main organs of the mice. In both groups, no mycobacteria were recovered from the liver. At 4 weeks after vaccination, both BCG and rBCG-ECD003 persisted in the spleen and the lung tissue, only the spleen tissue of the rBCG-ECD003 vaccination group showed a higher bacteria burden (p < .01; Figure 7a). On the contrary, at the 12^th week, only in one spleen sample from the rBCG-ECD003 was found a small amount of bacterial residue. In histological analysis, no obvious pathological changes were observed in the liver, spleen, and lung of the PBS group. Both mice infected with rBCG-ECD003 and BCG could be found with certain histopathological lesions in the liver at the 4^th and 12^th week, including hepatocellular necrosis, hepatocellular degeneration, and hepatic plate derangement. These adverse effects tends to recovery, while less hepatocellular necrosis could be seen at 12^th week(Figure 7c). Although immunization with BCG and rBCG-ECD003 resulted in slight hepatic injury in infected mice, damage to the basic structure of the liver was not observed in both groups. In contrast, significant pathological changes were not observed in the spleen. Both the BCG and rBCG-ECD003 groups exhibited only partial infiltration and slight alveolar septum thickening in the lungs at the 4^th week, which resembled the lung architecture of the PBS group. This finding suggests that there was modest tissue damage.

Figure 7. The safety analysis of the rBCG-ECD003 vaccine after immunization. The BALB/c mice were immunized with 1 × 10⁶ CFU of the parental BCG or rBCG-ECD003. At the 4th and 12th week, the mice were sacrificed, and the parental BCG and rBCG-ECD003 CFU recovered from lung (a), and spleens (b) The liver, spleen, and lung samples were collected and performed with H&E staining (c). The typical pathological changes in each group were shown.

Discussion

It is widely accepted that BCG-induced immunity may fade over time and may not be sufficient to resist the risk of Mtb reactivation or emerging Mtb infection in adults over a long period.¹⁵ Developing a novel effective TB vaccine is considered the best way to control this global TB epidemic. Unfortunately, lots of novel subunit vaccine candidates have failed to meet the desired effects, which may be related to its limited number of effective antigens against Mtb infection. Some speculated that a single antigen could tend to be caught in the trap of Mtb’s escape mechanisms and not be recognized by a broad population. This study is a step toward developing a recombinant BCG, rBCG-ECD003, which expresses a new composition of immunodominant multistage antigens, and testing its efficacy based on multiple criteria.

The specific feature of rBCG-ECD003 is its expressing fusion protein ECD003, which is composed of three multistage antigens: ESAT-6, CFP-10, and nDnaK. The early secreted antigens ESAT-6 and CFP-10 were absent from all BCG strains. In three mice models, the recombinant BCG that secret ESAT-6 and CFP-10 exhibited greater protection compared to BCG.¹⁶ However, it has been shown that during chronic TB infection, the ESAT-6-specific CD4⁺T appeared more differentiated and likely to undergo functional exhaustion, which may lead to loss of protective capacity in long term.¹⁷ A subunit vaccine H56, which contains early antigens Ag85B, ESAT-6, and a latent antigen Rv2660c, has shown greater protective efficacy in the late stage of Mtb infection than Ag85B-ESAT-6 vaccine proven the superiority of multistage antigens strategy.¹⁸ To prevent the loss of protection in the later stages of Mtb infection, we determined to combine the ESAT-6 and CFP-10 with a latent-stage antigen. Rv0350 is an essential gene coding the chaperone protein DnaK which helps in proper protein folding and bacterial growth in Mycobacteria.¹⁹ It was observed that DnaK had increased abundance in a dormancy model that mimics Mtb latent infection state in vivo.²⁰ This suggests that if a strong DnaK-specific immune response could be induced in LTBI, it may help to prevent the reactivation of tuberculosis. In order to further elevate the immunogenicity of DnaK, we select an epitope-rich region nDnaK This epitope-based vaccine has been widely used for vaccine development and has proven the superiority of such a strategy.²¹ Our previous study has demonstrated higher immunogenic of nDnaK. Thus, we hypothesize that a combination of these multistage antigens has potentially protective immune effects as an efficient vaccine against Mtb infection in the long term.

Multiple studies have underlined the necessity of IFN-γ in controlling Mtb expansion, and the enhanced production of IFN-γ has been associated with the vaccine’s protective efficacy.²² In contrast, Th2 cytokine IL-4 has been found to undermine bacterial containment.²³ In multiplex cytokine test assay, we observed an enhanced capacity of induction of IFN-γ in the rBCG-ECD003 group. Despite rBCG-ECD003 causing a rise in IL-4 at the 4^th week after immunization, the levels of IL-4 still remained remarkably low and this effect diminished in the long term (Figure 3). Our results demonstrated that rBCG-ECD003 tended to primarily induce a Th1-type immune response rather than a Th2-type response in the long term. Those findings support our hypothesis that overexpression of fusion protein ECD003 would improve the immunogenicity of BCG vaccine. In most of the early testing of the novel TB vaccine, robust IFN-γ induction was placed as the priority consideration factor for evaluating vaccine efficacy. However, due to the unsatisfying results of the powerful IFN-γ-inducing vaccine MVA85A, the role of IFN-γ remains controversial. One possible explanation for the ambiguous effects of IFN-γ is that it may serve as a threshold for protective immunity but cannot fully represent it.

Th1 cytokine IL-2 is considered to be instrumental in the expansion of T cells, which contribute to long-term protective immunity by stimulating the secondary expansion of memory T cells.⁹ In a study of Mtb72F/AS02A, the cynomolgus monkeys taking the Mtb72F/AS02A as a booster for BCG produced more IL-2 compared to the groups receiving BCG alone, companying with better protection.²⁴ Another Th1 cytokine TNF-α is considered to cooperate with IFN-γ to confer protection against Mtb, making it an important indicator for the prediction potential effectiveness of TB vaccines.⁸ IL-12 is a key immunoregulatory cytokine that stimulates Th1 T cell activation, especially IFN-γ production.²⁵ Except for the initial activating of the cell-mediated immunity, it helps to control chronic Mtb infection.¹⁰ In our study, we found that rBCG-ECD003 can induce robust secretion of multiple Th1 cytokines besides IFN-γ including IL-2, TNF-α, and IL-12 in mice (Figure 3). Those Th1 cytokines were considered to be the basis for controlling Mtb infection. However, these powerful and comprehensive Th1 cytokines profiles were not detected in the BCG group. Although BCG-vaccinated mice showed an elevated ability to secrete IL-2 and TNF-α when compared with the PBS group at the 12^th week, the levels remained weak. Consistent with the tendency of IL-4 levels, Th2 IL-6 and IL-10 in rBCG-ECD003 groups were temporarily elevated at the 4^th week but not shown at the 12^th week (Figure 3b). IL-6 and IL-10 have been associated with the risk of deteriorating cell-mediated immunity. During Mtb infection, IL-6 will lead to pathology while the production of IL-10 is conducive to immune evasion by Mtb.²⁶Based on current research, a higher anti-inflammation Th2 cytokines level may be more thought to be conducive to Mtb growth in vivo. Meanwhile, innate immune-related cytokines IL-17 and GM-CSF also contribute to the limitation of early Mtb infection.^11,12 The results showed that the levels of PPD-responding IL-17 and GM-CSF in mice’s spleen lymphocytes were higher in the rBCG-ECD003 group compared to the BCG group at the 4^th week, but they didn’t present significant difference at the 12^th week. Based on these findings, we conclude that the administration of rBCG-ECD003 can induce more powerful and comprehensive cytokine profiles than BCG and may promote the anti-mycobacterial defense in mice.

The great impact of CD4⁺T cells on immune responses against Mtb has been extensively discussed, with a focus on the capacity of different CD4⁺T cell subsets. Polyfunctional CD4⁺T cells represent enhanced cytokine-secreting ability and the function of aggregation at the site of infection. A correlation between polyfunctional IFN-γ⁺IL-2⁺TNF-α⁺ CD4⁺T cells and long-term protective immunity in mice has been reported.²⁷ However, the MVA85A vaccine, known for being great at inducing IFN-γ⁺IL-2⁺TNF-α⁺ CD4⁺T cells, did not boost the efficacy of BCG vaccination, implying that its actual role may differ from expectations.⁷ Likewise, another study revealed that while 80–90% of TB patients have detectable IFN-γ⁺IL-2⁺TNF-α⁺ CD4⁺T, only 10–15% of LTBI subjects showed the same.²⁸ Our data suggested that rBCG-ECD003 significantly upgraded the IL-2⁺ CD4⁺T cells and IFN-γ⁺IL-2⁺ CD4⁺T cells, and there was no expansion of the triple-express IFN-γ⁺IL-2⁺TNF-α⁺ CD4⁺T subsets (Figure 4). This bias in favor of inducing IL-2⁺ CD4⁺T and IFN-γ⁺IL-2⁺ CD4⁺T cells, which may be related to protection, was not presented in MVA85A. An analysis of Mtb-infected humans found that sputum smears-positive TB patients, who have a higher mycobacterial load, had lower levels of IL-2-producing CD4⁺T cells compared to LTBI and sputum smears-negative TB patients.²⁹ Thus, the results suggest that the rBCG-ECD003 appears to generate more effective CD4⁺T cell phenotypes and has the potential to confer long-term protection against Mtb infection. The pie charts of cytokine production profiles for CD4⁺T cells also reflected that rBCG-ECD003 could induce magnitude higher proportions of CD4⁺T cells expressing IL-2 alone or IFN-γ⁺IL-2⁺. In contrast, the cytokine production profiles induced by BCG were mainly dominated by IFN-γ⁺ CD4⁺T cells, which aligns with the findings of Komine-Aizawa S et al.³⁰ Additionally, IFN-γ⁺ CD4⁺T cells are considered to have an effector phenotype, while IL-2-expressing T cells tend to exhibit a central memory phenotype in BCG-vaccinated infants.³¹ However, the higher differentiation of effector phenotype CD4⁺T leads to a reduced capacity to enter the infected site.³² Apart from CD4⁺T cells, recent studies have highlighted the contribution of CD8⁺T cells to the protective immunity against TB.^30,33,34 Unfortunately, we did not observe a preference for the induction of polyfunctional CD8⁺T cells in the rBCG-ECD003 group which may be attributed to inferior activation of cross-priming capacity by mycobacteria.

Humoral immunity has historically been overlooked in the development of both clinical and preclinical TB vaccines. It was reported that parts of healthcare workers exposed to Mtb can produce protective antibodies that provide moderate defense against Mtb in mouse models, which highlighted the important function of antibodies.³⁵ Despite the significance of antibody-mediated immunity, the progress in developing TB vaccines that specifically target antibody responses has been limited. The higher levels of IgG specific to the multistage fusion protein ECD003 and PPD observed at the 8^th and 12^th week in the rBCG-ECD003 group indicate a stronger humoral immunity in the long term (Figure 5). Antibodies not only directly neutralize the bacteria, but may also enhance phagocytosis, phagolysosome formation, inflammasome activation, and cytotoxic activity of NK cells. Distinct from acting alone, humoral immunity may require interaction with CD4⁺T cells to fulfill its true role in protection against TB.³⁶ Since rBCG-ECD003 is a potent inducer of CD4⁺T cells, this strong humoral immunity may further enhance the protection provided by T-cell immunity, thereby restricting the Mtb growth of in vivo.

Considering the good performance of rBCG-ECD003 in multiple indicators, we hypothesize that it would surpass the protection provided by the parental BCG vaccine. Previous reports have shown a strong correlation between MGIA outcomes and the protection conferred by BCG in vivo after aerosol Mtb infection.^37,38 Thus, it has been widely used for the early evaluation of TB vaccine efficacy.^39,40 In our study, we found that vaccination with rBCG-ECD003 greatly enhances the ability of spleen lymphocytes to control Mtb growth in the long term. We observed a nearly 1.6-fold reduction in CFU compared to the BCG group. Despite the results from the multiplex cytokines assay showing the higher PPD-specific IFN-γ production in the rBCG-ECD003 group at both the 4^th week and 12^th week post-immunization, the greater inhibition of Mtb growth was observed only at the 12^th week post-immunization (Figure 6b). This phenomenon may be attributed to the potent TNF-α secreting ability in the rBCG-ECD003 groups at the 12^th week post-vaccination, as depicted in Figure 3b. Meanwhile, the lack of control of Mtb growth at the 4^th week may also be caused by the robust Th2 response and the high burden of mycobacteria load after immunization in mice. Therefore, it is believed that rBCG-ECD003 has the potential to enhance long-term protection, although further testing in more Mtb-challenging animal models is required to confirm its efficacy.

Due to the ESAT-6 and CFP-10 being proven virulence factors, the safety of the BCG vaccine may be compromised when they are introduced. According to the results from the mycobacteria burden assay, higher CFU numbers were only observed in the rBCG-ECD003 group at 4^th week post-vaccination (Figure 7a). At the 12^th week, the majority of mice in rBCG-ECD003 cleared the bacteria, indicating that immunocompetent mice could control bacterial growth. We also conducted the histological analysis to determine if the increased immunogenicity leads to undesirable pathology. The histopathology analysis presented in Figure 7c revealed that both BCG and rBCG-ECD003 had moderate adverse effects in vaccinated mice. These effects mainly manifested as hepatocellular degeneration, and the pathology tended to recover by 12 weeks. These findings are consistent with the results reported by Shoulah SA et al., which showed that subcutaneous vaccination of BCG vaccine in mice resulted in moderate adverse effects in the lungs and liver, indicating that BCG vaccine has mild virulence in mice models.⁴¹ Considering the well-established safety profile of BCG in humans and the effective control of bacterial growth in mice, we believe that these reactions observed in the two BCG vaccines were self-limiting. This finding confirms the safety of rBCG-ECD003, thereby supporting its potential use in human vaccinations in the future.

Overall, our study introduces a new recombinant BCG vaccine rBCG-ECD003, which incorporates three immunodominant multistage antigens into the parental BCG. The administration of the rBCG-ECD003 vaccine enhances immunogenicity and stimulates the production of robust multiple Th1 cytokines, along with long-lasting humoral immunity. In contrast to previous studies that primarily focused on triple cytokine-expressing CD4⁺ T cells, our findings demonstrate that rBCG-ECD003 predominantly induces other protective factors, such as the proportion increase of single IL-2⁺ and IFN-γ⁺IL-2⁺ CD4⁺T cells, which are associated with long-term protection. The rBCG-ECD003 vaccine also exhibits a greater capacity to control Mtb growth in vitro in the long term compared to the parental BCG, without obviously elevating virulence in mice. However, further testing is required to determine the exact level of protection provided by rBCG-ECD003 against Mtb infection in animal models. Considering these characteristics, we believe that rBCG-ECD003 holds promise for enhancing long-term protection and can contribute to the prevention of Mtb infection and TB reactivation in the LTBI population.

Author contributions

DF, RW, and KW designed the study. HL contributed to the epitope prediction. DF, XF, RW, BC, and CQ performed experiments. DF and ML analyzed and interpreted data. KW provided reagents. DF and KW wrote the manuscript. All authors contributed to the article and approved the submitted version.

Acknowledgments

We would like to thank the collaboration of the Laboratory Animal Center, Chinese Center for Disease Control and Prevention, for their help and support in the animal experiments.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was financially supported by the projects of 2018ZX10731301-002. The funders had no role in the study design, data collection and analysis, manuscript preparation or decision to publish.